Do Top Gainer And Loser Lists Make Money? A 13-Year Backtest
Published October 19, 2026 · 5 min read · Data: 2013–2025, about 40,000 trades per side
The Short Answer
No. Buying stocks off daily top-gainer or top-loser lists lost money after costs, and lost about twice as much as buying random stocks. Shorting the lists looked slightly positive on paper, but it does not survive higher costs.
On This Page
Why We Ran This
stkmrkt's own long picks were losing. Before trusting any rule for them, we tested whether "buy what is moving" has an edge at all, using the same data and the same costs for every version.
What We Tested
- The lists: each trading day from 2013 to 2025 we rebuilt the top 25 percentage gainers and top 25 percentage losers from stored daily prices.
- The trade: enter at the next day's open and hold five trading days.
- Four versions: buy the gainers ("chase"), short the gainers ("fade"), buy the losers ("fade") and short the losers ("chase").
- Hedged: each trade is adjusted for its market beta using SPY, so a rising market does not flatter the longs.
- Costs: estimated slippage on every trade, plus an estimated borrow fee on shorts.
- A control: the same trade on randomly chosen tradeable stocks (price at least $5, at least $2M traded per day). A list only has an edge if it beats this.
The Results
Average net return per trade, 2013–2025. The whiskers show the 95% range, so a bar whose whisker crosses zero is not distinguishable from no effect.
Bars are the average; whiskers are the 95% range. Gray bars are the random-stock controls.
| Trade | Trades | Avg Return | 95% Range |
|---|---|---|---|
| Buy the top gainers | 36,058 | -1.00% | -1.29% to -0.72% |
| Buy the top losers | 40,894 | -0.88% | -1.13% to -0.64% |
| Short the top gainers | 36,058 | +0.33% | +0.04% to +0.62% |
| Short the top losers | 40,894 | +0.21% | -0.03% to +0.46% |
| Control: random long | 40,894 | -0.42% | -0.52% to -0.32% |
| Control: random short | 40,894 | -0.25% | -0.35% to -0.14% |
Both long versions lose, and both lose about twice as much as the random-long control. Gainer and loser lists are full of lottery-ticket stocks that have already made their move, and buying them after the fact has been a bad habit in every period we looked at.
Why The Short Side Is Not A Strategy
Shorting the lists beat the control, and shorting the losers was ahead of random shorts in 13 of 14 years. That is still not a usable edge, for four reasons.
1. Costs decide it. The result flips sign as the cost assumption rises, and it disappears without the market hedge.
| Assumption | Avg Per Trade | 95% Range |
|---|---|---|
| Low costs, hedged | +0.40% | +0.15% to +0.64% |
| Middle costs, hedged (the main result) | +0.21% | -0.03% to +0.46% |
| High costs, hedged | -0.17% | -0.41% to +0.08% |
| Middle costs, not hedged | -0.11% | -0.43% to +0.23% |
2. A few trades carry it. Dropping the best 1% of trades turns the average into -0.26%. The single worst trade lost more than 350%, because a short can lose more than the amount invested.
3. It leans on 2024–2026. The edge over random shorts was small in most earlier years and large only recently:
Average net return per trade minus the random-short control, by signal year. 2026 is year to date.
4. It needs a bull market and a borrow. The trade averaged +0.29% when SPY was above its 200-day average and -0.25% below it. The test also assumes every stock can be shorted, and crashed or squeezed stocks are often impossible or very expensive to borrow.
The cleanest read is that the lists are a poor place to look for long ideas and an unreliable place to look for short ones.
What We Changed
stkmrkt's own day trades agreed with the backtest. Of 158 triggered day trades, the longs averaged -0.35% from entry to close, while the shorts averaged +1.54%. That is a small sample, so we treat it as consistent with the research, not as proof.
A stock that appears only on a gainer, loser or trending list is no longer eligible to be a long idea in our Day Trading Plan. A long needs support from a separate source, such as a scanner pick, a swing setup, a long-term screen or an earnings signal.
Limits Of This Test
How We Tested This, And What It Can't Tell You
- Survivorship. The stock universe is today's listings plus recently delisted names. Companies that went bankrupt earlier are mostly missing, which makes the short side look slightly worse than it was.
- Daily prices only. We cannot see the order of intraday highs and lows, and fills are idealised.
- Many cuts. Several splits on this page are exploratory. Trust the results that held in most years, beat the control and survived higher costs.
- Trade-weighted averages. Every figure averages individual trades, and the 95% ranges come from resampling blocks of dates, not single trades.
- Not advice. This is research on past data. See What Is Short Selling? for the risks of shorting.
Related Reading
This is research on past data and not a recommendation to buy, sell or short any security. Past results do not predict future returns.
Frequently Asked Questions
Does buying the day's top gainers make money?
No. In our 2013–2025 backtest, buying the top 25 gainers each day and holding five days lost 1.00% per trade on average after costs, which is worse than the random-stock control (-0.42%).
Does buying the day's top losers work?
No. Buying the top 25 daily losers averaged -0.88% per trade over five days after costs, again worse than random stocks.
Does shorting the top gainers or losers work?
Slightly on paper, but not as a strategy. Shorting the gainers averaged +0.33% and shorting the losers +0.21% per trade, yet the losers result turns negative at higher costs, depends on a few large trades, mostly comes from 2024 to 2026, and ignores that many of these stocks cannot be borrowed.
Is this investment advice?
No. It is research on historical data and not a recommendation to buy, sell or short any security.