Imagine you have a game slightly in your favor: on every bet, the odds are tilted a little to your side. Sounds like a winner, right? Now imagine that on every hand you bet almost all your money. Sooner or later a bad streak arrives —because it always does—, you lose several in a row, and you hit zero. Game over. You had an edge and still went broke.
This has a name: the gambler's ruin. Being right in the long run is worthless if a short streak knocks you out before the "long run" ever arrives. That is why the first rule of trading is not to win big: it is to not go broke.
Important note: this article is general education, not personalized financial advice. Trading carries a real risk of loss and nothing guarantees profit.
What position sizing is
Position sizing is simply how much money you put into each trade. The key question is not "how much can I win?", but "how much am I willing to lose if this trade goes wrong?". That is called risk per trade.
Many beginners think only about the gain and risk a huge slice of their account on a single "safe" idea. The problem is that no idea is safe, and one big trade that goes wrong can erase weeks of work.
The gambler's ruin, explained
Imagine two people with the same 2% edge on every bet. One risks 50% of their money per hand; the other risks 1%. The first might have a brilliant night... or vanish in three bad hands. The second moves slowly, but survives the bad streaks and is still at the table when the edge finally shows.
It does not matter how good your strategy is if a streak of bad luck removes you from the game before you can collect on it.
Bad streaks are not a possibility: they are a certainty. Even a strategy with a real edge will have runs of consecutive losses. Your job is not to avoid them (impossible), but to bet small enough to stay alive when they come.
Surviving matters more than maximizing
Here is the counterintuitive idea: the goal is not to squeeze every trade for the maximum. It is to make sure that no single trade, and no single streak, can finish you off. Whoever survives keeps compounding results over time. Whoever goes broke goes home, no matter how good their idea was.
A light touch on Kelly
There is a famous formula, the Kelly criterion, that tries to calculate what fraction of your capital to bet in order to grow sustainably. We are not going to prescribe it here, and for good reason: many professionals bet even less than Kelly suggests, because the formula assumes you know your edge precisely, and you almost never do. The practical lesson you can take away is simple: bet a fraction, never everything.
How AlphaLab handles this
AlphaLab does not tell you how much to bet —that would be personalized financial advice, and it is not— but it does help you make risk decisions honestly:
- It simulates real costs. Many strategies look profitable until you add the real cost of trading (spread, slippage, commissions). AlphaLab includes a realistic cost simulation, so the risk you see resembles real risk.
- It measures the worst case, not just the best. With tools like Monte Carlo simulations, AlphaLab shows you plausible losing streaks and drawdowns you might suffer, so you size your bet thinking about surviving those streaks, not just about the pretty result.
- It rewards robustness over spectacle. The system values strategies that hold up under hostile conditions over those that shine in the best possible scenario. That philosophy pushes, by design, toward survival.
Key takeaways
- Having an edge does not save you if you bet too much: the bad streak ruins you (gambler's ruin).
- Position sizing is decided by how much you can lose, not by how much you dream of winning.
- Bad streaks are inevitable; bet small enough to survive them.
- Bet a fraction, never everything; surviving matters more than maximizing each trade.
- This is education, not personalized advice: trading carries a real risk of loss.
If you want to see the risk of your strategies in honest numbers before you risk real money, you can try AlphaLab free for 14 days (card required, cancel anytime) at whop.com/alphalab-005b/alphalab-pro.