COT Report for Silver Traders: Concentration, Swap Dealers, and the Manipulation Question

Key takeaways

• No market in the COT dataset attracts more claims about a concentrated bank short than silver. The CFTC publishes the exact number those claims rest on, and almost nobody checks it.
• Across 609 weekly reports from January 2015 to September 2026, silver's four largest traders held a mean net short of 26.19% of open interest, against 20.31% in gold. Silver's concentration exceeded gold's in 537 of those 609 weeks, so the comparison people make is real.
• It is not exceptional. Platinum averages 33.41% and palladium 31.68%, both above silver, and silver's mean rank among the five metals is 2.98 out of 5. There is no platinum conspiracy literature.
• On the report dated 1 September 2026, silver sits at 26.5%, its 49.8th percentile. Gold sits at 27.0%, its 97.9th percentile. This week gold is the concentrated one.
• The largest silver short is Swap Dealers, not producers: 49,268 contracts, 41.12% of open interest. That reading is a 99.7th-percentile extreme, but most of the one-week jump was open interest falling 13%, not dealers selling.
• Managed Money is not crowded. Net long 12,170 contracts, a z-score of -0.39 and a COT Index of 36.5, which is below silver's own average.

Introduction

Silver occupies a strange position in the Commitments of Traders data. It is a mid-sized contract, 119,827 contracts of open interest against gold's 578,926, and yet it generates more argument per contract than anything else on the board. The argument is almost always the same one: a small number of banks hold an enormous short, and that short is why the price is not higher.

The useful thing about that claim is that it is testable. The CFTC has published a concentration ratio for every market, every week, for decades. It states precisely what share of open interest the four and eight largest traders hold, on a net basis, on each side. If a handful of institutions hold an unusual short in silver, that series is where it appears.

This article does two things. It explains how to read silver's COT report properly, in the same way our gold guide does for the larger metal. And it puts the concentration claim against the full published record, because the answer turns out to be more interesting than either side of the usual argument.

Every figure below comes from the CFTC disaggregated futures-and-options-combined report dated 1 September 2026, released Friday 4 September, and from the full weekly series back to January 2015.

Which CFTC Report to Use for Silver

Silver is a physical commodity, so it appears in the Disaggregated Commitments of Traders report. Not the Legacy report, which collapses everything into "commercial" and "non-commercial" and throws away the distinction this article depends on. Not the Traders in Financial Futures report, which covers currencies, rates and indices.

Use the futures-and-options-combined version. COMEX silver carries a real options book, and hedging in particular is expressed through options. The futures-only file misses it. If the formats are new to you, start with how to read the COT report.

The contract is SILVER - COMMODITY EXCHANGE INC., CFTC code 084691, the 5,000-ounce COMEX contract. Micro and cash-settled variants file separately and have shallower history.

The Four Groups in Silver, and What They Held

On the report dated 1 September 2026, against 119,827 contracts of open interest:

Managed Money also held 8,985 spreading contracts, and Swap Dealers 7,110.

Two things in that table are worth pausing on, because both are routinely misread.

The physical trade is not the big short. Producers and merchants, the miners and refiners and industrial users, are short 22,141 contracts. That is 18.48% of open interest, and it is the smallest of the three short books in absolute terms. If you have read that "the miners" are suppressing the silver price, the file says they hold under a fifth of the open interest short.

Swap Dealers are the big short. 49,268 contracts, 41.12% of open interest. This is the group most people mean when they say "the banks", and on the raw number they are right that it is the largest short in the market. What that position is, though, is the part that gets skipped. Swap dealers are intermediaries. They take the other side of over-the-counter exposure, structured products and ETF-related flow, and they lay off the resulting risk on the exchange. A dealer short on COMEX is frequently the mirror image of a client long somewhere the CFTC file does not show you. It is a byproduct of dealing, not a directional bet, which is why it is filed in its own category rather than with the speculators.

Note also that the classic Legacy "commercial" figure fuses the first two rows into a single net short of -44,238, or 36.92% of open interest. That fusion is where a great deal of confusion starts. A miner hedging production and a bank hedging a client swap are doing unrelated things, and the disaggregated report exists precisely so you do not have to treat them as one actor.

The Concentration Question, Measured

Here is the claim in its strongest form: a few banks hold a uniquely large short in silver. The CFTC series that answers it is Conc_Net_LE_4_TDR_Short, the share of open interest held net short by the four largest traders.

We pulled it for the whole metals complex, every week from 6 January 2015 to 1 September 2026, 609 reports.

* Copper's series in this format begins 8 February 2022, so n = 239 rather than 609.

The claim's true half. Silver is more concentrated than gold, and consistently so. Its mean is nearly six points higher, and silver's four-trader short concentration exceeded gold's in 537 of 609 weeks, 88.2% of the time. Anyone who has eyeballed the two series and concluded that silver's short is more concentrated than gold's has read the data correctly. That deserves saying plainly, because it is usually dismissed rather than checked.

The claim's false half. Silver is not the outlier. Platinum's mean concentration is seven points above silver's and palladium's is five points above. Silver's four-trader short exceeded platinum's in only 71 of 609 weeks (11.7%) and palladium's in 161 (26.4%). In the 239 weeks where all five metals report, silver had the lowest concentration of the complex 34 times and the highest only 18 times, for a mean rank of 2.98 out of 5. Dead middle.

So the structure people describe as evidence of a silver-specific conspiracy is a structure that is stronger in two other precious metals, neither of which attracts the same literature. A theory that explains silver has to explain why platinum, more concentrated in 88% of weeks, generates no comparable claim. The simpler reading is that concentration is a feature of small metal contracts with a handful of dealers intermediating them, and silver is a mid-sized example of a general pattern.

And this week, the ranking is inverted. Silver's 26.5% is its 49.8th percentile, which is to say completely ordinary for silver. Gold's 27.0% is its 97.9th percentile, close to its own record of 29.1%, and above silver's reading outright. Copper sits at its 89.5th. If you are hunting for unusual short concentration in metals on this report, silver is the wrong place to look. That is a genuinely useful piece of information and it is available to anyone, free, in a column most readers never open.

The Finding Nobody Is Discussing

While the argument about silver runs on, something considerably more unusual is happening across the whole complex. Swap dealer shorts, as a share of open interest, are at or near record highs almost everywhere.

Four of five metals sit above their 93rd percentile on the same report. Dealers are carrying an unusually large short across the complex at once, which is what intermediating a very large, very one-sided client long looks like from the exchange side.

Now the honest caveat, because it changes the size of the claim. Silver's jump from 36.15% to 41.12% in a single week looks dramatic and mostly is not. The dealers barely moved:

The dealer book shrank. The denominator shrank eight times faster. Almost the entire move in the ratio is open interest leaving the market, which is ordinary around contract roll, and not dealers pressing a short. Anyone quoting the 41% as fresh selling has mistaken a denominator for a numerator. The multi-year elevation is real and worth watching. The one-week spike is mostly arithmetic.

Where Managed Money Actually Sits

For all the noise about the short side, the speculative side of silver is quiet.

Managed Money is net long 12,170 contracts. Against a 52-week mean of 15,674 and a standard deviation of 8,901, that is a z-score of -0.39. The 156-week COT Index reads 36.5. Both say the same thing: fund positioning in silver is below its own average and nowhere near an extreme. It fell 1,065 contracts on the week and is up 1,103 over four weeks, so it is not trending hard either.

This matters for anyone trading the story rather than the data. "The funds are piled into silver" and "the funds have abandoned silver" are both false on this report. They are sitting slightly below normal, doing very little. The interesting positioning in silver right now is on the dealer side, and it is a complex-wide phenomenon rather than a silver one.

For contrast, gold's Managed Money net long is 140,811 with a z-score of +1.19, and copper's is 73,000 with a z-score of +0.97 and a COT Index of 93.6. Copper is the crowded metal on this report, not silver, and it is also the one metal where the dominant short really is hedged physical metal rather than dealer flow.

How to Trade Silver Positioning

Read the dealer line, but do not read it as a view. Swap dealer positioning tells you how much client flow has been intermediated. It is information about the market's plumbing, not about direction. Treating a dealer short as a bearish forecast is the single most common error in silver analysis.

Normalize before you react. Silver is thinner and more volatile than gold, so raw contract counts mislead badly across time. A z-score against the trailing 52 weeks and a COT Index over a long lookback are the minimum. Our z-score guide covers the method.

Use gold as the lead. Silver takes its macro direction from gold and adds an industrial demand cycle on top. A silver extreme that is not confirmed by anything in gold is usually a silver-specific squeeze setup rather than a macro turn, and those resolve faster and more violently.

Respect the thinness. At 119,827 contracts of open interest against gold's 578,926, silver moves further on the same flow. Extremes in silver are sharper and shorter-lived, which cuts both ways.

Watch open interest alongside every ratio. As the table above shows, a ratio can move five points without anybody trading. Whenever a share-of-open-interest figure jumps, check the denominator before you build a thesis on it.

Do not fade concentration. There is no evidence in the record that high four-trader concentration in silver precedes a price move in either direction, and we are not going to imply one. Concentration is a structural description of who is in the market. It answers the manipulation question. It is not a signal.

Where COTInsight Fits

The silver market page carries the current positioning read. Inside the app, silver sits on the same board as around 350 other weekly-reporting markets, ranked by extremity, with the z-score, 156-week COT Index, regime state, open-interest trend and divergence check on one screen. That is the view that makes "copper is at a COT Index of 93.6 and silver is at 36.5" a five-second observation rather than an afternoon of spreadsheets.

Ultimate adds the historical outcome statistics per z-score bucket, the twenty-year archive behind the tables in this article, the weekly AI commentary, and the TradingView indicator so the same z-score and COT Index sit on your silver chart.

To be clear about what it is not: there is no intraday data, no entry signals and no free tier. COT is weekly data with a three-day lag, and the product is built for swing and position timeframes.

Related guides: gold and copper. Related metals: gold, copper, platinum, palladium. Pricing is here.

Frequently Asked Questions

Which banks are short silver?

The CFTC does not name individual firms. It publishes categories and concentration ratios. On the report dated 1 September 2026, Swap Dealers as a group held 49,268 contracts short, 41.12% of open interest, and the four largest traders of any type held a net short equal to 26.5% of open interest. Swap dealers are largely bank intermediaries, so the category is the closest published answer, but the file identifies no firm and nobody outside the CFTC can name them from this data.

Is the silver short unusually concentrated compared to other metals?

Compared to gold, yes: silver's four-trader net short concentration averaged 26.19% against gold's 20.31% across 609 weeks from 2015 to 2026, and exceeded gold's in 88.2% of them. Compared to the wider complex, no: platinum averaged 33.41% and palladium 31.68%, both higher than silver. Silver's mean rank among the five metals is 2.98 out of 5.

Does a large commercial short mean the silver price is being suppressed?

The data cannot support that conclusion, and it contains a category error. The Legacy "commercial" figure fuses producers hedging metal with swap dealers hedging client exposure. On 1 September 2026 that combined net short was 44,238 contracts, of which the physical trade accounted for 16,677 and dealer intermediation for 27,561. A dealer short is normally the offset to a client long the report does not show. There is also no measured relationship in the record between concentration and subsequent price direction.

Why did the swap dealer share of open interest jump in the week to 1 September 2026?

Almost entirely because open interest fell. Swap dealer shorts went from 49,808 to 49,268 contracts, a decline of 540, while open interest fell from 137,784 to 119,827, a decline of 17,957 or 13.03%. The share rose 4.97 points on a shrinking denominator rather than on new selling.

Is Managed Money crowded in silver right now?

No. Net long 12,170 contracts on 1 September 2026, a z-score of -0.39 against the trailing 52 weeks and a 156-week COT Index of 36.5. That is below silver's own average. Copper, at a COT Index of 93.6, is the crowded metal on this report.

Which silver contract and report should I use?

SILVER - COMMODITY EXCHANGE INC., CFTC code 084691, the 5,000-ounce COMEX contract, read from the Disaggregated report in its futures-and-options-combined form. The Legacy report merges the categories that matter, and the futures-only file misses a real options book.

When is silver COT data published?

Every Friday at 3:30pm Eastern, reflecting positions as of the preceding Tuesday. The report dated 1 September 2026 was released on Friday 4 September.