What Is Overtrading, and What Happened When Researchers Measured It?
Overtrading means taking more trades than your edge and your costs can carry, and when researchers sorted 66,465 households at one discount broker by how much they traded, the busiest fifth earned almost the same gross return as the quietest fifth and a far worse net one.

Overtrading is taking more trades than your edge and your trading costs can carry: the count goes up, the cost of every round turn goes up with it, and in the research below the average decision got no better. It is not a regulated term, and the CFTC's futures glossary has no entry under that name, but the behaviour behind it has been measured on tens of thousands of real brokerage accounts.
The two studies that did the measuring sorted accounts by how much they traded, then compared what they earned before costs with what they earned after. The gap between those two numbers is the whole subject.
What is overtrading?
Overtrading is trading more often than your own strategy's edge justifies once the cost of each trade is deducted. There is no official definition. The CFTC's futures glossary defines round turn, day trader and churning, and carries no entry under overtrading at all. The word is a judgement about a trader's own activity rather than a rule anyone enforces.
What researchers do instead is measure something countable. Two proxies appear in the work below: portfolio turnover, which Barber and Odean calculate from the value a household bought in a month and the value it sold, each against the portfolio's starting value and then averaged, and the raw number of trades an account placed, which is how Odean separates his most active tenth. Neither one is overtrading by itself. A high turnover figure is only a problem when the trades it represents fail to pay for themselves.
The unit of cost is the round turn. The CFTC glossary defines it as one purchase plus one offsetting sale, in either order — both legs treated as a single completed transaction. The glossary's commission entry is the other half of that: a charge for the buying and the selling, or a fee for executing an order, which is to say a cost attached to the transaction rather than to the time you hold it. Commissions, exchange fees and whatever portion of the bid-ask spread you give up are charged on each round turn. Double your trade count and you double that line of your P&L whatever the market does, which is why frequency is a cost decision before it is a psychological one.
Is overtrading the same thing as churning?
No. Churning is a specific abuse with its own CFTC glossary entry, and the elements of it point at someone other than the account holder. There are four in the entry: the trading has to be excessive, the account has to be a discretionary one, the trades have to come from whoever controls that account, and the motive has to be the commissions the activity generates, at the expense of what the customer wanted.
A discretionary account, in the same glossary, is one where the holder has signed a written power of attorney, with a commodity trading advisor given as the example, letting someone else buy and sell without approval for each trade. Churning is therefore a claim about an agent's conduct. Overtrading, as traders use the word, is a claim about your own.
| Overtrading | Churning | |
|---|---|---|
| Whose account | Your own | A customer's discretionary account |
| Who places the trades | You | Whoever holds control of the account |
| Where it is defined | Nowhere in the CFTC glossary | Its own glossary entry |
| What makes it excessive | Trade count measured against your edge and your costs | Trading measured against the customer's interests |
| Motive in the definition | None stated | Generating commissions |
The practical consequence is that nobody is going to stop you. There is no definition of overtrading to breach, and on our own programs the rules count contracts and balances rather than trades, which is exactly why the arithmetic has to be yours to run.
What happened when researchers sorted investors by how often they traded?
Gross returns barely moved and net returns collapsed. Brad Barber and Terrance Odean studied positions and trades for 78,000 households at one large discount brokerage over the six years ending January 1997, published the result in the Journal of Finance in April 2000 under the title Trading Is Hazardous to Your Wealth, and sorted 66,465 of those households into five groups by average monthly turnover.
The spread between the groups was enormous. The quietest fifth turned over 0.19% of their portfolios in an average month; the busiest fifth turned over 21.49%, and the paper puts annual turnover for the households that trade most at more than 250%. Qualifying for that top group took at least 8.7% monthly turnover.
Then the comparison. Measured before costs, the five groups' annualised geometric returns spanned 18.5% to 18.7% — a range of 20 basis points across groups whose monthly turnover differed by a factor of more than a hundred. Measured after commissions and the spread, the busiest fifth earned 11.4% a year against 18.5% for the quietest. On the authors' own-benchmark measure the trading of the high-turnover households cost them 57 basis points a month, or 6.8% a year, relative to the low-turnover ones. Risk-adjusted returns went the same way: a footnote compares net Sharpe ratios between the extreme quintiles, 0.092 for the most active against 0.180 for the least, on the equity holdings the authors could see and no other assets.
That result was set up as a test between two theories. Under the rational-expectations argument of Grossman and Stiglitz, active traders should earn higher gross returns, because otherwise there is nothing to pay the costs with. Barber and Odean report that the prediction failed, and read the outcome as evidence for overconfidence instead. The sample is US common stocks held directly by households at one broker, so the figures describe those accounts and no others.
Did the most active accounts at least buy better than they sold?
They bought worse. Odean's Do Investors Trade Too Much?, in the American Economic Review for December 1999, took 10,000 accounts at a nationwide discount brokerage, selected at random from those that had traded at least once in 1987, and followed every trade through to December 1993 — 162,948 trade records, of which 97,483 could be matched to daily returns data: 49,948 purchases and 47,535 sales.
The test was simple. If a trader sells one security to buy another, the security bought has to outperform the one sold by enough to cover the cost of switching. In this sample that cost was about 5.9% for the round trip. Most of it is commission: an average of 2.23% of the purchase price to buy and 2.76% of the sale price to sell, roughly 5% for a sale-and-purchase pair, with the author's estimate of the effective bid-ask spread on top. Those are 1987-to-1993 discount-brokerage commissions, charged as a percentage of each stock trade, which is why the total reads in percent rather than in dollars.
A year on, 252 trading days after the trade, the securities these accounts had bought had returned 5.69% and the ones they had sold 9.00%. The gap runs the wrong way by 3.31 percentage points, before any costs are deducted at all. Among the tenth of investors who traded most, the same one-year gap was 2.69 points against them. So the extra activity was not a case of good decisions eaten by friction. On average the switch itself was a mistake, and the commissions were charged on top of it.
Two scope limits travel with those numbers: the trades studied are in NYSE, American Stock Exchange and Nasdaq securities, and accounts closed during the period were not replaced, which the author notes may bias the later years toward more successful investors.
Do those findings apply to futures day trading?
Partly, and the honest answer names the limit. Both studies cited here measure stock accounts at US discount brokers in the decade to January 1997: common stocks held directly by households in one, and NYSE, American Stock Exchange and Nasdaq securities in the other. The phrase day trading does not appear in either paper, and the return horizons Odean tests after a trade run from four months to two years rather than from an open to a close. Their percentages belong to that setting, and carrying 11.4% against 18.5% into a futures context would be inventing a result.
What does transfer is structural rather than empirical. Cost is charged per round turn, so total cost scales with trade count; a strategy's per-trade edge has to clear that cost before frequency helps; and the gap between a gross and a net result widens with activity. None of that depends on which contract you trade.
It holds on a simulated funded account as well. Our fees page is explicit that commissions and exchange fees hit a simulated balance exactly as they would hit a live one, and that they are deducted before any profit counts toward a payout or toward a winning day. For what the spread alone costs on each round turn in a named contract, scalping futures works that through against observed futures spread data.
How many trades a day is too many?
None of the sources behind this post states a number of trades that counts as too many. Barber and Odean's 8.7% monthly turnover is what put a household in their top quintile in a 1990s equity sample, and the CFTC glossary's churning entry attaches no number to the word excessive. What the research supplies is a method of comparison.
The arithmetic that replaces the missing number is per-trade expectancy against per-round-turn cost. If the average trade in your plan nets less than what a round turn costs you in commissions, fees and spread, each extra trade is a negative-expectancy bet and more of them makes the outcome worse rather than noisier. If it nets comfortably more, the count multiplies that figure — and multiplies the error in your estimate of it just as fast, because the estimate comes from however many trades you have actually closed. What decides the outcome is the sign of the per-trade result after costs. What win rate do you need to be profitable sets out how expectancy is calculated and where the inputs come from in your own records.
Which of a funded account's rules bound how often you trade?
None of them bound it directly, which is the part worth knowing before a busy session rather than after one. Three rules shape what a high trade count does to the account here, and not one of them counts trades.
First, nothing we currently sell carries a daily loss limit, so no daily tripwire closes a session early or interrupts a run of trades. Second, the one rule that can close the account on performance is the drawdown, assessed on closing balances: a line that trails your best close by a set amount, watched in real time, with a touch liquidating the account on the spot. Third, the contract limits cap the position you have open at any one moment, summed across instruments, and say nothing at all about how many times a day you open one.
Technique is bounded where frequency is not. Our prohibited conduct page rules out order rates and volumes aimed at the platform rather than at the market, and it sets a measurable line on very short trades: more than half your trades and more than half your profit have to come from trades held longer than ten seconds. Both of those constrain how a trade is executed. The count on its own breaks no rule on any program we sell, which leaves the cost of it as the only thing standing in the way.
What does a high trade count cost against the drawdown room?
Enough to matter, and the figure is a multiplication anyone can run. The drawdown article, updated 8 September 2026, puts the gap on a $50,000 Classic or Horizon account at $2,000, and our instruments page lists the E-Mini S&P 500 tick at $12.50 and the Micro S&P 500 tick at $1.25. So that gap is 160 E-mini ticks wide, or 1,600 micro ticks.
Now assume a market one tick wide and both legs of every trade taking liquidity, which makes one tick the spread cost of a round turn. On that assumption 160 round turns exhaust the entire gap on a single E-mini contract, with the market never once moving against the trade. Twenty round turns in a session is $250, an eighth of the room, and run at the four-mini ceiling that account size carries it is $1,000 in a day — half the gap, before a single trade has gone wrong.
Those figures price the spread and nothing else; commissions and exchange fees are charged on the same trades on top of them. Twenty round turns on one micro contract is $25 on the same assumption, which is the arithmetic case for trading a smaller contract while a strategy's trade count is still being settled. And if the honest question is whether the sessions have stopped being planned ones, our responsible trading page sets out a self-check for that.
FAQ
Did frequent traders lose money in these studies?
Not in absolute terms. In Barber and Odean's sample the busiest fifth of households earned a net annualised 11.4% while the quietest earned 18.5%, and both figures are positive. The average household earned 18.7% gross and 16.4% net, against 17.9% for a value-weighted index of NYSE, AMEX and Nasdaq stocks over the same years. The penalty for high activity was relative: about seven percentage points a year between the busiest quintile and the quietest, in a period when the index itself returned 17.9% annualised.
Is overtrading against the rules on a prop firm evaluation?
Not here. The only performance rule that can close an account on our programs is the drawdown, and there is no daily loss limit and no cap on how many trades you place. What our prohibited conduct page restricts is technique: high frequency and mass order entry, and very short trades, where more than half your trades and more than half your profit have to come from trades held longer than ten seconds.
What is a round turn, and why does it matter for overtrading?
The CFTC glossary defines a round turn as one purchase plus one offsetting sale, in either order, treated as a single completed transaction. It matters because it is the unit your costs are charged in. Commissions, exchange fees and the spread you cross all attach to the round turn, so they scale with trade count rather than with time held or with how far the market moves.
How can I tell whether I am overtrading?
Use the comparison Barber and Odean used: separate your gross result from your net result over a stretch of trading. Costs scale with count, so a respectable gross and a poor net points at the trade count between them. Then check the sign of your average trade after commissions, fees and spread. If it is negative, more trades is the thing making the record worse.
Sources
- Barber & Odean — Trading Is Hazardous to Your Wealth: The Common Stock Investment Performance of Individual Investors, Journal of Finance 55(2), April 2000
- Odean — Do Investors Trade Too Much? American Economic Review 89(5), December 1999, pp. 1279-1298
- CFTC — Futures Glossary (entries: Churning, Discretionary Account, Round Turn, Day Trader, Commission)
- Lumen Futures Help Center — No daily loss limit
- Lumen Futures Help Center — Drawdown explained
- Lumen Futures Help Center — Contract limits
- Lumen Futures Help Center — Prohibited conduct
- Lumen Futures Help Center — Platform, data and exchange fees
- Lumen Futures — Tradable instruments, tick size and tick value
- Lumen Futures — Responsible trading
Educational content about futures markets and simulated trading. Not investment advice, and not a solicitation to trade. Trading futures involves substantial risk of loss. Read the full risk disclosure.