Why 90% of Retail Investors Keep Losing to Algorithms — And How to Stop Being Exit Liquidity

Most retail losses aren’t caused by “evil bots,” but by playing a speed-and-microstructure game you can’t win—and by overtrading, poor execution, and leverage. This guide explains what algorithms actually do and gives a

TL;DR

• “90% lose” holds true in the context of risky leveraged products (options, futures, CFDs) and in high-frequency retail trading, but it is not a hard and fast rule for long-term investing. (think warning label instead of fact about every investor) (esma.europa.eu)
• When we talk about algorithms, we are not only referring to “AI traders” but more market making algorithms, execution algorithms, and arbitrage systems that are designed to act faster than humans.
• Retail doesn’t lose out mainly for reasons of: overtrading, leverage/short-dated position betting, poor order execution (spreads + slippage), and emotionally chasing crowd moves.
• If you don’t want to be exit liquidity, stop competing based on speed. Compete based on time horizon, fees, diversification, and disciplined execution.
• Implement a repeatable process: trade less, always use limit orders in liquid markets, avoid the open/close for discretionary trades, and either size your positions small enough to survive noise, or measure your executions using public disclosure (Rule 605/606). (sec.gov)

What “losing to algorithms” means (and what it doesn’t)

When people say “algorithms are beating retail,” they likely mean you are throwing trades against professional market participants whose systems are built to execute/make decisions in microseconds, manage order-flow predictively, and risk manage—three areas where humans consistently have no edge. That doesn’t automatically mean the market is “rigged.” In fact, a large body of research finds that certain kinds of algorithmic trading (especially automated market making) can improve liquidity and narrow bid-ask spreads—often reducing explicit trading costs. (link.springer.com)

The problem is that many retail strategies still lose money after you add the things that algorithms are best at exploiting: (1) predictable behavior, (2) poor execution, (3) overtrading, and (4) leverage.

About that “90%” number: use it as a warning label, not a debate trophy. You’ll see variations of “90% of retail investors lose.” As a blanket statement about all investors, it’s not reliable—many long-term, diversified investors do fine. But in leveraged, short-term products, regulators have documented very high loss rates. For example, European regulators have cited evidence that roughly 74%–89% of retail accounts typically lose money trading CFDs (contracts for difference). (esma.europa.eu)

So if your “investing” is actually frequent trading—especially with leverage—“90% lose” is a reasonable mental model for the default outcome unless you can prove you have a real edge, strong risk controls, and excellent execution.

Why retail keeps losing in algorithmic markets (the 7 mechanisms)

1. Overtrading turns small disadvantages into guaranteed underperformance
   A tiny, repeatable disadvantage is all the market needs. Classic studies of individual investors show that more active traders frequently underperform, which is consistent with a compounding effect of trading costs and behavioral blunders. (finance.martinsewell.com)
2. Leverage + short-dated options = a “negative expectancy” treadmill
   When you leverage, you not only amplify losses but you also can get forcibly liquidated, and often compound structural headwinds (spread, funding, decay, vol crush). This is why significant loss rates show up in a lot of leveraged (CFD) retail products, and many day-trading datasets. (esma.europa.eu)
3. Your execution quality is much worse than you think – spreads + slippage
   A lot of retail, perhaps innocent, butterfly around “entry” for hours then hit a market order at some of the worst possible moment… The SEC in their “Trading 101” educational resource make clear this idea – and moreover, that the trade off is between using a “market” (which has advantages of speed), and a “limit” (which has advantages of price control). The faster a market, sometimes that difference means it’s a plan vs. panic fill… (sec.gov)
4. It’s the “toxic flow” window
   Some times, there’s simply more “toxic flow” in terms of adverse selection (i.e. you are likely trading against better informed flow): the Open, news events / earnings, and sudden spikes in volatility (availability and turn.). Professionals widen spreads and lift liquidity at the times they are most nervous. Retail, oftentimes, does the opposite: they are “more” likely to trade in times of maximum emotional exuberance.
5. You’re adopting a “crowded narrative”
   Retail “comes together” on the same signal almost: “It broke out, finally, everyone saw it”, “Viral catalyst”, “This can’t go lower” and “It’s going to the moon”… A fast participant therefore doesn’t have to mind-read. They just need pattern-matching and inventory management.
6. You confuse being “right” with having a tradable edge
   The market can agree with your thesis and still take your money. If your average loss is larger than your average win (after costs), or if you can’t survive normal volatility, you can be “correct” and still go broke.
7. Hidden incentives and routing complexity can affect outcomes
   In the U.S., brokers and market centers operate under “best execution” expectations (FINRA Rule 5310) and various SEC disclosure regimes. But “best execution” is not the same as “best possible outcome for every single trade,” and the ecosystem includes wholesaling/internalization, routing decisions, and potential conflicts you should at least understand. (finra.org)

“Exit liquidity” explained in plain English

“Exit liquidity” means your buying or selling is supplying the other side of a smarter participant’s trade at a time that benefits them. You become the convenient counterparty when a pro wants to reduce risk, unwind inventory, or fade an emotional move.

How to stop being exit liquidity: a practical 10-step plan

1. Pick the game you’re actually playing (investing vs. trading). If you don’t have time to measure performance and execution, default to long-term, diversified investing.
2. Lower your trade frequency first. Before changing indicators, cut the number of discretionary trades you place and track whether your results improve.
3. Ban “forced urgency” trades: no chasing moves that already happened. If you missed it, you missed it.
4. Use limit orders by default for liquid stocks/ETFs, especially in fast conditions. Use market orders only when certainty of execution matters more than price. (sec.gov)
5. The first and last 10–15 minutes of the session? Avoid them for discretionary entries/exits unless you have a tested reason. Those windows typically have wider spreads and more price discovery noise.
6. Trade liquid instruments. Illiquidity is where slippage, gaps, and bad fills become “invisible fees.”
7. Size positions so you can survive normal volatility. If a normal intraday move can stop you out you’re trading too large (or your timeframe is mismatched).
8. Stop avoiding position-sizing logic by short-dated options. If you need the leverage of cheap money to “make it worth it” it probably isn’t worth it.
9. Measure execution quality, not just P&L. Compare your fill against the NBBO midpoint at the time you placed the order. Use this as a simple proxy for whether you ‘paid the spread’.
10. Audit your broker routing and disclosures at least once. Read the FARs and compare and contrast with peer brokers. Use public Rule 606 order-routing reports to understand where non-directed orders go, and use Rule 605 execution-quality disclosures where available to compare and contrast execution outcomes. (sec.gov)

A simple ‘retail edge’ framework: stop competing on speed

You can’t out-HFT an HFT, but you can choose edges that are structurally unavailable (or just unattractive) to many short-horizon algorithms—like longer holding periods, disciplined rebalancing, and minimizing costs and taxes.

• Time horizon edge: weeks/months/years instead of minutes.
• Fee edge: low turnover, fewer spreads paid, fewer mistakes.
• Behavior edge: pre-committed rules that curb FOMO/panic.
• Diversification edge: broad exposure vs single-name hero trades.

How to tell you’re actually improving (not just sholing up on facts):

1. Keep track of how you’re performing net of everything including commissions and spreads (estimate the latter too), how much borrow costs you incur, any theta/IV effects from optionality, and taxes possibly depending on your situation.
2. Keep a record of every trade that includes your thesis/explanation for entering, what sort of entry you did (market/limit), what time of day it was, what’s the liquidity of what you traded (spread e.g. of what tradeable market to volume), and what exit you plan on.
3. Tack on at least two execution metrics to your way of measuring yourself – did you get “price improvement” vs NBBO? If so, what’d the “effective spread” you paid for (as compared to what you got for) the “next best bid and offer” be? Think Rule 605 disclosures from the SEC are made to help render execution quality among brokers more “comparable” across market centers.
4. Then look into broker routing disclosures for your own broker. Order-routing disclosures according to Rule 606 should weigh heavily in your mind when picking a broker: 606 also lets customers get a few customer specific reports concerning ‘not held’ orders if requested.
5. Make a ‘minimum viable strategy’ test whereby you check if you can still stay profitable (or still meet your goal i.e get through one bar) cutting your trade frequency in half. Neither scoring higher nor averaging closer to the mean if in a standard-setting game means your strategy is vulnerable to depending on luck too much or could be overfit.

Where market structure is headed (and don’t regurgitate talking points when they talk of heady concerns):

U.S. equity market structure debates change over time. June 17, 2025, the SEC indicates the Commission withdrew a number of proposed rules on its Order Competition Rule page so if someone is arguing as if that proposal is pending still at the date you’re reading, it’s old news. Here’s the link: (sec.gov). Rule 665 amendments were adopted by the SEC too. They’re designed to be modernized and widened i.e. you can tell basically that much has changed there’s a need for this redesign or widening in brightness to come over/shine. (sec.gov).

Common mistakes that keep retail trapped (quick checklist)

• You measure success by win rate instead of expectancy (average win × win rate minus average loss × loss rate).
• You use stops to manage fear, not risk (tight stops + large size = death by noise).
• You trade the most during the most emotional market conditions.
• You switch strategies after drawdowns—right before mean reversion would have helped.
• You treat “commission-free” as “cost-free” and ignore spreads/slippage.
• You never audit execution quality or routing disclosures.

FAQ