In this study, I analyze 175 post-earnings momentum setups to find out what the biggest losing trades had in common, which variables showed up most often, and whether those patterns could be turned into a practical trading filter. You’ll see why 10%+ first-hour moves without 4-hour breakout confirmation stood out as a weak setup, how removing them changed win rate, profit factor, and net P&L, and the refined rule I’m now testing to reduce avoidable losses.
Post-earnings momentum trading sounds simple enough: a company reports earnings, the stock makes a strong initial move, and you trade in the direction of that momentum.
The problem is that some of the most convincing-looking earnings moves can also turn into the ugliest trades. In my dataset of post-earnings setups, the 10 largest losing trades all started with first-hour moves of at least 10%.
A whopping eight of those 10, or 80% of them, failed to produce a corresponding 4-hour breakout.
That led me to a much bigger question than simply identifying my worst trades:
Was there a measurable characteristic that could have warned me not to take them in the first place?
And get this… there was!
When I isolated every setup with a 10% or larger first-hour move but no 4-hour breakout, that entire 48-trade group was unprofitable.
Removing those setups increased historical net profit while reducing the number of trades, improving the win rate, and eliminating most of the largest drawdowns in the study.
10 Worst Post-Earnings Momentum Setups
Every one of the 10 largest losers had a first-hour move of at least 10%, while 8 of 10 failed to produce a 4-hour breakout.
| # | Stock | Direction | 1-Hour Move | 4H Breakout | Passive EOD | With -15% Stop |
|---|---|---|---|---|---|---|
| 1 | DOCS | Long | +68.37% | No | -21.45% | -$150.00 |
| 2 | CNXC | Short | -24.94% | Yes | -19.64% | -$150.00 |
| 3 | EGHT | Long | +13.16% | No | -17.50% | -$150.00 |
| 4 | MDB | Long | +22.07% | No | -16.68% | -$150.00 |
| 5 | INOD | Long | +14.04% | No | -16.50% | -$150.00 |
| 6 | AKAM | Long | +10.01% | No | -15.20% | -$150.00 |
| 7 | FISV | Short | -12.76% | No | -14.69% | -$146.90 |
| 8 | PHR | Long | +20.61% | No | -12.73% | -$127.30 |
| 9 | HPE | Long | +45.97% | Yes | -11.87% | -$150.00 |
| 10 | CRDO | Short | -13.50% | No | -11.74% | -$150.00 |
Quick Answer: How Successful Is Momentum Trading?
Momentum trading can be profitable, but the quality of the setup matters substantially. In my sample of 175 post-earnings momentum setups, trading every setup with a fixed $1,000 position, a 15% maximum stop loss, and a next-day end-of-day exit produced $4,799.30 in simulated net profit. After requiring a 4-hour breakout whenever the first-hour earnings move was 10% or greater, net profit increased to $5,309.20 across only 127 trades, while the win rate improved from 60.0% to 67.7% and the number of 15% stop losses fell from 12 to just three.
Key Post-Earnings Momentum Statistics
- 10 of the 10 worst trades in my dataset had first-hour earnings moves of at least 10%.
- 8 of the 10 worst trades did not produce a 4-hour breakout.
- The worst 10 trades had an average absolute first-hour move of approximately 24.5%.
- 48 setups had a first-hour move of at least 10% without a 4-hour breakout.
- Those 48 setups generated approximately -$509.90 using a $1,000 position, 15% stop loss and passive next-day EOD exit.
- The ≥10% / no-4-hour-breakout group had a win rate of only 39.6%.
- Trading every setup produced approximately +$4,799.30 across 175 trades.
- Applying the new 4-hour confirmation rule produced approximately +$5,309.20 across 127 trades.
- The refined strategy increased average return per trade from approximately 2.74% to 4.18%.
- The refined strategy improved the historical win rate from 60.0% to 67.7%.
- The number of 15% catastrophic stop losses fell from 12 to three.
- Jegadeesh and Titman found significant positive momentum returns over 3- to 12-month holding periods in their landmark momentum study.
- Canadian research on post-earnings announcement drift found a long-short earnings-surprise strategy generated more than 6% excess return over the following 60 days.
How I Analyzed the Worst Post-Earnings Momentum Trades
My dataset currently contains 175 completed post-earnings setups.
For this study, the simulated position size was fixed at $1,000 per trade. Entry occurred at the close of the first hourly earnings candle, with the position taken in the direction of that initial move.
I then used the stock’s maximum adverse excursion, or MAE, to model risk. If price moved more than 15% against the position, the trade was capped at a 15% loss, or $150 on a $1,000 position.
If that stop was not reached, I used the percentage change from the hourly candle close to the end of the following trading day.
The goal was not to simply delete the worst trades after seeing the results.
That would be hindsight.
The real goal was to identify a characteristic that existed before the trade was taken and could therefore potentially become an objective filtering rule.
Below are my findings.
Large First-Hour Moves Needed 4-Hour Confirmation
<10% move: eligible • ≥10% move: require a 4-hour breakout
Eight of the 10 Worst Trades Had No 4-Hour Breakout
Among the 10 largest losers, eight did not produce a 4-hour breakout.
Only two — HPE and CNXC — had higher-timeframe confirmation and still ultimately failed. That means 80% of the worst trades shared the same combination:
A large first-hour move, but no corresponding 4-hour breakout.
The hourly breakout by itself was much less useful. Six of the worst 10 actually had hourly breakouts. In other words, price could look technically strong on the first-hour chart and still fail badly afterward.
The 4-Hour Breakout Was the Key Divider
The problem appeared when that momentum did not extend into the larger timeframe, which makes intuitive sense.
The hourly chart tells me that the earnings release created a meaningful immediate reaction. The 4-hour breakout tells me whether that reaction is strong enough to push through a more substantial technical level.
For smaller moves, that extra confirmation may not be necessary. For very large moves, though, the data suggests it becomes a much more important qualifying feature.
The Worst Trades Started With Surprisingly Large Moves
One of the clearest patterns in the data was also one of the most surprising.
Every one of the 10 largest losing trades started with a first-hour earnings move of at least 10%.
Not most of them.
All 10.
The Biggest Losers Started With Bigger Moves
The size of those moves was also much larger than normal. The 10 worst trades had an average absolute first-hour move of approximately 24.5%, compared with about 12.5% across the full dataset.
Five of the 10 worst trades moved more than 20% during the first hour after earnings.
Some were extreme. DOCS moved more than 68%, HPE nearly 46%, CNXC roughly 25%, MDB more than 22%, and PHR more than 20%.
At first glance, that seems backwards.
Momentum traders are generally looking for strength. A stock moving 25% after earnings looks much stronger than one moving only 5% or 6%.
But the data suggests that an extremely large first-hour move can also create additional risk.
The reason may be fairly simple.
When a stock has already moved 20%, 30%, or 40% in a single hour, a large portion of the immediate earnings reaction may have already occurred.
The stock can become stretched away from its previous price range. Traders who entered earlier may start taking profits. Short sellers may cover their positions, while new traders begin chasing the move after much of the initial momentum has already happened.
That does NOT mean large earnings moves are automatically bad trades.
Some of the strongest setups in the dataset also started with very large first-hour candles.
The difference appears to be whether that initial move receives additional technical confirmation.
In other words, a 6% move may still have room to develop without immediately breaking a major higher-timeframe level. But when a stock has already moved 15%, 20%, or more, I want stronger evidence that buyers or sellers are still in control.
The larger the first-hour move became, the more important the 4-hour breakout appeared to be.
That led directly to the second major finding in the study.
Strong Fundamentals Did Not Protect the Worst Trades
I also wanted to know whether the largest losses were simply earnings setups with weak or contradictory fundamentals.
They were not.
Six of the 10 worst trades had technical and fundamental signals that generally agreed. Several had positive earnings results, revenue results, or forward guidance supporting the direction of the initial price reaction.
That means fundamental alignment by itself did not eliminate the risk of reversal.
Good Earnings Did Not Guarantee a Good Trade
Fundamentals can explain why the stock moved. They can help confirm that the market reaction is based on meaningful new information rather than random volatility.
But fundamentals do not necessarily tell me whether the stock has already moved too far, too quickly.
Price still has to confirm that traders are willing to continue pushing the stock in the same direction.
In this study, the 4-hour chart provided more useful information about the largest losing trades than fundamental alignment alone.
Sector Was Not the Main Problem Either
Technology and information technology stocks appeared several times among the worst trades, but those sectors also make up a significant portion of the overall dataset.
The losing trades were not isolated to one industry.
The Worst Trades Were Not Concentrated in One Sector
Consumer discretionary, healthcare, communication services, financial stocks and other groups also appeared among large reversals. That makes sector a considerably weaker explanation than first-hour move size and 4-hour confirmation.
There was also some evidence that elevated short float may contribute to violent reversals. Among the subset of worst trades where short-float data was available, several had double-digit short interest.
But the sample is too small for me to turn that into a rule. So for now, I view short float as context rather than a primary filter.
The Worst Trades Were Primarily Reversal Trades
Not surprisingly, the largest losers generally reversed after the initial earnings move.
But there is an important distinction here… Saying that losing momentum trades reverse is not useful by itself because I cannot know in advance whether a stock will reverse.
The useful finding is what happened before the reversal. In many of the worst examples, the sequence was:
Large first-hour earnings move → no 4-hour breakout → next-day reversal.
Don’t Predict the Reversal — Identify the Setup
That gives me something measurable.
Instead of attempting to predict a reversal, I can simply refuse to participate when the initial move is already unusually large and the higher timeframe fails to confirm it.
That turns an observation about losing trades into an objective trading rule.
My Refined Post-Earnings Momentum Rule
Based on these results, I tested the following version of the strategy:
- First-hour move below 10% → eligible.
- First-hour move of 10% or more → require a 4-hour breakout.
- 10%+ first-hour move without 4-hour confirmation → no trade.
The 15% catastrophic stop remains unchanged. Positions that do not hit the stop are held until the end of the following trading day.
This is important because I am not requiring a 4-hour breakout for every trade.
The dataset has repeatedly shown that smaller first-hour moves can produce clean continuation. Requiring higher-timeframe confirmation on all of them could eliminate profitable setups unnecessarily.
The filter only becomes mandatory once the first-hour move reaches 10%.
Before and After: What Happened When I Applied the Rule?
Trading every completed setup in the dataset produced the following simulation:
Trading Every Setup
Results using a $1,000 position, 15% catastrophic stop, and passive next-day EOD exit.
The strategy was already profitable. That is what makes the next result more interesting. After eliminating every ≥10% first-hour move that lacked a 4-hour breakout:
Results After Applying the 4-Hour Confirmation Rule
Setups below 10% remained eligible, while 10%+ first-hour moves required a confirmed 4-hour breakout.
The strategy took 48 fewer trades, a reduction of approximately 27%. Yet simulated net profit increased by approximately $509.90, or 10.6%.
Average expectancy increased from 2.74% to 4.18% per trade.
The win rate increased by 7.7 percentage points.
And perhaps most importantly from a risk-management perspective, the number of trades reaching the catastrophic 15% stop fell from 12 to three.
That is exactly what I want from a strategy filter. I do not want to add rules simply because they make the backtest look more sophisticated. A useful filter should remove a disproportionately weak group of trades while preserving — or ideally improving — total strategy performance.
That is exactly what happened here.
Fewer Trades Produced More Profit
One of the easiest traps in momentum trading is believing that more opportunities must mean more profit.
My results showed the opposite.
The 48 setups eliminated by the new rule were not merely lower-quality trades that reduced the average slightly.
As a group, they lost money.
By refusing to trade them, the simulated strategy made more money while using capital less frequently.
Fewer Trades, Better Results
This also increases the efficiency of each trade. Instead of generating approximately $27.42 of simulated profit for every $1,000 setup across the complete sample, the refined strategy generated approximately $41.80 per trade.
The goal of backtesting should not necessarily be to find more stocks to trade. It should be to deploy capital into the subset of setups with the strongest historical expectancy.
Why Momentum Still Needs Risk Management
None of this means the new rule eliminates losing trades. It does not.
Three eligible setups still exceeded 15% MAE and triggered the catastrophic stop.
That is exactly why I continue to separate entry filtering from risk management.
The filter attempts to eliminate a historically weak type of setup. The stop protects against the trades that pass every filter and still fail.
Academic momentum research points to the same broader reality.
Momentum has historically generated positive average returns, but Daniel and Moskowitz found that momentum strategies can experience infrequent but severe crashes, particularly during sharp market rebounds following periods of stress.
Better Entries Still Need a Maximum-Loss Rule
There are also practical frictions to consider. Research examining post-earnings drift has found that some of the largest theoretical opportunities occur in less-liquid stocks, where trading costs and market impact can absorb a substantial portion of paper profits.
A profitable backtest therefore should never be interpreted as guaranteed future trading performance. It should be treated as evidence for a hypothesis that can now be tested prospectively.
Conclusion – How to Eliminate the Worst Trades From Your Trading
Studying winning trades tells me what can work. Studying the worst trades tells me what can destroy an otherwise profitable strategy.
The biggest lesson from this study is not that I should avoid large earnings moves completely. The distinction appears to be confirmation amongst higher timeframes.
My current rule is therefore simple:
- Below 10% first-hour move: the setup remains eligible.
- 10% or greater: require a 4-hour breakout.
- No 4-hour breakout after a 10%+ move: do nothing.
That single filter removed 48 trades from my historical sample, increased simulated net profit from approximately $4,799 to $5,309, increased the win rate from 60.0% to 67.7%, and reduced catastrophic stop losses from 12 to three.
Those numbers do not prove the rule will work indefinitely. But this is now something that I will track going forward.
If future earnings setups continue showing the same relationship — particularly the weakness of 10%+ first-hour moves without higher-timeframe confirmation — then the evidence supporting the rule becomes considerably stronger.
Trading does not require eliminating every losing trade. It requires eliminating enough bad bets that the remaining opportunities have positive expectancy.
And today, I’ve discovered that, in my post-earnings momentum dataset, the 4-hour breakout appears to be one of the clearest ways I have found so far to do exactly that.
If you want to go deeper:
- Explore the Trading Statistics Hub to understand how different sectors behave across market cycles
- Study real setups inside the Trade Reviews section
- Learn the framework behind high-probability setups in the Post-Earnings Momentum Strategy
This is how you turn raw market data into repeatable trading edge.
More Trading Statistics…
Frequently Asked Questions
What is post-earnings momentum trading?
Post-earnings momentum trading attempts to trade continued price movement after a company releases earnings. The basic idea is that earnings information may not always be completely incorporated into the stock price immediately, allowing price to continue moving in the direction of the initial reaction.
This concept is related to the well-documented post-earnings announcement drift anomaly, in which stocks have historically shown a tendency to drift in the direction of earnings surprises after the announcement.
How successful is momentum trading?
Momentum trading has substantial historical academic support, but results depend heavily on strategy design, holding period, transaction costs and risk management. Jegadeesh and Titman documented significant momentum returns over 3- to 12-month holding periods, while other studies have documented continued returns following earnings surprises.
In my own 175-trade post-earnings dataset, the unfiltered strategy produced approximately $4,799 in simulated net profit using $1,000 positions. The refined version produced approximately $5,309 while taking 48 fewer trades.
What was the most common characteristic among the worst trades?
All 10 of the largest losing trades had first-hour earnings moves of at least 10%.
Eight of the 10 also lacked a 4-hour breakout, making the combination of a large initial move and missing higher-timeframe confirmation the strongest common pre-trade variable I found.
Should I avoid stocks that move more than 10% after earnings?
Not necessarily.
In my dataset, large earnings moves that also produced a 4-hour breakout remained profitable as a group. The historically weak category was specifically 10%+ first-hour moves without a 4-hour breakout.
Why does the 4-hour breakout matter?
A first-hour breakout shows that the earnings announcement created immediate momentum.
A 4-hour breakout provides additional evidence that the move is strong enough to break a higher-timeframe technical level rather than simply producing a short-lived spike.
My dataset suggests that this confirmation becomes particularly useful as the size of the initial earnings move increases.
Did strong earnings fundamentals prevent losing trades?
No.
Several of the largest losing trades had fundamentals that agreed with the initial direction of price. Strong EPS, revenue or guidance results did not guarantee that momentum would continue.
That suggests technical confirmation and fundamental analysis may provide different pieces of information.
Does the 15% stop loss improve returns?
The primary purpose of my 15% stop is not to maximize historical return. It is a catastrophic risk limit that caps the maximum planned loss on a $1,000 setup at approximately $150.
Under the refined strategy, only three of 127 historical trades triggered that stop, compared with 12 when every setup was traded.
Can this strategy eliminate every losing trade?
No, and that is not the goal.
No technical filter can reliably eliminate every future loss. The objective is to identify groups of trades with poor historical expectancy and avoid them while maintaining exposure to the setups that have performed better.
Is this rule proven?
No.
This is a finding from my historical dataset and should be treated as a hypothesis rather than a universal market rule. Because the filter was discovered by analyzing past results, the most important test is whether it continues to improve performance on new post-earnings setups that were not used to create the rule.
For me, that is now the next stage of the experiment.
References
Chordia, T., Goyal, A., Sadka, G., Sadka, R., & Shivakumar, L. (2009). Liquidity and the post-earnings-announcement drift. Financial Analysts Journal, 65(4), 18–32. https://doi.org/10.2469/faj.v65.n4.3
Chudek, M., Truong, C., & Veeraraghavan, M. (2011). Is trading on earnings surprises a profitable strategy? Canadian evidence. Journal of International Financial Markets, Institutions and Money, 21(5), 832–850. https://doi.org/10.1016/j.intfin.2011.06.004
Daniel, K., & Moskowitz, T. J. (2016). Momentum crashes. Journal of Financial Economics, 122(2), 221–247. https://doi.org/10.1016/j.jfineco.2015.12.002
Fink, J. (2021). A review of the post-earnings-announcement drift. Journal of Behavioral and Experimental Finance, 29, 100446. https://doi.org/10.1016/j.jbef.2020.100446
Jegadeesh, N., & Titman, S. (1993). Returns to buying winners and selling losers: Implications for stock market efficiency. The Journal of Finance, 48(1), 65–91. https://doi.org/10.1111/j.1540-6261.1993.tb04702.x
Leave a Reply