After the midterms, our clients started asking us the same question: what is actually going on in futures positioning?

We are unusually well placed to answer it. The Commitments of Traders report tells you nothing after Tuesday — and this week, Tuesday was before the election result. What we publish is daily and granular, by investor group, measured at every session's close.

So here is what the three sessions after the midterms actually looked like, and why the public picture could not have told you.

Why the Commitments of Traders Report Lags the Market

The mechanical fact is simple. The CFTC publishes the Commitments of Traders report once a week, on Friday at 3:30pm Eastern, using data from the close of the preceding Tuesday.

The Commission says so plainly in its own release schedule: "generally published each Friday at 3:30 pm Eastern Time, using the data from the immediately preceding Tuesday of that week."

So on any given day, the public picture of futures positioning describes a market that no longer exists.

All week — Wednesday, Thursday, and Friday morning — the newest report you can get describes the Tuesday of the previous week: eight days old on Wednesday, nine on Thursday, ten by Friday morning.

Friday afternoon you finally get this Tuesday's close, and by then three sessions of trading have already happened without anyone outside the market knowing where the books sit.

This Week That Matters More Than Usual

The midterms landed Tuesday night. Institutions spent Wednesday repositioning, Thursday de-risking and Friday rebuilding. Three sessions of real, consequential position changes in equity index futures, Treasuries, gold and crude.

The CFTC will tell you what that looked like as of Tuesday's close, before any of it happened.

We can tell you what it looked like Wednesday, Thursday and Friday.

What our Data Showed, Session by Session

Futures positioning by trader category, measured at each close, published the following morning.

Equity index futures

Managed money cut Russell 2000 net longs 49.7% over the three sessions, from 20,701 contracts to 10,417. Nasdaq-100 longs fell 47.5%. Large caps were not spared, but the cut was three to four times smaller: S&P 500 longs fell 13.1%, Dow longs 15.8%.

Other reportables did the opposite on the same contracts. They added 86.3% to S&P 500 net longs and 140.0% to Nasdaq-100 longs — buying where managed money was selling.

Treasuries

In rates, managed money covered shorts: the 30-year bond net short narrowed 20.5%, the 5-year 8.5%. Other reportables pressed theirs: the 5-year net short deepened 88.0%.

The largest single rate move in the set is easy to miss because it is spread across every group.

On the 2-year note, net longs rose 21,537 contracts over the three sessions — managed money +9,803, other reportables +8,030, nonreportable +3,704. Every category added.

That is the biggest three-day build in any Treasury contract we track that week, and it appears nowhere in the headline discussion of who sold what.

Gold

In three sessions producers and merchants more than tripled their gold net short, from -4,208 contracts to -14,394.

None of that is in the CFTC report yet. Some of it will not be in it until Friday, and by then it will describe Tuesday.

We can check that claim against the report itself

Saying we publish before the CFTC is easy. Being measured against the CFTC is the harder test, and it is the one that matters.

Directional accuracy by group

For every Tuesday observation where both our estimate and the actual Commitments of Traders figure exist, we compare direction: did the estimated weekly change in net positioning move the same way as the reported change? Averaged equally across products within each investor group, directional accuracy is:

• Managed money 67.97%

• Producer / merchant 71.71%

• Swap dealers 61.89%

• Other reportables 59.75%

• Nonreportable 62.02%

• Average of the five groups 64.67%

Measured on 29 products and 42,197 matched Tuesday observations.

Read that the way we read it internally: roughly two times in three, we call the direction of a group's weekly positioning change before the CFTC publishes it.

Where accuracy differs

The accuracy is not uniform, and pretending otherwise would be the fastest way to lose a sophisticated reader. Managed money is the easiest group to classify, large, visible and systematic, and we get direction right about 68% of the time.

Swap dealers are the hardest, at roughly 62%, because dealer books are hedged continuously and their reported positioning is the residue of two opposing flows.

How we track it

We keep improving these numbers. Directional accuracy is the metric our research team tracks against the published report, and it has risen as the model has been refined.

Every release is measured on matching Tuesday observations, on the same weeks, against the same ground truth, so the number is comparable across versions and we can see whether a change helped or hurt.

One limit worth stating: directional accuracy tells you whether we got the sign right, not whether we got the size right. Position sizing and trigger levels are a separate question, and that is what the flow forecasts dataset exists to answer.

Why Daily Matters, Not Just Faster Weekly

The obvious read is that we are the same information, sooner. That undersells it.

A weekly snapshot cannot show a sequence. If managed money cuts small-cap longs 50% over three sessions, the weekly report shows you a smaller number and no explanation. Was it one decision or three?

Did they sell into the Thursday panic or buy the Friday recovery? Did the other side of the trade arrive on the same day or the day after?

Those are different situations with different implications for what happens next. A position cut over three days because risk was being reduced looks nothing like a position cut in one session and partly rebuilt the next — and only one of those tells you whether the selling is finished.

The sequence is the information. A weekly print cannot carry it.

What This Means for Anyone Trading Around the Week

If you are positioning for the post-election regime, you are working from a picture of Tuesday's book while the market has already moved twice.

That is why we built this daily rather than weekly. The Indigo Panther futures flow product measures positioning by trader category at each session's close, and publishes it the next morning — before the session ahead, not three sessions after the one behind.

The gap between the two is not a technicality. It is the difference between knowing what happened and guessing at what is happening.

One Honest Caveat

Contract counts are not comparable across markets, and the precise contract totals are model-dependent. Direction and size of change are what we trust. Anyone quoting a single absolute number without the change beside it is over-reading the data — ours included.

Anyone quoting a single absolute number without the change beside it is over-reading the data — ours included.

What the Product Does Beyond Positioning

One clarification worth making, because it is easy to misread what we publish.

XTech Global Futures Flow Forecasts does not stop at describing where each group sits.

It identifies the price levels at which each group's positioning is likely to change, and the number of contracts they are expected to buy or sell at each of those levels — delivered at the close of each trade date, for the next session.

So the question is not only "where were they positioned on Thursday." It is "at what price does this group start selling, and how much." Those trigger levels are the mechanical flow that shows up on the tape, published before it happens rather than three sessions after.

Coverage, history and timing

Coverage is the top ~80 global futures products across all expiries, daily end-of-day, delivered 7:30am UTC on T+1.

The product description states deep history back to 1 January 2011 — 15+ years — and notes that the forecast models are trained on and validated against the same flow record that predicts the weekly COT report three trading sessions early.

The print describes a Tuesday close and does not land until Friday, so we publish the morning after the close it covers.

That is the difference between a snapshot and a flow product.

How we Measure It

Method: XTech Global Futures Flow (Indigo Panther) and Flow Forecasts, Unifier. Positions are inferred by investor category and measured at each session's close, then published the following morning. Net equals long minus short contracts.

The category names we use — managed money, swap dealers, producer/merchant, other reportables, and the nonreportable residual — are our decomposition, not the Commission's.

The CFTC's own financial-futures report uses a different taxonomy (dealer/intermediary, asset manager/institutional, leveraged funds, other reportables) that does not map one-to-one onto ours, so the figures here should not be read as CFTC category totals.

CFTC release timing quoted from the Commission's own Commitments of Traders FAQ.

Exponential Technology Inc. Real time supply and demand for the entire market.

Data Source: XTech Global Futures Flow Forecasts

The analysis draws on XTech Global Futures Flow Forecasts, which models both trader groups. The full methodology, figures, and results are in the white paper, "Measuring Market Impact and Inelasticity from Managed-Money and Producer Flow Forecast Data."