Why Institutions Win and You Don’t (The Premise Is Half Wrong)
Two things are worth establishing before the advice starts. Active managers don’t reliably beat the market — 79% of large-cap US equity funds lagged the S&P in 2025, and over 15 years no category shows majority outperformance. And bank traders mostly aren’t allowed to do what you do: proprietary trading at US banks has been prohibited since 2015, loosened since, but never restored. The real gap isn’t skill. It’s arithmetic.
They’re not better at predicting. Most of them have stopped trying to predict.
Hedge funds don’t win
In 2007 Warren Buffett offered to bet that over ten years, a low-cost S&P 500 index fund would beat any five hedge funds a professional could pick, measured net of fees and costs. Ted Seides of Protégé Partners took the bet, selecting five funds-of-funds that in turn held more than 100 underlying hedge funds — a deliberately broad sample rather than a single manager’s luck.
Over the decade to 2017, the funds returned roughly 36%. The index returned over 125%. Seides conceded with eight months still to run, writing that for all intents and purposes the bet was over and he had lost. (Institutional Investor)
Buffett’s argument for why wasn’t about talent, and it’s the part worth internalizing. Active investors, in aggregate, hold roughly the market’s positions — so in aggregate they must earn roughly the market’s return before costs, and less after them. That isn’t a claim about anyone being bad at their job. It’s an accounting identity. (Long Bets)
And it isn’t just one bet
A single ten-year window in a bull market is thin evidence, so it’s worth checking the systematic data. S&P Dow Jones has published the SPIVA scorecard since 2002, measuring active funds against their benchmarks using a survivor-bias-free database that includes funds which merged or closed.
In 2025, 79% of active large-cap US equity funds underperformed the S&P 500 — worse than 2024’s 65%, and the fourth-worst year for large-cap active managers in the scorecard’s 25-year history. Over 20 years the figure is around 92%. (SPIVA U.S. Scorecard)
The more damning number is persistence. Over 15 years there were no categories in which the majority of active managers outperformed, across domestic equities, international equities or fixed income — and top-quartile funds have close to random odds of staying top-quartile. Picking a manager who beat the market last year tells you very little about next year. (Index Fund Advisors, summarizing SPIVA)
A 2026 study sponsored by the Investment Adviser Association’s Active Managers Council argues SPIVA overstates underperformance: it counts funds rather than assets, so tiny funds nobody owns weigh as much as huge ones, and it compares against hypothetical benchmarks with no costs. Their fixed-income reanalysis moved 10-year underperformance from 71% down to 37% — a genuinely large correction. S&P’s response is that SPIVA deliberately measures the proportion of funds, not assets. Both are defensible. The equity picture survives the critique better than the bond picture does. (WealthManagement.com)
Seides made a reasonable counter, and it deserves airing: hedge funds don’t set out to beat the S&P. They aim for positive returns across market conditions, long and short, which makes a bull-market comparison somewhat apples-to-oranges. That’s true. It also doesn’t rescue the popular version of the claim — that institutional traders systematically outperform. In aggregate, net of fees, they didn’t.
Bank traders aren’t doing what you’re doing
Here’s the fact that reframes the whole question. The Volcker Rule — Section 619 of Dodd-Frank — prohibits banking entities from proprietary trading, defined as taking positions principally to profit from short-term price movements for the bank’s own account. It took effect on 1 April 2014, with full compliance required by 21 July 2015. (CFA Institute)
What remains permitted is the exemption list: market making, underwriting, risk-mitigating hedging, trading in government securities, and trading on behalf of customers. (CFTC fact sheet)
Read that list again with a retail trader’s eyes. Almost none of it is directional speculation. A market maker quotes both sides and gets paid the spread for providing liquidity — the position is a by-product of the service, not a view. Underwriting is fee business. Hedging is risk reduction by definition.
So when someone says bank traders beat retail traders, the comparison mostly isn’t happening. The bank desk is running a spread-capture and flow business with a structural edge. You’re trying to predict direction, which is the specific activity regulators pushed banks out of.
“Banned since 2015” is the short version, and it’s incomplete. A 2018 statute exempted smaller banks entirely — those under $10 billion in assets with trading assets below 5% of the total. A November 2019 final rule, effective January 2020, simplified the proprietary trading provisions and compliance requirements. A June 2020 rule relaxed the covered-fund restrictions on bank investment in hedge and private equity funds. The core prohibition on proprietary trading stands, but the perimeter around it has been redrawn more than once. Anyone telling you the rule is unchanged since 2013, or that it was repealed, is wrong in opposite directions. (OCC bulletin)
“But banks post record trading revenue”
They do, and it’s the obvious objection. JPMorgan reported a record trading haul as recently as April 2026. That isn’t a contradiction — it’s the point. Enormous trading revenue is exactly what a spread-and-flow business at scale produces, and it comes from volume of customer activity rather than from directional calls. A market maker’s best quarter is a volatile one with heavy two-way flow, whatever direction prices went. Record revenue is evidence of the model, not evidence that the desk out-predicted anyone.
The difference that actually explains it
Strip away the mystique and one variable does most of the work: how many times you get to be right.
Take a genuinely strong retail edge — a 55% win rate at 1:1 — and hold it constant. Change only the number of trades in the period, then measure how often the period ends negative.
| Trades in the period | Chance it loses money |
|---|---|
| 5 (a week, one a day) | 40.6% |
| 20 (a month, one a day) | 24.9% |
| 60 (a month, three a day) | 18.1% |
| 250 (a year, one a day) | 4.8% |
| 1,000 | 0.1% |
| 10,000+ | ~0% |
Identical skill in every row. At five trades a week, a genuinely profitable strategy shows a loss 41% of the time. At ten thousand trades it essentially cannot.
A market-making desk gets the bottom rows for free. It isn’t more certain about anything — it’s just sampling its edge often enough that variance stops mattering. You are permanently stuck in the top rows, where a real edge and a bad month are indistinguishable from the inside.
What they actually have
Beyond sample size, the genuine structural advantages are mundane and mostly unavailable to you:
- Order flow information. A market maker sees what customers are doing. That’s not a prediction — it’s data you can’t buy.
- Enforced, external risk management. A risk desk cuts a trader’s position whether or not the trader agrees. It’s the same mechanism as a prop firm’s drawdown limit binding when your own rule doesn’t — and it’s a feature, not a constraint.
- A salary. A bank trader’s rent doesn’t depend on this month’s P&L. That removes the pressure that produces most retail mistakes.
- Diversification across many uncorrelated books. The desk’s result is an average of dozens of positions, not the outcome of one idea.
- Cost of capital and execution that retail simply doesn’t get.
And what you have that they don’t
This is the part institution-worship gets wrong. The retail structure has real advantages, and they’re all about freedom rather than resources.
- No mandate and no benchmark. A fund manager measured against an index must stay invested. You can sit out for a week and nobody redeems.
- No size problem. Your order doesn’t move the market. Large funds can’t take small, good opportunities because they can’t get enough size on to matter.
- No redemption pressure. Nobody pulls your capital after a bad quarter, which is what forces institutional risk-shifting near reporting dates — the same tournament dynamic prop payout cycles recreate.
- You can wait. The single genuine retail edge is patience: you’re allowed to take zero trades today. Nearly every professional structure punishes that.
What follows from all this
- Stop trying to out-predict anyone. The prediction game is where variance dominates and your sample is tiny. If your approach needs you to be right more often than a professional, the arithmetic above says you won’t find that out for years.
- Increase your sample, carefully. More trades reduce variance only if the edge is real — otherwise you just reach ruin faster. Prove the edge on a small size first, then let sample size do its work.
- Judge yourself over 250 trades, not 20. At 20 trades a real edge loses a quarter of the time. Any conclusion you draw from a month is mostly noise, which is the same reason reading causes into individual losses invents rules.
- Buy the enforcement you lack. You have no risk desk. A platform-level lockout or a firm’s drawdown limit is the closest substitute.
- Use the advantages you actually have. Sitting out, trading small opportunities, and having no benchmark are worth more than any indicator.
The honest limits
“Institutions” is not one thing. Renaissance’s Medallion fund and a fund-of-funds average are both institutional. Some firms genuinely and persistently win; the aggregate does not. Both facts are true and people cite whichever suits the argument.
The Volcker Rule is US bank regulation, not a global law of nature. Non-bank prop firms, hedge funds and foreign entities all still take directional risk. The point isn’t that nobody speculates — it’s that “bank traders” specifically are mostly doing something else.
The variance model is deliberately simple. Independent trades, fixed win rate, 1:1 payoff. Real trades correlate and real edges drift. The direction is robust; the exact percentages are illustration.
The short version
The premise doesn’t survive checking. Over the decade to 2017 a basket of hedge funds picked by a professional returned about 36% against the S&P’s 125%, net of fees, and the manager conceded early — and the systematic data agrees, with 79% of active large-cap US funds lagging the S&P in 2025 and no category showing majority outperformance over 15 years — because active investors in aggregate must earn about the market return before costs and less after. Bank traders, meanwhile, have been barred from proprietary trading in the US since 2015 under the Volcker Rule; what’s exempted is market making, underwriting, hedging and customer business, almost none of which is directional speculation. The real gap is sample size: hold a 55% edge constant and it produces a losing period 41% of the time over five trades and essentially never over ten thousand. A market-making desk isn’t more certain than you — it samples its edge often enough that variance stops mattering, while you live where a genuine edge and a bad month look identical. Your advantages are the opposite of theirs: no mandate, no benchmark, no size problem, and permission to do nothing. (Long Bets)
Frequently asked questions
Do hedge funds beat the stock market?
In aggregate, no. Warren Buffett’s ten-year bet pitted an S&P 500 index fund against five funds-of-funds selected by Protégé Partners, holding over 100 underlying hedge funds between them. To 2017 the funds returned roughly 36% against the index’s 125% net of fees, and Protégé conceded eight months early. Individual funds do outperform, but the average does not, because active managers collectively hold something close to the market and pay far higher costs to do it.
Do banks still have proprietary trading desks?
Not in the US, for the most part. The Volcker Rule, Section 619 of Dodd-Frank, prohibits banking entities from trading principally to profit from short-term price movements on their own account, effective April 2014 with full compliance by July 2015. What’s still permitted is market making, underwriting, risk-mitigating hedging, government securities and trading on behalf of customers — mostly spread and fee businesses rather than directional speculation.
What advantage do professional traders actually have?
Mostly sample size and structure rather than better forecasting. Holding a 55% edge constant, a five-trade period shows a loss 40.6% of the time while a ten-thousand-trade period essentially never does — identical skill, different variance. Add order flow information, an external risk desk that cuts positions regardless of the trader’s opinion, a salary that doesn’t depend on this month, and diversification across many books, and the gap is structural rather than predictive.
What advantages do retail traders have?
Freedom, mainly. No mandate forcing you to stay invested, no benchmark to track, no redemption pressure after a bad quarter, and no size problem — your order doesn’t move the market, so small opportunities that are useless to a large fund are available to you. The biggest one is permission to take no trades at all. Nearly every professional structure punishes inactivity; yours doesn’t.
How many trades before I know if my strategy works?
More than most people assume. With a real 55% edge at 1:1, a 20-trade sample still ends negative about a quarter of the time and a 60-trade month about 18% of the time. You need somewhere in the region of 250 trades before a losing result becomes genuinely unlikely at under 5%. Any conclusion drawn from a single month is mostly noise, which is why judging a strategy monthly tends to produce changes that make it worse.
Related on this site: why most traders blow their accounts · what share of traders actually make money · why your journal invents rules · true cost rankings
Variance figures come from 40,000 simulated periods per trade count at a 55% win rate and 1:1 payoff, with independent trades; real results correlate, so treat them as illustrating the mechanism. Regulatory descriptions refer to US law and were checked against agency sources in September 2026. Nothing here is financial advice. Futures trading carries substantial risk of loss.
