Momentum Scalping vs Momentum Trading: How Each Works, and the SPY Data

Michael Whitman · Trading since 2017, full-time since 2020 · YouTube creator, 137K subscribers · Published 2026-09-08 · Updated 2026-11-01 · Reviewed 2026-11-01

Momentum scalping is jumping on a strong, fast move and getting out seconds or minutes later; momentum trading rides a move for days to months. Both bet that a move keeps going. On SPY, the S&P 500 fund, the biggest five-minute bars of autumn 2026 were followed by another bar the same way 45.87% of the time, and one round trip of about $0.026 a share was 11.83% of a median five-minute move.

What each one is

Momentum scalping waits for a burst: a large candle on a one- or five-minute chart, often on news, a break of the morning high, or a jump in volume. The trader enters in the direction of the burst and aims to take a few cents or ticks before it stalls, with a stop just behind the candle. A position lasts seconds to minutes, and a busy day can hold dozens of them. The general style is covered on the scalping page.

Momentum trading buys strength and holds while it lasts. The signal is a price that has risen over weeks or months, a new high, or a moving-average trend, and the exit comes when the move stops, so one position can last days, weeks or months. Momentum trading covers that side.

How momentum scalping works, step by step

1. Pick one very liquid instrument with a spread of a cent or one tick, because the target is small.
2. Wait for the burst. A bar several times the size of a normal bar, in the direction of the session’s move so far.
3. Enter on the close of that bar or the break of its high (its low for a short).
4. Set the exit before the fill: a target of a few cents to a few ticks and a stop just past the other end of the burst bar.
5. Get out when the next bar stalls, whether or not the target came.

Everything rests on step 2 being followed by more of the same. That is a testable claim, and the original data below tests it on real five-minute SPY bars.

Where they differ

How much movement each trade needs. Momentum takes whatever a real move offers, and over weeks a typical move in SPY was measured in dollars. A momentum scalp works inside moves of a few cents to a dollar, while paying the same per-share cost to get in and out.

How many round trips you pay. A scalper may pay for dozens of round trips a day; a momentum trader pays for a handful a month.

What each needs from execution. Momentum tolerates a slightly late fill. Scalping does not, because a fraction of a cent of slippage can be a large share of the entire target.

A worked example

The biggest mid-session jump in the sample. Of 60 SPY sessions from 31 July to 23 October 2026, the largest rise in one five-minute bar, leaving out each session’s opening bar, came on 28 September 2026: the bar from 1:10 to 1:15 PM Eastern closed at $768.89, up $2.27 from the previous bar’s close of $766.62. That is the textbook momentum-scalp signal, about eleven times the median five-minute move of $0.205 in those sessions.

Buy 100 shares at that close. The round-trip cost at $768.89 is an assumed one-cent spread plus the SEC and FINRA sale fees: $0.01 + 0.0000206 × $768.89 + $0.000166 = about $0.026 a share, or $2.60.

What happened next. The next bar closed at $767.285, $1.605 lower. Selling there loses $160.50 on the shares plus $2.60 of cost, $163.10. Waiting three bars, to the 1:30 PM close of $767.06, loses $183.00 plus $2.60, $185.60. The burst reversed instead of extending.

One case proves nothing; it was picked for being the biggest, not the most typical. The count of all 460 strong bars below is the real answer, and it leans the same way.

Which one to use: Momentum Trading or Scalping?

Use momentum-style holding when your costs are ordinary retail costs. Over a month, a round trip was well under 1% of a typical SPY move, so the result depends on your decisions rather than on your fee schedule.

Use scalping only when your cost per round trip is close to zero next to a five-minute move, which usually means one very liquid instrument, a tight spread and a fee structure most retail accounts do not have.

Do not treat a big five-minute bar as a reason on its own. In SPY the bar after one went the same way less often than after an ordinary bar, so a momentum scalp needs an edge from somewhere other than the burst itself.

The original data

Do strong five-minute bars keep going? Yahoo Finance five-minute SPY bars for the 60 full regular sessions from 31 July to 23 October 2026 give 4,680 moves, each bar’s close against the previous close (the first bar of a session against the open). About the biggest tenth, 470 moves, were $0.64 or more in either direction. For each one with bars left in the session, the test compared the close one, three and six bars later with the strong bar’s close, in the strong bar’s direction.

For comparison, after any five-minute move the next bar went the same way 2,201 times in 4,579 (48.07%). One strong bar was followed by a flat bar, as were 41 of all the moves. So the strongest bars were followed through slightly less often than ordinary ones, and the median strong bar gave a few cents back. Sixty sessions of one fund is a small sample and the gap is a couple of points, so read this as no help from the burst, not as a reliable reversal signal.

Every strong bar is listed with the closes one, three and six bars later in the strong-bar table, so the count can be redone or extended to another stretch.

Two longer measurements on SPY answer the momentum trader’s side. Both use Yahoo Finance bars downloaded on 3 October 2026: daily adjusted closes from 29 January 1993 to 2 October 2026 (dividends reinvested), and 4,680 five-minute bars from 60 full sessions, 10 July to 2 October 2026.

First, did a move continue over the next equal period? Each period’s return was paired with the next one, and the table counts how often the two went the same way. If moves were unrelated, the expected share depends only on how often the market rose in that period, and that is the comparison column. Five-minute and 30-minute pairs stay inside one session; daily and longer periods are back-to-back blocks counted back from 2 October 2026, so no period is used twice. The p column uses the normal approximation to the binomial.

No horizon shows moves carrying on more often than chance. The five-minute figure is the only clear departure, and it runs the other way: consecutive five-minute moves reversed slightly more often than they continued, a correlation of -0.035. Days and weeks lean the same way by a small margin (p = 0.033 and 0.048, borderline with six periods tested), and months sit inside the noise. A one-year row is in the published file too, but with 32 pairs it is too small to read.

The one place a trend did separate what followed was the trailing year. At each month-end from January 1994 to August 2026, SPY’s return over the past twelve months was positive 322 times and the next month rose in 220 of them, 68.32%. After the 70 month-ends with a negative year it rose in 35, 50.00%. A two-proportion test gives p = 0.004. Treat it with care: the 70 negative months come from a few bear markets, so they are not 70 independent cases, and next-month swings after a negative year were wider (a standard deviation of 6.23% against 3.72%).

Three more funds and a later window give the same short-term answer. To check that the five-minute figure was not a quirk of one fund or one stretch of weeks, the same count was run on QQQ (the Nasdaq-100 fund), IWM (Russell 2000) and DIA (the Dow), using Yahoo Finance five-minute bars for the 60 full sessions from 30 July to 22 October 2026, downloaded 22 October 2026.

Consecutive five-minute moves went the same way 47.95% of the time in QQQ (4,511 pairs), 48.06% in IWM (4,409) and 48.40% in DIA (4,461), against about 50.0% expected if moves were unrelated. The two-sided p values are 0.006, 0.010 and 0.032. All four funds leaned toward reversal at five minutes, and none toward continuation.

Daily moves since each fund listed went the same way 49.91% of the time in QQQ, 48.98% in IWM and 49.26% in DIA, each a little under the 50.46%, 50.21% and 50.32% expected, with the Yahoo closes checked against Nasdaq’s own. Every row is in the four-fund table.

Second, how big is a round trip next to a typical move? Take 100 SPY shares at the 2 October 2026 close of $769.64. Crossing an assumed one-cent spread costs $1.00. On the sale, the SEC’s Section 31 fee of $20.60 per million (from 4 April 2026) is $1.5855, and FINRA’s Trading Activity Fee of $0.000166 a share is $0.0166. Both apply to sales only; the SEC charges the exchange and FINRA charges the broker, and this example assumes both are passed on to you. That is $2.60, or about $0.026 a share, spread included, before any commission. The spread is an assumption: under SEC Rule 612 quotes move in steps of $0.01, or $0.005 for the stocks with the tightest average spreads, so it cannot be narrower than that step.

The intraday moves come from the 60 five-minute sessions and the longer ones from the 252 sessions from 1 October 2025, as absolute close-to-close changes over every window of that length. The longer windows overlap heavily: the three-month median rests on 190 overlapping windows inside one year. The intraday trading page measures the same cost problem with bar ranges by time of day; this table extends it out to months. Both measurement tables are published with every cost row.

When it fails

The first failure is scalping on a retail cost base, and it looks like a skill problem when it is an arithmetic one. Entries hit the levels, exits hit the targets, stops are honored. The account still shrinks, because a large share of every small gain is handed back at both ends of every trade. Take 100 round trips of 100 SPY shares that win and lose the median five-minute move of $0.22 equally often: the gross result is zero, and the cost at $2.60 a round trip is $260.

The second is running momentum where moves do not continue. On SPY, days and weeks leaned slightly toward reversal (a borderline result), so buying strength on a short horizon was buying into a small headwind, not a tailwind. At five minutes the lean was clearer and it held in every fund tested: buying a strong five-minute bar in QQQ meant betting on a move that repeated its direction 47.95% of the time.

A third is chasing the burst. The bar that looks most like momentum is the one most often followed by a give-back: after SPY’s strongest five-minute bars, the median next bar moved $0.0375 the other way.

Questions people ask

Is scalping the same as momentum trading?

No. Scalping is a holding period, seconds to minutes; momentum is a reason for a trade, that a move keeps going. Momentum scalping joins them: a momentum entry with a scalper’s exit, so it carries the scalper’s cost per trade.

What timeframe do momentum scalpers use?

Usually one-minute or five-minute charts, or tick charts. A regular US stock session of 390 minutes holds 78 five-minute bars, the bars measured on this page.

Do big five-minute candles keep going?

Not in this sample. After SPY’s biggest tenth of five-minute moves from July to October 2026, the next bar went the same way 45.87% of the time, against 48.07% after any bar.

Is momentum scalping good for beginners?

The arithmetic is hardest at this speed. A round trip of about $0.026 a share was 11.83% of a median five-minute SPY move, so a beginner pays a large share of each small gain before skill counts.

Related

Momentum trading covers holding a move while it runs. Scalping covers the very short holding period. And day trading covers the wider category both belong to.

This page is educational, not financial advice. Test every idea on your own charts before risking money.