What the numbers mean

Small traders below the CFTC reporting threshold ('Non-Reportable'), long minus short.

Positive

Small traders hold more longs than shorts: on balance they're betting the currency rises vs USD.

Negative

Small traders hold more shorts than longs: on balance they're betting it falls vs USD.

Size

Contract counts differ a lot between currencies. Compare sizes with the % of OI version.

How to use it: Context only. Small specs are often late to a trend, so heavy small-spec positioning with specs can mean a crowded trade.

Non-Reportable long − short

Speculator net position divided by all contracts outstanding. It makes currencies of very different sizes comparable.

Above +20%

A heavy long position relative to the market.

−10% to +10%

A fairly balanced market.

Below −20%

A heavy short position relative to the market.

How to use it: Use it to compare how committed specs are across currencies, e.g. in the screener.

nc_net ÷ open interest × 100

Commercial net position divided by open interest.

How to use it: Compare hedger commitment across currencies.

c_net ÷ open interest × 100

Small-trader net position divided by open interest.

nr_net ÷ open interest × 100

All futures contracts currently open in this currency, from every group. A measure of how much money is in the market.

How to use it: Context for size. It drops sharply around quarterly expiry (March, June, September, December).

Open interest (all contracts outstanding)

Of all speculator positions (long + short), the percentage that are long.

Above 50%

More speculator longs than shorts.

50%

Evenly split.

Below 50%

More speculator shorts than longs.

How to use it: An easy-to-read version of the net position: 70% means 7 in 10 spec contracts are bets on the currency rising.

nc_long ÷ (nc_long + nc_short) × 100

Days between this report and the one before it for this contract.

7

Normal weekly report.

More than 7

A week is missing for this contract. Weekly changes span more than one week. Values are never filled in.

Days since previous row. 7 is normal; more than 7 means a missing week.

How much the small-trader net position moved since last week.

Positive (+)

Small traders got more bullish this week (added longs and/or cut shorts).

Negative (−)

Small traders got more bearish this week (added shorts and/or cut longs).

Big vs usual

A change much larger than recent weeks is a strong signal; a tiny one is noise.

nr_net this week − nr_net previous available week

Speculator long contracts this week minus last week. Positive means new buyers arrived; negative means longs were closed.

How to use it: Together with the change in shorts, this decides the change type (fresh longs, long liquidation, etc.).

nc_long this week − previous available week

Speculator short contracts this week minus last week. Positive means new sellers arrived; negative means shorts were closed (covered).

How to use it: Together with the change in longs, this decides the change type.

nc_short this week − previous available week

The change in the total number of contracts outstanding.

Rising with the move

New money is entering. Confirms the week's conviction.

Falling with the move

Positions are being closed. The move is exits, not new bets.

Big drop in Mar/Jun/Sep/Dec

Usually the quarterly contract roll. Don't read it as a signal.

open interest this week − previous available week

The same 0–100 index measured over the last 156 weeks. It shows whether today's position is unusual in a longer history.

90 – 100

Extreme high. The most bullish they've been in this window. Crowded: the trend may be late.

80 – 90

Stretched high. Strongly bullish, but little room left before extreme.

50 – 80

Upper half. Bullish trend zone, the best place to follow the move, especially if rising.

20 – 50

Lower half. Bearish trend zone, especially if falling.

10 – 20

Stretched low. Strongly bearish, little room left before extreme.

0 – 10

Extreme low. The most bearish in this window. Crowded short: watch for a turn.

vs the 6-month index

6M high but 3Y middling = strong recent move with room to run. Both extreme = a truly historic crowd.

How to use it: Context, not part of the score. Use it to judge whether a 6-month extreme is also a long-term extreme.

(nc_net − min₁₅₆) ÷ (max₁₅₆ − min₁₅₆) × 100 · 3 years

Where the hedgers' net position sits in its own 26-week range, from 0 (most short) to 100 (most long).

90 – 100

Hedgers the most long in 6 months, usually while specs dump the currency. Bullish at a spec extreme short.

10 – 90

No signal on its own.

0 – 10

Hedgers the most short in 6 months. Bearish at a spec extreme long.

How to use it: Only matters at spec extremes: it's the second half of the reversal-watch test.

COT Index (26w) of c_net

Where the small traders' net position sits in its 26-week range (0 = most short, 100 = most long).

How to use it: Context. Small traders piling in at an extreme adds to the 'crowded' warning.

COT Index (26w) of nr_net

Larry Williams' commercial index. Like the commercial COT index, but first divides the commercial net by open interest, so it isn't fooled by the market growing or shrinking.

Above 80

Commercials unusually long for their size: a bullish contrarian hint.

20 – 80

No signal.

Below 20

Commercials unusually short: a bearish contrarian hint.

How to use it: A cross-check on the commercial index at turning points.

COT Index (26w) of c_net ÷ oi × 100 (Williams commercial index)

Whether last week's move was new money (conviction) rather than old positions closing.

+1

Fresh longs last week, and the net moved at least 5% of its 6-month range.

0

Short covering, long liquidation, or fresh positions too small to count.

−1

Fresh shorts last week, and the net moved at least 5% of its 6-month range.

+1 fresh_longs, −1 fresh_shorts, else 0. 0 when |d_nc_net| < 5% of the 26-week nc_net range (CONVICTION_MIN_MOVE).

The headline number: flow + index + conviction, from −3 (strongly bearish) to +3 (strongly bullish) for the currency against USD.

+2 to +3

Strong bullish positioning. A candidate for the long side of a pair.

+1 / −1

A mild lean. Useful only against a strong opposite currency.

0

No edge, or forced to 0 because the regime is extreme.

−2 to −3

Strong bearish positioning. A candidate for the short side of a pair.

How to use it: Pair a high-score currency with a low-score one. Trade only when the gap is 3 or more.

flow_score + index_score + conviction_score (−3 to +3). Forced to 0 when the regime is extreme.

How many reports in a row the pair has shown the same bias. Display-only: it never changes the gap, bias or score.

New

The bias appeared this week. One week is not yet a signal: wait for a second report.

Flipped

Long last week, short now (or the reverse). The two-week rule says do not flip on one week of data.

Held 2w+

The bias has survived at least one more report. This is the confirmation the two-week rule asks for.

How to use it: Prefer pairs whose bias has held for two reports or more. Treat new and flipped biases as a watch list.

Consecutive reports, back from the latest, where bias(gap) is unchanged. A missing gap ends the run.

Whether the 4-week speculator flow and the 4-week price move point the same way. Display-only: it never changes the score or bias.

Confirmed

Specs are buying and price is rising (or selling and falling). The positioning signal has price behind it.

Diverging

Specs are buying into a falling price, or selling into a rising one. Price disagrees: wait for it to turn before acting on the bias.

Price flat

Price moved less than 1% while specs moved. Not yet confirmed either way.

How to use it: Price has the final say. A diverging currency is a weaker leg, even with a strong score.

sign(flow_score) vs close_tue ÷ close_tue 4 reports ago − 1 (ECB rate; under ±1% counts as flat)