I run three trend-following strategies on S&P futures. Two are momentum-based, one is a breakout system. All three have been getting chopped to pieces since January.
I thought it was me. Bad entries, late exits, position sizing too aggressive. Spent two weekends backtesting different parameter sets. Nothing helped.
Then I pulled the actual data on intraday reversals and it clicked.
The S&P 500 has reversed intraday direction 28 times in the last 3 months. That's the highest count since 2015 and the second highest since the 2008 financial crisis peak of 35. Nearly half of all trading sessions in this window erased the opening gap or opening move entirely.
My systems weren't broken. The regime changed.
What an "Intraday Reversal" Actually Means
I'm defining it the same way most quant desks do: a session where the market opens in one direction (gap up or strong first-30-min move), then closes in the opposite direction by end of day. Gap up, close red. Gap down, close green.
28 out of roughly 63 trading sessions. That's a 44% reversal rate.
For context, the long-run average is somewhere around 20 to 25%. We're nearly double that. And the reversals aren't small. Several have been full-percentage-point swings from the open.
If you're running any system that buys strength or sells weakness in the first hour, you've been donating money to mean-reversion traders for three months.
Why This Is Happening
My best guess, and it's just a guess based on what I see in the data:
Headlines are driving the tape more than fundamentals. Geopolitical news, tariff threats, Fed speculation. The market gaps on overnight sentiment, then reverses once the actual flow starts during regular hours.
The implied daily move on the S&P is around 1.77% right now, which is the highest since April 2025. But actual sessions where the market moved more than 1.75% in one direction? Only 2 in the last 3 months. The market is pricing in massive moves that aren't materializing in a sustained direction. It's moving a lot, but going nowhere.
That's a chop regime. And chop kills trend followers.
How I'm Adapting
I've made three changes to my systems since I saw this data.
Delayed entries. I stopped entering in the first 30 minutes. My breakout system now waits until 10:30 AM ET before confirming a directional bias. This single change cut my reversal losses by roughly 40% in backtesting against the last 3 months.
Tighter profit targets. In a trending regime, I let winners run. In a reversal regime, I take profits at 60 to 70% of the average true range instead of trailing. The math is simple: if half the sessions reverse, trailing stops just give back the profit.
Added a mean-reversion overlay. When the opening 30-minute candle is in the top or bottom 20% of the prior day's range AND volume is below average, I fade the open instead of following it. This is the opposite of what my trend systems do. Running both in parallel has smoothed the equity curve significantly.
None of this is complicated. The hard part was admitting the regime had shifted instead of blaming my execution.
The Bigger Picture for Funded Traders
If you're trading a funded account, drawdown rules don't care whether the market is trending or chopping. You get the same max loss limit either way.
That's the problem. A system calibrated for a trending market will blow through its drawdown limit in a reversal regime before you realize what happened. I've seen it happen on two of my funded accounts this quarter. Not blown, but uncomfortably close.
The fix is recognizing the regime early. 28 reversals in 3 months isn't hidden information. It's right there in the price data. If you're not tracking reversal rate as a metric, add it. It takes about 20 lines of Python.
Here's the rough logic:
# Count intraday reversals over N sessions
reversals = 0
for session in last_n_sessions:
open_direction = 'up' if session.open > session.prev_close else 'down'
close_direction = 'up' if session.close > session.open else 'down'
if open_direction != close_direction:
reversals += 1
reversal_rate = reversals / len(last_n_sessions)
# If reversal_rate > 0.35, you're in a chop regime
When the rate crosses 35%, I reduce position size by half and switch to mean-reversion entries. When it drops below 25%, I go back to trend following at full size.
Simple. Not sexy. But it's kept my funded accounts alive this quarter while pure trend followers are getting stopped out daily.
What Comes Next
Historically, extreme reversal regimes don't last forever. The 2008 peak of 35 reversals eventually gave way to one of the strongest trends in market history. The 2015 spike preceded a breakout that ran for months.
The current chop will resolve. It always does. Some catalyst, probably geopolitical or Fed-related, will pick a direction and the trend followers will get paid again.
Until then, I'm fading opens, taking quick profits, and keeping my drawdown tight. The data says this is the wrong time to be a hero with directional bets.
Top comments (0)