">
AI Day Trader
Day trading strategy

Intraday Momentum Strategy

Research on stock index funds has found a pattern in the trading day: the direction the market moves in its first half hour tends to carry into its last half hour. The intraday momentum strategy trades the final 30 minutes on that basis.

The StreetAlpha AI Day Trader runs this strategy live, and logs every call before its outcome.
Watch it live

What is intraday momentum?

The opening and the close are the two busiest parts of the session. The idea behind this strategy is that they are linked. If the market rises from yesterday's close through the first half hour, it has been somewhat more likely than not to rise again in the final half hour, and the same for declines.

Several explanations have been offered. Some investors only act on the morning's news late in the day. Others have to adjust or hedge positions before the close, and their trading follows the direction the day has already taken. Whatever the cause, the effect is small. It is a slight tilt in the odds, not a reliable forecast.

Because the edge is modest, the strategy is simple. It makes one decision per ticker at 3:30 PM Eastern and holds until the closing bell, with a stop in case the last half hour goes sharply the other way.

The rules StreetAlpha uses

Many versions of this strategy exist. These are the exact rules the AI Day Trader applies, on 5-minute bars, to SPY, QQQ, IWM and two stocks chosen each morning.

The morning moveThe price at 10:00 AM Eastern compared with the prior day's close.
ThresholdThat move must be at least 0.15% in either direction. A smaller move is treated as no signal.
Decision time3:30 PM Eastern.
DirectionThe same as the morning move: long if the first half hour was up, short if it was down.
Held untilThe close at 4:00 PM Eastern.
InvalidationThree times the average range of a 5-minute bar away from the entry.
TargetTwice the distance to the invalidation. It is set deliberately far, so in most cases the call simply runs to the close.
LimitsOne per ticker per day.

A worked example

First half hourLast half hourPrior closeInvalidationEntry9:3010:003:304:00

Say an index fund closed yesterday at 400.00. At 10:00 this morning it is at 401.00, up 0.25%. That is above the 0.15% threshold, so the morning counts as an up move.

At 3:30 the fund is trading at 402.20. A long is logged at that price. Its 5-minute bars have been averaging a range of 0.30, so the invalidation is three times that below the entry, at 401.30. The target is twice that distance above, at 404.00.

The call is held to the close at 4:00. If the fund finishes at 402.90, that is the exit. The call only closes early if price reaches 401.30 or 404.00 first.

The numbers in this example are made up to show the mechanics. They are not a real trade.

Live results

Every intraday momentum setup the rules find is recorded when it happens and followed to its outcome, whether or not the AI takes it. Nothing is added or removed afterwards.

Sessions tracked1
Setups the rules found4
AI took1
AI passed3

This is a young record. Tracking began on October 5, 2026. With 4 graded setups so far, these figures can swing a long way on a single trade and say little yet about how the strategy performs.

GradedProfitableTotal result
Every setup taken41 of 4−1.16%
The AI's calls10 of 1−0.92%

Of the 4 setups that have played out, 1 reached the target, 2 were invalidated and 1 ran out of time. By direction: longs 0 of 2 profitable, −1.08%; shorts 1 of 2 profitable, −0.09%.

The AI's most recent intraday momentum calls

DateTickerSideHow it endedResult
2026-10-05PBRLongInvalidated−0.92%

Tracked since October 5, 2026, through October 5, 2026. Results are the percent move from entry to exit, with every call the same size, added up, after an assumed cost of 0.02% per trade. "Every setup taken" counts each setup the rules found whether or not the AI took it. Results are hypothetical and no trades are placed.

When it tends to work, and when it does not

Tends to work when

  • The day has kept the direction it set in the morning.
  • It is a busy session with above-average volume, when the effect has been found to be stronger.
  • The move is market-wide, with the main index funds all pointing the same way.

Tends to fail when

  • The morning move has already been fully reversed by mid-afternoon.
  • A headline late in the day changes the picture.
  • The morning move was small and the day has been quiet.

How the AI decides whether to take one

When the rules produce an intraday momentum setup, the AI is told it is a bet that the last half hour follows the first. It checks whether the day still looks the way the morning said it would.

What counts in its favor

  • Price at 3:30 still on the same side of the open and of VWAP as the morning move
  • Net options premium moving in the same direction
  • The index ETFs aligned with each other

What counts against it

  • A morning move that has fully reversed by 3:30
  • A late headline
  • Options flow or large block levels pointing the other way

Each decision is published with its reasoning. When the AI passes, the page later shows what taking the setup would have returned, so its judgement can be checked.

Common mistakes

Expecting large moves. The last half hour of an index fund usually moves a fraction of a percent. This is a small, frequent effect.

Applying it to any stock. The research behind it is on broad index funds. Individual stocks are driven much more by their own news.

Ignoring what happened in between. If the market has spent the afternoon undoing its morning move, the premise is weaker.

Holding past the close. The pattern is about the final half hour of the same day. It says nothing about the next morning.

Questions

What is intraday momentum?

It is the tendency, found in research on stock index funds, for the market's return in the first half hour of the trading day to predict the direction of its return in the last half hour.

Does the first half hour predict the last half hour?

Studies have found that it has, modestly, on broad index funds over long periods. The effect is small and far from certain on any single day, and patterns like this can weaken once they are widely known.

Why would the morning predict the close?

Suggested reasons include investors who act on morning news only late in the day, and traders who must hedge or rebalance before the close and do so in the direction the day has already moved.

Which securities does it apply to?

The research is on broad index funds such as those tracking the S&P 500. StreetAlpha applies the rule to its index ETFs and its daily stock picks, and reports the results for each.

Keep going

Watch the AI Day Trader live, read the frequently asked questions, or see the other strategies it runs: Opening Range Breakout, VWAP Pullback, Failed Breakout Reversal, Gap and Go.