Richard Donchian

As of: 2026-07-18T23:28:59Z

Task: T0625 | 078-richard-donchian | B-philosophy

Source note: Donchian left no book, the original 1957 article is not openly digitized, and the complete newsletter run was not recovered. This reconstruction therefore ranks the surviving article excerpts and rule sheets first, then named colleague testimony, contemporaneous fund evidence, and modern strategy-family research. Modern channel defaults, Turtle rules, and volatility-sizing conventions are not treated as Donchian's own.

Core Worldview

Richard Donchian's mature worldview begins with a refusal to decide that a market is inherently too expensive or too cheap. Price can keep moving after it looks unreasonable, so the trader's task is not to call the final top or bottom. It is to recognize that a trend has begun, join it with limited risk, and remain until contrary market evidence appears. His 1957 formulation says a good trend method should cap the loss without capping the gain: reverse or close when an opposing trend becomes evident, but maintain the position while the existing trend persists (Ostgaard's documented 1957 excerpt).

That doctrine was learned through failure, not asserted from theory. In a reproduced 1982 Forbes account, Donchian recalled that an early, valuation-driven cocoa short worked, encouraged him to short supposedly overvalued coffee, and was followed by a ruinous rise in coffee. His lesson was that there is no absolute price ceiling or floor. The same account says Futures, Inc. later traded as low as $0.04 before dissolution; that is a reported low before dissolution, not an audited lifetime return or terminal price (1982 Forbes excerpt).

The mature philosophy is therefore reactive rather than predictive, but it is not the pure price-only machine of later legend. The Twenty Guides combine trend direction with volume, relative strength, chart formations, market background, capitalization, activity, and statistical reports. Barbara Dixon's 1978 presentation also warns that successful use requires interpretation and that even the major trend can be unclear (Dixon, 1978 scan). Donchian's worldview is best described as disciplined trend participation inside a broader technical judgment framework.

The Edge - What Markets Misprice And Why

Donchian's operational edge was persistence. Once movement becomes established, it often continues long enough for an asymmetric process to work: many false starts are cut off, while a minority of sustained moves remain open. He did not need to know the eventual destination. A moving-average penetration, range break, volume expansion, or relative-strength signal was evidence that price had already begun to reveal the opportunity.

His surviving work does not provide a developed behavioral-finance theory for why trends exist. Modern evidence supplies plausible mechanisms but should not be retrofitted as his own explanation. Moskowitz, Ooi, and Pedersen find one- to 12-month return persistence followed by partial longer-horizon reversal across 58 liquid futures and forwards, consistent with initial underreaction and delayed overreaction; they also find both information shocks and futures-market pressure contribute (Moskowitz, Ooi and Pedersen, 2012). Hurst, Ooi, and Pedersen add herding, central-bank behavior, and capital frictions as possible sources of delayed adjustment (Hurst, Ooi and Pedersen, 2013). These studies support the family resemblance of the edge, not Donchian-specific alpha.

Diversification completes the logic. No single commodity is guaranteed to trend, and the few exceptional moves cannot be identified in advance. Donchian said he departed from the era's commodity specialists by looking across markets together, combining broad participation with cutting losses and following trends (Ostgaard / Jobman 1980 excerpt). The edge is thus not a magic entry parameter. It is persistence plus payoff asymmetry, repeated across enough markets for the outliers to matter.

Process: Idea Sourcing To Sell Discipline

Idea Sourcing And Research

Ideas began with observable market action across a broad commodity universe. Donchian manually studied historical prices, moving averages, ranges, chart patterns, volume, relative strength, and market background. His 1960 article High Finance in Copper is important precisely because it joined fundamental copper context to technical trading rules; its publisher record does not support the later claim that he rejected all fundamental information (Financial Analysts Journal DOI record).

Research was iterative. He tested rules over long histories, compared parameters, and accepted that the chosen lengths were not sacred. In the surviving posthumous transcript of his 5/20 article, he says five and 20 days were not necessarily the best lengths and invites tests of exponential or weighted averages and different price inputs. His practical stopping rule was not perfection but a comparatively simple method that made money on balance over a long test and could then be followed consistently (5/20 article transcript).

Valuation And Entry

There is no intrinsic-value step in the mature trading process. Valuation may inform background, but it does not authorize a top or bottom call. Entry requires market confirmation. The Twenty Guides favor strong markets on the long side and weak ones on the short side, volume-backed moves out of inactivity, and with-trend entries after a one-day countertrend move rather than chasing the preceding three-day move (Dixon, 1978 scan).

The famous 5/20 method is not a bare crossover. The 20-day average defines the basic trend, but a cross counts only after a commodity-specific penetration threshold and comparison with earlier penetrations. Alternative rules use 15- and 25-close extremes. The faster five-day average mainly governs filtered liquidation and reinstatement. Signals are delayed according to weekday and holiday rules rather than executed uniformly at the cross (5/20 article transcript, computerized decision tree). The actual design tries to filter noise while remaining responsive to a trend change.

Sizing

The defensible sizing evidence is qualitative. Donchian advises light commitments when the market position is uncertain, adequate capital, and stops related to the chart formation. He also advises a larger long than short position when a broad uptrend makes the percentage arithmetic asymmetric (Dixon, 1978 scan). No primary source recovered for this task specifies a universal percentage risk per trade, volatility-normalized unit, portfolio heat ceiling, or leverage target. Those features belong to later systematic descendants and should not be back-attributed to him.

Portfolio Construction

Broad diversification is a requirement, not decoration. Donchian warned that following the 5/20 method in only one or a few futures increased risk inordinately. The portfolio supplies multiple independent chances to encounter a large trend and reduces dependence on one contract's noise (5/20 article transcript).

But the original portfolio doctrine is incomplete by modern standards. No reliable evidence defines correlation buckets, sector caps, equal-risk allocation, or a formal aggregate drawdown limit. Donchian's portfolio construction can be reconstructed as broad multi-commodity exposure with lighter commitments in uncertainty; it cannot be responsibly converted into an exact modern risk budget.

Sell Discipline

The sell discipline expresses the philosophy most clearly: losses are bounded, winners are not assigned a target, and exit follows evidence rather than comfort. Stops can protect profits or limit losses, and market orders are preferred for closing because certainty of exit matters more than an ideal price (Dixon, 1978 scan). In the 5/20 method, an opposing 20-day signal can reverse the position, while filtered five-day penetrations can take it temporarily out and later reinstate it (computerized decision tree).

This requires surrendering part of an open profit. Waiting for reversal evidence means neither buying the exact low nor selling the exact high. Donchian treats that lag as the cost of avoiding a profit target that would amputate the rare long sweep.

Risk Management

Donchian's risk system has four recoverable layers. First, define a loss-limiting exit before hope takes over. Second, reduce commitment when trend evidence is ambiguous. Third, diversify widely enough that one contract cannot determine survival. Fourth, hold adequate capital for repeated whipsaws and delayed exits. These principles are explicit; exact portfolio mathematics are not (Dixon, 1978 scan, 5/20 article transcript).

The distinction between strategy return and investor return is essential. A later CTA study finds that, from 1994 to 2012, gross excess returns were positive while investor net excess returns were statistically indistinguishable from zero; it also documents voluntary-reporting, backfill, survivorship, and graveyard biases (Bhardwaj, Gorton and Rouwenhorst, 2014). This later industry evidence is not a judgment on Donchian's rules alone. It is a warning that turnover, fees, selection bias, and vehicle structure can consume a plausible gross edge.

Futures, Inc. supplies the harsher historical warning. The reproduced Forbes account says it fell from a $10 offer to a reported $0.04 low before dissolution, even though that same failure helped drive Donchian away from valuation calls (1982 Forbes excerpt). A sound maxim does not prove sound sizing, continuous adherence, or a client-suitable vehicle.

Temperament And Psychology

The required temperament is patient, probabilistic, and resistant to public opinion. The trader must buy after an advance has begun, short after a decline has begun, tolerate looking late, accept frequent small losses, and remain with a winner when taking the profit would feel safer. The Guides put loss limitation above every other rule, counsel against immediate action on a widespread view, and recommend waiting for clear movement rather than forcing marginal trades (Dixon, 1978 scan).

Discipline is harder than rule discovery. Ed Seykota, who worked with Donchian and coded the 5/20 system, said the Twenty Guides were insufficient to define a complete system and contained internal conflicts; he also recalled that Donchian took some signals and skipped others (Seykota FAQ). That testimony makes the psychological lesson more credible, not less: even the author of the discipline could find consistent execution difficult.

Humility about parameters is part of temperament too. A trader should neither optimize endlessly nor worship the original setting. The goal is a robust method whose behavior and drawdowns the operator can actually tolerate.

Evolution Over Career

The 1934 Twenty Guides were compiled for stocks; no contemporaneous publication record was located. Their chart-pattern and Dow Theory vocabulary shows an early hybrid technician. In 1957, Donchian articulated the asymmetric trend principle through moving-average and swing-trading examples. In 1960, High Finance in Copper joined commodity fundamentals to a short-range technical rule. The 5/20 method was used from the early 1960s and published in detailed form in 1974; Dixon's 1978 article then republished the older Guides for commodity traders (Ostgaard's historical source map, Dixon, 1978 scan, High Finance in Copper DOI).

Implementation later became collective. Dixon wrote many letters and operated the 5/20 system through 1989, so later results may reflect her interpretation as well as Donchian's authorship (FIA Hall of Fame). Seykota converted parts of the method into code and exposed the gap between principles and a complete algorithm (Seykota FAQ).

The time horizon evolved too. In a later recollection, Seykota says Donchian's original two-week rule stopped working unchanged and that viable horizons lengthened materially; he also says Donchian did not always follow it while disciplined people in his office appeared to fare better (Seykota 2020 recollection). This is retrospective testimony, not a controlled test, but it directly contradicts the idea of timeless magic parameters.

What Donchian Explicitly Rejected

Donchian rejected declaring a market too high or too low, forecasting the ultimate extent of a move, taking a quick profit merely because it was available, letting a loss grow, acting immediately on crowd opinion, and concentrating a systematic program in one or a few contracts (1982 Forbes excerpt, Dixon, 1978 scan, 5/20 article transcript).

He did not reject judgment, fundamentals, or adaptation outright. The Guides require interpretation, his copper article used fundamental material, and he openly invited better parameters. The accurate rejection is narrower: fundamental valuation should not overrule market evidence by forcing a premature top or bottom call.

Regimes Where It Thrives Versus Struggles

The philosophy thrives when price adjustment is persistent: extended commodity shocks, slow macro repricing, sustained bull or bear markets, and simultaneous but not perfectly correlated trends across several contracts. Hurst, Ooi, and Pedersen explain that trend strategies have historically helped most in extreme markets that unfolded over months or years, giving a lagging system time to change direction; that does not make them instant crash insurance (Hurst, Ooi and Pedersen, 2013).

It struggles in sideways markets, abrupt reversals, gaps through stops, and short-lived moves that reverse before the entry and execution delays can recover their costs. Diversification also weakens when many contracts become one correlated macro position. Donchian's own answer to uncertainty was smaller commitment and selectivity, not a claim that whipsaw could be eliminated.

The broader evidence is supportive but conditional. Szakmary, Shen, and Sharma find positive net mean returns for all tested dual-moving-average and channel parameterizations in at least 22 of 28 commodity markets over a 48-year sample, with pooled robustness tests (Szakmary, Shen and Sharma, 2010). Lempérière and coauthors find persistent long-horizon trend evidence but significant weakening of shorter trends (Lempérière et al., 2014). Neither result establishes that Donchian's exact settings, costs, or execution remain optimal.

Tensions Between Stated Philosophy And Actual Behavior

The first tension is mechanical discipline versus discretion. Donchian's legacy is systematic, yet Dixon says the Guides require interpretation, Seykota found them incomplete and internally conflicting, and Seykota recalled that Donchian skipped some signals (Dixon, 1978 scan, Seykota FAQ). The published philosophy is reproducible at the principle level, not as one unambiguous full trading program.

The second is price-only legend versus hybrid practice. Donchian rejected valuation as a timing veto, but his Guides still use volume, background, chart formations, statistical reports, and non-price judgment. High Finance in Copper also integrated fundamentals (High Finance in Copper DOI).

The third is timeless simplicity versus parameter and implementation dependence. The surviving 5/20 tree contains penetration thresholds, historical comparisons, delayed execution, and exceptions; Donchian himself said its lengths were not necessarily best (5/20 article transcript, computerized decision tree). Later channel defaults and Turtle rules simplify a lineage; they do not reproduce this method.

The fourth is intellectual influence versus verified investor outcome. Modern research supports diversified trend following as a strategy family, and Donchian's influence on Seykota, Dixon, and later systematic traders is unusually well documented. But there is no continuous audited Donchian return series, Futures, Inc. suffered a reported near-total decline, and later CTA evidence shows how fees and reporting bias separate gross strategy performance from client experience (1982 Forbes excerpt, Bhardwaj, Gorton and Rouwenhorst, 2014).

The durable inheritance is therefore a decision architecture, not a sacred indicator: do not impose a valuation ceiling on price, demand observable trend evidence, risk less when evidence is weak, diversify broadly, cut adverse positions, leave favorable positions uncapped, and remain humble about parameters, costs, and personal adherence.

As of 2026-07-19, the single best-documented executed campaign is Futures, Inc.'s diversified short book in the second quarter of 1951. A contemporaneous corporate notice supplies the period, direction, four market groups, $25,904 of profit, a 59.27% rise in net asset value (NAV), and the 17.81% drawdown that followed [single-source: September 1951 corporate notice]. No other recovered Donchian campaign combines that many elements. It is a vehicle-level result, not a personal trade or a commodity-by-commodity ledger (September 1951 Commercial and Financial Chronicle scan; searchable FRASER text).

The archive does not support eight conventionally complete trade tickets. This chapter therefore ranks four executed Futures, Inc. campaigns, one actual but position-unattributed fund interval, two published historical contract tests, and one multi-market backtest. The last three are included because they are Donchian's best quantified system evidence, but they are not live account profits. No sufficiently documented Donchian personal campaign or Shearson/client-account campaign was found.

Returns below are simple changes in the source's quoted price or NAV unless stated otherwise. They are not annualized. Futures losses can exceed the initial investment, so a profit divided by stated margin is not a portfolio return (CFTC futures-market primer).

[single-source] means that a figure comes from one numerical source lineage; a scan and its OCR are two access paths, not independent corroboration.

1. The Single Best: Diversified Short Book, Q2 1951

Context, discovery and thesis. Futures, Inc. was created to trade multiple commodities long or short while retaining reserves. In the three months ended June 30, 1951, the vehicle was positioned for falling prices. The surviving notice says profits came largely from shorts in cocoa, cotton, cottonseed oil, and grains. It does not identify a discretionary forecast, mechanical trigger, delivery month, or separate thesis for each market. The sound inference is narrower: Donchian expressed one bearish regime view through a diversified short portfolio rather than one heroic bet (1951 corporate notice).

Size and structure. The result belongs to Futures, Inc. Exact contracts, contract counts, notional exposure, margin, capital allocation, entry prices, and market-level weights were not disclosed. It would therefore be wrong to call cocoa, cotton, cottonseed oil, or an unspecified grain the largest position.

Entry, path and drawdown. The notice gives only quarter boundaries. NAV rose from $2.75 to $4.38, a recomputed 59.27%. July and August losses then reduced NAV to $3.60 by August 31, a 17.81% decline from the June peak. Even after that reversal, NAV remained 30.91% above the March starting value [single-source: September 1951 corporate notice]. Those later losses are the only recoverable path evidence; the source does not say whether they came from holding the same shorts, reversing them, or trading different markets (original scan, p. 3).

Exit and P&L. The reported corporate profit was $25,904 for the quarter. The wording does not specify realized versus unrealized components, fees, taxes, distributions, or cash flows. Nor does it establish a June 30 liquidation. Accordingly, $25,904 and +59.27% are vehicle-period measures, not closed-trade P&L for the named shorts.

What it teaches. Diversification let one directional theme appear in several markets, but it did not eliminate reversal risk. This dossier ranks first because the archive reveals both the gain and what happened next. The immediate 17.81% giveback is also a useful antidote to treating trend capture as a smooth process.

2. Opening Cocoa and Soybean Shorts, 1948-49

Context, discovery and thesis. Shares were first offered at $10 in October 1948, and initial operations covered November 15 through December 31. Before launch, Donchian argued that postwar commodity prices were generally too high and proposed selective short sales, diversification, and large cash reserves. That was a macro/valuation thesis coupled with trend observation, not yet the later pure-reactive legend (November 1948 launch report).

Size and structure. The fund traded a portfolio of commodity futures, predominantly short. Cocoa and soybeans contributed materially, but the source does not disclose delivery months, contracts, weights, margin, or either market's profit share. It also does not establish that all other positions made money.

Entry, path and drawdown. NAV was $11.34 on December 31 and $12.52 in January 1949, up 13.4% and 25.2%, respectively, from the $10 offer. A $0.20 dividend was declared [single-source: February 1949 corporate report], but the surviving notice does not establish whether it should be added to the $12.52 snapshot; a 27.2% total-value claim would therefore be conditional. No intraperiod drawdown was recovered (February 1949 corporate report).

Exit and P&L. The corporation reported $3,953 of net realized and unrealized profit after taxes through December 31. That is not cocoa-only P&L, soybean-only P&L, or proof that every position was closed. A 1982 Forbes-derived recollection says cocoa fell from roughly 30 cents to 19 cents and made substantial money, but the accessible reproduction supplies no account record, size, or exact exit. It may describe this same campaign and is not independent confirmation of the precise prices (Forbes excerpt reproduction).

What it teaches. The first winner combined a broad bearish view with diversified implementation. Its importance is intellectual as well as financial: the same retrospective story says a later coffee short punished the belief that price could simply be judged too high or too low. The mature method moved toward following observed trends.

3. Long Cocoa and Coffee, 1953-54

Context, discovery and thesis. From approximately October 15, 1953 through January 15, 1954, Futures, Inc. went long cocoa and coffee. The direction matters: a scrambled OCR rendering can suggest otherwise, but the original scan says profits came mostly from long positions based on the trend-following methods then used by the fund (January 1954 corporate report).

Size and structure. This was a two-market Futures, Inc. campaign. Contracts, delivery months, units, margin, portfolio weights, entry prices, and attribution between cocoa and coffee were not published.

Entry, path and drawdown. The fund's share bid recovered from $1.59 to $2.39 and its ask from $1.73 to $2.60 in less than three months. Those changes are 50.31% and 50.29% [single-source: January 1954 corporate report]. They are quoted-share recoveries, not commodity returns, NAV returns, or the P&L of the two futures positions. No maximum drawdown is available. The very low starting quote also shows why the opening 1948-49 winner cannot stand in for a successful lifetime fund record.

Exit and P&L. Management reported taking profits on cocoa and coffee shortly before publication, then entering shorts in corn, soybeans, and wool. That supplies a genuine exit event and a rapid cross-market rotation, but no cash profit or percentage campaign return. The outcome of the new shorts was not recovered and is not counted as another great trade.

What it teaches. This is the clearest contemporaneous evidence that Donchian's vehicle could reverse its original bearish orientation, follow rising trends, realize gains, and rotate into falling ones. It ranks below 1951 only because its observable 50% share-price move cannot be tied to a NAV ledger or contract-level P&L.

4. Short Sugar, Long Wheat, Soybeans and Eggs, July 1957

Context, discovery and thesis. In July 1957, Futures, Inc. combined short No. 4 world-sugar contracts with long wheat, soybean, and egg positions. The simultaneous long and short book fits Donchian's cross-market trend approach better than a single macro call. A contemporaneous report says these positions, especially the sugar shorts, largely caused the month's share-price rise (August 1957 corporate report).

Size and structure. The source names direction and markets, but not contract months, number of contracts, entries, exposure, margin, or weights. The phrase “largely due” supports contribution, not full attribution.

Entry, path and drawdown. The share bid rose from $3.38 on June 30 to a $4.72 high on July 24 and closed July at $4.53. That is a 39.64% peak move and a 34.02% month-end move [single-source: August 1957 corporate report]. By September, the fund reported a $92,551 net gain for the ten months through July 31 and a NAV increase from $2.93 to $4.53, or 54.61% [single-source: September 1957 corporate report]; because that is a ten-month vehicle result, it cannot be assigned wholly to July's four-market book (September 1957 update). No campaign drawdown is disclosed.

Exit and P&L. No trade exits or market-level P&L were recovered. At fiscal year-end, the fund reported $293,041 of net assets and $136,387 of net earnings; NAV rose from $2.93 to $5.17 including a $0.15 distribution, which the source called an 81.5% gain [single-source: December 1957 corporate report]. That full-year record is corroborating fund context, not the P&L for the sugar, wheat, soybean, and egg positions (December 1957 fiscal-year report).

What it teaches. A trend portfolio need not hold one market direction. The evidence also demonstrates why attribution discipline matters: the archive can establish that named positions drove a monthly move without proving that the full ten-month or fiscal-year earnings belonged to them.

5. The 1954-55 Recovery - Actual Fund Result, Unidentified Trades

Context, thesis, size and structure. Futures, Inc. recovered further after the cocoa-and-coffee exit, but the surviving March 1954-to-March 1955 comparisons do not name positions, directions, or signals. NAV rose from $2.59 to $3.96, or 52.90%; from September 30, 1954 to March 31, 1955 it rose from $3.39 to $3.96, or 16.81%. Assets increased from $101,576 to $170,332, while shares outstanding increased from 30,006 to 43,025 [single-source: June 1955 corporate report] (June 1955 report). The additional share offering was tied to daily NAV [single-source: March 1955 offering notice], so asset growth partly reflected subscriptions rather than investment performance (March 1955 offering notice).

Entry, path, exit and P&L. The period boundaries are accounting snapshots, not entries and exits. No contracts, costs, cash flows, drawdown, absolute trading profit, or realized/unrealized split were recovered. The NAV changes are the only defensible percentage measures.

What it teaches. This interval is included as a guardrail. It is a strong actual result but a weak “trade”: without positions, even correct NAV arithmetic cannot identify what Donchian bought or sold. It ranks above the simulations only because real shareholders experienced the fund interval.

6. December 1959 Copper - Best Quantified System Illustration

Context, discovery and thesis. Donchian's 1960 article, High Finance in Copper, studied a channel rule that compared the current price with the ranges of the preceding two calendar weeks. It was an in-sample historical exercise in daily copper futures, not the later generalized four-week rule and not a verified live trade (article DOI; Park and Irwin academic review).

Size and structure. The study used one December 1959 copper contract, $1,000 of stated margin, and $51.50 of cost per round trip. The accessible review does not disclose the number of round trips, notional exposure, reserve capital, or whether margin requirements changed.

Entry, path and drawdown. Signals came from breakouts of the preceding two weeks' ranges. Exact trade dates, fills, path, intratrade equity, and maximum drawdown were not recovered from accessible evidence.

Exit and P&L. The reported net historical gain was $3,488. That equals 348.8% of the stated margin [single-source numerical reproduction: Park and Irwin table], but margin is not account equity or full notional capital. The figure therefore cannot be presented as a 348.8% investment return. The study supplied no benchmark, out-of-sample test, or parameter-selection audit.

What it teaches. The example shows why breakouts can turn persistent price movement into an explicit rule. It also shows the danger of reporting a leveraged margin multiple without the capital and drawdown required to survive the path.

7. December 1960 Copper - a Second Historical Contract Test

Context, thesis and structure. The same two-week channel method was applied to one December 1960 copper contract. The stated structure again used $1,000 margin and $51.50 per round trip; it remains a historical calculation rather than verified Donchian, Futures, Inc., or client-account execution (Park and Irwin table).

Entry, path, drawdown and exit. The accessible review reports neither exact signals and fills nor drawdown and closing date. It also does not reveal how many round trips were charged, so gross P&L cannot be reconstructed.

P&L. Net gain was $1,390, or 139.0% of stated margin [single-source numerical reproduction: Park and Irwin table]. The same denominator warning applies. Combining the 1959 and 1960 profits produces $4,878, but no defensible compounded return follows because the evidence supplies no common account, cash balance, timing, or margin-call path.

What it teaches. A second positive contract illustration is better than one selected example, but two in-sample cases are not a live track record. Ed Seykota later recalled that the short copper horizon decayed and that Donchian did not take every system signal, further separating published rules from actual execution (Seykota retrospective).

8. The 1961-June 1974 Diversified 5/20 Study

Context, discovery and thesis. Donchian's five- and 20-day moving-average framework sought long commodity trends while using conditional entry, exit, and reinstatement rules. A hosted transcript of the 1974 article preserves the detailed rule set and the insistence on diversification (1974 article transcript).

Size and structure. The historical study covered 28 commodities from January 1961 through June 1974 on a one-contract basis. An accessible secondary treatment describes 20 profitable markets, eight losing markets, and a portfolio profitable in nine of 14 calendar years; soybeans contributed most overall but lost in seven years, including 1967-70 (Cynthia Kase, Trading with the Odds). Source summaries sometimes label the period 1961-73, while the facsimile lineage gives January 1961-June 1974; the longer exact range is used here and the discrepancy is flagged.

Entry, path and drawdown. Signals came from the 5/20 rule tree, not a bare crossover. Market-level annual losses establish a rough path, but no trade ledger, aggregate equity curve, maximum drawdown, starting capital, collateral rule, or portfolio rebalancing convention was recovered.

Exit and P&L. A historical reconstruction with a facsimile reports more than $250,000 of aggregate profit on the one-contract basis [single-source facsimile lineage] (Wolfgang Schimmel, “The Donchian Puzzle”). That is a dollar backtest result, not 250,000%, a CAGR, a Futures, Inc. return, or Donchian's personal profit. Without starting equity and costs, no valid percentage return can be calculated. Seykota's first-person account says his computerized interpretation appeared to make money for a client, but supplies no client, dates, ledger, amount, or return and says Donchian skipped some signals (Seykota FAQ).

What it teaches. Diversification mattered because even the most profitable market could lose in half the calendar years. The evidence supports a research program and a strategy lineage, not a claim that Donchian executed every signal or earned the simulated total.

What the Ranking Excludes

The famous short-cocoa recollection and later disastrous coffee short are useful origin stories, but neither has enough position detail to stand apart from the 1948-49 campaign or qualify as a quantified loss. The 1970 example of shorting at $40 with a $43 stop was explicitly instructional, not a named execution. The reported 1982 scale of approximately $27 million managed and nearly $1 million of personal trading profit derives from one Forbes profile available through derivative reproductions; it contains no campaign ledger, starting-equity denominator, or auditable return.

Account boundaries are equally important. The four executed campaigns above belong to Futures, Inc., not automatically to Donchian personally. The copper and 5/20 records are historical tests. Later 5/20 operation cannot be assigned solely to him because Barbara Dixon wrote many of the market letters and ran the system through 1989 (FIA Hall of Fame). Targeted searches found no reconstructable Shearson/client campaign with instrument, entry, size, exit, drawdown, and P&L.

Skill, Luck, Survivorship, and Confidence Limits

Three conclusions survive the incomplete archive. First, Donchian's best documented actual gains were portfolios, not isolated all-in trades. Second, the record repeatedly distinguishes vehicle outcomes from market-level attribution: NAV can be exact while the winning contracts remain unknown. Third, the 1951 reversal, the weak long-run fund evidence, and the in-sample nature of the copper studies make smooth hero narratives untenable.

The strongest evidence of skill is repeated directional flexibility: short portfolios worked in 1948-49 and 1951, long cocoa and coffee worked in 1953-54, and the 1957 book held longs and shorts simultaneously. Diversification and the willingness to rotate are observable across separate contemporary notices. Luck and selection remain inseparable from the published winners, however. The archive omits most losing trades, the copper examples were selected in-sample, the 5/20 result lacks an equity path, and no complete hit rate or audited lifetime return exists. Ranking the recoverable successes is therefore a study of documented campaigns, not proof that their outcome was repeatable ex ante.

The result is a deliberately narrow answer. The 1951 short book is the best documented executed campaign; December 1959 copper is the best quantified historical rule illustration; and the 5/20 study is the broadest published research program. None supplies the complete sizing, fill, cost, cash-flow, and drawdown data required for an audited trade-level return.

Richard Donchian's cleanest lesson did not come from a winning trend. It came from the danger of learning the wrong thing from a win. A profitable early commodity short apparently strengthened his belief that prices had an intrinsic ceiling; the next short became a formative loss. The mature Donchian method—follow price, cut losses, diversify, and avoid declaring any price inherently high or low—was in large part an answer to that error.

The evidence is much less complete than the lesson. No continuous audited return series, trade ledger, margin record, liquidation statement, or complete shareholder-distribution history for Futures, Inc. was located. Contemporaneous notices provide point-in-time fund figures; a 1982 Forbes interview survives only in reproductions; later colleagues supply retrospective testimony. Accordingly, this chapter distinguishes Donchian personally, Futures, Inc., brokerage or customer accounts, historical rule tests, and Barbara Dixon's later operation of the 5/20 system.

Loss map and evidence boundaries

The point record is still revealing. It shows repeated drawdown-and-recovery cycles, not one monotonic collapse. It also shows why share price, NAV, fund assets, and investor total return must not be merged.

1. The 1929 loss: a turn from stories toward price

The Donchian Foundation's memorial biography says that, after Yale and a return to the family rug business, Donchian suffered personal financial losses in the 1929 crash. It connects those losses to his intensive study of technical analysis and his view that chartists made more sense than conventional market commentators (Donchian Foundation, n.d.). Barbara Dixon's 1978 presentation of his trading guides gives the same broad sequence: market losses after the crash preceded his interest in technical analysis (Dixon, 1978).

This is a formative mistake, but not a reconstructable trade. Neither source identifies his securities, entry dates, leverage, realized loss, or remaining capital. Later accounts escalate the language from losses to losing most or all of his investments, without an account statement. The conservative conclusion is that he suffered meaningful personal losses, not that he was wiped out.

The behavioral root cause is also inferential. The crash exposed the weakness of relying on narratives or static valuation without an explicit exit. The guides, reportedly compiled in 1934 but not verifiably published then, emphasized limiting losses, using stops, reducing commitments when uncertain, and treating capital adequacy as part of survival (Dixon, 1978). Chronology makes the crash a plausible catalyst for those controls; it does not prove that each rule followed a particular 1929 position.

2. The dangerous winner: cocoa success and the coffee short

The most important error in Donchian's own reported telling began as a success. A reproduced 1982 Forbes interview says he shorted cocoa around 30 cents and watched it fall toward 19 cents. He then treated that outcome as validation of a valuation judgment: commodity prices could simply be too high. He next shorted coffee around 20 cents, only to see it rise toward a dollar. His retrospective lesson was that price has no absolute high or low (Forbes reproduction, 1982). These price points and the causal recollection are [single-source]; the open web copy is a reproduction, not the original magazine scan.

A contemporaneous notice partly corroborates the setup, not the loss. Futures, Inc.'s initial November 15–December 31, 1948 period produced $3,953 of after-tax realized plus unrealized profit from short commodity futures [single-source]; cocoa and soybeans both contributed materially. The notice also reported Donchian's continuing belief that general commodity prices were too high (Chronicle, 1949). It therefore supports the anchoring mechanism: an early rewarded short book reinforced a broad bearish valuation thesis. It does not isolate the cocoa profit, identify the later coffee trade, or prove that both positions were in Futures, Inc.

The missing fields matter. No contract month, date, size, margin, margin call, exit, realized loss, or account statement was recovered. “Coffee rose fivefold” describes the reported commodity-price path, not a fivefold account loss. Futures losses are nonlinear with position size, margin, rolls, and exit timing. Calling this Donchian's largest loss would therefore be invention.

The process change is clearer than the P&L. Static cheap/dear judgment gave way to a reactive rule: do not fight a persistent move merely because its level looks extreme. That rule did not eliminate discretion, but it removed the premise that a price must reverse because it had crossed an intuitive threshold (Forbes reproduction, 1982).

3. Futures, Inc.: capital impairment, recoveries, and another drawdown

Futures, Inc. was offered publicly in October 1948 at $10 per share. The prospectus summary described a selective long-or-short commodity program with diversification and reserve capital (Chronicle, 1948). The surviving snapshots then show severe capital impairment. The best same-metric primary interval is a 78.04% fall from the $12.52 January 1949 NAV to $2.75 on March 31, 1951 [single-source]. Because the record includes at least a $0.20 distribution and lacks the intervening path, even this is NAV impairment—not a continuous investor total return or a proven maximum drawdown (Chronicle, 1949; Chronicle, 1951).

• By March 31, 1951, NAV was $2.75—72.5% below the offer price before any adjustment for distributions [single-source]. A profitable quarter lifted NAV to $4.38, a 59.27% point-to-point gain, before July and August losses reduced it to $3.60, a 17.81% decline that surrendered 47.85% of the prior dollar-per-share gain (Chronicle, 1951).
• The bid reached $1.59 on October 15, 1953, 84.1% below the $10 offer before distributions. It recovered to $2.39 by January 15, 1954, a 50.31% quoted-price rebound [single-source] (Chronicle, 1954). A bid is not NAV or total return.
• NAV rose from $2.59 on March 31, 1954 to $3.96 a year later, while shares outstanding also increased; fund-asset growth therefore mixed performance and issuance [single-source] (Chronicle, 1955).
• In the fiscal year ended September 30, 1957, NAV rose from $2.93 to $5.17 and the fund paid a $0.15 dividend [single-source] (Chronicle, 1957). That large recovery did not restore the original $10 point value.

The 1958 record adds another reversal. A March regulatory notice covered 54,300 additional shares to be sold at the market, showing that the fund was still seeking capital after its strong 1957 (Chronicle, 1958). A later notice reported March 31 net assets of $299,234, versus $193,355 a year earlier, and NAV of $4.83 versus $3.32. Yet its June 13 offering price was $3.99; because that price was formula-linked to the daily redeeming bid, it signals renewed weakness but is not an exact NAV-to-NAV drawdown [single-source] (Chronicle, 1958).

These are vehicle-level near-death indicators, not proof of one catastrophic trade. The notices do not identify which positions caused the early-1950s impairment, how much came from fees or distributions, or what a continuously invested shareholder received. The proper inference is that the fund repeatedly lost large portions of point value and repeatedly recovered part of them. Its trend-oriented program did not spare it from whipsaw or path-dependent capital strain (Chronicle, 1951; Chronicle, 1954).

4. Closing the fund to new sales: adaptation, not liquidation

On March 31, 1960, Futures, Inc. shifted from open-end to closed-end operation, withdrew the unsold balance of its offering, and halted new sales while continuing redemptions at NAV. The notice quoted Donchian's reasons: protect current shareholders from dilution, preserve the value of an existing tax loss, and refine a new program that he believed worked better in a roughly $100,000–$500,000 fund than in a much larger pool. On March 22 the shares were quoted at $1.87 bid and $2.03 ask, versus $1.65/$1.79 at year-end 1959 [single-source] (Chronicle, 1960).

This is the best documented organizational response to poor cumulative economics. The fund did not then liquidate: it continued to redeem shares and trade on a closed-end basis. The decision nevertheless reveals three constraints that clean strategy descriptions omit—scale, tax-loss economics, and dilution from new subscriptions. Donchian was adapting the vehicle around the program rather than assuming a rule that worked in one capital base would transfer unchanged to another (Chronicle, 1960).

The eventual ending is murkier. The Forbes reproduction says the stock traded as low as four cents before dissolution [single-source] (Forbes reproduction, 1982). Schimmel's archival reconstruction quotes former colleague Brent Elam recalling later efforts to buy shares around $0.75–$1.20 and liquidate the company [single-source], but no original tender, dissolution certificate, final NAV, or distribution schedule was recovered (Schimmel, 2011). The four-cent figure is therefore a reported interim low, not a terminal price or auditable lifetime return.

5. The gap between rules and behavior

Donchian's later systems contained explicit defenses against his early mistakes, but witnesses say his own implementation was not consistently mechanical. Seykota recalls that Donchian sometimes took signals and skipped others; followers who applied the rule more consistently appeared to fare better. He also says the published rules could conflict, requiring a programmer to choose non-conflicting subsets rather than merely copy a complete algorithm (Seykota FAQ, n.d.). In a later account, Seykota again distinguishes Donchian's stated two-week copper rule from his inconsistent personal adherence (Seykota, 2020). Elam similarly remembered tension between Donchian's judgment and model outputs (Schimmel, 2011).

This is a reported implementation conflict, not evidence of deception or a documented client breach. The witnesses do not identify the skipped signals, their counterfactual P&L, or the affected accounts. It is nevertheless an important error of omission: the available evidence suggests that Donchian sometimes omitted trades his own framework called for. No defensible “missed fortune” can be calculated (Seykota FAQ, n.d.; Schimmel, 2011).

The behavior also exposes an incompleteness problem. A hosted transcript of the 5/20 method includes penetration tests, prior-signal comparisons, delayed execution, holiday exceptions, close-outs, reinstatements, and diversification requirements (Donchian 5/20 transcript, 1974). Donchian acknowledged that five and 20 days were not necessarily optimal and that the action rules could improve. Dixon's guides likewise say successful application requires individual interpretation (Dixon, 1978). Mechanical discipline reduced some behavioral freedom; it did not remove specification judgment.

6. What failed inside the methods

The 5/20 historical study was profitable in nine of 14 years, which also means five losing years. Eight of 28 markets lost money. Soybeans lost in seven of 14 years, including 1967–1970, even though the market contributed strongly to the aggregate result [single-source] (Kase, 1996). These figures describe a historical fixed-contract study, not a live Donchian account. They demonstrate the expected cost of the approach: individual markets and years can whipsaw even when a diversified portfolio works.

Seykota later judged that Donchian's original two-week copper rule no longer worked and that viable horizons had lengthened, in his experience, from weeks toward months (Seykota, 2020). That is Seykota's retrospective assessment, not Donchian's admission that trend following failed. It points to parameter decay: once market structure, participation, or noise changes, a once-useful horizon may become too fast.

There is also a research-design failure boundary. Park and Irwin classify Donchian's 1960 copper study among the earliest technical-rule studies and explain that much early research did not adequately control data snooping, parameter selection, risk, or out-of-sample performance (Park and Irwin, 2004). Later CTA research warns that voluntary databases, backfill, survivorship, graveyard bias, and fees can make investor results look better than the experience actually available ex ante (Bhardwaj, Gorton and Rouwenhorst, 2014). Those studies do not measure Donchian's accounts; they explain why his historical system totals should not be upgraded into an audited track record.

Behavioral roots and process changes

Errors of omission, legal search, and attribution

No documented Donchian equivalent of “passing on a named stock that rose tenfold” was found. The only supportable omission class is skipped system signals, and the record lacks the signal dates and counterfactual result. Turning that into a ranked missed-trade list would manufacture precision.

No named SEC enforcement, CFTC enforcement, NFA discipline, lawsuit, arbitration, bankruptcy, client complaint, or criminal matter involving Richard D. Donchian or Futures, Inc. was found in the searched digital sources. This is a bounded negative, not a clean-record certification. Futures, Inc. began in 1948, the CFTC was created only in 1974, and the first modern federal CPO/CTA operating rules arrived in 1979 (CFTC, n.d.). Pre-digital state, exchange, predecessor-agency, and arbitration records are incomplete.

Finally, later activity cannot repair the fund record by attribution. Dixon wrote many later letters and operated the 5/20 system through 1989 (FIA, n.d.). Seykota tested and modified rules. Brokerage customer or model accounts were not automatically Futures, Inc. Donchian's intellectual influence was broad, but results produced by collaborators, later parameters, or different vehicles are not his personal audited recovery.

Luck, skill, and the durable lesson

Luck helped create the original mistake: a correct cocoa direction rewarded a weak absolute-value premise. Skill lay in extracting a better rule from the subsequent failure. Donchian replaced prediction at a fixed price level with conditional reaction, made loss limitation explicit, treated diversification as necessary, and admitted that parameters and rules could improve (Forbes reproduction, 1982; Dixon, 1978; Donchian 5/20 transcript, 1974).

Those improvements did not produce immunity. Futures, Inc. suffered severe point-value impairment, recovered repeatedly, changed structure, and was later reported dissolved. Short horizons decayed; individual markets and years lost; the human operator reportedly skipped signals (Chronicle, 1960; Forbes reproduction, 1982; Seykota, 2020; Kase, 1996). The durable lesson is therefore narrower and more useful than the legend: systematic trend following can discipline a trader's worst instincts, but only if the system is specified, followed, diversified, capitalized, and continuously tested against changing markets.

Richard Donchian left no located book, memoir, public speech archive, or authenticated recording. His usable first-person record is unusually dispersed: three bylined securities articles from 1949-1951, a 1957 article preserved only in quotation, original 1960, 1970 and 1974 article scans, a compilation of trading guides transmitted by Barbara Dixon, and two magazine profiles. The 30 excerpts below are therefore short evidence fragments rather than a quotation anthology.

Every excerpt is under 25 words, and aggregate verbatim use from each underlying work is also no more than 25 words. That stricter work-level ceiling treats mirrors and reproductions as one source family. Dates identify the underlying work, not the date of a later host. The early articles matter because they complicate the familiar portrait of Donchian as a price-only technician: before codifying trend following, he wrote conventional valuation, operating-improvement, credit-cycle, and asymmetric-payoff analysis.

Valuation, fundamentals, and what can go wrong

1. “high earnings and low price-earnings ratios are no assurance” — Donchian, 1949 (FRASER issue text, printed pp. 1391 and 1433). In “Are Stocks as Cheap as They Seem?” he used 1920-1921 history to challenge a superficially cheap market.

2. “Such earnings ... can evaporate rather rapidly as prices readjust” — Donchian, 1949 (same article). Reported profits could include inventory gains and abnormal demand; a low multiple did not make those profits durable.

3. “not necessarily safe when prices are sky-high” — Donchian, 1949 (same article). This is an early warning against applying a valuation threshold without considering the commodity-price regime.

4. “better service, lower costs” — Donchian, 1950 (FRASER issue text, printed p. 894). His Western Union thesis began with mechanization and operating leverage rather than chart action.

5. “Any appreciable increase in earning power would build up per share earnings rapidly.” — Donchian, 1950 (same article). Capital structure turned operating recovery into a potentially leveraged equity outcome.

6. “building a foundation ... over the longer-term” — Donchian, 1950 (same article). The sentence led to a specific price and dividend-status forecast, showing how much more predictive his pre-trend prose could be.

Uncertainty and asymmetric positioning

7. “‘danger ahead’ warnings for stocks” — Donchian, 1951 (FRASER issue text, printed p. 2442). Tightening money and falling bonds led him to reverse the optimism of his prior Western Union recommendation.

8. “as many of us are most of the time” — Donchian, 1951 (same article). He acknowledged that investors could reasonably be confused rather than certain about market direction.

9. “Heads you make more,” — Donchian, 1951 (same article). His proposed mix of warrants and cash was designed as a convex payoff, although the numerical scenario was a forecast, not a realized return.

Loss limitation and staying with the trend

10. “automatically limit the loss” — Donchian, 1957 (Ostgaard reproduction and citation, citing Commodity Year Book, p. 35). Automatic loss limitation was the first duty of a sound trend-following method.

11. “without limiting the gain” — Donchian, 1957 (same reproduced passage). Losses and gains were deliberately asymmetric: one was bounded, the other given room.

12. “reversed—or at least closed out—at a limited loss” — Donchian, 1957 (same reproduced passage). Opposing trend evidence, not a narrative defense of the position, triggered the response.

13. “maintained as long as the trend continues” — Donchian, 1957 (same reproduced passage). A continuing trend supplied its own reason to remain invested; the source assigns no profit target.

Short selling, prejudice, and controls

14. “valuable money-making tool” — Donchian, 1970 (FRASER issue scan, printed pp. 438-439). In his complete bylined article, he argued that properly timed short selling deserved the same analytical legitimacy as buying.

15. “People tend to fear whatever they do not understand” — Donchian, 1970 (same scan). He treated aversion to short selling partly as an education problem.

16. “predetermined distance” — Donchian, 1970 (same scan). The phrase describes placement of a stop-buy order above the short-sale price; his $40/$43 illustration was instructional, not a documented trade.

17. “trees do not grow to the skies” — Donchian, 1970 (same scan). Even a structurally rising market still created falling-price periods that a two-sided manager could use.

The compiled trading guides

Barbara Dixon’s October 1978 article says Donchian first compiled the guides in 1934, recovered a complete copy in 1966, and considered them useful four decades later. It does not prove publication in 1934 or sole authorship of every inherited maxim. The wording below is therefore attributed to the Donchian compilation as transmitted by Dixon, not to a recovered 1934 publication.

18. “LIMIT LOSSES, ride profits” — Donchian compilation, 1934; published by Dixon, 1978 (original magazine page, printed p. 44). The capitalized imperative ranks payoff discipline above the remaining guides.

19. “Light commitments are advisable” — Donchian compilation, 1934/1978 (same original page). Uncertain market position called for less exposure, not forced conviction.

20. “Judicious use of stop orders” — Donchian compilation, 1934/1978 (same original page). Stops could protect profits, constrain losses, or initiate positions when related to chart formation.

21. “Buy strong acting, strong background commodities” — Donchian compilation, 1934/1978 (same original page). Relative strength and broader market background both entered the decision.

22. “Watch for volume climax” — Donchian compilation, 1934/1978 (same original page). The technical guide specifically places this clue after a long move, not as a context-free signal.

Diversification as a break with specialization

23. “a diversified approach” — Donchian in Darrell Jobman’s 1980 profile (original magazine page, printed p. 42). Jobman explicitly introduces the passage with “Donchian says,” making it authenticated reported speech rather than Donchian-authored prose.

24. “they were worlds apart” — Donchian, 1980 (same original page). Cocoa, cotton, and grain specialists had operated in separate domains.

25. “look at all the commodities together” — Donchian, 1980 (same original page). His innovation was portfolio-wide observation rather than mastery of only one market.

26. “cutting losses short and going with a trend” — Donchian, 1980 (same original page). Diversification, loss control, and trend participation appear in one reported formulation.

The lesson from early failure

27. “Trends persist.” — Donchian in William Baldwin’s 1982 Forbes profile (derivative reproduction, original March 1, 1982, p. 143). It is the shortest surviving summary of his method, but the original page image remains unrecovered.

28. “I was bearish” — Donchian, 1982 (same reproduction). The remembered starting disposition preceded a profitable cocoa short and a disastrous attempt to repeat the valuation call in coffee.

29. “never to be a price trader” — Donchian, 1982 (same reproduction). His rule rejected declaring an absolute price ceiling or floor.

30. “no such thing as too high a price or too low a price” — Donchian, 1982 (same reproduction). This is reported recollection, not permission to ignore stops, exposure, or opposing trend evidence.

Annotated index of primary and near-primary materials

Bylined articles and authored market writing

1. “Are Stocks as Cheap as They Seem?”, March 31, 1949 — Donchian tests low multiples and high yields against the 1920 collapse; the strongest recovered evidence of his early valuation and regime analysis.
2. “Western Union,” March 2, 1950 — A conventional company thesis based on mechanization, cost reduction, capital structure, and a multi-year operating recovery.
3. “Tri-Continental Warrants Plus Plenty of Cash,” June 14, 1951 — A barbell recommendation intended to preserve downside protection while retaining leveraged upside.
4. “Trend-Following Methods in Commodity Price Analysis,” 1957 — The foundational loss-limited, gain-unlimited passage survives through a later historical study with a precise Commodity Year Book page citation; no open original scan was found.
5. “High Finance in Copper,” November-December 1960 — Original journal scan of Donchian’s hybrid fundamental and two-week technical study; the DOI record independently fixes the bibliographic identity.
6. “Short Selling Is More Common Than Generally Realized,” August 13, 1970 — Complete two-page bylined article explaining everyday short-sale analogies, stop-buy protection, historical price direction, and short selling’s market function.
7. “Donchian’s Five- and 20-Day Moving Averages,” December 1974 — Original Commodities page and issue for the system’s premise and detailed rules; it controls over the edited posthumous 1995 reproduction.
8. Dixon, “Donchian’s 20 Guides to Trading Commodities,” October 1978 — Original magazine page preserving the compiled guides and the 1934/1966/1978 transmission history; Dixon’s commentary is not Donchian testimony.

Corporate statements, newsletter, and interviews

9. Futures, Inc. closed-end announcement, March 24, 1960 — A contemporaneous report quoting Donchian on capacity, dilution, tax-loss preservation, and suspending new sales; it is a corporate news item, not a signed article.
10. Commodity Trend Timing, approximately 1960-1979 — The weekly letter is probably Donchian’s largest direct-written corpus, but no public run or catalog finding aid was recovered. Later issues cannot automatically be assigned solely to him because Dixon wrote many letters and operated the 5/20 system (FIA biography).
11. Jobman, “Richard Donchian: Pioneer of Trend-Trading,” September 1980 — Original interview page with explicit “Donchian says” framing for the diversification passage.
12. Baldwin, “Rugs to Riches,” March 1, 1982 — Profile/interview source for trend persistence and the cocoa/coffee lesson; the accessible text is a derivative reproduction of Forbes p. 143.
13. Schimmel, “The Donchian Puzzle,” July 2011 — The most useful archival reconstruction and source map for the 1960 article, 1974 system description, newsletter holdings, and Brentin Elam recollections; it is not itself Donchian voice.

Speeches, recordings, podcasts, and false positives

No authenticated Donchian speech transcript, panel transcript, award-acceptance remarks, audio, video, podcast, or oral history was located. The 1983 Managed Accounts Report award is biographically documented, but no acceptance text or recording surfaced (Donchian Foundation biography). That absence should remain visible rather than being filled with later admirers’ language.

Ed Seykota’s accounts of coding the 5/20 method, conflicting rules, skipped signals, and horizon decay are important colleague testimony, but they are Seykota’s words (FAQ; 2020 recollection). The unattributed “trend is your friend,” modern Turtle rules, ATR sizing, percentage-risk limits, and three-line software channels were also excluded. None was recovered in a Donchian original.

What the record shows

The corpus reveals development rather than a timeless slogan. In 1949-1951 Donchian reasoned from earnings quality, operating improvement, capital structure, money conditions, and payoff design. By 1957 the central doctrine had become more reactive: limit the loss automatically, avoid capping the gain, and hold while the trend persists. The compiled guides retained judgmental inputs—volume, relative strength, market background, trend lines, and uncertain-position sizing—so “systematic” did not mean one context-free formula.

The archive also establishes three boundaries. First, the guides were compiled in 1934 but are only recoverable in Dixon’s 1978 publication; “first published in 1934” is unsupported. Second, later letters and system operation involved Dixon, so the whole newsletter record cannot be treated as Donchian-only prose. Third, the 1980 interview is now verifiable in its original magazine scan, while the 1982 Forbes wording still depends on a reproduction. The latter belongs in the corpus with an explicit provenance warning, not with the confidence of a scanned bylined article.

The safest synthesis is therefore narrower than the legend. Donchian repeatedly advocated loss control, open-ended winners, broad market coverage, less exposure under uncertainty, and willingness to follow price instead of declaring it irrational. His own earlier articles show that he also understood fundamentals and valuation. The quotations establish what he wrote or was reported to have said; they do not prove that he executed every signal, originated every maxim, or produced a continuous audited return record.

As of: 2026-07-19 Task: T0629 | Investor: 078-richard-donchian | Code: F-key-writings

Corpus verdict

Richard Donchian left no recovered investment book, memoir, speech transcript or public archive of signed letters. His usable corpus is instead a small set of bylined articles, plus a much larger but publicly unavailable weekly newsletter. The seven substantial bylined works recovered in this search begin with fundamental stock analysis in 1949 and end with an explicit diversified trend system in 1974. That sequence matters: Donchian did not simply replace judgment with a price channel. He moved from regime-aware valuation and company analysis toward rules that made loss control, reversal and diversification operational, while still allowing fundamental judgment to qualify a technical result (1949 article; 1960 article; 1974 article).

The record has two material gaps. First, an open original copy of his 1957 Commodity Year Book chapter was not found, so its treatment below is limited to an independently verified citation, catalog evidence and the surviving excerpt rather than reconstructed detail (Park and Irwin; Google Books catalog; Stig Ostgaard). Second, a later collaborator wrote many issues of the postwar newsletter and transmitted Donchian's Twenty Trading Guides. Consequently, neither body can be labeled wholly Donchian-authored without issue-level evidence (Barbara Dixon article; FIA profile). These are bounded public-search findings, not proof that private or unindexed material does not survive.

Works by Donchian

1. “Trendfollowing Methods in Commodity Price Analysis” (1957)

Classification and access. This is the pivotal bibliographic item: a Donchian chapter beginning on page 35 of the Commodity Year Book. The table of contents spells Trendfollowing as one word; Park and Irwin independently list the title, author, year and pages 35-47, while the catalog confirms the volume. No open original scan was recovered. Ostgaard preserves the page-35 passage and supplies the best accessible historical discussion, but his essay is a later practitioner source rather than a substitute facsimile (Park and Irwin; Google Books catalog; Stig Ostgaard).

Central thesis. A useful commodity method should define trend continuation and reversal while automatically limiting losses and leaving gains uncapped (Stig Ostgaard).

Key ideas supported by the recovered evidence:

1. Price trends can persist long enough to support systematic participation (Stig Ostgaard).
2. A method must bound losses without imposing an equivalent ceiling on gains (Stig Ostgaard).
3. Moving averages and swing rules can turn trend evidence into operating decisions (Park and Irwin).
4. A trader should remain positioned while the trend remains intact (Stig Ostgaard).
5. Opposing trend evidence should close or reverse the position rather than invite prediction-based delay (Stig Ostgaard).

Best pages. Begin with page 35, the only page for which the central formulation was recovered. Treat pages 36-47 as a priority archive request, not as permission to invent section-level guidance.

2. “High Finance in Copper” (1960)

Classification and access. This is the strongest complete primary work: a bylined Financial Analysts Journal article, volume 16, issue 6, printed pages 133-142. The complete original scan and DOI record are available (original scan; DOI record).

Central thesis. Copper should be analyzed through supply-demand structure, geopolitics and futures mechanics, then traded with a simple two-calendar-week breakout rule whose leverage is constrained by small predetermined losses (original scan).

Key ideas:

1. African and Chilean production risks can shift the balance between surplus and scarcity (original scan).
2. Inventory changes and prices tend to move inversely, making stock changes a useful fundamental input (original scan).
3. Futures produce much more leverage than copper shares, increasing both opportunity and ruin risk (original scan).
4. A basic rule buys above the prior two calendar weeks' highs and sells below their lows (original scan).
5. A Friday or pre-holiday qualification filters marginal signals (original scan).
6. The opportunity tables are illustrations, not promises of attainable profit (original scan).
7. Fundamental judgment can refine a mechanical rule, but cannot remove the need for stops (original scan).
8. The method is deliberately described as crude, and the historical examples do not guarantee repetition (original scan).

Best pages. Printed pages 133-135 explain market structure; page 139 compares opportunity and risk; page 141 states the rule and worked examples; page 142 supplies the caveats and split long-run/short-run outlook. Tax and straddle passages describe 1960 law, not current advice (original scan).

3. “Donchian's 5- and 20-Day Moving Averages” (1974)

Classification and access. The original appeared in the December 1974 issue of Commodities, despite the archive identifier's later futures-us label. The accessible original pages 10-11 control wherever visible, but the ending page is unresolved and the open preview omits part of the continuation. A 1995 posthumous edited republication supplies readable continuity and is labeled rather than silently backdated (original page; edited republication).

Central thesis. A diversified commodity program can use the 20-day average for the basic trend and the five-day average for exit and reinstatement, with adaptive penetrations and execution exceptions to reduce weak signals (original page).

Key ideas:

1. The 20-day average supplies the principal signal; 15- and 25-close comparisons can qualify it (original page).
2. The required penetration scales with the market's price rather than using one fixed point amount (original page).
3. The five-day average controls exit and later re-entry after a basic trend is established (original page).
4. Rules for early reversal, one-day delay, weekday and holiday timing address execution edge cases (edited republication).
5. A position may be justified by an active trend even without a fresh signal, but only with a stop (edited republication).
6. Broad commodity diversification is integral because results vary sharply across markets (edited republication).
7. Five and twenty days are practical parameters, not natural constants (edited republication).
8. A durable simple method is preferable to endless optimization, and the rules remain open to improvement (edited republication).

Best sections. Prioritize the rule tree, historical-results discussion, diversification warning and closing parameter caveat. This is not a naive five-day/20-day crossover: penetration, exception, exit and reinstatement rules are material. Kase's later reading reports that the nearly fourteen-year, 28-market study was profitable overall in 20 markets, losing in eight, and profitable in nine of fourteen years; those figures are her secondary reconstruction, not an audited account result (edited republication; Cynthia Kase).

4. “Short Selling Is More Common Than Generally Realized” (1970)

Classification and access. This is a complete two-page bylined article in The Commercial and Financial Chronicle, August 13, 1970, printed pages 438-439. Donchian was identified as a CFA and Hayden, Stone's director of commodity research (original issue PDF).

Central thesis. Short selling is a normal economic transaction, not an inherently destructive act, and a stop-protected short can be a disciplined way to participate in a decline (original issue PDF).

Key ideas:

1. Selling now for later delivery has familiar analogues in ordinary commerce (original issue PDF).
2. Stock and commodity shorts differ in borrowability, mechanics and then-current tax treatment (original issue PDF).
3. A stop-buy order can predetermine the acceptable loss on a short (original issue PDF).
4. An unprotected long position is not automatically safer than a controlled short (original issue PDF).
5. Short covering creates eventual buying demand and can add liquidity (original issue PDF).
6. A two-sided method can switch between long and short rather than remain passive through a decline (original issue PDF).

Best pages. Page 438 is best for definitions, mechanics and loss control; page 439 is best for switching, liquidity and the market-function argument. The straddle discussion is historical tax context only.

5. “Are Stocks as Cheap as They Seem?” (1949)

Classification and access. This bylined Commercial and Financial Chronicle article survives in its complete issue scan and OCR on printed pages 1391 and 1433 (FRASER).

Central thesis. Low reported multiples and high yields can be deceptive when earnings depend on abnormal demand, elevated commodity prices and inventory gains that may reverse (FRASER).

Key ideas:

1. Ratios should be interpreted within the profit and commodity-price regime that produced them (FRASER).
2. Inventory gains reduce the durability of reported earnings (FRASER).
3. A historical analogue can expose what a normalized profit environment might look like (FRASER).
4. A conventional ten-times-earnings rule is not inherently conservative near a cyclical peak (FRASER).
5. The defensible multiple may fall as the commodity-price level rises (FRASER).
6. Historical comparison is a scenario test, not a certain forecast (FRASER).

Best pages. Read page 1391 for the valuation problem and page 1433 for the comparative table and conclusion.

6. “Western Union” (1950)

Classification and access. This one-page entry in the Chronicle's “The Security I Like Best” feature identifies Donchian as a registered investment adviser and president of Futures, Inc. (FRASER).

Central thesis. Mechanization, labor savings, network assets, new services and a compact capital structure could turn Western Union's weak recent record into a leveraged multi-year earnings recovery (FRASER).

Key ideas:

1. New transmission centers could reduce operating cost (FRASER).
2. Telefax and TWX offered potential volume growth (FRASER).
3. Transition costs depressed recent results but might not recur (FRASER).
4. Labor reduction created operating leverage (FRASER).
5. Tax relief and stable rates could support margins (FRASER).
6. A small share count magnified the per-share effect of recovery (FRASER).
7. The stated horizon was two or three years, not an immediate catalyst (FRASER).

Best section. The article is short enough to read in full. Its closing forecast is the core of the thesis, but it is a contemporaneous recommendation, not evidence that the forecast was realized.

7. “Tri-Continental Warrants (Buy Plus Plenty of Cash)” (1951)

Classification and access. This second “Security I Like Best” selection occupies printed pages 2442 and 2456 of the June 14, 1951 Chronicle (FRASER).

Central thesis. Under tightening money and elevated equity risk, an investor could keep roughly 80% in cash or government bonds and 20% in perpetual Tri-Continental warrants, seeking bounded portfolio loss with leveraged upside (FRASER).

Key ideas:

1. A prior company thesis should be revised when the credit regime changes (FRASER).
2. Company quality and monetary conditions are separate analytical questions (FRASER).
3. A small warrant allocation can retain substantial equity sensitivity (FRASER).
4. Perpetual warrants preserve optionality without near-term expiry (FRASER).
5. The cash component constrains total portfolio downside (FRASER).
6. Scenario arithmetic is more informative than one point forecast (FRASER).
7. The structure is an early barbell, but its warrant leverage still carries loss risk (FRASER).

Best pages. Page 2442 explains the regime change and allocation; page 2456 contains the scenario arithmetic.

Access-limited and mixed-authorship corpus

The weekly newsletter

Schimmel reports an approximately weekly 1960-1979 run of roughly four pages per issue, or about 4,000 pages, under Trend Timing Comments, with a bound set in the Institute for Financial Markets archive. The Donchian Foundation and Seykota biography use Commodity Trend Timing, while Dixon's 1978 article uses the shortened Trend Timing. No original masthead, public full-text run, accession record or institutional finding aid was recovered; Schimmel's custody claim was not independently confirmed. The safest provisional label is therefore the Trend Timing weekly commodity letter, also reported under the two longer variants (Wolfgang Schimmel; Donchian Foundation; Seykota resource page; Barbara Dixon article).

Authorship must remain issue-specific. Dixon says she wrote the letter when Donchian did not, and FIA reports that she wrote many later letters and operated the 5/20 system through 1989 (Barbara Dixon article; FIA profile). Claims about the whole run therefore cannot safely be placed in Donchian's sole voice.

The Twenty Trading Guides

The earliest recovered publication is Dixon's October 1978 article. It says Donchian first compiled the guides for stock traders in 1934, recovered a complete copy in 1966 and later considered most applicable to commodities. It does not establish a separately published 1934 Donchian work. The defensible description is “a Donchian compilation, as transmitted and annotated by Barbara Dixon” (Barbara Dixon article). Its durable themes include skepticism toward crowded opinion, smaller commitments under uncertainty, stop-based loss control, asymmetric recovery arithmetic, relative strength and volume/chart context. Read both printed pages 44-45, while separating the guides from Dixon's commentary.

Other reported bodies remain unlocated and should remain separate: untitled Hemphill, Noyes technical letters from approximately 1933-35; the later Security Pilot stock service sold to brokerages; and approximately two years of postwar Shearson, Hammill Market Outlook letters. A contemporary notice confirms Donchian joined Shearson, Hammill in 1946, but no original letter or reported Wall Street Journal excerpt was recovered (Seykota resource page; 1946 staff notice).

Best works about Donchian, ranked

1. Wolfgang Schimmel, “The Donchian Puzzle” (2011). The strongest source-critical reconstruction identifies the newsletter archive, distinguishes titles, interviews Brent Elam and Trish Foshe, reproduces a 1974 rule facsimile and exposes ambiguity between published and implemented rules. Its limitations are practitioner authorship, retrospective testimony and a Scribd access path rather than an open publisher archive (Schimmel).
2. Darrell Jobman, “Richard Donchian: Pioneer of Trend-Trading” (1980). This is the best accessible contemporaneous profile/interview and gives explicit speaker framing, especially on diversification. It is brief, celebratory and not a performance audit (original issue).
3. William Baldwin, “Rugs to Riches” (1982). This late-career Forbes profile is unusually candid about early trading mistakes, fund deterioration and scale. Only a derivative reproduction was recovered, and several figures follow one source lineage, so it is a lead rather than independently audited evidence (reproduction).
4. Barbara Dixon, “Donchian's 20 Guides to Trading Commodities” (1978). A direct collaborator's transmission is the best evidence for the guides and for later newsletter authorship boundaries. It mixes a Donchian compilation with Dixon's own explanation and must be cited accordingly (original article).
5. Stig Ostgaard, “On the Nature and Origins of Trend Following” (2008). The best intellectual genealogy preserves the 1957 citation and excerpt and places Donchian among earlier trend ideas. It is a brief, non-peer-reviewed practitioner working paper (Ostgaard).
6. Ed Seykota's recollections. A former collaborator gives unusually frank first-person evidence on computerization, conflicting rule sets and the difference between published signals and Donchian's own adherence. The material is retrospective, aggregated from a FAQ and unsupported by an account ledger (Seykota FAQ).
7. Cynthia Kase, Trading With the Odds. This is the strongest located empirical reading of the 5/20 study, particularly its cross-market and cross-year dispersion. It is a later secondary interpretation, not the original test file or live performance record (Kase).
8. Perry Kaufman, Trading Systems and Methods. This serious textbook offers the best modern method context and helps distinguish later formalization from Donchian's own wording. Its relevant treatment is small and cannot establish original authorship (Wiley).
9. Barry Jaffarian, ed., A Survey of the Managed Futures Industry. The book situates Donchian within managed-futures history and institutional development. Its biographical material is derivative and broad rather than a close reading of his writings (survey).
10. Don M. Chance, Managed Futures and Their Role in Investment Portfolios. This CFA Research Foundation study provides sober industry history and portfolio context. Its Donchian discussion follows earlier secondary lineages and adds little person-specific primary evidence (CFA Institute PDF).

Authorship traps and recommended reading path

Do not turn modern three-band “Donchian Channels,” Turtle 20/10 or 40/20 systems, ATR sizing, or a collaborator's decision tree into Donchian-authored text without a primary bridge. Likewise, the 1995 edited 5/20 reprint is not new writing, Jobman and Baldwin are works about him, and compiling guides in 1934 is not the same as publishing them then. Schimmel also shows why a printed rule and the rule actually clarified to Elam may differ (Schimmel).

For the shortest defensible route, read the 1960 article first because it reveals Donchian's hybrid fundamental and technical method; then the 1974 article for the mature system; then the 1949 and 1951 articles for the regime and asymmetric-payoff roots. Use the 1957 excerpt as a bridge, not a recovered full chapter. Finish with Schimmel for provenance and adverse evidence, Jobman for contemporaneous voice, and Dixon for the boundary between Donchian's ideas and collaborator transmission.

As of: 2026-07-19

Task: T0630 | 078-richard-donchian | G-mental-models

Evidence stance: Donchian left several different decision architectures, not one complete universal system. The 1957 trend principle, 1960 two-week copper rule, 1974 5/20 moving-average method, early security analysis, and Twenty Trading Guides must be separated. Labels below distinguish Donchian-authored ideas, a Donchian compilation transmitted by Barbara Dixon, and this chapter's reconstructions. Modern three-line channels and the Turtle package of breakout, volatility-sizing, pyramiding, and exposure rules are later implementation layers, not recovered Donchian rules (modern platform convention, Original Turtle Rules).

Named Heuristics And Frameworks

The names in this section are analytical handles. Unless expressly stated, they are not phrases Donchian coined.

1. Trend persistence is a state, not a price target - direct

Donchian's mature method asks whether an observable trend remains in force, not where price ought ultimately to stop. The recovered excerpt from his 1957 chapter says the position should remain while the trend continues and should close or reverse when an opposing trend becomes evident (Ostgaard's sourced 1957 excerpt). This turns an unknowable forecast into a conditional state: trend intact, trend opposed, or temporarily out.

The model is reactive but not passive. A trader defines evidence that a trend exists, accepts entering after the move has begun, and surrenders the fantasy of selling the exact high. What matters is the path after entry, not a point estimate of terminal value.

2. Loss-limited, gain-open asymmetry - direct

The payoff objective is asymmetric: terminate adverse positions while leaving favorable positions without a profit target. Donchian's 1974 article treats quick profit-taking as dangerous in commodity trading and builds the 5/20 method around trend continuation, reversal, temporary exit, and reinstatement rather than a fixed gain objective (original 1974 page, readable edited republication).

This does not mean every loss is small or every trend is captured. Gaps can defeat an intended stop, delayed signals surrender part of a move, and repeated false signals can cumulate. It means the rules are designed so that a rare long trend is allowed to become much larger than one ordinary losing trade.

3. Confirmation before commitment - direct

Donchian did not treat a line crossing as sufficient evidence. The 5/20 method requires market-specific penetration units, comparisons with earlier penetrations, 15- and 25-close qualifications, and timing exceptions. The 20-day average defines the basic direction; the faster five-day average mainly handles exit and reinstatement (original 1974 page, Elam's later decision-tree implementation).

The mental model is a hurdle, not an indicator: weak evidence should not receive full commitment merely because it is new. The 1960 copper rule expresses the same idea differently. It buys above the preceding two calendar weeks' highs and sells below their lows, with an extra qualification when the controlling extreme occurred on the last trading day (Donchian's 1960 original). These are two distinct rule families, not interchangeable descriptions of one channel.

4. Strong-long, weak-short selection - compilation via Dixon

The Twenty Guides favor strong-acting, strong-background markets for longs and weak markets for shorts. They also watch for rising volume after dullness, relative activity, chart structure, and statistical background. Barbara Dixon's earliest recovered publication says Donchian compiled the guides in 1934 and rediscovered a copy in 1966; that establishes a reported compilation history, not a recovered 1934 publication (Dixon's 1978 original).

This framework ranks candidates rather than predicting fair value. It also shows that Donchian's practice was never simply price-only. Volume, market background, capitalization, formations, and fundamental statistics could affect selection even when price governed the actual commitment.

5. Light commitment under uncertainty - compilation via Dixon

The Guides advise smaller commitments when the market position is unclear and concentrating attention on clearly defined moves. They pair that selectivity with stop orders, adequate capital, and patience (Dixon's 1978 original). This is a qualitative risk heuristic: confidence changes exposure, but it does not supply a numerical position-size formula.

The missing mathematics matter. No recovered Donchian source specifies percentage equity at risk, volatility-normalized units, maximum leverage, portfolio heat, correlation caps, or a maximum-drawdown stop. Supplying those controls is necessary for a live portfolio, but calling a modern choice Donchian's rule would manufacture evidence.

6. Diversification as both search engine and survival rule - direct

Donchian warned that using the 5/20 method in one or a few selected futures increased risk inordinately. Broad diversification was integral because no trader knows in advance which market will supply the long trend that pays for many false starts (edited 5/20 republication). His reported 1980 explanation similarly emphasizes looking across commodities, limiting losses, and following trends rather than specializing in one contract (Jobman's original profile).

The framework has two jobs: spread contract-specific error and create more opportunities to encounter an outlier. It does not prove that contract count equals diversification. Without volatility weights or correlation limits, many nominal positions can still become one inflation, dollar, growth, or liquidity bet.

7. Fundamentals map the terrain; price triggers the trade - reconstructed

Donchian's mature lesson was not that fundamentals are useless. His 1960 copper article analyzes mine supply, inventories, demand, geopolitics, leverage, and equity relationships before presenting its automatic rule. It even proposes a copper-specific inventory signal, while insisting that futures risk requires stop protection (1960 original).

His earlier equity work makes the hybrid clearer. In 1949 he normalized earnings against commodity-price and inventory regimes rather than trusting a low headline multiple (1949 original). In 1951 he revised a security thesis when monetary conditions changed and paired roughly 80% cash or government bonds with 20% perpetual warrants (1951 original). The defensible reconstruction is that fundamentals supplied context and scenarios, while market evidence prevented valuation from becoming a timing veto.

8. Vehicle capacity is part of the strategy - reconstructed from direct evidence

A method does not operate outside its account, costs, taxes, and scale. In 1960 Futures, Inc. stopped new sales and converted to closed-end operation. The notice cited dilution, preservation of a tax loss, and Donchian's view that a new program suited a roughly $100,000-$500,000 pool better than a much larger fund (contemporaneous 1960 notice).

This is not proof that trend following has one universal capacity ceiling. It is direct evidence that Donchian treated vehicle size and shareholder economics as design constraints. A backtest that omits implementable contract size, margin reserves, subscription flows, fees, and tax consequences is not yet an investment product.

9. Parameter humility and usable simplicity - direct, edited republication

Donchian said five and 20 days were not necessarily the best lengths and invited testing of other averages and price inputs. His stopping rule for research was practical: prefer a comparatively simple method that worked on balance over a long test and could be followed consistently, then change it only when better evidence appeared (edited 5/20 republication).

This is neither parameter worship nor perpetual optimization. The operative model is robust-enough simplicity: test a family of reasonable choices, reject brittle precision, freeze the chosen rules, and distinguish a model change from a discretionary override.

A Reconstructed Decision Checklist

This is an operational reconstruction from the sources, not a recovered Donchian checklist. It deliberately leaves missing fields visible.

1. Choose one architecture. Decide ex ante whether the test concerns the two-calendar-week copper reversal rule, the filtered 5/20 system, or the judgmental Guides. Do not blend their most attractive parts after seeing results (1960 original, original 1974 page, Dixon's 1978 original).
2. Define the investable universe. For 5/20, use enough liquid contracts to make broad diversification feasible. State exchanges, contract months, liquidity thresholds, delivery exclusions, and whether shorts are permitted. Donchian specified breadth but not these modern screens (edited 5/20 republication).
3. Freeze the data convention. Record close, five- and 20-day averages, penetration unit, relevant 15-, 20-, 25-, and 60-session comparisons, signal date, and holiday schedule. For breakout research, specify whether ranges use complete calendar weeks, prior highs/lows, or a later rolling channel. State the continuous-contract and roll method (Elam decision tree, 1960 original).
4. Classify market state. Mark the basic trend long or short and the actual position long, short, or temporarily out. Add volume, relative strength, and fundamental background only if the chosen architecture permits them (edited 5/20 republication, Dixon's 1978 original).
5. Require the stated confirmation. In 5/20, a close must pass the relevant penetration and historical-close test. In the copper rule, price must exceed the prior two complete weeks' range, including the last-day exception. A naive moving-average cross or generic 20-day channel is a different model (original 1974 page, 1960 original).
6. Schedule execution before the next bar. The 5/20 publication uses one-day, weekday, and holiday delays; Elam's later tree makes those exceptions programmable. Record generated signal, scheduled action, actual fill, and any deviation (Elam decision tree).
7. Choose and label a risk budget. Donchian supports lighter commitments under uncertainty, stop use, adequate capital, and diversification. The implementer must supply percentage risk, contract scaling, leverage, margin reserve, gap stress, sector limits, and correlation limits as modern additions, not hidden Donchian rules (Dixon's 1978 original, edited 5/20 republication).
8. Predefine exit, reversal, and reinstatement. A qualifying opposing 20-day signal reverses the basic trend. A sufficiently large five-day penetration can move the position temporarily out and later reinstate it. Stops protect individual positions, but no fixed profit target should amputate the strategy's payoff asymmetry (Elam decision tree, Ostgaard's sourced 1957 excerpt).
9. Audit the portfolio, not just each trade. Monitor aggregate directional and macro exposures, stressed margin, costs, rolls, and drawdown. One contract in each of many markets is not automatically balanced risk.
10. Audit the operator separately from the model. Preserve every generated signal and override. Ed Seykota reported that the Guides did not define a complete conflict-free algorithm and that Donchian took some signals while skipping others (Seykota FAQ). Selective execution must not be compared with a fully systematic backtest.
11. Review only on a declared schedule. Test reasonable neighboring parameters and out-of-sample periods, then document any rule change. A changed horizon creates a new strategy lineage, not proof that the old rule still works (edited 5/20 republication, Seykota's 2020 recollection, Park and Irwin).

The minimum journal would contain market and contract, data convention, basic signal, actual state, qualifying thresholds, scheduled and actual execution, stop, size, costs, portfolio exposure, override and reason, and post-trade outcome. That ledger is necessary to make the rules auditable; the surviving record instead contains an incomplete rule description, a collaborator operationalization, and retrospective adherence evidence (Elam decision tree, Schimmel's provenance reconstruction, Seykota FAQ).

Failure Modes Of The Model

Whipsaw becomes behavioral failure

In a range, breakouts and moving-average reversals can produce many valid small losses. The 5/20 historical study reportedly lost in five of fourteen years and eight of 28 markets; those are secondary fixed-contract test figures, not live account returns (Kase's secondary analysis). After enough losses, the operator may skip the next signal—the very trade that could supply the long outlier. Seykota's testimony makes that adherence risk concrete without identifying the counterfactual P&L of any skipped Donchian trade (Seykota FAQ).

Stops define intent, not maximum realized loss

The Guides and Donchian's 1970 short-selling article use stops to predetermine acceptable risk, and the latter prefers a stop-protected short to an unprotected position (1970 original). But gaps, limit moves, illiquidity, and delayed execution can create a worse fill. Leverage and correlated positions can then convert several modest signal losses into margin pressure. Because Donchian left no complete sizing or aggregate-risk formula, loss limitation at the signal level should not be confused with a proven portfolio loss ceiling.

The rule can be incomplete or internally ambiguous

The 5/20 description contains thresholds, lookback comparisons, timing exceptions, exits, and reinstatements. Brent Elam later translated clarifications into a decision tree; that implementation is collaborator evidence, not a Donchian-authored document (Elam decision tree, Schimmel's provenance reconstruction). Dixon says the Guides require individual interpretation, while Seykota found conflicts. A backtest must therefore publish its conflict hierarchy rather than claim it simply tested “Donchian.”

Parameters and market structure decay

Seykota later said the original two-week horizon ceased working unchanged and that viable horizons lengthened in his experience (Seykota's 2020 recollection). This is retrospective collaborator testimony, not a controlled Donchian experiment. It nevertheless fits Donchian's own parameter humility. Early technical-rule studies also often lacked strong controls for data snooping, parameter choice, risk, and out-of-sample validation (Park and Irwin). A historically profitable rule can represent a durable family resemblance, an exhausted horizon, or an optimized artifact; the available Donchian record cannot fully separate them.

Vehicle economics can erase strategy economics

Futures, Inc. suffered severe point-value impairment before later recoveries and was eventually reported dissolved. The often-repeated four-cent low comes from a reproduced 1982 profile and is not a terminal price or lifetime total return (1982 profile reproduction). Later CTA research documents how fees, voluntary reporting, backfill, survivorship, and graveyard bias can separate gross strategy results from investor experience; that research is industry context, not Donchian account evidence (Bhardwaj, Gorton and Rouwenhorst).

Evidence laundering creates a false complete system

The most persistent error is to combine an original Donchian entry, Dixon commentary, Elam programming choices, Seykota adaptations, Turtle risk rules, and a modern platform channel, then attach the composite performance to Donchian. Barbara Dixon wrote many later newsletters and operated the 5/20 system through 1989, so later output is not automatically Donchian-only (FIA's Dixon profile). Likewise, modern trend research can support a strategy family without verifying Donchian's parameters, discretion, costs, or accounts.

Skill is the architecture; luck is the realized path

Donchian's skill was converting a forecasting problem into repeatable conditions: confirmation, reversal, loss control, diversification, and parameter review. Luck still determines which contract trends, whether the chosen horizon matches the regime, whether a gap defeats the planned exit, and whether early success rewards a sound premise. The reproduced 1982 profile says a profitable cocoa short encouraged the much worse idea that coffee simply had to be expensive; that favorable first outcome reinforced the wrong causal model (1982 profile reproduction). A diversified process can improve exposure to favorable paths, but neither a winning historical illustration nor a surviving maxim proves that skill alone produced the result.

Transferability For An Individual Investor

An individual can reproduce the decision architecture: observe rather than predict; require confirmation; predefine adverse evidence; remain open to a large favorable move; trade both directions where legally and operationally feasible; diversify across genuinely different risks; reduce commitment when evidence is weak; and keep a complete rule-and-override ledger. Public daily data and simple code can calculate ranges and moving averages more reliably than Donchian could by hand.

An individual cannot reproduce the historical program exactly. The recovered sources do not supply the full newsletter archive, continuous contract construction, delivery-month selection, roll timing, original fills, complete cost schedule, percentage-risk formula, leverage policy, correlation limits, margin reserve, discretionary-override protocol, or audited account ledger. Small accounts also face coarse integer contracts and may be unable to obtain the breadth Donchian considered essential (Schimmel's archival reconstruction, edited 5/20 republication).

The practical solution is transparent adaptation. Use current liquid instruments, define a contemporary risk budget and gap stress, state every roll and cost assumption, test neighboring parameters, and label each addition as the implementer's own. The 20/10 and 55/20 breakouts, volatility-based N units, and pyramiding belong to the later Turtle system; the three-line rolling high/low/midpoint channel is a modern platform convention, not a recovered Donchian specification (Original Turtle Rules, TradingView convention).

The durable model is therefore narrower than the legend and more useful: price is allowed to prove the trader wrong, losses are meant to be finite, gains are not assigned a ceiling, broad search creates chances to find an outlier, and every rule remains subordinate to evidence about implementation, capacity, and decay. No named enforcement or litigation matter was found in the bounded digital search, but Donchian's activity largely predates modern CPO/CTA rules and the surviving archive is too incomplete to certify an event-free legal history (CFTC history, Schimmel's archival reconstruction).

Evidence current to: July 19, 2026
Status: Deceased (April 24, 1993)
Central judgment: Donchian's influence is far more verifiable than his investment returns. Inherit the decision architecture, not the legend, fixed parameters, or performance claims.

Executive Brief

Richard Donchian did not invent the human impulse to follow trends. His more consequential achievement was to make that impulse investable and transmissible: a public liquid-commodity vehicle, diversified long and short exposure, and rules that converted observable price behavior into repeatable entry, reversal, and loss-control decisions. Futures, Inc. was publicly offered in October 1948, although managed-futures histories commonly use its 1949 operating start as the field's birthday (1948 notice; 1949 report). That vehicle innovation is documented. A durable investment record is not.

The heroic return narrative fails an institutional evidence test. The strongest executed episode is a diversified short book with company-reported profit of $25,904 [single-source] in the second quarter of 1951; stated net asset value rose from $2.75 to $4.38, then fell to $3.60 [single-source]. The same-metric record from $12.52 in January 1949 to $2.75 in March 1951 implies 78.0% impairment [single-source calculation] before that recovery (1949 report; 1951 notice). No continuous audited total-return series, distribution history, personal ledger, client composite, or liquidation account was recovered. The four-cent low, $27 million under management, and roughly $1 million personal profit remain claims from one derivative reproduction of a 1982 Forbes profile, not audited facts (1982 reproduction).

His method also resists a price-only caricature. Donchian's 1949-1951 equity articles normalized earnings, examined operating leverage and capital structure, and adapted cash-plus-warrant exposure to monetary conditions (1949 article; 1950 article; 1951 article). His 1960 copper study integrated inventories, production, geopolitics, leverage, and a two-week price rule, while warning that a historical profit did not promise repetition (original article). The mature 5/20 framework likewise used filters, delayed execution, exits, reinstatements, diversification, and parameter humility rather than a naive crossover (1974 page; edited continuation).

Implementation was collective. Barbara Dixon wrote many later letters and operated the 5/20 system; her 1978 article is the earliest recovered publication of the Twenty Guides and says their application requires interpretation (Dixon article; FIA biography). Ed Seykota found incomplete and conflicting instructions, programmed selected subsets, and remembered Donchian skipping signals (FAQ; recollection). Donchian therefore deserves credit for product design, directional flexibility, diversification, codification, and learning from failure—not for every modern channel, Turtle rule, collaborator refinement, or trend-following result.

The transferable core is a decision architecture: participate only after observable confirmation, cap a wrong path, keep a favorable path open, search broadly enough for rare outliers, and test the operator as severely as the rule. It is strongest in persistent trends and vulnerable in ranges, gaps, correlation spikes, and crowded short-horizon implementations. Donchian's documented 1993 death makes this a historical reconstruction, not a current recommendation (Foundation chronology; contemporary obituary). For a modern allocator, the correct due-diligence question is therefore not whether a product displays a Donchian channel. It is whether the manager has specified position sizing, correlated exposure, leverage, liquidity, roll treatment, costs, overrides, and an investor-level net record. The historical materials leave those controls incomplete. Modern users must build and validate them rather than fill the gaps with descendant conventions.

Ten Transferable Lessons, Ranked

1. Limit the losing path without capping the favorable path

Donchian's most durable idea is payoff asymmetry, not a particular moving-average length. The surviving 1957 excerpt describes accepting many bounded reversals so a sustained move can become materially larger than any one loss (1957 excerpt and genealogy). That structure can survive a low hit rate; it cannot survive uncontrolled leverage, gaps, or execution failure.

Transfer: Define invalidation and loss capacity before entry, but exit winners because the evidence changes rather than because a round-number gain feels sufficient. Boundary: A stop is an instruction, not insurance. Modern implementations must add gap stress, liquidity assumptions, aggregate exposure, and a capital-at-risk limit that Donchian's recovered writings do not specify.

2. Let observable confirmation outrank an absolute-price opinion

The 1960 copper study paired deep fundamental context with a price rule and showed why an apparently expensive or cheap level does not identify the end of a trend. The 1982 profile similarly presents valuation anchoring as a costly early lesson (copper article; 1982 reproduction). Price did not replace analysis; it adjudicated timing when analysis and market behavior diverged.

Transfer: Translate a thesis into an observable trigger and predefine what contrary behavior would falsify it. Boundary: Confirmation arrives late by design. It will surrender the first and last part of a move and can be disastrous if users quietly convert delayed confirmation into unlimited tolerance.

3. Treat diversification as a search mechanism, not decoration

Donchian emphasized scanning many commodities because the few sustained moves could pay for numerous false starts, a logic he reiterated in a contemporaneous 1980 interview (Jobman interview). Diversification here increases the probability of holding an outlier; it is not merely variance smoothing.

Transfer: Apply one coherent opportunity definition across genuinely different markets or business cases. Boundary: Contract count is not risk diversification. Shared inflation, dollar, liquidity, or deleveraging shocks can make nominally different positions one bet, so correlation, sector, and portfolio-heat controls must be supplied today.

4. Choose one rule architecture before testing it

Donchian's two-week reversal, 5/20 method, and Twenty Guides are related but distinct (copper article; 1974 page; Dixon article). Combining their best-looking clauses after observing outcomes creates a strategy that never existed. The 1974 method also contains penetration tests, execution delays, exceptions, exits, and reinstatement—not one crossover (edited continuation).

Transfer: Version the exact decision tree, data inputs, exceptions, and fill assumptions before evaluating results. Boundary: A readable slogan is not a reproducible strategy. If two researchers cannot derive the same orders from the specification, any backtest precision is cosmetic.

5. Audit the operator separately from the rule

Seykota's testimony that Donchian sometimes skipped signals turns discretion into a separate source of risk and possible return (FAQ; recollection). A sound rule can be defeated by selective compliance; a weak rule can appear superior when overrides are remembered selectively.

Transfer: Log every signal, order, override, reason, and counterfactual outcome. Evaluate base rules, overrides, and actual execution separately. Boundary: Retrospective colleague accounts are not an account ledger. The appropriate conclusion is that implementation risk existed, not that discretion definitively helped or hurt.

6. Treat every parameter as provisional

Donchian's own mature presentation warned against assuming 5 and 20 days were universally optimal (1974 page; edited continuation). Seykota later recalled that horizons lengthened as shorter rules became less effective (recollection). Independent reviews of technical-rule studies also warn about data snooping, transaction costs, risk adjustment, and weak out-of-sample design (Park and Irwin review).

Transfer: Prefer robust parameter regions, multiple samples, walk-forward tests, realistic costs, and simple economic explanations to one historical optimum. Boundary: Constant retuning can turn humility into overfitting. Change a parameter only through a prospective governance process, not because recent losses are uncomfortable.

7. Use fundamentals to map the regime and price to govern timing

The early equity work and the copper article show a hybrid investigator: inventories, production, monetary conditions, business economics, and leverage informed the opportunity set; market action governed commitment and reversal (1949 article; 1950 article; 1951 article; copper article). Calling Donchian purely mechanical erases this evolution.

Transfer: Separate a slow regime model from a fast execution model, and state which one has authority when they conflict. Boundary: A fundamental story can become a license to ignore the stop, while a price rule can encourage trading without understanding contract economics. The two layers need an explicit conflict protocol.

8. Reduce commitment under weak evidence—and supply the missing sizing system

The Twenty Guides discuss light commitment, capital preservation, pyramiding cautions, and diversification qualitatively. They do not recover a complete percentage-risk, volatility, leverage, correlation, or maximum-drawdown policy (Dixon article).

Transfer: Scale exposure with signal quality only after imposing non-negotiable portfolio constraints. Normalize risk across instruments, cap correlated clusters, reserve liquidity, and model limit moves. Boundary: Do not launder later Turtle volatility units or modern portfolio heat back into Donchian. These are necessary contemporary complements, not recovered Donchian rules.

9. Treat vehicle economics as part of the strategy

Futures, Inc. eventually closed to new sales while converting structure because new capital could dilute tax assets and alter shareholder economics (1960 vehicle notice). Fees, commission arrangements, reserves, liquidity, subscriptions, redemptions, capacity, and tax treatment determine whether a model edge reaches an investor.

Transfer: Reconcile gross model results to net investor experience, including collateral return and every layer of friction. Boundary: Managed-futures databases can suffer selection, survivorship, backfill, and fee distortions; later industry evidence cannot reconstruct Donchian's missing ledger (CTA evidence).

10. Separate influence, system studies, and investor returns

Donchian's intellectual lineage is strong: a public commodity vehicle, decades of rule transmission, collaborators, and recognizable descendants (1948 notice; Dixon article; FIA biography). The evidentiary bridge from that influence to lifetime investor wealth is weak. A fixed-contract historical study is not a fund return; AUM is not profit; a partial NAV sequence is not total return; a collaborator-operated model is not Donchian's personal performance (Kase analysis; 1982 reproduction; 1951 notice).

Transfer: Maintain separate ledgers for ideas, simulations, model accounts, client composites, personal capital, and vehicle shareholders. Boundary: This discipline may yield a less exciting biography, but it prevents a famous framework from becoming fabricated alpha.

Style Taxonomy

diversified-managed-futures; long-short-trend-following; breakout-and-moving-average; rules-based-with-discretion; hybrid-fundamental-technical; loss-limited-gain-open; broad-commodity-search; parameter-adaptive; capacity-aware; pre-modern-CTA.

Do not classify Donchian as purely price-only, the author of the complete Turtle system, the inventor of every modern channel display, or a manager with a verified continuous audited record. The Turtle package includes volatility units, pyramiding, portfolio caps, and 20/10 and 55/20 breakouts, while the current platform channel adds an upper/lower midpoint convention not established by Donchian's recovered original text (Turtle rules; modern platform definition).

Regime Dependence

These are mechanism-based expectations, not reconstructed Donchian returns. Modern time-series-momentum research supports the broader existence of trend persistence, but it does not authenticate Donchian's exact rules or record (time-series momentum study).

Skill, Luck, Criticism, and Verification

The strongest evidence of skill is architectural: Donchian launched an unusually early public vehicle, traded both directions, treated losses as the price of finding outliers, diversified the search, published testable rules, revised methods, and transmitted them to practitioners (1948 notice; Jobman interview; 1974 page). This is invention and institution-building skill even if it cannot be converted into a trustworthy CAGR.

Luck and selection are inseparable from the surviving performance stories. Trend strategies depend on rare sustained moves; selected copper, cocoa, or coffee examples emphasize the very outliers the method seeks. A later reconstruction finds substantial market and year dispersion even within the fixed-contract 5/20 study (Kase analysis). Historical totals depend on contracts, sample windows, data, roll rules, fills, costs, collateral, and capital denominators that were not recovered. Favorable regimes can therefore be mistaken for manager skill, while one-contract simulations can be mistaken for scalable returns.

The principal criticism is not that the architecture lacks value. It is that the legend outruns the ledger. Futures, Inc. suffered severe documented impairment [single-source calculation]; the derivative four-cent claim is neither a verified terminal NAV nor lifetime shareholder return; collaborator authorship is mixed; rules contain ambiguity and discretion; and crucial sizing, portfolio-risk, roll, and cost conventions are absent (1949 report; 1951 notice; 1982 reproduction; Dixon article; Seykota FAQ). No named enforcement or legal matter surfaced in bounded current searches, but that is not a clean-record certification. Pre-1974 federal coverage applied to an evolving list of commodities, the 1974 Act broadened jurisdiction, the first CPO/CTA operating rules arrived in 1979, and NFA began regulatory operations in 1982 (pre-CFTC history; 1970s history; NFA history). Offline predecessor-agency, exchange, state, court, and arbitration files remain incomplete.

Closest and Most-Opposite Investors Already in the Canon

Closest

• Ed Seykota is the direct collaborator and computational descendant. Both emphasize price response, diversified futures, and loss asymmetry; Seykota adds systematic testing, portfolio heat, and client-psychology governance.
• Richard Dennis is the closest formal breakout descendant. Dennis and the Turtles add volatility-normalized sizing, pyramiding, portfolio caps, and standardized channel machinery that should not be back-attributed to Donchian.
• William Eckhardt shares complete-system thinking and parameter humility, while adding robust statistics, volatility sizing, utility, and explicit anti-overfit discipline.
• Larry Hite institutionalized the diversified, loss-limited, rare-large-winner architecture with more explicit sizing, correlation, and no-override governance.
• Michael Marcus is the closest hybrid cousin: commodity fundamentals plus technical confirmation, but with more concentration, discretion, and market-tone judgment.

Most opposite

• Jack Bogle expresses humility through passive beta, minimal turnover, and no tactical timing rather than active long/short futures, reversal, and stops.
• Warren Buffett relies on business cash flows, quality, permanent capital, and very long holding periods rather than liquid price confirmation and futures margin.
• Benjamin Graham buys valuation dislocation with a margin of safety; Donchian joins established price movement. Both codify rules and diversify, but the triggers point in opposite directions.
• Nick Sleep seeks a few rare, concentrated, low-turnover quality compounders rather than repeated entries across a broad market search.
• John Templeton enters at maximum pessimism through valuation-led contrarian baskets, the mirror image of waiting for directional confirmation.

Unresolved Questions

• Complete 1957 text: physical Commodity Year Book holdings covering 1957 exist at the New York Public Library and Library of Congress. A lawful copy of pages 35-47 could replace the surviving excerpt with the full argument.
• Newsletter corpus: no public accession record, finding aid, issue inventory, or complete run resolves the titles, dates, missing issues, corrections, performance tables, or Donchian-versus-Dixon authorship.
• Lifetime vehicle return: audited statements, dividends, subscriptions, redemptions, fees, final NAV, dissolution documents, and liquidation distributions remain missing.
• Account boundaries: Futures, Inc. shareholders, Donchian's personal capital, brokerage clients, newsletter models, historical studies, and Dixon-operated accounts have not been reconciled.
• Original 1974 continuation: the visible original page and posthumously edited republication still require a sentence-level comparison against the complete printed issue.
• System ownership: the rule-level contributions of Donchian, Dixon, Brentin Elam, and Seykota remain only partly separable.
• Portfolio risk: no complete specification for percentage risk, volatility normalization, leverage, margin reserve, sector/correlation caps, portfolio heat, or maximum drawdown was recovered.
• Backtest mechanics: contract selection, rolls, delivery exclusions, corrected prices, ambiguous signals, fills, commissions, slippage, collateral, and reserve capital remain unspecified.
• Offline adverse record: National Archives Record Group 180 contains predecessor-agency correspondence, case, regional, and enforcement files that need name-and-address-level review (record-group guide).
• Original 1982 profile and estate: the original Forbes page, exact birth date, probate, estate disposition, corporate records, and archive transfers remain unrecovered.

Bottom Line

Donchian's durable edge was not foresight about the next commodity. It was a system for admitting uncertainty: wait for evidence, exit a failed path, retain exposure to a persistent one, and search broadly enough for positive skew to matter. His durable warning is equally important. Without an exact rule book, risk system, execution record, and investor-level ledger, a good architecture can be turned into a false performance history. The Canon should preserve both halves.

As of: 2026-07-19. Ranked by authority and role in the chapter. This Task A map covers exactly the 25 unique external URLs cited in profile.md. The point record is strongest from 1948 through 1957; later return, AUM and dissolution claims are much less complete.

Task A - Profile (T0624)

1. Commercial and Financial Chronicle - February 21, 1949 - Best contemporaneous foundation record: October 1948 public offer, $10 opening NAV, late-1948 profit, January 1949 NAV/dividend, structure and early positions; OCR must be checked against the scan.
2. Commercial and Financial Chronicle - May 4, 1950 - Contemporaneous SEC-notice report for 55,000 shares, Futures Distributors, 40 Wall Street and the use of proceeds to trade commodities and futures long or short.
3. Commercial and Financial Chronicle - January 25, 1954 - Contemporaneous evidence of the bid recovering from $1.59 to $2.39 and of cocoa/coffee trend positions; proves a severe decline from the $10 offer.
4. Commercial and Financial Chronicle - March 14, 1955 - Records the additional 94,000-share public offer at a $4.42 daily-NAV-linked price.
5. Commercial and Financial Chronicle - June 6, 1955 - Supplies exact 1954-1955 assets, shares, NAV and position comparisons; supports arithmetic and vehicle-structure checks.
6. Commercial and Financial Chronicle - June 20, 1957 - Latest strong fund snapshot located: March and May 1957 assets/NAV, share count and dividend-policy context.
7. Richard D. Donchian Foundation - founder biography - Best official chronology for family, education, employment, military service, brokerage affiliations, newsletter, recognition and death; a memorial foundation source, not an independent return record.
8. Richard Donchian - High Finance in Copper, DOI record - Primary identifier for Donchian's November-December 1960 Financial Analysts Journal article, showing his combination of commodity fundamentals and technical rules.
9. Financial Analysts Journal volume 16, issue 6 - Publisher issue metadata corroborating the copper article's author, title, date and placement; full text requires access.
10. Twenty Trading Guides PDF - Barbara Dixon's October 1978 article reproducing 11 general and nine technical guides and clarifying that they were compiled, not demonstrably published, in 1934.
11. CMT Association - July 2011 newsletter and Annual Award - Institutional recognition, life chronology and concise four-week-rule description; partly draws on the Foundation.
12. Wolfgang Schimmel - The Donchian Puzzle - Most important adverse archival reconstruction: separates hybrid practice, system hypothetical and fund record, and flags missing documents; a mirror of a July 2011 trade-journal article whose original publisher PDF was unavailable.
13. FIA Hall of Fame - Barbara S. Dixon - Independent trade-association record confirming Dixon's 1970-1973 work, letter-writing role and operation of the 5/20 system through 1989.
14. Stig Ostgaard - On the Nature and Origins of Trend Following - Historical synthesis and bibliography locating the 1957, 1960, 1974, 1980 and 1982 materials; a useful source map rather than a substitute for originals.
15. New York Times - Richard B. Donchian; Commodities Specialist, 87 - Contemporaneous obituary for age and death timing; its “B.” conflicts with the well-supported Davoud identity and is treated as a headline error.
16. Forbes - 1982 excerpt on Futures, Inc. failure - Derivative excerpt of William Baldwin's “Rugs to Riches,” supporting the $10 offer, coffee/cocoa origin story, reported $0.04 low before dissolution and the dissolution claim; single-source until the original is recovered.
17. Forbes - 1982 profile/AUM mirror - Longer derivative presentation of the reported $27 million AUM and fee/personal-profit figures; explicitly not treated as an audited return source.
18. Chris Perruna - Donchian's 5- and 20-Day Moving Averages - Substantial derivative reprint of the 1974 system article, useful for rules and test period but not independent performance evidence.
19. CME Group - The Miraculous Growth of Managed Futures - Exchange history distinguishing the conventional 1949 industry date from the prior-year discretionary-account approval.
20. CME Group - What Are Managed Futures? - Exchange primer situating Donchian in the later CTA model and clarifying how the modern universe expanded beyond physical commodities.
21. A Survey of the Managed Futures Industry, second edition - Industry history describing Futures, Inc. as the first commodity fund and using the 1948 date; its Donchian notes ultimately rely on the Foundation and obituary.
22. CFTC - History of the CFTC in the 1970s - Primary chronology for the agency's 1974 creation and first Part 4 CPO/CTA operating rules in 1979, explaining but not curing the disclosure gap.
23. Moskowitz, Ooi and Pedersen - Time Series Momentum - Academic evidence across 58 liquid instruments; supports the time-series-momentum family, not Donchian's missing returns.
24. Hurst, Ooi and Pedersen - A Century of Evidence on Trend-Following Investing - Long-history practitioner research on diversified trend following, retained with affiliation and backtest caveats.
25. Lemperiere et al. - Two Centuries of Trend Following - Historical study finding persistent long-horizon trend effects and weakening shorter-horizon results; validates neither Donchian's execution nor investor outcomes.

Task A evidence limitations

• No continuous audited Futures, Inc. total-return series, complete offering document, annual report, fee schedule, shareholder ledger or liquidation filing was located.
• The contemporaneous record establishes an October 1948 public offer. The recurring 1949 date is best treated as a later industry convention, not the actual first-offer date.
• The opening NAV gain was brief. Mid-1950s prices/NAV were far below the $10 offer, and Forbes later reported $0.04 before dissolution; missing distributions prevent an exact lifetime total return.
• The 1961-June 1974 result is a fixed-contract historical system study with incomplete denominator and cost detail. It is not a personal or fund CAGR.
• The $27 million 1982 AUM and associated fee/personal-profit figures derive from one Forbes profile accessible only through derivative mirrors. AUM is not return and does not belong automatically to Futures, Inc.
• The complete 1960-1979 newsletter run and original 1957 and 1974 articles were not publicly recovered; later descriptions and reprints may omit context or revisions.
• Dixon's operation of the 5/20 system through 1989 does not prove that Futures, Inc. survived to 1989 or that results belonged to one continuous vehicle.
• Academic trend studies use later data, instruments and portfolio construction. They support a general mechanism, not Donchian-specific alpha.
• No named enforcement action was located, but the vehicle's pre-1979 history and incomplete archives make that a limited negative finding, not proof of a spotless record.

Task B - Investment Philosophy (T0625)

Ranked by authority and role in the chapter. This Task B map covers exactly the 14 unique external URLs cited in investment-philosophy.md; repeated appearances of one URL in the chapter count once here.

1. Barbara Dixon - Donchian's Twenty Trading Guides - Best surviving near-primary statement of the general and technical principles, including loss limitation, lighter commitments under uncertainty, stops, relative strength, countertrend entry and the need for individual interpretation; it says the guides were compiled, not verifiably published, in 1934.
2. Richard Donchian - 5- and 20-day moving averages - Substantial posthumous article transcript establishing the philosophy, penetration filters, historical-close tests, execution delays, diversification requirement and Donchian's explicit parameter humility; a hosted transcript rather than the original magazine scan.
3. Computerized Donchian 5/20 decision tree - Surviving programmer's operationalization of the entry, exit, reinstatement, timing and holiday exceptions; demonstrates that the method was not a bare moving-average crossover.
4. Stig Ostgaard - Trend Following: On History, Verity, and Seeking the Noumenon - Historical source map preserving the documented 1957 loss-limiting/gain-unlimited passage and the 1980 multi-commodity diversification quotation; secondary synthesis, not a return audit.
5. Ed Seykota - FAQ aggregation - First-person collaborator evidence that the 5/20 system appeared economically viable, the Twenty Guides did not define a complete conflict-free algorithm, and Donchian took some signals but skipped others.
6. Ed Seykota - From Trend Following - Retrospective first-person evidence on the two-week copper rule, horizon decay and the gap between Donchian's stated system and personal adherence; useful but not an audited test.
7. Forbes 1982 Donchian excerpt - Most revealing account of the shift from valuation calls to reactive trend following and of Futures, Inc.'s reported $0.04 low before dissolution; a reproduction rather than the original scan.
8. FIA Hall of Fame - Barbara S. Dixon - Independent trade-association confirmation that Dixon wrote many later letters and operated the 5/20 system through 1989, limiting sole-author and sole-operator claims for Donchian.
9. Richard Donchian - High Finance in Copper DOI - Primary identifier for the 1960 Financial Analysts Journal article and the evidence boundary around his hybrid fundamental/technical copper work; full text remains access-controlled.
10. Moskowitz, Ooi and Pedersen - Time Series Momentum - Peer-reviewed evidence for one- to 12-month return persistence, longer-run partial reversal and possible information/hedging-pressure mechanisms across 58 futures and forwards; strategy-family evidence only.
11. Hurst, Ooi and Pedersen - Demystifying Managed Futures - Modern multi-horizon explanation of trends, diversification, extreme-market behavior, volatility scaling, costs and fees; practitioner-affiliated and not Donchian-specific.
12. Szakmary, Shen and Sharma - Trend-following trading strategies in commodity futures - Peer-reviewed abstract for 48 years and 28 commodity markets, reporting broad net profitability and data-mining robustness across tested dual-moving-average and channel rules.
13. Lemperiere et al. - Two centuries of trend following - Long-history strategy-family evidence distinguishing persistent long-horizon trends from weakened shorter horizons; validates neither Donchian's parameters nor his execution.
14. Bhardwaj, Gorton and Rouwenhorst - CTA performance and persistence - Peer-reviewed adverse industry evidence on fees, voluntary reporting, backfill, survivorship and graveyard bias; later-period context for separating gross strategy plausibility from investor outcomes.

Task B evidence limitations

• The original 1957 article, original 1974 article, complete newsletter run and a continuous audited Donchian trading record were not recovered on the open web.
• The Twenty Guides require interpretation and, according to Seykota, do not themselves define a complete internally consistent system.
• No defensible primary rule was located for percentage risk per trade, volatility-normalized sizing, correlation caps, portfolio leverage or maximum drawdown. Later Turtle and CTA controls are not attributed to Donchian.
• The 5/20 rules are a filtered conditional tree. Describing them as a simple crossover materially misstates the evidence.
• Modern three-line channels, 20/10 Turtle rules, ATR sizing, midpoint exits and platform defaults are intellectual descendants, not established Donchian-authored rules.
• Seykota's recollections are uniquely valuable first-person evidence but remain retrospective. Dixon's surviving article is near-primary and hosted on a practitioner site.
• The accessible 1982 Forbes text is a reproduction. It supports a reported low before dissolution, not a terminal share price or lifetime investor return.
• Academic trend studies support or challenge the broader strategy family. They do not establish Donchian's personal alpha, exact costs, execution quality or client returns.

Task C - Greatest Trades (T0626)

Ranked by authority and role in the chapter. This Task C map covers exactly the 20 unique external URLs cited in greatest-trades.md; repeated appearances of one URL in the chapter count once here. Several URLs are different access paths to the same source family and are not treated as independent corroboration.

1. Commercial and Financial Chronicle - September 17, 1951 scan - Strongest campaign record: Q2 profit, named short markets, NAV rise, and the subsequent two-month drawdown; visually checked against the scan; all figures are one single-source numerical lineage.
2. Commercial and Financial Chronicle - September 17, 1951 searchable text - Searchable access path for the same single-source corporate notice; OCR is not independent evidence.
3. Commercial and Financial Chronicle - November 1, 1948 - Contemporaneous opening thesis, portfolio design, selective-short plan and reserve policy.
4. Commercial and Financial Chronicle - February 21, 1949 - Single-source opening campaign result: dates, $3,953 after-tax realized plus unrealized profit, NAV path, dividend, short bias, and cocoa/soybean contribution.
5. Commercial and Financial Chronicle - January 25, 1954 - Single-source contemporaneous long cocoa/coffee campaign, share-price path, profit-taking, and rotation into corn, soybean and wool shorts; direction checked against the original scan.
6. Commercial and Financial Chronicle - March 14, 1955 - Single-source share-offering context establishing that AUM growth included issuance and cannot be treated wholly as investment performance.
7. Commercial and Financial Chronicle - June 6, 1955 - Single-source March and September assets, shares outstanding, and NAV for the 1954-55 recovery.
8. Commercial and Financial Chronicle - August 19, 1957 - Single-source July 1957 share-price path and attribution to short No. 4 world sugar plus long wheat, soybean and egg positions; visually checked against the scan.
9. Commercial and Financial Chronicle - September 16, 1957 - Single-source ten-month $92,551 net gain, NAV change, total assets and comparison dates; not a market-level P&L statement.
10. Commercial and Financial Chronicle - December 9, 1957 - Single-source fiscal-year net assets, net earnings, NAV, dividend and reported percentage gain; visually checked to prevent assets/earnings reversal.
11. Richard Donchian - High Finance in Copper DOI - Primary bibliographic identifier for the November-December 1960 article; full text remains access-controlled.
12. Park and Irwin - The Profitability of Technical Analysis - Single-source numerical reproduction of the two-week copper rule's markets, period, costs, stated margin and 1959-60 net gains; table visually checked.
13. Richard Donchian - 5- and 20-day moving averages - Hosted transcript of the 1974 article for the detailed rule tree, diversification instruction and historical-study context; not an original scan.
14. Cynthia Kase - Trading with the Odds - Secondary synthesis for market/year dispersion in the 5/20 study and soybeans' uneven contribution; period labeling conflicts with the facsimile lineage.
15. Wolfgang Schimmel - The Donchian Puzzle - Historical reconstruction and facsimile lineage for the more-than-$250,000 5/20 result; useful adverse source, not an audited account record.
16. Ed Seykota - FAQ aggregation - Retrospective first-person collaborator evidence on computerization, apparent client viability, rule conflict and Donchian's skipped signals; no account ledger.
17. Ed Seykota - From Trend Following - Retrospective evidence that the short copper horizon decayed and published rules differed from personal adherence.
18. Forbes 1982 excerpt reproduction - Derivative source for the cocoa/coffee origin story and later fund failure; exact cocoa prices and later income/AUM figures remain single-source.
19. FIA Hall of Fame - Barbara S. Dixon - Trade-association record establishing Dixon's letter-writing and 5/20 operating role, preventing automatic sole attribution to Donchian.
20. CFTC - Futures market basics - Regulator primer for the futures-contract and margin context used to distinguish a margin multiple from a full-capital return.

Task C evidence limitations

• No complete trade ledger, audited Futures, Inc. total-return series, personal-account statements, Shearson/client composite, offering-document run, fee record or liquidation file was recovered.
• The 1951 campaign is the most complete, but its $25,904 is vehicle-period profit and the source does not disclose realization status, cash flows, costs, taxes, contracts or market-level attribution.
• Share quotes and NAV snapshots are not interchangeable. Quoted moves cannot be assigned directly to one futures position, while AUM growth can include subscriptions.
• The 1959-60 copper results and 1961-June 1974 5/20 result are historical tests. Stated-margin multiples are not capital-normalized returns, and the 5/20 study lacks starting equity, costs, drawdown and a complete ledger.
• The exact 5/20 aggregate dollar result follows one facsimile/source lineage. The accessible article transcript and Kase synthesis help with rules and dispersion but do not independently audit that total.
• The accessible 1982 Forbes material is derivative. It does not independently establish trade size, exact exits, personal-account returns or a continuous fund record.
• Dixon's later operation and Seykota's recollections prevent simple sole-author or all-signals-executed claims. Neither source supplies an audited Donchian account.
• No named enforcement action or lawsuit was found in targeted searches, but archive gaps and the vehicle's early period make that a bounded negative finding, not proof of absence.

Task D - Mistakes and Losses (T0627)

Ranked by authority and role in the chapter. This Task D map covers exactly the 21 unique external URLs cited in mistakes-and-losses.md; repeated appearances of one URL count once here.

1. Commercial and Financial Chronicle - September 17, 1951 - Strongest discrete loss record: $2.75 opening-period NAV, the Q2 rise to $4.38, and the July-August decline to $3.60; a single contemporary corporate notice, not a complete return series.
2. Commercial and Financial Chronicle - March 24, 1960 - Best documented organizational response: closed-end conversion, suspended sales, continued NAV redemptions, quoted prices, tax-loss preservation, dilution concerns, and Donchian's stated small-fund capacity rationale.
3. Commercial and Financial Chronicle - February 21, 1949 - Contemporaneous opening-period profit and short-bias record; corroborates the valuation-anchoring setup but attributes profit jointly to cocoa and soybeans and does not document the coffee loss.
4. Commercial and Financial Chronicle - January 25, 1954 - Contemporaneous $1.59 bid low and $2.39 recovery point; proves severe quoted-price impairment while leaving distributions and lifetime return unresolved.
5. Commercial and Financial Chronicle - June 23, 1958 - March 1958 net-assets/NAV comparison and June 13 offering price; documents renewed weakness but also explains why a linked offering price is not a same-metric NAV drawdown.
6. Commercial and Financial Chronicle - November 1, 1948 - Contemporaneous $10 public-offer and original selective long/short, diversification, and reserve-capital design.
7. Commercial and Financial Chronicle - June 6, 1955 - Exact 1954-55 NAV, total-assets, and share-count comparison, showing why asset growth mixed performance with issuance.
8. Commercial and Financial Chronicle - December 9, 1957 - Fiscal-1957 NAV recovery and dividend; establishes that the fund path included large recoveries as well as losses.
9. Commercial and Financial Chronicle - April 7, 1958 - Regulatory notice for 54,300 additional shares, showing that the fund was still seeking capital after the 1957 recovery.
10. Richard D. Donchian Foundation - founder biography - Institutional memorial account of personal losses in the 1929 crash and the subsequent turn to technical analysis; supplies no loss amount or account record.
11. Barbara Dixon - Donchian's Twenty Trading Guides - Near-primary 1978 presentation connecting post-1929 losses to technical study and documenting later stop, exposure, capital, patience, and interpretation controls.
12. Forbes 1982 Donchian excerpt - Reported first-person cocoa/coffee lesson and later four-cent low/dissolution claim; accessible only as a derivative reproduction and treated as one single-source lineage.
13. Wolfgang Schimmel - The Donchian Puzzle - Adverse archival reconstruction and Brent Elam oral history concerning discretion, model conflict, later share repurchases, and attempted liquidation; retrospective and not an audited fund record.
14. Ed Seykota - FAQ aggregation - First-person collaborator evidence on skipped signals, apparently better results from consistent followers, and incomplete or conflicting published rules.
15. Ed Seykota - From Trend Following - Retrospective account of inconsistent adherence and decay in the original two-week horizon; Seykota's judgment, not a Donchian admission or account audit.
16. Richard Donchian - 5- and 20-day moving averages - Hosted article transcript for rule complexity, diversification warnings, parameter humility, and admitted scope for rule improvement.
17. Cynthia Kase - Trading with the Odds - Secondary reproduction of year/market dispersion in the 5/20 study; numerical family remains single-source and describes a historical test, not live P&L.
18. Park and Irwin - The Profitability of Technical Analysis - Academic review placing Donchian's copper study among early technical-rule research and defining data-snooping, optimization, risk, and out-of-sample limits.
19. Bhardwaj, Gorton and Rouwenhorst - CTA performance and persistence - Later-period academic evidence on fees and voluntary-database biases; context for evaluating system claims, not Donchian-specific performance.
20. CFTC - History of the CFTC in the 1970s - Official regulatory chronology bounding the negative legal search and the pre-modern-rule archive gap.
21. FIA Hall of Fame - Barbara S. Dixon - Trade-association record confirming Dixon's later letter-writing and system-operation role, preventing automatic attribution of her results to Donchian or Futures, Inc.

Task D evidence limitations

• No continuous audited Futures, Inc. total-return series, complete distribution history, trade ledger, margin record, original 1982 Forbes scan, dissolution filing, final NAV, or shareholder liquidation schedule was recovered.
• Share quotes, NAV, offering prices, fund assets, and investor total return are different measures. The chapter uses point-to-point arithmetic only within defensible boundaries and labels each material single-source number.
• The cocoa/coffee episode preserves Donchian's reported behavioral lesson, but dates, contracts, size, margin, loss, exit, and account identity remain unknown.
• The reported $0.04 was an interim low before dissolution, not a documented terminal price or auditable lifetime investor loss.
• Skipped signals and conflicting rules come from Seykota and Elam. They are evidence of implementation tension, not Donchian admissions, documented client breaches, or calculable errors of omission.
• The 5/20 dispersion and copper-horizon evidence concern historical tests or retrospective judgments. Neither establishes live Donchian account P&L.
• No named legal or regulatory matter was found in the searched digital sources, but pre-digital, state, exchange, predecessor-agency, and arbitration records are incomplete.

Task E - In His Own Words (T0628)

1. Commercial and Financial Chronicle - “Are Stocks as Cheap as They Seem?”, March 31, 1949 - Complete bylined article showing Donchian’s early earnings-quality, valuation and regime analysis.
2. Commercial and Financial Chronicle - “Western Union,” March 2, 1950 - Complete bylined operating-recovery thesis based on mechanization, costs and capital structure.
3. Commercial and Financial Chronicle - “Tri-Continental Warrants Plus Plenty of Cash,” June 14, 1951 - Complete bylined barbell recommendation designed for asymmetric upside and downside.
4. Stig Ostgaard - Trend Following: On History, Verity, and Seeking the Noumenon - Source map preserving the documented 1957 loss-limiting passage and exact Commodity Year Book page citation.
5. Richard Donchian - High Finance in Copper original scan - Original journal pages for the hybrid fundamental and two-week technical study.
6. Richard Donchian - High Finance in Copper DOI - Independent primary bibliographic identifier for the November-December 1960 article.
7. Commercial and Financial Chronicle - Futures, Inc. closed-end announcement, March 24, 1960 - Contemporaneous report quoting Donchian on capacity, dilution and suspending sales.
8. Commercial and Financial Chronicle - “Short Selling Is More Common Than Generally Realized,” August 13, 1970 - Complete scanned bylined article on short selling, stops and two-sided market participation.
9. Donchian’s Five- and 20-Day Moving Averages original page - Original December 1974 Commodities page, controlling over the edited posthumous reproduction.
10. Barbara Dixon - Donchian’s Twenty Trading Guides original page - Original October 1978 publication and controlling evidence that the guides were compiled, not demonstrably published, in 1934.
11. FIA Hall of Fame - Barbara S. Dixon - Trade-association evidence that Dixon wrote many later letters and operated the 5/20 system, limiting Donchian-only attribution.
12. Darrell Jobman 1980 Donchian interview original page - Original Commodities page with explicit speaker framing for the diversification quotation.
13. William Baldwin 1982 Forbes excerpt reproduction - Derivative access to the reported trend-persistence and cocoa/coffee statements; original Forbes p. 143 remains unrecovered.
14. Wolfgang Schimmel - The Donchian Puzzle - Strongest archival reconstruction for the 1960/1974 works, newsletter holdings and Brentin Elam recollections; not Donchian voice.
15. Richard D. Donchian Foundation biography - Family-foundation chronology for the newsletter, career and 1983 award; celebratory rather than critical.
16. Ed Seykota - FAQ aggregation - First-person colleague evidence used to exclude Seykota’s system criticisms and recollections from Donchian’s own corpus.
17. Ed Seykota - From Trend Following - Later oral-history evidence on the two-week rule and changing horizons; Seykota’s words, not Donchian’s.

Task E evidence limitations

• No authenticated Donchian speech transcript, recording, podcast, oral history, award-acceptance text, memoir, book or signed public letter archive was recovered.
• The original 1957 article, complete Commodity Trend Timing run and 1982 Forbes page remain unavailable in open publication-grade form; the 1960, 1974, 1978 and 1980 magazine pages were recovered.
• The 1934 date applies to compilation of the Twenty Guides, not a demonstrated publication; Dixon’s 1978 article is the earliest recovered publication and contains her own commentary.
• The original December 1974 5/20 page controls over the edited posthumous 1995 reproduction; unmatched later wording is not silently backdated.
• Aggregate quoted words from every underlying work, including all mirrors in one lineage, are capped at 25; every individual excerpt is also under 25 words.
• Modern Turtle rules, software-channel defaults, ATR sizing, fixed-percentage risk, generic maxims and colleague testimony are not back-attributed to Donchian.

Task F - Key Writings (T0629)

Ranked by authority and role in the chapter. This Task F map covers exactly the 24 unique external URLs cited in key-writings.md; repeated appearances of one URL count once here.

1. Richard Donchian - High Finance in Copper original scan - Complete original 1960 journal article and strongest recovered primary work; visually checked across printed pages 133-142.
2. Richard Donchian - Donchian's 5- and 20-Day Moving Averages original page - Original December 1974 Commodities page and controlling text for the mature rule set, diversification and parameter caveats.
3. Commercial and Financial Chronicle - Short Selling Is More Common Than Generally Realized - Complete original August 13, 1970 issue PDF; printed pages 438-439 cover mechanics, stop protection and two-sided participation.
4. Commercial and Financial Chronicle - Are Stocks as Cheap as They Seem? - Complete bylined 1949 article on earnings quality, valuation and commodity/inventory regimes.
5. Commercial and Financial Chronicle - Western Union - Complete bylined 1950 operating-recovery thesis based on mechanization, costs, services and capital structure.
6. Commercial and Financial Chronicle - Tri-Continental Warrants Plus Plenty of Cash - Complete bylined 1951 barbell recommendation built around cash and perpetual warrants.
7. Park and Irwin - The Profitability of Technical Analysis - Independent academic citation for the title, year and pages of the unrecovered 1957 chapter, plus method context; not a substitute for its full text.
8. Google Books - Commodity Year Book catalog - Catalog and table-of-contents confirmation for the volume containing the 1957 chapter; no open chapter scan was recovered.
9. Stig Ostgaard - On the Nature and Origins of Trend Following - Best accessible source for the 1957 page-35 excerpt and intellectual genealogy; later practitioner analysis rather than a facsimile.
10. High Finance in Copper DOI - Independent bibliographic identifier for the November-December 1960 Financial Analysts Journal article.
11. Barbara Dixon - Donchian's 20 Guides to Trading Commodities original page - Earliest recovered publication of the guide compilation and primary evidence for Dixon's commentary and newsletter role.
12. FIA Hall of Fame - Barbara S. Dixon - Trade-association evidence that Dixon wrote many later letters and operated the 5/20 system, limiting sole Donchian attribution.
13. Wolfgang Schimmel - The Donchian Puzzle - Strongest archival and source-critical reconstruction for newsletter custody, rule ambiguity, collaborator interviews and adverse evidence.
14. Richard D. Donchian Foundation - founder biography - Institutional memorial chronology corroborating the long newsletter run; celebratory, title-conflicted and not issue-level authorship evidence.
15. Darrell Jobman - Richard Donchian: Pioneer of Trend-Trading original page - Best accessible contemporaneous interview/profile, especially for diversification, but brief and non-auditing.
16. William Baldwin - Rugs to Riches reproduction - Derivative access to a candid 1982 Forbes profile; original page and independent support for several figures remain unrecovered.
17. Ed Seykota - FAQ aggregation - Retrospective first-person collaborator evidence on computerization, rule conflicts and implementation; no account ledger.
18. Cynthia Kase - Trading With the Odds - Strongest located secondary empirical reading of the 5/20 study's market and year dispersion.
19. Perry Kaufman - Trading Systems and Methods - Serious modern technical context for Donchian-style methods; not evidence of Donchian's original wording.
20. Don M. Chance - Managed Futures and Their Role in Investment Portfolios - CFA Research Foundation industry history and portfolio context; its Donchian material is derivative.
21. Barry Jaffarian, ed. - A Survey of the Managed Futures Industry - Broad managed-futures history and institutional context; biographical treatment is derivative rather than a close reading.
22. Ed Seykota - Richard Donchian resource page - Best accessible chronology separating the Hemphill letters, later Security Pilot service, and postwar Market Outlook; retrospective and not issue-level proof.
23. Commercial and Financial Chronicle - Shearson, Hammill staff notice - Contemporary confirmation that Donchian joined the firm's staff in 1946; does not recover or authenticate any Market Outlook issue.
24. Richard Donchian - 5- and 20-day moving averages edited republication - Readable continuity for material missing from the open 1974 preview; posthumously edited in 1995 and not treated as verbatim original copy.

Task F evidence limitations

• No Donchian-authored investment book, memoir, speech transcript, recording, oral history, signed-letter archive or public issue-level inventory of his newsletters was recovered.
• The original 1957 chapter was not recovered. Its treatment is limited to an independent bibliography and a surviving page-35 excerpt; no unobserved section structure is reconstructed.
• The newsletter appears under three title variants, with no recovered original masthead or finding aid to resolve them. The approximately 4,000-page run and archive custody are reported by Schimmel but were not independently verified.
• Dixon wrote many later newsletters. The Twenty Guides are therefore described as Donchian's compilation as transmitted and annotated by Dixon, not as a recovered 1934 publication or a 1978 Donchian article.
• The accessible original 1974 Commodities pages control over the 1995 edited republication, but the open preview omits part of the continuation. Modern channel formulas, Turtle rules, ATR sizing and collaborator decision trees are not silently back-attributed.
• Kase's 5/20 figures concern a historical study, not audited live performance. Schimmel, Seykota and Baldwin provide retrospective or derivative evidence rather than account records.

Task G - Mental Models (T0630)

Ranked by authority and role in the chapter. This Task G map covers exactly the 22 unique external URLs cited in mental-models.md; repeated appearances of one URL count once here.

1. Richard Donchian - High Finance in Copper original scan - Complete original 1960 article for the two-calendar-week reversal rule, leverage, stop use, fundamental context and retrospective-test caveats.

2. Richard Donchian - Donchian's 5- and 20-Day Moving Averages original page - Controlling original page for the mature 5/20 architecture, penetration filters and basic trend state; the open preview omits part of the continuation.

3. Richard Donchian - 5/20 edited republication - Readable continuity for execution exceptions, diversification, exit/reinstatement, parameter humility and simplicity; posthumously edited in 1995.

4. Barbara Dixon - Donchian's Twenty Guides original pages - Earliest recovered publication of the Donchian compilation, with Dixon's interpretation and reported Donchian comments; not a recovered 1934 publication.

5. Stig Ostgaard - On the Nature and Origins of Trend Following - Best accessible source for Donchian's 1957 page-35 trend-continuation, reversal and asymmetric-loss formulation; later practitioner synthesis.

6. Brentin Elam - Computerized Donchian 5/20 decision tree - Collaborator operationalization of penetration, exit, reinstatement and calendar exceptions after Donchian clarified ambiguities; not Donchian-authored.

7. Richard Donchian - Are Stocks as Cheap as They Seem? - Original evidence for regime-aware earnings normalization and the boundary against a price-only legend.

8. Richard Donchian - Tri-Continental Warrants Plus Plenty of Cash - Original security-specific cash/warrant barbell and monetary-regime adaptation; not a futures sizing formula.

9. Richard Donchian - Short Selling Is More Common Than Generally Realized - Complete original article for short-sale mechanics, stop protection and two-sided participation.

10. Commercial and Financial Chronicle - Futures, Inc. closes to new sales - Contemporaneous direct evidence that capacity, dilution and tax assets constrained the vehicle.

11. Darrell Jobman - Richard Donchian: Pioneer of Trend-Trading original page - Contemporary interview/profile evidence for broad market search, diversification and asymmetric trend participation.

12. Ed Seykota - FAQ aggregation - First-person collaborator evidence on incomplete/conflicting guides, computerization and skipped signals; retrospective and without a trade ledger.

13. Ed Seykota - From Trend Following - Retrospective evidence on short-horizon decay and Donchian's inconsistent adherence; Seykota's adaptation is not back-attributed.

14. FIA Hall of Fame - Barbara S. Dixon - Trade-association confirmation that Dixon wrote many later newsletters and operated the system through 1989.

15. Wolfgang Schimmel - The Donchian Puzzle - Source-critical reconstruction for Elam provenance, missing records, collaborator evidence and attribution limits; a mirrored trade-journal article.

16. Cynthia Kase - Trading With the Odds - Secondary analysis for losing-market and losing-year dispersion in the fixed-contract 5/20 historical study; not live performance.

17. Park and Irwin - The Profitability of Technical Analysis - Academic evidence on early rule studies and data-snooping, parameter, risk and out-of-sample limitations.

18. William Baldwin - Rugs to Riches reproduction - Derivative access to Donchian's valuation-anchoring lesson and the reported Futures, Inc. four-cent low; single-source and not lifetime return.

19. Bhardwaj, Gorton and Rouwenhorst - CTA performance and persistence - Later industry evidence on fees and database biases; context rather than a Donchian account study.

20. CFTC - History of the CFTC in the 1970s - Primary regulatory chronology explaining why the negative legal search cannot certify the largely pre-CFTC record.

21. Curtis Faith - Original Turtle Trading Rules - Retrospective original-Turtle account for the Dennis/Eckhardt 20/10 and 55/20 breakout, volatility-unit, stop, pyramiding and portfolio-cap package; descendant evidence, not Donchian authorship.

22. TradingView - Donchian Channels - Current platform definition of upper, lower and midpoint lines, used only to distinguish modern display convention from recovered Donchian text.

Task G evidence limitations

• No complete Donchian-authored system joins the 1957 principle, 1960 copper rule, 1974 5/20 method and Twenty Guides. They remain separate architectures.
• The original 1957 chapter, complete newsletter run, continuous trade ledger, audited return series and full original 1974 continuation were not recovered.
• The Twenty Guides are a reported 1934 compilation first recovered in Dixon's 1978 publication. Elam's decision tree, Seykota's adaptations and Dixon's later system operation are collaborator evidence.
• Donchian did not specify percentage-risk sizing, volatility units, pyramiding, portfolio heat, correlation limits, leverage ceilings, maximum drawdown, roll conventions or a formal override protocol.
• Kase's figures and the 1960/1974 historical totals are strategy studies, not audited Donchian, Futures, Inc. or client-account results.
• Modern channel formulas, Turtle rules and academic trend results support a lineage or strategy family, not Donchian's exact rules or performance.

Task H - Synthesis (T0631)

Ranked by authority and role in the synthesis. This Task H map covers exactly the 31 unique external URLs cited in synthesis.md; repeated appearances of one URL count once here.

1. Commercial and Financial Chronicle - Futures, Inc. public offer - Contemporary primary notice establishing the October 1948 public offer, before the conventional 1949 industry date.
2. Commercial and Financial Chronicle - Futures, Inc. operating report - Contemporary primary notice for opening operations, initial NAV and the early point-value record.
3. Commercial and Financial Chronicle - Futures, Inc. 1951 results - Contemporary corporate notice for the diversified short-book profit and NAV recovery; one company-reported snapshot, not an audit.
4. William Baldwin - Rugs to Riches reproduction - Derivative access to the reported four-cent low, AUM, personal-profit and early valuation-loss claims; the original page remains unrecovered.
5. Richard Donchian - Are Stocks as Cheap as They Seem? - Original bylined evidence for earnings normalization, commodity regimes and the boundary against a price-only legend.
6. Richard Donchian - Western Union - Original bylined operating-leverage and capital-structure analysis.
7. Richard Donchian - Tri-Continental Warrants Plus Plenty of Cash - Original bylined evidence for monetary-regime adaptation and a cash-plus-warrants barbell.
8. Richard Donchian - High Finance in Copper original scan - Complete original article for the hybrid fundamental-technical process, two-week rule, leverage and testing caveats.
9. Richard Donchian - Donchian's 5- and 20-Day Moving Averages original page - Controlling original page for the mature rule architecture; the open preview omits part of the continuation.
10. Richard Donchian - 5/20 edited continuation - Readable posthumously edited continuity for exceptions, exits, reinstatement, diversification and parameter humility.
11. Barbara Dixon - Donchian's Twenty Guides original pages - Earliest recovered publication of the compilation and primary evidence that application required interpretation.
12. FIA Hall of Fame - Barbara S. Dixon - Trade-association evidence that Dixon wrote many later letters and operated the 5/20 system.
13. Ed Seykota - FAQ aggregation - Retrospective first-person collaborator evidence on computerization, incomplete or conflicting rules, and implementation.
14. Ed Seykota - From Trend Following - Retrospective evidence on changing horizons and Donchian's inconsistent signal adherence; no ledger.
15. Richard D. Donchian Foundation biography - Memorial chronology for death and career; useful but celebratory.
16. New York Times - Richard B. Donchian obituary - Contemporary death notice corroborating deceased status and date.
17. Stig Ostgaard - On the Nature and Origins of Trend Following - Best accessible source for the 1957 page-35 excerpt and trend-following genealogy; not a full facsimile.
18. Darrell Jobman - Richard Donchian: Pioneer of Trend-Trading - Contemporary interview/profile evidence for diversification and the broad-market search.
19. Park and Irwin - The Profitability of Technical Analysis - Independent academic review for data-snooping, parameter, risk, cost and out-of-sample limitations.
20. Commercial and Financial Chronicle - Futures, Inc. closes to new sales - Contemporary evidence that capacity, dilution and tax assets constrained the vehicle.
21. Bhardwaj, Gorton and Rouwenhorst - CTA performance and persistence - Later industry evidence on fee and database biases; context, not a Donchian account study.
22. Moskowitz, Ooi and Pedersen - Time Series Momentum - Modern academic evidence for the broader strategy family; not authentication of Donchian's exact method or record.
23. CFTC - History before the CFTC - Official chronology for the evolving pre-1974 federal commodity-jurisdiction boundary.
24. CFTC - History of the CFTC in the 1970s - Official chronology for expanded jurisdiction and the first CPO/CTA operating rules.
25. NFA - History - Official chronology establishing the 1982 start of NFA regulatory operations.
26. New York Public Library - Commodity-price collection - Current catalog evidence of physical Commodity Year Book holdings covering 1957.
27. Library of Congress - Commodities price indexes guide - Current catalog lead for the Commodity Year Book series and a targeted 1957 chapter request.
28. National Archives - Record Group 180 - Official guide to predecessor-agency correspondence, case, regional and enforcement records requiring offline review.
29. Cynthia Kase - Trading With the Odds - Secondary analysis of dispersion within the fixed-contract 5/20 historical study; not cited as live performance.
30. Curtis Faith - Original Turtle Trading Rules - Retrospective Turtle-system account used only to separate its sizing, pyramiding and portfolio package from Donchian's recovered rules.
31. TradingView - Donchian Channels - Current platform definition used only to distinguish the upper/lower/midpoint display convention from Donchian's recovered original text.

Task H evidence limitations

• No continuous audited Donchian, client or Futures, Inc. total-return series was recovered; NAV snapshots, model studies, personal-profit claims, AUM and collaborator-operated results remain separate evidence classes.
• The original 1957 chapter, complete newsletter run, full original 1974 continuation, original 1982 Forbes page, complete vehicle records and account-level trade ledgers remain unrecovered.
• The method evolved across distinct architectures and collaborators. Dixon, Elam, Seykota, Turtle, modern channel and academic contributions are not silently back-attributed to Donchian.
• No complete sizing, volatility, leverage, correlation, drawdown, contract-roll, cost, collateral or override policy was recovered.
• The negative adverse/legal search is bounded and cannot certify a clean record; offline predecessor-agency, exchange, state, court and arbitration files remain incomplete.
• Donchian died in 1993. The synthesis is a historical reconstruction, not a current recommendation or representation of a living investor's views.