Briese built the COT Index and Movement Index still used across the industry, and shows the actual math behind both. The catch: the CFTC's public data is riddled with lag and contract-size errors he spends a full appendix teaching you to fix by hand.
The one idea
The Commitments of Traders report shows what commercial hedgers, large speculators (funds), and small traders are actually holding each week. Briese's core claim is that commercials are the smart money: they buy on weakness and sell on strength because their trading is tied to a physical business (a producer hedging a crop, a refiner hedging fuel), not to forecasting. When their net position reaches a historical extreme, it tends to mark a turn.
To make that usable he converts raw net positions into two indicators. The COT Index rescales a trader group's current net position to a 0-100 percent range based on where it sits versus its own high and low over a look-back period (three years by default). The COT Movement Index measures how fast that index is changing over six weeks. Used together, an extreme index reading flags where a move might exhaust, and a 40-point six-week swing in the Movement Index flags when it actually has.
The book's second half applies this market by market (currencies, stock indexes, metals, energy, rates, grains, softs), then closes with a full simplified 35-market backtest and an appendix on why almost everyone else's COT backtests are quietly wrong.
What it actually teaches
Briese's system is not sequential so much as four layered components you build up in order:
NET POSITION
Start with a trader group's net position: long contracts minus short contracts, taken straight from the weekly report. Commercials, large speculators (funds), and small traders each get their own net position line. This is the raw input for everything else.
COT INDEX
Rescale net position to a 0-100 percent bullish scale: COT Index = 100 x (Current Net - Minimum Net) / (Maximum Net - Minimum Net), where Minimum and Maximum are the group's most negative and most positive net positions over the look-back window. Briese's default look-back is 3 years (36 months), a figure he says its originator, Curtis Arnold, chose only because that was all the data he had. A worked example from the book: commercial net position -44,892, 36-month min -112,469, max +40,186, giving a COT Index of 44 percent. A reading above 90 percent is treated as a commercial buying climax (a top setup); a reading below 5 percent (the book uses a slightly tighter threshold at tops than the 10 percent used at bottoms, to offset the market's upward bias) is treated as a selling climax (a bottom setup).
MOVEMENT INDEX
Take the six-week rate of change of the COT Index itself. Because the Index runs 0-100, the Movement Index runs -100 to +100. Briese's rule: a +40 point commercial surge within six weeks tends to end a corrective pullback in an uptrend; a -40 point plunge tends to end a reaction in a downtrend. When a market blows through a 40-point signal without reversing, he treats that failure as a warning that the larger trend itself is turning, not just the swing within it.
COMBINE AND CONFIRM
Layer the two: an absolute-extreme COT Index reading (90/5 percent) flags range boundaries and market turns, while the Movement Index catches intermediate swings the absolute level misses, especially mid-trend when the Index sits pinned near one extreme for months. He also flags that COT signals should be paired with a price-based trigger (a breakout, a technical setup) rather than traded alone, since fading an extreme by itself is bottom-picking.
What it looks like Monday morning
A reader who wants to actually run this pulls the free weekly Commitments of Traders report from the CFTC site, builds a rolling net-position series for the markets they trade, and calculates the COT Index and 6-week Movement Index for the commercial category first. The immediate use is as a screen, not a trigger: when a market's commercial COT Index crosses 90 or drops under 5, or the Movement Index swings 40 points in six weeks, that market goes on a watchlist for a technical entry, rather than becoming a trade on its own.
The harder Monday-morning task is the data itself. Before trusting any historical backtest against this data, a reader needs to account for the CFTC's contract-size changes (grains switched from 1,000-bushel to 5,000-bushel contracts in 1998, the S&P 500 contract halved in 1997, and half a dozen others Briese lists by name) and the real publication lag (COT data was only released weekly starting October 2000; anything using older weekly releases as if they were available same-day is testing on data that didn't exist yet).
| Market | Fast MA | Slow MA | Gross Profit | Trades | Profit/Trade |
|---|---|---|---|---|---|
| S&P 500 | 5 | 14 | $321,200 | 137 | $2,345 |
| Crude Oil | 10 | 14 | $66,920 | 92 | $727 |
| British Pound | 6 | 12 | $43,344 | 122 | $355 |
| Gold | 14 | 15 | $15,820 | 113 | $140 |
| Cocoa | 20 | 21 | $1,110 | 93 | $12 |
Totals across all 35 markets: $1,716,245 gross profit, 3,502 trades, $490 average profit per trade; no slippage, commission, or stop loss included, and the author states results before October 2000 (pre-weekly reporting) were only marginally profitable.
In the real world, you cannot peek ahead to tomorrow's newspaper in making today's trades.Stephen Briese, The Commitments of Traders Bible
Where it fails
- The book's own backtest only really works after October 2000. Briese tests a simple two-moving-average crossover on the fund COT Index across 35 markets and reports all 35 profitable (before costs, no stop loss), for a combined $1,716,245 hypothetical profit over 3,502 trades. But he states plainly that applying the same rules to pre-2000 data shows only marginal profits, because the report was released bimonthly or twice-monthly before then and the lag alone (3 to 10-plus days) breaks same-week signal generation. That means the edge he demonstrates is conditioned on seven years of data (2000-2007), not the 20-plus years of history the book otherwise leans on.
- The 90/5/40 thresholds come from one edition and one look-back choice that even the author calls arbitrary. The three-year look-back for the COT Index was picked by Curtis Arnold because, in Briese's own words, 'that's all the data I had to begin with,' not because it was optimized. Briese shows a 13-week look-back produces a completely different, choppier signal pattern on the same market. A reader locking into 90/5/40 as fixed rules is trading someone else's arbitrary window.
- The data cleanup burden is a book-length project by itself, and it is not optional. Appendix B runs eight pages of contract-size and reporting-lag fixes a reader has to apply by hand: grain contracts need a 0.20 multiplier before 1998, S&P 500 futures need a 2x multiplier before November 1997, the NYBOT Dollar Index needs a 0.5x adjustment from mid-1992, lumber has four separate contract-size eras with different multipliers, and pre-1992 mid-month CFTC data is flagged by the CFTC itself as containing uncorrectable errors. Skip any of this and a historical COT chart is silently wrong, sometimes by a factor of five.
- Backtests only show winners, and the author admits it. Briese explicitly says he limited his 35-market test to seven years rather than 20-plus because a longer track record 'easily masks' multi-year drawdowns in whichever markets were picked, and states he did not test for drawdown size at all. The trading rules chapter is presented as a proof of concept, not a system: no stop loss, no slippage or commission, and the author calls the model 'not intended for actual trading.'
- The free data source the book points readers to may no longer be maintained. The appendix directs buyers to CommitmentsOfTraders.ORG for the author's personally cleaned historical database, described as available to book purchasers for free updates. That offer is tied to a single proprietor's website nearly two decades after the 2008 printing; a reader should verify it still exists and still updates before relying on it rather than building the CFTC adjustments themselves.
- Commodity index traders complicate the 'commercial equals smart money' premise the whole book rests on. Briese flags that the CFTC folds long-only commodity index funds into the commercial category, and that in a May 2007 sample these index traders held 34 percent of long open interest across 12 markets, comparable to traditional hedgers. Since index funds are structurally long and not counter-trend value traders, the commercial net position in index-heavy markets mixes two groups with opposite motives, which the COT Index cannot separate.
Who it is for
Buy it if
You already trade futures with a discretionary or technical system and want a genuinely independent, non-price data set to screen setups and confirm trend exhaustion, and you are willing to do the contract-adjustment homework in Appendix B before trusting any historical chart. It also rewards a reader who wants the actual formula behind the popular COT Index and Movement Index rather than a vendor's black-box version.
Skip it if
You want a plug-and-play trading system: the book's own test explicitly disclaims that use, has no stop loss, and only holds up cleanly in the post-2000 weekly-release era. It also is not for a reader trading equities or crypto with no futures exposure, since every worked example and the entire data-cleaning appendix is built around CFTC futures contracts.
As an Amazon Associate I earn from qualifying purchases. It costs you nothing extra, and it does not change what gets recommended or what the verdict says.