Is Futures Trading a Zero-Sum Game? Yes, Before Costs

Futures trading is a zero-sum game before costs and a negative-sum game after them. Every contract has one buyer and one seller, so each dollar the long gains is a dollar the short loses (zero-sum means all participants' gains and losses add up to exactly zero), and both sides pay commissions and fees on top. In economic terms the answer is softer: hedgers use futures to offset risks held elsewhere and can lose on the hedge while coming out ahead overall.

Key takeaway

Futures are exactly zero-sum before costs: the clearing house collects each day's losses and pays the same amount out as gains, so gross profit and loss across all accounts sums to $0. Commissions, exchange and clearing fees then make the market negative-sum, while the bid-ask spread is a transfer from liquidity takers to liquidity providers. A day trader profits only by beating the average counterparty by more than spread plus fees, which makes cost control part of strategy.

Why are futures zero-sum before costs?

Futures are zero-sum before costs because a contract only moves money between the two traders who hold it. Buying one creates no share of a business, no dividend and no interest, only a promise between a long and a short, and every open long is matched by exactly one open short. The count of those matched contracts not yet closed out or settled at expiry is the open interest.

Studying E-mini S&P 500 transactions tagged with trader IDs, Baron, Brogaard and Kirilenko (2014) wrote in a CFTC working paper: "Trading in the E-mini is a zero-sum game: one trader's profits come directly at the expense of the opposite trader."

Costs break the symmetry in one direction. Brokers charge commissions, the exchange and clearing house charge fees, and the National Futures Association collects an assessment fee, so money leaves the pool of traders on every contract.

How does the clearing house keep futures zero-sum?

The clearing house, or central counterparty, keeps futures zero-sum by standing between every buyer and seller and settling gains and losses in cash every day. Through novation (the original contract between two traders is replaced by two contracts, each facing the clearing house), the clearing house becomes buyer to every seller and seller to every buyer, so its own position is always flat.

Daily settlement is a federal regulatory requirement. Under 17 CFR 39.14, a derivatives clearing organization must "effect a settlement with each clearing member at least once each business day," and settlement includes "Payment and receipt of variation margin for futures, options, and swaps." Variation margin is the cash each position gains or loses when marked to market at the day's CME settlement price, and is separate from the initial and maintenance margin posted to hold a position.

In the ledger in the next section, the day settles at 6,512.00. The hedger's short pays $2,400 into the clearing house, which pays $1,950 to the speculator and $450 to the market maker and ends at exactly $0. Real payments are netted per clearing member (the firm that clears trades for its customers), but the arithmetic is identical.

The clearing house is the buyer to every seller and the seller to every buyer, and it neither gains nor loses on settlement. Every dollar a winning account receives came out of a losing one the same day.

What does a zero-sum futures day look like?

A zero-sum futures day looks like the ledger below: three traders' gross results cancel to $0.00, and fees turn the total into -$26.20. The contract is the E-mini S&P 500 (ES), worth $50 per index point with a 0.25-point tick worth $12.50, as described by Kirilenko, Kyle, Samadi and Tuzun (2014). All prices, the one-tick spread and all fees are illustrative.

A hedger protecting a stock portfolio sells 4 ES at the bid, 6,500.00, to a market maker. A day trader (the speculator) buys 4 at the ask, 6,500.25, from the market maker, who is now flat. Later, with the market higher, the speculator sells 4 at the bid, 6,510.00, back to the market maker. The day settles at 6,512.00 with the hedger short 4, the market maker long 4 and the speculator flat: open interest of 4.

Gross P&L = sum over each trade of signed quantity × (settlement price - trade price) × $50, with buys positive and sells negative. Net P&L = gross P&L - contract sides traded × all-in cost per side. Per-side costs are illustrative bundles of commission, exchange, clearing and regulatory fees, not any broker's or CME's published rates.

Participant	Trades (illustrative)	Gross P&L at 6,512.00	Costs (sides × per side)	Net P&L
Hedger	Sell 4 @ 6,500.00, hold short	-$2,400.00	4 × $1.00 = $4.00	-$2,404.00
Speculator	Buy 4 @ 6,500.25, sell 4 @ 6,510.00	+$1,950.00	8 × $2.25 = $18.00	+$1,932.00
Market maker	Buy 4 @ 6,500.00, sell 4 @ 6,500.25, buy 4 @ 6,510.00, hold long	+$450.00	12 × $0.35 = $4.20	+$445.80
Total	4 long vs 4 short still open	$0.00	$26.20	-$26.20

The average trader nets about -$8.73 (-$26.20 / 3). The $26.20 in fees is the only money that left the system, paid to brokers, the exchange and clearing house, and the NFA.

Remove the costs and the ledger is perfectly zero-sum. Put them back and the group as a whole loses exactly what it paid in commissions and fees, which is the average trader's starting handicap.

Is the bid-ask spread a cost or a transfer?

The bid-ask spread is both: a real cost to the trader who crosses it and a zero-sum transfer to the trader whose quote was hit. Measured against the mid-price (halfway between bid and ask), the hedger paid a half-spread of 0.125 points on 4 contracts, $25, and the speculator paid two half-spreads, $50. The market maker collected all three, $75, and made another $375 holding 4 long from the 6,510.125 mid to the 6,512.00 settlement, for its $450. Stripped of spread, the speculator's directional call earned $2,000 (mid 6,500.125 to mid 6,510.125).

The spread is not a fee that leaves the system: it is a transfer from traders who demand immediacy to the traders who provide it. That is why paying it on every entry and exit is a real cost for a day trader even though the market stays zero-sum.

Kirilenko et al. (2014) describe the liquidity provider's business model: "A textbook market maker will try to buy at the bid price, sell at the offer price, and capture the bid-ask spread as a profit." The risk sits in inventory: Baron, Brogaard and Kirilenko (2014) read their E-mini results as consistent with the idea that "Mixed and Passive HFTs earn the bid-ask spread in the short-run but are adversely selected on a longer time scale," meaning better-informed traders tend to trade against their quotes just before prices move against them. The spread shown on the DOM (depth-of-market) ladder is the price a taker pays for immediacy.

Who is on the other side of your futures trade?

The other side of any single futures fill is anonymous, but in the E-mini S&P 500 it is almost always a professional, and overnight positions sit mostly with asset managers, dealers and hedge funds. In the best-documented sample, the E-mini audit trail of May 3-5, 2010, Kirilenko et al. (2014) found that on an average day 15 high-frequency trading accounts produced 34.22% of volume, 189 market makers 10.49%, 3,504 opportunistic traders 30.79%, and fundamental buyers and sellers 24.00% combined. Small traders (fewer than 10 contracts a day) averaged 6,065 accounts, about 51.1% of the 11,875 active accounts, and produced 0.50% of volume.

Positions are tracked by the CFTC's weekly Commitments of Traders (COT) reports. For financial futures, the Traders in Financial Futures (TFF) report sorts large traders into four groups defined in the CFTC's TFF explanatory notes: Dealer/Intermediary (firms that "earn commissions on selling financial products, capturing bid/offer spreads and otherwise accommodating clients"), Asset Manager/Institutional (pension funds, endowments, insurers, mutual funds), Leveraged Funds (hedge funds and commodity trading advisors) and Other Reportables (mostly hedging business risk). Traders below the reporting threshold are nonreportable. The ES picture as of Tuesday September 22, 2026, with open interest of 1,890,653 contracts and 419 reportable traders, is below; percentages exclude spreading positions (long and short in different months), about 9.7% of open interest.

TFF category (ES)	Long, % of OI	Short, % of OI	Net contracts
Asset Manager/Institutional	59.2%	9.7%	+934,455
Dealer/Intermediary	8.9%	42.4%	-634,203
Leveraged Funds	6.4%	26.2%	-375,574
Other Reportables	2.6%	4.2%	-29,026
Nonreportable (small traders)	13.2%	7.7%	+104,348
Total			0

Institutions are the natural longs, dealers and leveraged funds the main shorts, and small traders a modest net long. A retail day trader is mostly trading against professionals whose costs are far lower.

Net positions sum to exactly zero. Asset managers hold the largest net long and dealers the largest net short, but the notes warn that "staff classifies traders, not their trading activity," so a leveraged fund's short may itself be a hedge. The snapshot follows the September quarterly expiry (ES open interest fell 555,866 contracts that week), and composition changes weekly.

Micro E-mini S&P 500 (MES, $5 per point) looks different. In the same report only 27 MES traders were reportable, and nonreportable small traders held 54.9% of long and 61.8% of short open interest, against 13.2% and 7.7% in ES. An overnight micro position is, in aggregate, mostly held against other small traders; an ES position mostly against institutions.

Are futures zero-sum in economic terms?

Futures are not zero-sum in economic terms, because many participants hold an offsetting exposure outside the futures account. The ledger's hedger lost $2,404 on 4 short ES, but the short protected an index-tracking stock portfolio of 4 × $50 × 6,500 = $1,300,000. A 12-point rise is 0.1846%, which lifts that portfolio by $2,400 (illustrative, ignoring dividends and basis, the gap between futures and index prices). Combined, the hedger is down only the $4.00 of fees, the price of locking the portfolio's value for the day. Baron, Brogaard and Kirilenko (2014) make the same point about firms in their data: "a loss in the E-mini market does not imply the trading firm loses money overall."

Zero-sum describes the futures ledger, not the people in it. The hedger's "loss" bought insurance and cost four dollars overall, which is why hedgers keep taking the other side of speculators who are trying to profit.

Keynes (A Treatise on Money, 1930) and Hicks (Value and Capital, 1939) developed the theory of normal backwardation: hedgers who are net short pay speculators a risk premium, so futures prices sit below the expected future spot price. The evidence is mixed. Gorton and Rouwenhorst (NBER Working Paper 10595, 2004) found fully collateralized commodity futures earned returns and Sharpe ratios (return per unit of volatility) similar to equities, consistent with a paid risk premium, while Gorton, Hayashi and Rouwenhorst (NBER Working Paper 13249, 2007) tied those premiums to inventory levels and rejected hedging pressure as an important determinant. The theory concerns commodity hedgers, not a premium an index-futures day trader can count on.

What does zero-sum mean for a day trader?

For a day trader, zero-sum means every round trip starts behind by the spread plus fees: 1.36 ticks in ES at the illustrative costs here. With a one-tick spread and round-turn fees of $4.50, an ES round trip starts $17.00 behind ($12.50 + $4.50), or 0.34 index points. In MES, with the same 0.25-point tick worth $1.25 and an illustrative $1.00 round turn, the hurdle is $2.25, or 1.8 ticks. Fixed fees weigh more on micros: at 6,500 the ES fee is 0.138 basis points (hundredths of a percent) of $325,000 notional, the MES fee 0.308 basis points of $32,500.

Baron, Brogaard and Kirilenko (2014) found that small traders lost an average of $3.67 per contract, marked one minute after each trade, in August 2010, the most of any group, while fundamental traders came out $0.19 per contract ahead. The widely repeated claim that 90% of futures traders lose has no primary source we could trace; the evidence that does exist is collected in what percentage of day traders lose money.

Commissions, market orders that pay the spread and slippage beyond the best price all come out of the same edge; futures trading costs, commissions and fees breaks them down line by line for funded accounts. Funded traders also pay evaluation and activation fees to the firm outside the market, covered in how prop firms make money.

Is the futures market rigged because it is zero-sum?

The futures market is not rigged by being zero-sum, because zero-sum describes the accounting, not the fairness. The clearing house guarantees both sides of every contract, and settlement moves cash mechanically at a published price. Measured intermediation costs are small: Baron, Brogaard and Kirilenko (2014) estimated the effective cost of high-frequency trading to other E-mini traders in August 2010 at 0.22 basis points, about $1.10 per contract. Consistent winners exist (the median HFT firm in the same study earned an annualized Sharpe ratio of 4.3 on gross trading revenue), which shows skill and speed are rewarded, not that outcomes are fixed. Whether stops get targeted is a separate question, examined in is stop hunting real.

Does a trade copier change the zero-sum math?

A trade copier does not change the zero-sum math; it repeats whatever a strategy nets after costs on every account it copies to. Thor, this blog's own product, copies one master account to many accounts server-side, and each copy pays its own commissions, fees and spread and gets its own fill, which can differ from the master's. A strategy that clears the 1.36-tick ES hurdle keeps its margin on every account, and one that does not loses on every account. A copier cannot make a losing strategy pay, so fix the edge and the costs first.

Go deeper

Frequently asked questions

Is the stock market zero-sum like futures?

No. Shares are claims on companies' earnings and dividends, so shareholders as a group can gain together when businesses grow, while futures only move price changes between a long and a short. Short-term stock trading against other traders behaves much more like the futures case.

Are options a zero-sum game too?

Yes, before costs: every option buyer's gain is the option writer's loss, and the premium passes from one to the other. Fees make the total negative, and a hedger using options can still come out ahead economically.

Is futures trading just gambling?

No, although a speculator with no edge faces similar math. Futures let hedgers shed price risk and help set prices, whereas a casino game has a negative expected value fixed by the house. A trader without an edge still loses the fees and spread on average, much as a gambler loses the house edge.

Do market makers trade against my position?

They take the other side of many fills, but a market maker quotes both sides for spread income rather than betting against any particular trader. A market maker that buys your sell order aims to sell to the next buyer, and holding your position is inventory risk, not its goal.

Does the clearing house profit when traders lose?

No. It earns clearing fees on every contract whoever wins, and variation margin passes through it without becoming its revenue.

If a losing trader cannot pay, do winners still get paid?

Yes, the clearing house guarantees settlement, so winners are paid even when a loser defaults. The defaulting account's clearing firm must cover the shortfall, and the clearing house's own financial safeguards stand behind that firm.

Does zero-sum mean half of futures traders win?

No, zero-sum is about dollars, not headcount. A few large winners can balance many small losers, or the reverse, and fees push the total below zero whatever the split.

Is prop firm trading zero-sum?

The futures trades are, but a prop firm evaluation adds fees that sit outside the market's zero-sum pool. When an evaluation runs on a simulated account, no market counterparty exists at all, and the only real cash flows are between the trader and the firm.

Can I see who took the other side of my futures trade?

No, fills are anonymous, and after clearing your contract faces the clearing house rather than the original trader. The CFTC's Commitments of Traders reports show only aggregate positions by trader category, generally published each Friday at 3:30 pm ET with data from the preceding Tuesday.