Strategy teardown · part 1 of 3

We systematized Qullamaggie's momentum strategy. Does the raw signal work?

Kristjan Kullamägi turned a small account into tens of millions trading momentum breakouts — and he published the rules. We built a faithful systematic version on 1,500 US stocks. The result captures his exact signature, small losses and big winners. But raw, it barely makes money.

By ridingyo2026-06-20 Original strategy: Kristjan Kullamägi (qullamaggie.com)

What this is. A clean-room systematic implementation of Kullamägi's publicly published rules, for analysis — no original content or code is copied; he is credited and linked. Daily data on the current S&P 1500, 2015–2026. This is part 1 of 3: build it faithfully and see if the raw signal has an edge.

The strategy, in his words

Kullamägi trades a handful of timeless setups. The core one — the breakout:

Scan for the stocks up the most over 1-, 3- and 6-month windows — the current leaders.
Wait for a big move (30–100%+) to settle into a tight, orderly consolidation that surfs the rising 10- and 20-day moving averages.
Buy the breakout out of that base. Stop at the day's low, no wider than the stock's average daily range.
Sell a third to a half after 3–5 days, then trail the rest with the 10/20-day MA — exit on the first close below it.

Plus the Episodic Pivot (an earnings/news gap-up) and a parabolic short (which is an intraday setup we cannot faithfully test on daily bars — so it's out of scope here). His own summary of the edge: a 25–30% win rate is fine — the money is in small losses and the occasional 5–20x winner.

How we systematized it

We turned the discretionary rules into mechanical ones: rank the universe by a composite 1/3/6-month return and keep the top 15% (the leaders); require an uptrend (price above a rising 10/20-day EMA) and a volatility-contracted base; enter on a breakout to a new 20-day high (or an EP gap-up on volume); stop at an ADR-based level; size each position so a stop costs ~0.75% of the account; and trail the winners out on the first close below the 10-day EMA. No discretion, no market-timing overlay — the raw signal, exactly as written.

The result: the right shape

First, the good news. The distribution of trade outcomes is exactly his signature:

Histogram of trade outcomes in R multiples — many small losses, a thin right tail out to 14R — Trade outcomes in R (multiples of the amount risked). A 37% win rate, a cluster of small losses, and a thin right tail reaching +14R — "small losses, big winners," reproduced mechanically.

The mechanism is real: most trades are small losers, a few winners run for 5, 10, 14× the risk, and those carry the system. This is what a genuine momentum/trend edge looks like.

...and the bad news: raw, it barely works

Now the equity curve. Over eleven years the strategy is a regime roller-coaster:

Equity curve: choppy, spikes +80% in 2021 then gives back 48% in 2022, ends near flat — Account equity (log scale). It explodes higher in the 2021 momentum boom, then surrenders nearly half in the 2022 momentum crash — and finishes only modestly above where it started.

Metric	Value
CAGR	1.7%
Sharpe	0.19
Max drawdown	−47.6%
Win rate	37%
Profit factor	1.03
Trades (11y)	2,288

A profit factor of 1.03 and a −48% drawdown is not a strategy you can trade. The edge is in there — the right tail is real — but it is buried under whipsaw losses and, above all, a complete lack of market timing. The system happily buys breakouts into the teeth of a bear market, and 2018, 2020 and 2022 punish it for it.

What this tells us

Part 1 verdict

The signal has the right shape — but the edge is in the wrapper.

The mechanical breakout reproduces Kullamägi's win/loss profile but not his returns. The gap is everything he does around the signal: sizing up in strong markets and sitting in cash in bad ones, selecting the cleanest setups by hand, and taking partial profits. Raw, the signal is a coin flip with a fat tail and a brutal drawdown.

And we are flattering it. Our universe is the current S&P 1500, so every stock that was delisted, acquired or went to zero is invisible — classic survivorship bias, which makes this result an optimistic upper bound. The honest number is worse than what you see above.

Next — part 2

How much is survivorship, and where does the drawdown come from?

We pull the result apart: quantify the survivorship inflation, locate the drawdown in time and in market regime, test it out-of-sample, and deflate the headline for every parameter we tried. Then part 3 rebuilds it with the discipline — a regime filter, profit-taking, volatility sizing — to see whether the legend's edge can be made tradeable by a system.

Check the math yourself

Tool
Position Size & Risk of Ruin — survive a 37%-win-rate system with a −48% drawdown Tool
Deflated Sharpe Ratio — what does a momentum backtest survive after the search?

Educational analysis, not investment advice. A methodology case study, not a recommendation to trade any strategy or instrument, and not a judgement of the original trader. Simulated and optimized results have severe limitations — survivorship bias here means the real figures are worse — and do not predict future performance. See the full disclaimer.