Trading With the Insiders: Larry Williams’ COT Report Method

Every market has a group of participants who know more than you do. In commodities, that group is not a secret cabal — it is a published list. Every Friday afternoon, the U.S. Commodity Futures Trading Commission releases a report that breaks down exactly who is holding which futures contracts: the producers and merchants who actually use the commodity, the big speculative funds, and everyone else. The catch is that almost nobody reads it correctly, and for decades almost nobody read it at all.

Larry Williams changed that. He was the first trader to treat the Commitments of Traders (COT) report as a serious, systematic trading tool, and he wrote the first book devoted to it — Trade Stocks and Commodities with the Insiders: Secrets of the COT Report. His central idea is disarmingly simple: the data tells you what the smartest money in any market is doing, and most of the time you should be doing the same thing. This guide walks through what the report is, how Williams reads it, the COT Index he built to make it tradable, where it breaks down, and how a crypto trader gets an analogous “who is positioned where” read even though Bitcoin has no CFTC hedger column.

What the COT Report Actually Is

The COT report is a weekly snapshot of open interest in U.S. futures markets, broken down by category of trader. The CFTC requires any entity holding a position above a reporting threshold to identify itself, and it aggregates those positions into groups. The report is published every Friday at 3:30 PM Eastern, but — and this matters enormously — the data reflects positions as of the close of business the previous Tuesday. By the time you read it, the snapshot is already three days old.

There are several versions of the report. The classic Legacy report splits the market into three buckets, and this is the version Williams built his work around. A more granular Disaggregated report (introduced after the 2008 commodity boom) and a Traders in Financial Futures report exist for finer analysis, but the three-category Legacy framework is the right place to start because the logic is clearest. Those three buckets are Commercials, Large Speculators (reportable non-commercials), and Small Speculators (the non-reportable remainder).

The Three Categories — And Why It Matters Who Is Who

The entire method rests on understanding the motive behind each group’s positions. They are not all trying to do the same thing, and that is precisely why their behavior carries information.

Commercials are the hedgers: the farmer, the grain elevator, the oil refiner, the jewelry manufacturer, the bank with currency exposure. These are the people Williams calls the insiders. They are in the physical business, they know the real supply-and-demand picture, and they use futures to lock in prices rather than to gamble on direction. Crucially, their trading is often contrarian by nature: a producer sells more futures as prices rise (locking in a good price) and buys them back — or stops selling and goes net long — as prices collapse to levels they consider cheap. That mechanical behavior means commercials tend to accumulate near bottoms and distribute near tops, which is exactly what a great trader does.

Large Speculators are the managed-money funds, CTAs, and trend-following systems. They are smart, well-capitalized, and largely momentum-driven. They pile in with a trend and reach their largest long position right when a market is most extended — near the top — and their largest short right near the bottom. They are not dumb money, but their structural role as trend followers means they are usually heaviest on the wrong side at the turn.

Small Speculators are the non-reportable traders — effectively retail and small accounts. As a group their timing is poor: they buy enthusiasm and sell fear. Williams treats them as a sentiment fade, similar to how he and others use other crowd indicators alongside tools like the Williams %R oscillator for short-term timing.

Category	Who they are	Motive	What an extreme signals
Commercials	Producers, processors, merchants — the physical-market “insiders”	Hedging real exposure; buy when cheap, sell when dear	Record net-long → bullish; record net-short → bearish. Follow them.
Large Speculators	Funds, CTAs, trend-following systems	Ride momentum; heaviest at the extreme of a move	Crowded long → trend exhaustion ahead; crowded short → squeeze risk. Mild fade.
Small Speculators	Retail and small non-reportable accounts	Emotion-driven; chase price, capitulate at lows	Heavily one-sided → that side is usually wrong. Fade them.

The Core Thesis: Follow the Commercials, Fade the Crowd

Williams’ thesis can be compressed to a single sentence: do what the commercials do, and bet against what the small speculators do. The two large groups (commercials and large specs) are almost always positioned against each other — one’s long is the other’s short — so when the commercials reach an unusually large net-long position while the speculators are unusually net-short, you have a market where the most informed participants are buying from the least informed. Historically, that configuration has clustered around major price lows. The mirror image — commercials heavily net-short into a parabolic rally while small specs pile in long — has clustered around major tops.

This is the same instinct that runs through almost every enduring trading philosophy: be a buyer when informed money is accumulating and a seller when the crowd is euphoric. The COT report just happens to put real, audited position data behind the idea instead of leaving it to gut feel. It is, in spirit, a smart-money tracking tool — only the “smart money” here is named and counted by a government agency.

The Problem With Raw Numbers — And Williams’ Fix

Here is where most people who glance at the COT report go wrong. The raw net position — say, commercials are net-short 120,000 contracts of gold — is almost meaningless on its own. Is 120,000 a lot? It depends entirely on the size of the market, which grows and shrinks over the years, and on what “normal” looks like for that particular commodity. A net-short of 120,000 might be an extreme in one decade and routine in another.

Williams’ key contribution was to normalize the data into something comparable across markets and across time. He built the COT Index: take a group’s net position, look back over a fixed window (he favored about three years, roughly 156 weeks), find the lowest and highest net position in that window, and express today’s reading as a percentage of that range.

The formula is a classic stochastic-style scaling:

COT Index = (Current Net − Lowest Net in lookback) ÷ (Highest Net − Lowest Net in lookback) × 100

The result is a number from 0 to 100. A reading near 100 means the group is at the most net-long it has been in three years; a reading near 0 means the most net-short. Now “120,000 contracts” becomes “the commercials are at a 94 — nearly their most bullish posture in three years.” That is something you can act on, and you can compare it directly to gold, crude, soybeans, or the euro.

Reading the Extremes

Williams looks for the index to reach extremes, not for it to wiggle in the middle. The practical rules of thumb most COT practitioners use:

• Commercial COT Index above ~80–90: the insiders are heavily net-long relative to their own history — a bullish setup. Start hunting for longs.
• Commercial COT Index below ~10–20: the insiders are heavily net-short — a bearish setup. Look for shorts or get defensive.
• The signal is strongest when the three groups disagree sharply — commercials at one extreme while small specs sit at the opposite extreme.
• A mid-range reading (40–60) carries little information. The method only speaks when positioning is stretched.

A Worked Example: Reading a COT Extreme Into a Bias

Imagine you are following crude oil futures. Price has been grinding lower for three months and sentiment is grim; the financial press is talking about a glut. You pull the latest COT report and run the index over a three-year lookback. Suppose you find:

• Commercials net-long; three-year COT Index reads 91 — their most bullish posture since the last major low.
• Large speculators have flipped heavily net-short; their index reads 12.
• Small speculators are net-short and capitulating; their index reads 8.

This is the textbook bottoming configuration. The insiders who handle physical barrels are aggressively accumulating, while both the trend-following funds and the retail crowd are leaning short into weakness. Your bias swings from neutral-to-bearish toward bullish. Notice the language: it is a bias, not an entry. The COT does not tell you that the low is today. It tells you that the risk/reward has tilted in favor of the longs and that you should now be looking for a timing trigger — a higher low, a bullish reversal bar, an oscillator turning up — rather than fighting the setup with shorts.

If, two weeks later, price prints a clean swing low and turns up, you act on the long side with conviction because the positioning backdrop agrees. If price keeps falling, the COT bias tells you to be patient and treat dips as opportunities rather than as confirmation to short. That is how a positioning tool is supposed to be used — as a filter on which direction you are allowed to trade, paired with a separate timing method.

The Limitations You Must Respect

The COT method is powerful, but it is also widely misunderstood and misapplied. Williams himself is careful about its boundaries, and so should you be.

It is positioning, not timing. This is the single biggest mistake. An extreme commercial reading can persist for weeks — sometimes a month or two — while price drifts further in the “wrong” direction before turning. Commercials are not trying to pick the exact low; they are accumulating value over a zone. If you treat a COT extreme as a market-order trigger, you will get run over. Pair it with a timing tool and proper risk control.

The report lags by three days. Tuesday’s data, Friday’s release. In a fast market, a lot can happen in three days. The COT is a slow, structural lens, not a tactical one. It is useless for intraday decisions.

It works best on real futures — especially commodities. The commercials-as-insiders logic is strongest where there is a genuine physical business hedging genuine exposure: grains, energy, metals, softs, and to a degree currencies and rates. It is weaker or meaningless in markets where there is no true commercial hedger with superior information.

Extremes are rare and the sample is thin. Genuine three-year extremes might occur only a handful of times in a market. That is a feature — the signal is selective — but it also means you cannot backtest it like a high-frequency strategy. Sizing matters: when the setup is this rare and this strong, it deserves real capital, which is exactly the kind of decision an expectancy and Kelly calculator can help you frame so you are not over- or under-betting your edge.

Everyone can see it now. When Williams started, the COT was an obscure data dump. Today it is dissected by countless traders and services, which can blunt the edge in the most-watched contracts. The logic still holds at true extremes, but do not expect the easy, lonely signals of the 1970s.

Crypto Has No COT — But It Has Better Positioning Data

Bitcoin and Ethereum perpetual futures don’t file a CFTC Commitments of Traders report. There is no government-audited “commercials” column for crypto. But crypto traders are, in a sense, luckier: the exchanges publish positioning and flow data in real time, with no three-day lag and no Tuesday cutoff. The trick is knowing which crypto metrics map onto Williams’ categories.

• Perpetual funding rate is the closest thing crypto has to a crowd-sentiment gauge. When funding is deeply positive, longs are paying shorts to keep their positions open — the market is crowded long, much like small specs piling into a top. Deeply negative funding signals crowded shorts and squeeze potential. An extreme in funding is a fade signal, the same way an extreme small-spec COT Index is.
• Open interest tells you how much leverage has been built into the move. Rising price on rising OI is fresh positioning; a sharp OI drop on a price drop is a flush — the crypto analogue of speculators capitulating.
• Long/short account and position ratios published by exchanges separate the retail-heavy crowd from the larger accounts, giving you something close to the small-spec-versus-large-player split.
• Exchange flows — coins moving onto or off exchanges — hint at accumulation versus distribution by larger holders, the nearest stand-in for “what are the insiders doing.”

Put those together and you can build a COT-style mental model for Bitcoin: are the leveraged speculators crowded and euphoric (fade), or are they flushed and fearful while larger holders quietly accumulate (lean long)? That is precisely the read the AIO Perps Flow tooling is designed to surface — funding, open interest, and long/short positioning condensed into a single positioning picture, updated live rather than three days late.

How to Actually Use It in a Trading Plan

Whether you trade soybean futures or a Bitcoin perp, the workflow Williams pioneered translates cleanly:

• Establish the bias from positioning first. Run the COT Index (or, in crypto, read funding/OI/long-short extremes). If positioning is mid-range, the tool is silent — trade on other methods.
• Wait for an extreme and a disagreement. The best setups have informed money and crowd money pointed in opposite directions.
• Add a timing trigger. Never buy a COT low blindly. Combine it with price-structure or oscillator confirmation, as you would when stacking it with a volatility breakout and money-management framework.
• Size for a rare, high-conviction event — and define your risk. These signals are infrequent. When one fires with confirmation, it warrants more capital than a routine setup, but only with a hard stop, because even insiders can be early.

Larry Williams’ great insight was not that markets have insiders — everyone suspects that. It was that in regulated futures, the insiders are published, and a little arithmetic turns that disclosure into a tradable edge. The instruments have changed; crypto replaces the CFTC report with live funding and flow data. The principle is identical: find out what the informed money is doing, find out what the crowd is doing, and when those two diverge to an extreme, side with the informed money.