How to Use COT Reports for Forex Trading (Without Getting Lost in Spreadsheets)

COT reports are one of the most underused fundamentals tools in retail forex. Here's how to actually read them, what to look for, and how to use them in real trades.

The Commitment of Traders report is one of the most underrated fundamentals tools in forex, and one of the most badly explained. Most articles about COT either turn it into a spreadsheet exercise that nobody finishes reading, or they oversell it as a magic indicator that prints money. Both miss the point.

I use COT data on most of my trade ideas, and we built the COT page in TradeSave+ around how I actually read it. So this is opinionated. Take it as a working trader's read rather than a textbook explainer.

What COT actually is

The CFTC publishes a report every Friday afternoon (US time) showing the aggregated positions of large traders in regulated futures markets. For forex, the relevant categories are non-commercial (speculators, mostly hedge funds and commodity trading advisors) and commercial (corporate hedgers).

You care about the non-commercial side. They're the directional money. Their long positions, short positions, and the net position (longs minus shorts) tell you how the smart speculative money is positioned in each currency.

The report covers data through the previous Tuesday, which means there's a 3-day lag. That sounds bad. In practice it doesn't matter, because COT is a positioning indicator, not a timing indicator. You're not using it to time entries. You're using it to bias your direction over the next few weeks.

Why most retail traders use it wrong

The two most common mistakes I see:

1. Looking at the absolute net position number. "AUD is +50,000 net long, that's bullish." The absolute number tells you almost nothing. AUD's typical range might be -80,000 to +120,000 depending on the year. EUR's range is wider. JPY's is wider still. So +50,000 means different things for different currencies. You need the relative position within that currency's own historical range.

2. Trading COT as a contrarian signal. "Speculators are very long AUD, that means a top is in." Sometimes. Often it doesn't. When speculative longs are extreme AND the trend is mature AND price is hitting a major level, the contrarian read works. When you're early in a trend, extreme positioning just means the trend has institutional support and is likely to continue. So the contrarian framing is a 2-out-of-3 trap.

What to actually look at

Three things, in order of importance.

1. Net position in z-score terms

Take the current net position. Compare it to the rolling history (typically 1, 2, or 5 years). Convert to a z-score so you know how extreme the positioning is in standard deviations from the mean.

A z-score above +1.5 or below -1.5 is statistically meaningful. Above +2 or below -2 is genuinely extreme. The TradeSave+ COT page shows z-scores and percentile ranks alongside the raw numbers because the raw numbers are usually misleading on their own.

2. Net position change over time

The direction matters more than the level. AUD might be at +30,000 net long, which is mid-range historically. Boring. But if six weeks ago it was at -50,000 short and the net has been climbing every week, that's an active trend in positioning. Hedge funds are accumulating. That's bullish even though the absolute number isn't extreme.

The opposite case: AUD at +90,000 long, near the historical high, but the net has flatlined for four weeks. Positioning is extreme but the momentum has stalled. That's where contrarian reads start to work.

You want to see the rate of change of positioning, not just the snapshot.

3. Divergence between two currencies in a pair

This is the one that actually translates to trades.

If AUD net positioning is climbing and JPY net positioning is dropping, AUDJPY has a positioning tailwind. If GBP net positioning is dropping fast and CAD net positioning is rising fast, GBPCAD has a positioning headwind for longs.

The bigger the divergence, the cleaner the read. Most pairs will have neutral or mild positioning bias. Once or twice a quarter you get a setup where both legs of a pair are at extreme opposite ends, and those are the cleanest fundamental conditions for a directional trade.

Layer this on top of technical structure. COT divergence on its own isn't a trade. COT divergence at a clean technical retest zone is a high-conviction setup.

What COT doesn't tell you

Worth saying because too many retail courses oversell COT.

It doesn't time entries. The report is weekly, lagged 3 days, and positioning can stay extreme for months before reversing. You'll lose money trying to scalp off it.

It doesn't override technicals. If COT says EURUSD should be bearish but price is breaking out above a major resistance with volume, the technicals are usually telling you something COT can't see.

It doesn't override news. If FOMC turns hawkish and USD rallies hard, the COT positioning from the previous Tuesday is stale. New information beats old positioning.

The right framing: COT is a directional bias tool. It tells you which way to lean. It doesn't tell you when to pull the trigger.

How I use it in practice

For each pair I'm watching, I check three things:

Z-score of net positioning on each leg. Are either of them extreme?

Rate of change on each leg over the last 4-6 weeks. Which way is the momentum in positioning?

Divergence between the two legs. Does the pair have a clean fundamental tailwind in one direction?

If a pair has clean COT divergence (z-score gap of around 2 or more between the two legs) AND the rate of change supports the bias (one accumulating longs, the other accumulating shorts) AND there's a clean technical setup, that's an A+ trade for me.

If COT is mid-range and ambiguous, I'll still take technical setups, just with smaller size and a faster exit.

If COT is screaming the opposite direction of the technicals, I usually skip the trade entirely. The cost of trading against extreme positioning is usually worse than the opportunity cost of missing it.

Why most retail tools don't surface this well

The CFTC publishes the data free at cftc.gov. Raw downloads are CSV files that take work to make sense of. Most third-party COT pages show you the absolute net number, maybe a chart of how it's changed, and stop there. That's enough to glance at once a week. It's not enough to actually use COT as a layer in a multi-factor read.

The TradeSave+ COT page is built around the three-things-that-matter list above: z-scores and percentile ranks for each currency, rate-of-change over rolling windows, and pair-level divergence reads so you can see at a glance which crosses have a fundamental tailwind. Every trade you log gets auto-tagged with the COT context active at entry, so you can later filter your trade history by positioning regime, and cross-reference the live risk-on / risk-off score and see whether your edge holds in extreme positioning vs neutral.

7-day free trial if you want the COT layer connected to your actual trades.

The mental shift

The thing that changed my COT reading wasn't a tool. It was treating COT as positioning context rather than as a signal. A COT reading on its own isn't actionable. A COT reading combined with where price is on the chart, what the regime is, and where the rate differential is, that's actionable.

Most retail traders bounce off COT because they tried it as an indicator and it didn't work. It doesn't work as an indicator. It works as a layer in a multi-factor read.

Once you're using it that way, the reports start to feel less like a homework assignment and more like a piece of edge that the algos can't easily replicate.