Risk Management for Prediction Market Traders

Most retail traders blow up not from bad strategy, but from bad risk management. They find an edge, scale up too fast, hit a losing streak, panic, and lose everything they made plus more. This guide covers the boring discipline that separates traders who survive years from traders who flash and burn.

## Why Risk Management Matters More Than Strategy

A 60% win rate with terrible risk management loses money. A 52% win rate with great risk management makes money slowly but reliably. The math is brutal: a 50% drawdown requires a 100% gain to recover. A 75% drawdown requires 300% gain.

Most traders fixate on finding "the perfect strategy" while ignoring the system that prevents catastrophic losses. The order matters: risk management first, strategy second.

## The Five Pillars of Risk Management

### 1. Position Sizing

Never risk more than 1-5% of your bankroll on a single trade. Most beginners size 10-20% per trade because edges look "obviously profitable". They go bust on routine losing streaks.

Detailed approach in our [Kelly Criterion guide](/blog/kelly-criterion-position-sizing) — the short version: cap any single position at 5% of bankroll, period. No exceptions for "high confidence" or "guaranteed wins". Both are illusions.

### 2. Maximum Total Exposure

Even with good per-trade sizing, opening 20 simultaneous trades exposes you to correlated risks. If all 20 are election-related, they share the same outcome driver.

Rule: never have more than 25% of bankroll in active positions at any time. If you have 5 great ideas, average position size = 5%. If you have 20, average = 1.25%. Forces you to prioritize your best ideas.

### 3. Stop Losses

For every position, decide before entering: at what point are you wrong? At what loss do you exit?

For Polymarket positions, common stop loss approaches: - Fixed percentage: exit if down 25% from entry - Fixed price: exit if YES price drops below $0.20 (regardless of entry) - Time-based: exit if no progress within 30 days - News-triggered: exit if specific event happens

Pick one and follow it. The mistake isn't choosing the wrong stop — it's not having one and "hoping" losing positions recover.

### 4. Bankroll Segregation

Trading bankroll should be separate from: - Living expenses (rent, food, bills for 6+ months) - Emergency fund (3-6 months of essentials) - Retirement savings (401k, IRA, equivalent) - Other investments (stocks, real estate)

If your trading account hits zero, you must be inconvenienced, not bankrupt. The first $5k you risk should be 100% acceptable to lose without affecting your life.

This sounds obvious. Most retail traders break it. Don't.

### 5. Drawdown Limits

Define maximum tolerable drawdown before you start. Common framework: - 10% drawdown: review strategy, reduce position sizes 50% - 20% drawdown: pause trading for 1 week minimum, audit decisions - 30% drawdown: stop entirely, return after 30 days break

The point isn't to prevent any drawdown — those are inevitable. The point is to react systematically when they happen, not emotionally.

## The Specific Risks of Prediction Markets

Beyond generic trading risks, prediction markets have unique exposures:

### Resolution Risk

A market might resolve in an unexpected way due to: - Ambiguous resolution criteria - Oracle disagreement (UMA disputes on Polymarket) - Cancellation by the platform - Edge cases not anticipated in market design

Real example: a "Will X happen by Date" market resolves NO because X happened one day late. Was the bet wrong? Sort of. Did the trader make money? No.

Mitigation: read resolution criteria carefully BEFORE trading. If ambiguous, don't trade or trade smaller.

### Liquidity Risk

You can enter a position easily. Exiting is harder when: - Market becomes one-sided (everyone wants to sell, no buyers) - News crashes liquidity overnight - Market approaches resolution (spreads widen)

Mitigation: only trade liquid markets. Avoid contracts with sub-$10k 24h volume. For more on liquidity assessment, see our [CLOB and liquidity guide](/blog/understanding-clob-order-book).

### Platform Risk

The platform itself can fail: - Polymarket has had hours-long outages - Smart contract bugs (extremely rare but possible) - Regulatory action against the platform - Frontend errors that show wrong prices

Mitigation: don't keep all capital on one platform. Withdraw profits periodically. Have a plan for what happens if the platform is unreachable for 24-48 hours.

For Polymarket-specific platform issues, the Predite platform has automatic outage detection and bot pausing — but it's not a substitute for understanding the risk yourself.

### Concentration Risk

Building large position in single market means single-event ruin. Even with good per-trade sizing, if 40% of your bankroll resolves on Election Night, you're not diversified.

Mitigation: spread positions across uncorrelated markets. Politics + crypto + sports + entertainment, not all politics.

### News Risk

A position thesis can be destroyed by news arriving between your entry and resolution: - Polling shows unexpected shift - Player gets injured - Crypto protocol gets exploited - Geopolitical event changes everything

Mitigation: track news related to your positions. Have alerts set up for material events. Be willing to exit when your thesis breaks, even at a loss.

## The Psychological Side

Risk management is 80% psychology, 20% math. The math is simple. Following it under stress is hard.

### Loss Aversion

Humans feel losses twice as strongly as gains of equal size. After a losing trade, the urge to "make it back" leads to: - Bigger position sizes (revenge trading) - Lower-quality setups (forcing trades) - Wider stops or no stops (refusing to take more losses)

All three accelerate losses. The healthy response after a loss is to do exactly what you would have done if it had been a win — follow your process.

### Sunk Cost Fallacy

You bought YES at $0.50. It's now $0.30. Your average cost is "still" $0.40 if you buy more. Average down? Or accept the loss?

The sunk cost is gone either way. The question is: at $0.30, is YES a good trade? Forget what you paid. Forget what you wished. Look at current price vs your current best estimate of probability.

### Anchoring

You decided this market was "worth $0.60" when it traded at $0.50. Now it's $0.45. You still think it's worth $0.60. Is your estimate updated for the new information that moved the price?

Usually it should be. Markets move because new information arrives. If you stubbornly hold your old estimate, you're trading against the consensus update. Sometimes that's right. Often it's anchoring.

### FOMO

You see a trader on Twitter post a 5x winner. Your bankroll suddenly feels small. You want to find your own 5x winner. So you start trading bigger.

The trader you saw probably has 10 losing trades you didn't see. And even if they're a 5x specialist, that doesn't mean you should change your sizing.

For more on the psychology side, see our [trader psychology guide](/blog/trader-psychology-prediction-markets).

## Risk Management in Practice

Day-to-day workflow that implements all the above:

### Before each trade 1. What's my edge? (Skip if under 3pp) 2. What size? (5% max, half-Kelly default) 3. What's my exit? (Stop loss and take profit defined) 4. Am I exceeding 25% total exposure? (If yes, pass) 5. Am I exceeding 5% in this category? (Diversification check)

### During the trade - Don't watch the price obsessively. You'll be tempted to micro-manage - Set price alerts for stop loss and take profit - Trust your process — you decided the rules pre-trade for a reason

### After each trade - Record outcome in trade journal (Predite tax report can help) - Was the outcome consistent with your probability estimate? - Was your size appropriate given your actual edge? - What would you do differently next time?

### Weekly review - Total exposure within limits? - Drawdown approaching trigger levels? - Any pattern in losses (specific market type, time of day, recent news)? - Anything to adjust in process?

### Monthly review - Net P&L - Win rate - Average win vs average loss (ratio) - Sharpe ratio if calculable - Calibration check: when you said 70%, did you win 70% of those trades?

## Common Failure Modes

**Going all-in on "the one opportunity"**: There's never one opportunity that's worth your full bankroll. There are infinite opportunities over time. Survive to see them.

**Doubling down to recover**: Mathematically guarantees ruin on losing streak. Never bet more after a loss than you would have bet without it.

**Trading bigger after wins**: Variance is variance. Three wins in a row doesn't mean your edge has increased. Stick to the system.

**Skipping stop losses**: "I'll exit when it gets back". You won't. Average position holding time on losing trades without stops: forever, until it goes to zero.

**Confusing leverage with edge**: Bigger position size doesn't create edge. It amplifies whatever your real edge (or lack of) is.

## Bottom Line

Risk management is the unglamorous foundation that makes everything else work. You can have the best probability estimates in the world, but if you bet 20% per trade and hit a routine losing streak, you go bust.

Size small. Stop losses on every position. Bankroll segregated. Track everything. Review weekly. Adjust monthly.

Boring? Yes. But trading is supposed to be boring. If it feels exciting, you're probably about to lose money.

For an integrated approach combining risk management with edge identification, our [+EV trading guide](/blog/what-is-positive-ev-trading) ties the strategy and the discipline together.