Capital for Polymarket arbitrage: How much to start

Short answer: most practitioners start with between $1,000 and $10,000 of deployable pUSD. That range balances execution, fees, and the minimum edge sizes you’ll see on Polymarket. This guide walks through why that floor exists, how to size trades, and the practical constraints that push a beginner toward the $1k–$10k range.

Key takeaways

• Expect a working minimum of roughly $1,000 to participate reliably in intra-market arbitrage; $5k–$10k is common for repeatable, automated flows.
• The edge is the arithmetic gap between $1.00 and the sum of best-ask legs; you must net that against taker fees, slippage, and settlement timing.
• Maker fees are zero; taker fees vary by category (0%–1.8%) — include them in sizing math.
• Small accounts suffer from tick-size, order-book depth, and minimum-lot effects more than larger accounts.
• Always enumerate resolution, slippage, partial-fill, fee, and smart-contract risks when calling a spread "mathematical."

Why $1k is a practical lower bound

Two sets of frictions create a real-world minimum: discrete pricing/ticks and taker fees plus book depth.

Tick size and minimum meaningful edge

On Polymarket most markets use a $0.01 tick (tightening to $0.001 near extremes). If the cheapest combined best asks create an edge of $0.02 and your strategy requires filling both legs, you need enough notional for that $0.02 to be economically worthwhile after fees and execution risk. With very small capital the per-trade absolute gain is tiny and can be consumed by the smallest practical fee or a single partial fill.

Order-book depth and slippage

Thin books mean your market/FAK orders lift the book and incur slippage. With $100 you might clear only a few cents of book on each leg; with $1,000 you commonly reach deeper resting liquidity and capture larger, repeatable edges.

Fees and minimum returns

Maker fees are zero on Polymarket, but taker fees vary by category (currently up to 1.8% in some categories). That taker fee is applied to the leg you hit as a taker. If your edge is 1–2%, a taker fee of 1% can halve or eliminate profitable opportunities. Starting capital needs to be large enough that absolute profit exceeds expected fees and the cost of failed attempts.

Why many traders prefer $5k–$10k

Consistency and automation

When you automate, you want trades to be meaningful even after occasional partial fills and short-lived price moves. $5k–$10k puts you in a band where common intra-market spreads (2–3% on liquid markets historically) translate into enough USD per trade to justify the automation, monitoring, and infrastructure.

Ability to diversify

With more capital you can run multiple lines across different markets and market states simultaneously. Diversification reduces single-market idiosyncratic risk: a few losing fills won’t wipe out your account if other lines are working.

Builder Program and relayer limits

If you plan to participate in Polymarket’s Builder Program later, understand the tiered relayer limits for routing orders. Limits affect throughput and therefore the amount of capital you can effectively deploy in automated strategies. Check Polymarket settings for credential steps if you scale.

Sizing a single arbitrage leg: a simple worksheet

1. Estimate raw edge: Edge = 1.00 − Σ bestAsk(outcomes).
2. Subtract expected taker fees (use the market category fee band). If you don’t know the exact fee, assume a conservative upper bound from the documented fee range.
3. Estimate expected slippage and partial-fill loss (expressed as a percent or absolute USD). For small books use a higher slippage estimate.
4. Required capital per trade = (Minimum acceptable USD profit) / (Net edge after fees and slippage).

Example outline (numbers illustrative only):

• Raw edge: 2% on a $1,000 notional = $20.
• Taker fees: 1% on legs = $10 (approx).
• Estimated slippage/partial fills: $5 expected.
• Net expected profit: $5. That’s small but repeatable. If you want $50 typical profit per attempt, you need ~10× the notional.

Note: do not treat these figures as guaranteed returns. They are a sizing framework.

Practical constraints that inflate capital needs

• Tick-size effects: very small accounts hit quantisation noise.
• Minimum meaningful bet size: when a single cent or tick is the only available improvement, very small accounts can’t express the needed granularity.
• Settlement and resolution timing: locked capital until UMA resolution or dispute reduces capital turnover. Faster turnover requires more parallel capital.
• Geo and account restrictions: Polymarket restricts orders by IP and jurisdiction; ensure you are eligible under Polymarket’s published policy.

Risk checklist (always run through this before you deploy capital)

• Resolution risk: UMA disputes can delay or change outcomes; funds can be locked longer than expected.
• Slippage and partial fills: FAK orders may partially fill; partial fills can leave you directionally exposed.
• Fee risk: Taker fees vary by category and can materially reduce small edges.
• Smart-contract and counterparty risk: trades use CTF and the Polymarket relayer; these are vectors for technical failure.
• Regulatory and geographic risk: Polymarket blocks certain jurisdictions and forbids VPN evasion.

Never call a spread strictly "risk-free" without listing the above items.

How this affects your trading

If you have under $1k in deployable pUSD, focus on learning and manual execution rather than aiming for automated, high-frequency flows. Use that time to practice reading order books, understanding tick-size effects, and sizing FAK orders.

Between $1k and $5k you can execute opportunistic intra-market arb manually and occasionally automate single-line strategies. Expect longer times between meaningful fills and treat each trade as an experiment.

From $5k–$10k you can build reliable, repeatable flows, diversify across markets, and absorb occasional execution losses. That band is where many semi-professional arbitrageurs start to scale.

If you plan to scale beyond that, factor in relayer throughput, Builder Program tiers, and the operational costs of monitoring and running bots.

Closing thoughts

Capital for Polymarket arbitrage is less about a magic number and more about matching your objectives to market realities. $1k–$10k is a practical starting band because it balances tick-size, book depth, and fee drag against realistic edges. Size trades conservatively, always include taker fees and slippage in your math, and keep an explicit checklist for resolution and operational risks. The arithmetic of edge matters, but so does the operational context you bring to exploit it.

For a complete primer on techniques and execution, see the guide: /guides/polymarket-arbitrage-complete-guide and the technical overview of the market engine at /guides/polymarket-clob-explained.

Frequently asked questions

Is $1,000 enough to profit from Polymarket arbitrage?

You can begin with $1,000, but expect limited throughput and higher sensitivity to tick size, slippage, and taker fees. $1,000 is commonly treated as a practical lower bound for testing strategies; many traders scale to $5k–$10k for repeatability.

How do fees affect minimum capital sizing?

Maker fees are zero, but taker fees vary by category (0%–1.8%). Taker fees directly reduce the net edge on the leg you take. Always subtract expected taker fees from your raw edge when sizing capital — if fees approximate the edge, the trade may be unprofitable.

Can I start with under $1,000 if I only do endgame trades?

Endgame trades (prices near 0.95–0.99) can be executed with smaller capital, but they carry higher catastrophe and resolution risk and are not guaranteed. If you attempt endgame strategies with limited capital, accept that each trade is higher risk and less diversified.

Do I need a Builder account or special relayer credentials to start?

No. You can trade through the public Relayer and a regular wallet. Builder tiers exist for third parties routing orders with attribution and higher rate limits; they matter only when you scale throughput beyond the default relayer constraints.

How should I think about position sizing per trade?

Size each trade so that the expected USD profit after estimated taker fees and slippage meets your minimum acceptable return. Use conservative slippage estimates for thin books and avoid allocating your entire capital to a single trade.