Trader Playbook: Market Making on Prediction Markets

# Trader Playbook: Market Making on Prediction Markets **Market making on prediction markets is one of the most reliable edges available to sophisticated traders — if you know how to manage spreads, inventory, and information flow simultaneously.** Unlike directional betting, market makers profit from the bid-ask spread by continuously quoting both sides of a contract, capturing the difference while staying roughly delta-neutral. Done correctly, this approach can generate consistent returns of 15–40% annually on deployed capital, independent of which outcome actually resolves. This playbook is built for power users who already understand the basics and want a systematic, repeatable framework for running a market making operation on platforms like Polymarket and Kalshi. --- ## What Makes Prediction Markets Ideal for Market Making? Prediction markets have structural properties that traditional financial markets lack — and those quirks create **unique market making opportunities**. First, contracts have binary or bounded payoffs (typically $0 or $1). This caps your maximum loss per contract and makes position sizing more predictable than, say, options on equities. Second, most prediction markets are **chronically underliquidity** — order books are thin, spreads are wide, and retail flow dominates. That's the market maker's paradise. Third, unlike equity markets where high-frequency firms have millisecond advantages, prediction markets still reward **information-driven spread setting** and **smart inventory management** over raw speed. A solo power user with the right framework can compete effectively. Key structural advantages for market makers: - Wide natural spreads (often 3–8 cents on mid-price) compared to <0.01% in equity markets - Low adverse selection from retail-dominated flow - Binary payoff structure simplifies delta hedging - Events with long resolution windows allow slow, patient liquidity provision --- ## The Core Market Making Framework: Spreads, Inventory, and Edge Every market making operation rests on three pillars. Master all three or the strategy breaks down. ### Setting Your Quoted Spread Your **quoted spread** is the gap between your bid and your ask. On a contract you believe has a true probability of 55%, you might quote 52–58 cents. The 6-cent spread means you earn ~3 cents per side per trade — before adverse selection costs. Spread calculation formula: ``` Minimum Spread = (Adverse Selection Cost × 2) + (Operational Cost per Trade) ``` In practice, on active Polymarket contracts, competitive spreads run **2–5 cents** on popular markets and **5–15 cents** on thin markets. Start wider and tighten as you gather data on flow toxicity. ### Managing Inventory Risk **Inventory risk** is the enemy of every market maker. If you quote both sides and the market moves against your accumulated position, you bleed. Here's how professionals handle it: 1. **Set hard inventory limits** — never hold more than X contracts net long or short in a single market 2. **Skew quotes dynamically** — if you're long 500 shares, lower your bid and raise your ask to attract sell flow 3. **Use correlated markets as hedges** — if you're long "Democrat wins Senate" on one platform, short a correlated contract elsewhere 4. **Time your positions** to resolution — longer-dated contracts give you more rebalancing runway For a practical deep dive into automating this logic, the [AI Market Making Playbook: Trading Prediction Markets](/blog/ai-market-making-playbook-trading-prediction-markets) is required reading for anyone moving beyond manual operations. ### Estimating Your True Edge Before quoting any market, you need a **probability estimate independent of the current market price**. This is your "fair value." Your edge is the gap between your fair value and where you can profitably quote. Sources for building fair value models: - Historical base rates (e.g., incumbents win X% of elections under Y conditions) - Polling aggregates and forecast models - Implied probabilities from correlated markets on other platforms - LLM-assisted signal extraction from news — see this [real case study on LLM-powered trade signals with limit orders](/blog/llm-powered-trade-signals-with-limit-orders-a-real-case-study) for a practical example --- ## Automation: The Non-Negotiable for Serious Market Makers Manual market making is viable for learning, but it does not scale. You cannot monitor 20 markets, reprice quotes every 30 seconds, and manage inventory manually without making costly errors. **Automation handles:** - Continuous quote refresh based on inventory and fair value updates - Automatic spread widening when news events create adverse selection risk - Position flattening when inventory limits are breached - Cross-platform arbitrage capture when prices diverge The comparison between the two dominant platforms matters here. If you're deciding where to deploy your market making bot, this [full guide on automating Polymarket vs Kalshi in 2026](/blog/automating-polymarket-vs-kalshi-in-2026-full-guide) breaks down API capabilities, rate limits, and fee structures in detail. ### Step-by-Step: Building a Basic Market Making Bot 1. **Define your market universe** — start with 5–10 high-volume markets in one category (politics, sports, crypto) 2. **Build a fair value model** — even a simple one based on base rates beats quoting blindly 3. **Set inventory parameters** — max net position, skew triggers, and emergency flatten thresholds 4. **Connect to platform API** — use REST for order management and WebSocket for real-time book updates 5. **Implement quote refresh logic** — reprice every N seconds or on every book change above threshold delta 6. **Add adversarial filters** — pause quoting when volume spikes abnormally (news events, resolution signals) 7. **Log everything** — P&L per market, fill rates, inventory drift, and spread capture metrics 8. **Iterate on fair value** — backtest your estimates against resolutions weekly [PredictEngine](/) provides pre-built infrastructure for many of these steps, reducing time-to-deployment significantly for power users who don't want to build from scratch. --- ## Platform Comparison: Polymarket vs Kalshi for Market Makers Choosing the right venue matters. Here's a structured comparison for market makers specifically: | Feature | Polymarket | Kalshi | |---|---|---| | **Order Book Type** | CLOB (Central Limit Order Book) | CLOB | | **API Access** | Public REST + WebSocket | Public REST + WebSocket | | **Maker Fee** | 0% (zero maker fee) | Variable, ~0–1% | | **Taker Fee** | ~2% of winnings | ~1–3% of winnings | | **Contract Types** | Binary, multi-outcome | Binary, series contracts | | **Typical Spread (popular markets)** | 2–5 cents | 3–7 cents | | **Typical Spread (niche markets)** | 8–20 cents | 10–25 cents | | **Volume (daily, 2025 avg)** | $10M–$50M+ | $2M–$15M | | **Regulatory Status** | CFTC-exempt (offshore) | CFTC-regulated | | **Best For** | High-volume, fast markets | Regulated, institutional flow | **Key insight:** Polymarket's zero maker fee structure is explicitly designed to incentivize market makers. If your fill rate is high enough, the spread capture net of fees is materially better than Kalshi for pure market making. However, Kalshi's regulated status attracts different — sometimes more predictable — order flow. --- ## Advanced Tactics for Power Users ### Adverse Selection Defense The biggest threat to a market maker isn't a slow day — it's getting picked off by informed traders who know something you don't. **Adverse selection** occurs when your counterparty has better information than your model. Defense tactics: - **Monitor news velocity** — if a market's Twitter/X mention rate spikes, widen spreads or pause immediately - **Track order flow imbalance** — if you're consistently getting filled only on one side, someone is directionally trading against you - **Set asymmetric spreads** for markets with binary news risk (e.g., FDA decisions, election nights) - Use the [algorithmic NLP strategy compilation for power users](/blog/algorithmic-nlp-strategy-compilation-for-power-users) to build automated news monitoring into your quoting system ### Cross-Platform Spread Capture One of the highest-Sharpe strategies for market makers is **cross-platform arbitrage** — quoting one side on Polymarket and the opposite on Kalshi (or another venue) when the same underlying event trades at different prices. This isn't pure arbitrage (resolution risk exists if platforms resolve differently), but the spread capture is often 5–15 cents per round trip on correlated contracts. For a systematic approach, review the [cross-platform prediction arbitrage quick reference for Q2 2026](/blog/cross-platform-prediction-arbitrage-quick-reference-q2-2026). ### Inventory Recycling via Mean Reversion When inventory builds up in one direction, rather than panic-flattening at a loss, experienced market makers use **mean reversion strategies** to work the position back to neutral over time. This means actively quoting the contra side more aggressively until inventory normalizes. This approach works best in markets with no imminent resolution catalyst. The [mean reversion strategies guide for a $10k portfolio](/blog/mean-reversion-strategies-advanced-tactics-for-a-10k-portfolio) covers the mechanics in detail, including entry/exit rules and position sizing logic. ### Specialization by Market Category Generalist market makers get destroyed by specialists. Pick a niche where your information edge or model quality is above average: - **Politics/Elections** — high volume, well-studied base rates, rich external data sources - **Sports** — fast resolution, correlated with traditional sportsbook lines, good for model validation - **Crypto** — high volatility requires wider spreads but also generates more fee revenue - **Entertainment** — often mispriced by casual bettors, lower adverse selection risk --- ## Risk Management Rules Every Market Maker Needs Even the best strategy blows up without hard risk limits. These are non-negotiable: - **Maximum daily loss limit:** 2–3% of total deployed capital. Auto-pause all quoting when breached. - **Single market concentration:** Never more than 15% of capital tied to one contract. - **Resolution day protocol:** Widen spreads to 2× normal within 24 hours of any market's resolution. - **Correlated exposure cap:** If three markets all resolve on the same political event, treat them as one position for sizing purposes. - **Counterparty risk:** On decentralized platforms, smart contract risk is real. Don't deploy more than you'd accept losing to a protocol exploit. --- ## Metrics That Actually Matter for Market Makers Stop tracking P&L in isolation. These are the metrics that tell you whether your operation is healthy: | Metric | What It Measures | Target Range | |---|---|---| | **Spread Capture Rate** | % of quoted spread actually captured per fill | 40–70% | | **Fill Rate** | % of quotes that get filled | 15–35% | | **Adverse Selection Ratio** | % of fills that move against you within 1 hour | <30% | | **Inventory Turnover** | How often net position flips per day | 2–5× | | **Sharpe Ratio (monthly)** | Risk-adjusted returns | >1.5 | | **Return on Deployed Capital** | Net P&L / capital at risk | 1–4% monthly | Track these weekly. A rising adverse selection ratio is an early warning sign that informed flow is entering your markets before you've updated your fair value. --- ## Frequently Asked Questions ## What is market making on prediction markets? Market making on prediction markets means continuously quoting both a buy price (bid) and a sell price (ask) on binary event contracts, profiting from the spread between them. Market makers provide liquidity to other traders and earn the bid-ask spread in exchange for bearing short-term inventory risk. It differs from directional trading because your primary edge comes from flow and spread capture, not from predicting outcomes. ## How much capital do I need to start market making on prediction markets? Most power users start with $5,000–$25,000 of deployed capital to market make across 10–20 active markets simultaneously. Below $5,000, position sizing constraints limit how many markets you can quote meaningfully, which reduces diversification and increases single-market risk. Above $100,000, you start needing more sophisticated automation and risk systems to manage inventory efficiently. ## What is the biggest risk for prediction market market makers? **Adverse selection** — getting consistently filled by traders who have better information than your model — is the primary risk. The second biggest risk is inventory concentration: if you accumulate a large one-sided position and the market resolves against you, no amount of spread capture covers the loss. Robust inventory limits and news-monitoring systems are essential defenses. ## Can I automate market making on Polymarket and Kalshi simultaneously? Yes, and running on both platforms simultaneously is actually a best practice because it allows cross-platform arbitrage and diversifies your fill sources. Both platforms offer public APIs with REST and WebSocket endpoints. The main challenge is managing combined inventory across platforms and ensuring your fair value model updates propagate to both quote engines in real time. [PredictEngine](/) supports multi-platform automation natively. ## How do I handle market making around major news events? The standard protocol is to **pause or significantly widen quotes** (2–4× normal spread) when a major news event that affects your markets is imminent or breaking. This reduces fill probability but protects against adverse selection from traders who processed the news faster than your model. Resume normal quoting only after you've updated your fair value estimate based on the new information. ## Is prediction market market making legal? Legality depends on your jurisdiction and the platform. Kalshi is CFTC-regulated and open to eligible US contract participants. Polymarket operates offshore and restricts US-based users per its terms of service. Always verify your jurisdiction's rules and the platform's terms before deploying capital. Nothing in this article constitutes legal or financial advice. --- ## Start Building Your Market Making Edge Today Market making on prediction markets rewards patience, systematic thinking, and continuous iteration. The traders generating consistent 20–30% annual returns on this strategy aren't smarter than everyone else — they've built better models, harder risk rules, and more disciplined automation than their competition. If you're ready to move from manual trading to a fully automated market making operation, [PredictEngine](/) gives you the infrastructure, signal tools, and platform integrations to deploy faster and trade smarter. Whether you're quoting politics markets, sports events, or crypto outcomes, the playbook above gives you the framework — PredictEngine gives you the edge to execute it at scale. Start your free trial today and put this playbook into action.