Paper Trading Guide for Event Contracts

A strategy that returns 8% in a spreadsheet backtest can lose money the first week it touches a live order book — not because the edge was fake, but because the backtest never had to deal with a stale price, a thin book at exactly the moment you wanted to size up, or the very real urge to abandon your rules after two losses in a row. Paper trading closes that gap. It runs your strategy against live, unresolved Polymarket and Kalshi markets in real time, with no capital at risk, so the first time your rules meet reality isn't also the first time your money is.

Why paper trade instead of going straight to backtesting results

Backtesting and paper trading answer different questions. A backtest tells you whether a rule set would have worked on data you already have — fast to run, but it can't test your live discipline, and it's easy to unknowingly build in lookahead bias or assume fills at prices that were never actually available in size. Paper trading tells you something a backtest can't: whether you can execute the strategy correctly under real, unresolved uncertainty, with genuine unknowns and genuine temptation to deviate from the plan. The two are complementary — backtesting prediction markets against historical order-book data is the right first filter for whether a strategy has any edge at all; paper trading is the second filter for whether you can actually run it.

What to validate during a paper trading run

Paper trading is only useful if it's honest about the frictions a live account would face. Four things specifically need checking, not just P&L:

1. Fee-adjusted P&L, not gross P&L

A simulated trade that "wins" $0.08 per contract on Kalshi looks different once you subtract the actual taker fee — round_up(0.07 × contracts × price × (1 − price)) — from every fill. Strategies that look profitable gross and unprofitable net are extremely common in event contracts specifically, because fees are a meaningfully larger share of a $0.01-$0.99 contract than they are of, say, a $50 stock.

2. Execution against real depth

It's easy to log a simulated fill at the best quoted price without checking whether that price actually had enough size behind it to fill your full order. Compare your paper fills against the order book snapshot at that exact moment — if you were "filled" for 500 contracts at a price where only 80 were resting, your paper results are fiction.

3. Sizing discipline under real conditions

A bankroll management rule (see bankroll management for event trading) is easy to follow on paper when nothing is at stake. Paper trading still surfaces a version of this problem: does your rule set actually produce sane position sizes across correlated markets, like the 500+ active 2026 midterm contracts, or does it accidentally let you stack correlated exposure past your intended cap?

4. Drawdown behavior

Run the paper strategy long enough to hit a losing streak. How does the equity curve behave? Does your drawdown rule actually trigger the way you designed it to, or does the strategy's logic quietly keep sizing up through a bad stretch?

How long to run it

A handful of trades tells you almost nothing — a strategy can look great or terrible off pure variance in a small sample. Aim for at least 30-50 simulated trades, or better, a full natural event cycle: an election night with dozens of resolving markets at once, a complete slate of Fed decisions across a quarter, or a full run of weekly economic releases. The goal is a sample large enough that the result reflects your edge and discipline, not the luck of which five trades happened to land first.

Question	Backtesting answers it	Paper trading answers it
Did this rule set have edge historically?	Yes	Partially — only over the live window tested
Will I actually follow my own sizing rules?	No	Yes
Are my fill-price assumptions realistic?	Only as good as the historical data	Yes — tested against live depth
Does my drawdown rule actually trigger?	Can be simulated	Tested under live pressure

When to graduate to live capital

Move to a funded account once — and only once — sizing, fees, drawdown rules, and execution assumptions have all held up across your full paper run, not just the P&L line. A paper strategy that made money but only because you quietly ignored your own position caps twice hasn't actually been validated; it's told you the rules don't hold under your own hand, which is exactly the thing live capital would expose at a much higher cost. This applies equally to a directional strategy or a two-sided quoting approach — see market making explained for how liquidity-providing strategies specifically need paper validation of fill rates before any capital is at risk.

Where PolyMarketMaker's Simulation tier fits

Running a disciplined paper trade manually — tracking fees, checking fills against real depth, logging every trade with reasoning — is tedious enough that most traders skip half of it and get an inflated read on their own strategy. PolyMarketMaker's Simulation tier is built around exactly this workflow: live Polymarket and Kalshi order-book data, real fee schedules applied automatically to every simulated fill, and the same order-book ladder, depth charts, and candle tools you'd use live, all running against paper capital. It's the same terminal, the same data, the same fee math — the only thing missing is real money on the line, which is the point. PolyMarketMaker Simulation runs $149/mo; Live Trading with the full automated quoter and safety rails is $299/mo once you're ready to graduate.

Common paper trading mistakes

The single most common failure is a fee-free paper environment — traders track gross wins and losses, skip the fee deduction "for now," and get a strategy that looks solidly profitable until real fees are applied on day one live. The second most common is trading a small, cherry-picked sample of easy-looking setups rather than every trade the rules would have signaled, which quietly filters out the losing trades a disciplined system would have taken too. Both mistakes produce the same result: a paper track record that doesn't survive contact with a live account. The fix for both is mechanical — apply real fees to every simulated fill, and take every signal the rules generate, not just the comfortable ones.

FAQ

What is paper trading in prediction markets?

Simulating trades against live, real-time Polymarket or Kalshi order books without risking real capital, to validate a strategy's execution and your own discipline before funding a live account.

How is paper trading different from backtesting?

Backtesting runs against historical data and is fast but can hide lookahead bias and unrealistic fills. Paper trading runs live against unresolved markets in real time, testing decision-making under genuine uncertainty.

How long should I paper trade before going live?

At least 30-50 simulated trades, or a full event cycle like an election night or a quarter of Fed decisions, so results reflect real edge rather than a small lucky or unlucky sample.

Does paper trading account for fees and slippage?

Only if you apply the venue's actual fee formula to every simulated fill and check that fill against real order-book depth — skipping either step overstates results.

Can I paper trade Polymarket and Kalshi together?

Yes, and it's specifically useful for cross-venue strategies like arbitrage, since fee schedules and liquidity differ enough between the two that a strategy validated on one doesn't automatically transfer to the other.

This article is for educational purposes only and is not financial advice. Trading involves risk of loss.