Picking the right contract is only half the decision. How much you stake on it is what decides whether a run of bad luck ends your participation.
A bankroll is the total money you have set aside for trading and can afford to lose. Position sizing is the discipline of deciding how much of that bankroll to put on any single market. The core idea is simple. Even a genuinely good decision can lose, losses cluster, and staking too much on each trade means a normal losing streak can wipe you out before any edge has a chance to show. Sizing small and consistently is what keeps you in the game long enough for the maths to matter, and it never turns a bad bet into a good one.
A bankroll is the amount you have deliberately set aside for trading and could lose entirely without affecting your rent, your bills, or your peace of mind. It is not your savings, not borrowed money, and not money earmarked for anything else. Defining it clearly is the first act of risk control, because every sizing decision is expressed as a fraction of that number. If you cannot say what your bankroll is, you cannot size a position responsibly, and you are far more likely to chase losses with money you needed.
Responsible sizing means risking a small, fixed fraction of your bankroll on any one market, often a low single digit percentage, rather than an amount that feels right in the moment. Fixed fractional staking has a useful property. As your bankroll shrinks your stakes shrink with it, which slows the damage of a bad run, and as it grows your stakes grow in proportion. The exact fraction is a personal choice, but the principle is that no single market should be able to hurt you badly, however confident you feel.
Risk of ruin is the probability that a series of losses drains your bankroll to the point where you can no longer continue. It rises sharply as your stake per trade grows. Stake ten percent of your bankroll per market and a fairly ordinary losing streak can end you. Stake one or two percent and the same streak is survivable. This is not pessimism, it is arithmetic. Losses do not arrive evenly spaced, they cluster, and sizing has to assume the bad cluster will come.
Some sizing methods, such as the Kelly criterion, try to maximise long run growth by sizing in proportion to your estimated edge. The catch is that your edge is an estimate and is often smaller or more uncertain than you think. Because overestimating edge leads directly to oversizing, many careful participants deliberately stake a fraction of any theoretical optimum. Sizing as if you might be wrong is not timid, it is the honest response to the fact that prices can be right when you are not.
The clearest way to see why size matters is to follow one bankroll through a run of losses. Assume you begin with a bankroll of one thousand dollars, you stake the same flat fraction of it on every market, and a contract that resolves against you and is held to expiry costs the whole stake. A losing streak then simply multiplies the bankroll by one minus your stake fraction, once for each loss. The arithmetic below is exact under those assumptions, and it shows the gap between a small stake and a large one widening fast. None of these figures is a prediction of results, and a losing streak of this length is unremarkable rather than unlucky.
| Stake per market | After 3 losses | After 5 losses | After 10 losses |
|---|---|---|---|
| 1% of bankroll | 97.0% | 95.1% | 90.4% |
| 2% of bankroll | 94.1% | 90.4% | 81.7% |
| 5% of bankroll | 85.7% | 77.4% | 59.9% |
| 10% of bankroll | 72.9% | 59.0% | 34.9% |
| 20% of bankroll | 51.2% | 32.8% | 10.7% |
Method: each figure is the original bankroll multiplied by one minus the stake fraction, raised to the number of losses, assuming flat staking on the starting bankroll and that each losing market costs the full stake. Exact under those assumptions and illustrative only, not a calculation for any specific market or platform. As of June 2026.
Two things jump out. At one or two percent a five loss streak barely dents the bankroll, so you live to make the next decision. At twenty percent the same streak takes two thirds of it, and ten losses leave almost nothing. The person staking twenty percent does not need to be unlucky to be ruined, only ordinary, because losses cluster and a run of five or ten in a row arrives sooner than intuition suggests. This is the whole argument for keeping each stake small, and it sits underneath every point about expected value and the risk of loss.
Imagine two people who make identical decisions with a small genuine edge. One stakes 2 percent of a bankroll per market, the other stakes 20 percent. A run of five losses in a row, which is unremarkable, costs the first about a tenth of the bankroll and is easily survived. It can take the second below the point of no return. Same decisions, same luck, different survival, decided entirely by size.
An illustration of how sizing changes outcomes, not a recommended stake or a prediction of results.
You cannot control whether an outcome happens, and you cannot control the price you face. The one variable that is entirely yours is how much you stake. That is why experienced participants treat sizing as the most important decision they make, more important than the specific market they choose, because it is the decision that determines whether they are still trading after a normal run of bad luck.
The deep problem sizing addresses is that good decisions still lose. A market priced at sixty cents that resolves no was not necessarily a bad bet, it was a bet that lost, and that distinction matters enormously. If you judge your sizing by individual results you will lurch between overconfidence after wins and panic after losses. If you size by a fixed rule, you remove the emotion and let the process survive the inevitable cold streaks.
Bankroll discipline also protects you from the most expensive mistake in any speculative activity, which is chasing. After a loss the urge to win it back quickly with a bigger stake is powerful and almost always destructive. A predefined sizing rule takes that decision out of your hands in the moment when you are least able to make it well. The rule is there precisely because your judgement is worst right after a loss.
It is worth being clear about what sizing cannot do. It cannot turn a market you misread into a winner, and it cannot manufacture an edge that is not there. If your estimates are no better than the price, careful sizing simply means you lose money more slowly, with fees and spreads grinding the bankroll down. Sizing keeps you solvent so that a real edge, if you have one, has time to show. It is a survival tool, not a profit machine.
Finally, sizing is inseparable from honesty about the money. A bankroll only works as a concept if it is genuinely money you can afford to lose, kept separate from the money your life depends on. The moment you are sizing positions against money you need, no fraction is small enough, because the real risk is no longer financial, it is to your wellbeing. That is where responsible play and bankroll discipline meet.
The practical strength of a sizing rule is that you set it once, when you are calm, and then follow it when you are not. Deciding in advance that no single market gets more than a small fixed fraction of your bankroll removes the in the moment argument with yourself, the one you tend to lose. Writing the rule down, even informally, makes it harder to quietly abandon after a couple of wins or a frustrating loss.
It also helps to review the bankroll itself on a fixed schedule rather than after every result. Topping it up impulsively after a bad run is just chasing by another name, and letting it ride after a good run can quietly push your stakes higher than you intended. A periodic, unemotional review keeps the number honest and keeps your sizing anchored to money you can genuinely afford to lose. None of this guarantees a profit. It guarantees only that a normal run of bad luck will not end your participation, which is the whole point. This page is general information, not financial advice, and you should make your own decisions or consult a qualified professional.
The simplest approach is flat staking, where every market gets the same fixed fraction of the bankroll, say one or two percent, regardless of how confident you feel. It is crude but robust, and it is the method behind the table above. A close cousin is fixed fractional staking, where the stake is a set percentage of the current bankroll rather than the starting one, so stakes shrink automatically through a bad run and grow back through a good one. Both share the same protective feature, which is that no single market can take more than a small slice, however sure the outcome seems.
A more ambitious idea is the Kelly criterion, a formula first published by the physicist John L. Kelly Jr in 1956 in the Bell System Technical Journal. Kelly sizes each stake in proportion to your edge, the gap between your own probability estimate and the price, divided by the odds on offer. In theory, staking the full Kelly fraction maximises the long run growth rate of a bankroll. The appeal is obvious, because it ties the size of a bet directly to how good the bet is, putting more on a strong edge and less on a weak one.
The catch is that full Kelly is violent. It is commonly described in betting and quantitative trading practice as carrying expected peak to trough drawdowns of around half the bankroll at some point in any long sequence, even when the edge is real (per widely cited Kelly analyses, as of June 2026). Worse, the formula is only as good as the edge you feed it, and most people overestimate their edge. Because overstating the edge leads straight to oversizing, a large share of experienced participants deliberately stake a fraction of the Kelly amount, often a half or a quarter. Half Kelly is frequently cited as capturing most of the theoretical growth while cutting the swings sharply. The figure below shows the same Kelly signal scaled down in the common way.
One limit of every sizing method deserves stating plainly, because it is where overconfidence does the most damage. None of them can rescue a bet with no edge. If your probability estimate is no better than the price, Kelly tells you to stake nothing, and any positive stake loses money in expectation once fees and the spread are counted. Sizing decides how fast you win or lose given an edge. It cannot manufacture the edge, and it cannot turn a market you misread into a winner. You can compare published costs across venues in our cross platform fees table, because those costs come straight off the top of any edge you think you have.
For most people the honest conclusion is modest. A small flat fraction, set in advance and reviewed on a schedule rather than after every result, gives almost all of the protection with none of the false precision of an edge estimate you cannot really trust. If you do use Kelly, sizing well below the full figure is the standard response to the fact that your edge is uncertain. Whichever you choose, the rule is the same one this whole page rests on, that no single market should be able to do you serious harm. This is general information, not financial advice, and you should make your own decisions or consult a qualified professional. The same discipline runs through responsible play and through every platform you might use.
No sizing rule makes trading safe, and a careful stake on a poor decision is still a loss. Treat your bankroll as money you can afford to lose, never as a way to win back what is gone. Stake only what you can afford to lose, never to chase a loss, and never on borrowed money. If it stops feeling like a free choice, step back. In the United States you can call or text the helpline on 1-800-GAMBLER or visit ncpgambling.org.
A bankroll is the total amount of money you have deliberately set aside for trading and can afford to lose entirely without affecting your bills, savings, or wellbeing. Every sizing decision is expressed as a fraction of that number, so defining it clearly comes first.
Position sizing is deciding how much of your bankroll to put on a single market. Responsible sizing risks a small fixed fraction per trade rather than an amount chosen by feeling, so that no single market can do serious damage.
Risk of ruin is the probability that a run of losses drains your bankroll to the point where you can no longer continue. It rises sharply as your stake per trade grows, which is the main reason careful participants keep each stake small.
Kelly is a formula that sizes a stake in proportion to your estimated edge to maximise long run growth. Because edge is usually uncertain and easy to overestimate, many people stake only a fraction of the Kelly amount to avoid oversizing.
No. Sizing controls how fast you can lose and helps you survive bad streaks, but it cannot create an edge or turn a misread market into a winner. If your estimates are no better than the price, careful sizing only means you lose more slowly.
There is no single correct figure, and this page is not advice. The widely shared principle is to risk only a small fraction of a bankroll you can afford to lose on any one market, and to size as if your edge might be smaller than you think.
The rules change fast. Get the changes that affect you, plain and current, not tips.
Independent. Every claim dated and sourced. No platform pays for its place.