Polymarket Risk Management 101

Almost nobody blows up on Polymarket because they were wrong too often. They blow up because they were wrong big — one oversized position, one market they were sure about, one bad day with no limit to stop the bleeding. Risk management is the unglamorous discipline that prevents that. Here's the practical version.

Why prediction markets punish bad sizing

Polymarket prices are probabilities between $0 and $1. A market at $0.90 is the crowd saying "90% likely." That implies you'll be right often — which is exactly the trap. Win nine times for small gains, then bet too much on the tenth and lose it all, and you're underwater despite a 90% hit rate. Binary outcomes are unforgiving: the position either resolves to $1 or to $0. Sizing, not accuracy, is what protects you from that asymmetry.

1. Bankroll and position sizing

Start by defining your bankroll: the total amount you've allocated to prediction-market trading and are prepared to lose. Everything else is a percentage of that number, not of your net worth and not of your hopes.

Then cap each position. A common, conservative approach is to risk a small fixed fraction of bankroll per market — say 1–3% — so that no single resolution can meaningfully dent you. Traders who scale stake with conviction sometimes use a fractional Kelly approach (betting a fraction of what the Kelly criterion suggests), but the key word is fractional: full Kelly is far too aggressive for markets where your probability estimates are themselves uncertain.

2. Exposure limits

Per-position caps aren't enough on their own. Twenty positions at 3% each is 60% of your bankroll at risk — and if those markets are correlated (think several markets keyed to the same election or event), they can all move against you together. Set a total-exposure limit on how much of your bankroll can be deployed at once, and watch for correlated exposure where "diversified" positions are really one bet in disguise.

3. Loss limits and circuit breakers

A daily loss limit caps how much you can lose in a single session before you stop. A circuit breaker goes further: it automatically halts all trading when something looks wrong — a string of losses, an unusual price move, a data feed that's gone stale. The point is to remove your discretion at exactly the moment your discretion is worst: when you're down and tempted to "make it back."

This is also the clearest argument for automation. A human won't reliably honor a loss limit at 2 a.m. after a rough run. Software will — if you built the breaker and didn't give yourself an override.

4. The discipline layer: emotion and exits

The hardest risks to manage are the ones in your own head. Three patterns ruin more accounts than any market ever has:

• Revenge trading — sizing up to recover a loss, which is how a bad day becomes a bad month.
• Holding losers — refusing to exit a position because selling makes the loss "real."
• Cutting winners early — banking a small gain out of anxiety while letting losers run.

The fix is to decide exit rules in advance and follow them mechanically — take-profit and stop conditions, time-based exits when a thesis hasn't played out, and a plan for markets that resolve while you hold them. This is exactly the kind of rule that benefits from being automated, because automation doesn't get scared or greedy.

Position sizing: a worked example

Abstract rules are easy to nod along to and hard to follow. Here's the concrete version. Say your bankroll is $2,000 and you cap each position at 2%. That's a $40 maximum stake per market — no matter how confident you feel. Find ten opportunities you like? That's $400 deployed, 20% of bankroll, leaving plenty of room to survive a bad streak.

Now watch what undisciplined sizing does to the same account. You're "sure" about one market and stake $600 — 30% of bankroll — at a price of $0.80. The outcome resolves against you. You're down $600, and you now need a 43% gain on the remaining $1,400 just to get back to even. One position erased what a dozen disciplined trades would take weeks to build. That asymmetry is the entire argument for caps.

Traders who vary stake by conviction sometimes use a fractional Kelly rule — betting, say, a quarter or a half of what the full Kelly criterion suggests. The reason to stay fractional is humility: Kelly assumes you know your true edge, and on prediction markets your probability estimate is itself a guess. Betting full Kelly on an overestimated edge is a fast route to ruin. When in doubt, size down.

Diversification that isn't: correlation risk

Spreading across many markets feels safe, but it only helps if those markets are actually independent. Five positions tied to the same election, the same team, or the same macro event aren't five bets — they're one bet wearing five jackets. When the underlying event breaks the wrong way, they all lose together, and your "diversified" book takes a single concentrated hit.

The fix is to think in terms of themes, not just positions. Group correlated markets and apply an exposure limit to the group as a whole, not only to each market individually. Real diversification means uncorrelated outcomes — different events, different drivers — not just a longer list of tickers.

A simple risk checklist

1. Define your bankroll — money you can afford to lose, written down.
2. Set a hard per-market cap (e.g. a small % of bankroll) and never override it.
3. Set a total-exposure limit and watch for correlated positions.
4. Set a daily loss limit and a circuit breaker that enforces it automatically.
5. Write your exit rules before you enter, and follow them mechanically.
6. Review weekly: did you respect every limit? If not, the limit or the discipline needs fixing.

Frequently asked questions

How much of my bankroll should I risk per Polymarket trade?

A common conservative range is 1–3% of bankroll per market, capped hard so no single resolution can meaningfully dent you. The exact figure depends on your edge and risk tolerance, but the principle is fixed: set a ceiling in advance and never override it.

What is a circuit breaker in trading?

A circuit breaker is a rule that automatically halts trading when something looks wrong — a string of losses, an abnormal price move, or stale data. Its value is that it acts at the exact moment your own judgment is worst: when you're down and tempted to chase the loss back.

Why do most prediction-market traders lose money?

More often from poor sizing and emotional decisions than from bad predictions. Revenge trading, holding losers, and cutting winners early destroy more accounts than any single wrong call. Mechanical limits and pre-decided exits are the defense.

Does automation reduce risk?

It doesn't reduce market risk, but it can enforce your risk discipline — honoring loss limits and exits that a human won't reliably follow under pressure. The risk of loss remains; what improves is consistency. See how to automate Polymarket trading for how the guardrails fit together.