THE POWER OF OUTLIER TRADES.

In momentum swing trading, a small number of outsized winners produce most of the profit.

Most trades are small losses, small wins or scratches. A few run far enough to pay for all of them, and then some. Marios Stamatoudis puts it as 10–15 big winners making a great year.

That changes what the job actually is. You watch the stocks where outliers are most likely to show up, take every valid setup in them, and never cut a potential outlier short.

Two tempting habits work against that. One is capping every trade with a fixed target. The other is sitting out after a losing run, which is often exactly when the next big winner turns up.

Why a few trades make the year

A strategy with a low win rate and small, fixed losses has a built-in cost. Every losing trade costs about one unit of risk, and there are a lot of them.

That cost is the price of being in the game. The winners pay for it, and a few winners pay for almost all of it, because winning trades have a long right tail. Most winners make a few R, and a handful make ten or twenty times their risk.

The traders this blog draws on describe the same shape. Qullamaggie has said most of his money comes from 10–20% of his trades, and that his approach stays profitable at a 25–30% win rate as long as losses stay small and the big winners get to run. William O'Neil made a similar point decades ago: only one or two stocks in ten turn out to be truly outstanding.

The stock market itself is lopsided in the same way. Hendrik Bessembinder's study found that about 4% of stocks produced all of the market's gains from 1926 to 2016.

If the gains are that concentrated, your job is less about finding lots of good trades and more about being in the few great ones.

Outliers in R are partly made at entry

An outlier in R comes from two things: how far the stock moves, and how little you risked to be in it.

R is the move divided by the stop distance, so the same move is worth very different amounts depending on your entry. A 15% move on a 1% stop is 15R. The same move on a 5% stop is 3R.

So a big R-multiple doesn't need an extraordinary stock. A good move plus a tight, well-placed stop can produce one. That's one more reason the traders this blog draws on work so hard on low-risk entries. Qullamaggie's are an opening-range high with a stop at the low of the day, never wider than the stock's average daily range.

Marios Stamatoudis: put yourself where outliers happen

If a handful of trades make the year, the real question is how you end up in them.

Marios Stamatoudis, a top performer in the 2023 US Investing Championship, has written about this on X. His view, paraphrased: a great year comes from roughly 10–15 big winners, so the work is to put yourself where outliers are more likely to show up. He makes three points about how.

1. Your trade frequency has to overlap your universe

Here's his example. You track 300–400 stocks and take about 60 trades a month. Even if every trade is in a different name, you only touch 15–20% of your list in a month (60 ÷ 400 = 15%, 60 ÷ 300 = 20%).

So if you just take whichever setups look best on the day, whether one of those trades lands on the month's outlier is close to random.

2. Shrink the universe, but don't cut the outliers

The obvious fix is a smaller list, but that only helps if the future outliers survive the cut.

Marios' numbers go like this. If 20 of your 400 names become outliers, 5% of the list is an outlier. Cut to 50 names and keep 10 of those 20, and the share rises to 20%, even though you lost half the outliers.

We worked through what that means for a trader picking names at random from each list. These are our own calculations on his example numbers:

Hypergeometric probabilities, each trade in a different name. The 20-trades-a-month figure is our illustration, not Marios'.

The catch is the filter. A cut like "great fundamentals only" can throw away real outliers. Marios points to OKLO, LWLG and AXTI as big movers that didn't have the fundamentals such a filter would demand (OKLO turns up again in the examples below).

3. Phenomena over setups

His third point is the one that ties the others together. Big moves come from phenomena: a theme, a story, a catalyst, a change in fundamentals, or several at once. A perfect chart pattern in a stock nobody cares about rarely becomes an outlier.

He notes that many of the biggest moves came from setups he'd only rate about 2.5 to 3 out of 5. So instead of hunting for textbook patterns, he weighs behavioural, fundamental and technical evidence together. Knowing the story behind a move also gives you the conviction to size up when it matters, which price alone didn't give him.

Qullamaggie says much the same in fewer words: earnings are fuel, so find stocks with a reason to go up before you look for patterns.

In practice, the best way to shrink your universe without losing outliers is to filter for the phenomena themselves. That means the strongest relative strength, the leading themes and sectors, fresh catalysts and earnings surprises.

Habit 1: don't cap the whole position

A fixed profit target looks sensible. Take 3R and move on.

And it does lock in gains on trades that pop and then fade back. But it also caps every trade at the same size, including the rare one that would have run to 20R or 50R. If a few trades make the year, a target caps the very tail this whole style is built on.

There's a better answer to the pop-and-fade problem: sell a piece into strength and let the rest run. That's the first of Qullamaggie's selling habits, and Oliver Kell's version is that if you don't sell into strength, you'll end up selling into weakness.

How much you sell matters, though. The more you sell into strength, the higher your win rate will be, but the more you cap the outlier trades that would have ridden the moving averages. In our selling techniques test of 544 breakouts, trailing the whole position on the 21-day EMA made +311.8R, with 24 trades of 10R or more. Selling half at 2R lifted the win rate from 15% to 38%, but it halved the 10R-plus trades to 12 and cost 92R.

Top traders notice the same pattern. Martin Luk and Qullamaggie both accept win rates of about 25–35% so their winners can ride a moving average: Luk trails on the 9-day EMA, and Qullamaggie sells a third to a half and trails the rest on the 10- or 20-day.

A partial banks some profit on the trades that fade, while the rest, trailed on a moving average, stays in the trades that keep going. And if you do use a target, keep it on a piece of the position and keep it distant. Try the options on your own trades with a backtest of exit rules on past trades.

Habit 2: don't sit out after a losing streak

The natural reaction to three losses in a row is to sit the next few out. It feels prudent.

But a losing streak tells you nothing about the next trade. The setups you skip after a streak come from the same distribution as the ones you take, and one of them may be the outlier that pays for the streak.

Big winners also tend to arrive in bunches, when a theme gets going or the market turns, and that's often right after a run of losses. So sitting out a stretch of signals can mean missing several of them at once, at exactly the moment it feels most sensible to wait.

Market conditions, on the other hand, are information, and they're a better reason to trade smaller. Cutting size in a hostile market keeps you in every valid setup while limiting the damage. See drawdowns and losing streaks for sizing rules that keep you in the game without skipping the winners.

Habit 3: keep taking the setup

You can't tell at entry which trade will become the outlier. On the day it was bought, each of the four trades below looked like any other breakout, with a normal stop and no promise of more.

So the only way to be sure you're in the big one is to take every valid setup in your universe, the same way, every time. That includes the ones that look a little ordinary, since Marios found many of his biggest moves came from setups he'd only rate 2.5 to 3 out of 5.

It also means handling each trade the same way once you're in. A small first stop, a piece sold into strength, and the rest trailed until the trend breaks. If you do that every time, the outlier takes care of itself when it shows up.

And if a 30% win rate already feels like a lot of losing, why the best swing traders win only 25–35% of the time shows what that sequence really looks like.

How the trail kept you in

The four examples below all used the same simple plan, from how to trade breakouts. The stock was bought as it traded through the top of its base, with the stop at the low of the day (never closer than a third of the stock's average daily range). A third was sold at the close of day three, and the rest was trailed on the 21-day EMA and sold on the first close below it.

That's all. No target, no prediction of how far it would go. Watch what the trail did in each one.

A five-month trend: TXG in May 2026

10x Genomics (TXG) had already run 66% in a month before it rested for 26 sessions in a base about 24% deep. In late May it traded up through the top of that base, with a stop only 2.3% away.

The day-three third was sold for 8.8R, a good trade on its own. The rest rode the 21-day EMA for more than four months without a single close below it, and went on the first one in October for 94.8R. A 3R target would have been out on the first day, and the same entry trailed on the faster 9-day EMA was sold seven sessions in for 6.2R. The whole trade came to +66.1R.

Wild swings, one rule: MRVL in April 2026

Marvell (MRVL) built a 20-session base about 25% deep through March. It broke out at the open in early April, and the stop was 1.9% away.

The third sold on day three made 4.6R. The rest climbed steadily through April and May, then went through some violent days in June, which is where a nervous trader would have bailed. But none of those days closed below the 21-day EMA, so the trail stayed put. It gave back a good chunk from the top before the first close below it in early July, and still made 76.3R on the rest. That give-back is the price of a trail, and it's a small one. The trade came to +52.4R.

A fast one: OKLO in September 2025

Oklo (OKLO) is one of the names Marios points to as an outlier a fundamentals filter would have thrown out. It had run 75% in two months, then built an 18-session base about 27% deep. In mid-September it gapped open above the base, with a 2.4% stop.

This one moved fast, and the day-three third was already worth 10.4R. The stock then pulled back hard in late September, but held above the 21-day EMA and ran again into October. The rest was sold on the first close below the average five weeks in, for 27.6R, more than double what a 9-day EMA trail made on the same entry. The trade came to +21.9R.

A slow stock: TEL in May 2025

Outliers don't need a wild stock. TE Connectivity (TEL) moved only about 3% on an average day, and in May 2025 it broke out of a 12-session base about 12% deep. The stop at the low of the day was just 1.2% away.

The day-three third made 3.7R. After that the stock just kept creeping higher, with one big gap in July, and never closed below the 21-day EMA for five months. The rest went on the first close below it in October for 33.8R, on a price move of only about 39%. A 9-day EMA trail had sold the whole thing within two weeks for 3R. The tight stop made the move big in R, and the patient trail made it an outlier. The trade came to +23.8R.

Ten trades, one year

Our own breakout test shows how extreme this gets. In the 544-trade breakout test, the whole plan made +258R. The best 10 trades, less than 2% of the total, made 294R on their own. Take them out and the same plan finishes about 36R in the red.

Those ten weren't flukes in an otherwise good system. They were the system. The other 534 trades, added together, lost money.

Top traders find the same thing in their own records. Christian Flanders, who made +433% in the 2024 US Investing Championship, went back through his data in July 2026. His top 10 trades made 127 of his 129 risk-adjusted R in 2025, so without them his year was break-even. In 2024 it was 228 of 244.

That's the same shape Qullamaggie describes, with most of his money coming from 10–20% of his trades, and the same reason Marios counts 10–15 big winners in a great year. Ten trades won't sound like much until you check your own journal and find they're the whole year.