A Polymarket stop loss bot sounds simple: enter a high-probability position, then sell if its price falls. The dangerous assumption is that a price threshold automatically produces an exit. It does not. A reliable automation needs two separate systems: a trigger that decides when risk conditions have changed, and an execution policy that decides how, when, and whether to send a sell order.

That distinction matters most when the market is moving quickly or the order book is thin. A displayed price can be stale, a midpoint can be untradeable, and a sell order can be partially filled or rejected by the conditions you set. The objective is not to create a fictional guaranteed stop. It is to define the loss, liquidity, and execution compromises you are willing to accept before the market forces you to decide under pressure.

A stop is a trigger plus an execution policy

Treat these as separate objects in your rules. The trigger is an observation: for example, “consider exiting if the best executable bid is 75 cents or lower.” The execution policy is the response: for example, “sell up to 100 shares with a 73-cent limit, only if the spread is no more than 3 cents; accept an immediate partial fill; retry once if shares remain; otherwise stop automation and alert me.”

This is the right mental model because a stop price is not a promised fill price. FINRA explains this general execution principle in the context of volatile markets: a stop can trigger an order without guaranteeing its eventual execution price, while a stop-limit adds price control but can fail to execute. That is an analogy about order mechanics, not a Polymarket-specific rule or regulation. See FINRA’s discussion of stop-order execution.

The five prices that should not be confused

• Displayed price: a value shown in an interface, such as a last trade or midpoint. It is useful context, not necessarily the price at which you can sell now.
• Best bid: the highest current price a buyer is offering. For a seller, this is the first executable price level.
• Trigger price: the rule threshold that starts exit evaluation. It should be based on a defined data field, not an ambiguous screen price.
• Order price: the limit you actually submit. It specifies the worst price you will accept for the shares it covers.
• Fill price: the price or weighted average prices your order actually receives. It can differ from the trigger and can be split across several book levels.

Polymarket uses a central limit order book. Buyers take the best ask, while sellers receive the best bid; the midpoint or last-traded value may therefore differ from the available execution price. Its documentation also advises traders to inspect book depth for larger orders. Read the underlying mechanics in Polymarket’s Prices & Orderbook guide.

Work through a 90-cent entry and 75-cent exit example

Assume you buy 100 YES shares at 90 cents, for a $90 position cost before any applicable fees. You want to reduce risk if the market weakens. “Exit at 75 cents” is incomplete because it does not say which price is observed, what price is acceptable, or what happens when there is not enough demand.

A more operational rule could be: trigger an exit review when the real-time best bid is at or below 75 cents. At that moment, calculate the sellable quantity available at bids of 73 cents or better. If at least the chosen minimum quantity is available and the bid-ask spread is 3 cents or less, submit a 73-cent sell limit for the remaining shares. If only part fills immediately, follow the selected partial-fill policy. If no acceptable fill arrives within the timeout, cancel any live remainder and stop for review.

Suppose the displayed midpoint is 75 cents, but the book is 72 cents bid and 78 cents ask. A 75-cent trigger based on the midpoint may fire, yet a seller can immediately receive only 72 cents if they are willing to take the bid. If the 72-cent bid contains just 30 shares, selling 100 shares aggressively may consume lower bids as well. That is slippage: the difference between the price you expected from the trigger and the price achieved across actual fills.

Spread and depth answer different questions

The spread is the gap between the best bid and best ask. A wide spread is a warning that the cost of immediate execution is high or that the market is uncertain. Depth is the quantity available at each price level. A narrow spread does not prove that your entire position can exit near the best bid; there may be only a small number of shares waiting there.

Before placing an exit, estimate the proceeds from selling your intended quantity through the bid levels at or above your minimum acceptable price. This estimate is still conditional. Polymarket notes that an order-book estimate is a snapshot, and price protection can result in a partial fill or no fill when the book changes. The relevant mechanics are documented in Polymarket’s Place Orders guide.

Choose an exit policy deliberately

There is no universally correct choice between certainty of action and price protection. The right policy depends on the position size, remaining downside, market liquidity, and whether you can monitor an incomplete exit. The mistake is leaving that choice implicit.

Partial fills: FAK versus FOK

A Fill-and-Kill (FAK) order fills whatever quantity is immediately available under its terms and cancels the remainder. A Fill-or-Kill (FOK) order requires the entire quantity to fill immediately or it cancels in full. Polymarket documents both behaviors. FAK is often the more practical risk-reduction tool because it can reduce exposure even when the full position cannot exit at once. But it creates a new state: a smaller residual position that needs an explicit follow-up rule.

• Use FAK when reducing some exposure promptly is more important than exiting every share at exactly the same time. Define whether the bot may retry the residual amount and how many times.
• Use FOK when an incomplete exit would create an unacceptable position or operational problem. Accept that the result may be no exit at all.
• Do not interpret either instruction as a liquidity guarantee. Both depend on the live book when the order reaches it.

Marketable limit versus unbounded urgency

For an automated stop, a limit price provides a hard floor: sell only at or above the specified price. A limit can be made immediately executable by setting it at a price that can meet current bids, but it should still incorporate a defined worst acceptable price. Sending an order without a clear price boundary may solve urgency while making slippage unbounded in the rule design. Conversely, a tight limit can preserve price discipline but leave the position open as the market falls.

Build liquidity and failure controls into the rules

A stop-loss bot should be conservative when its inputs are unreliable. Polymarket’s real-time streams provide order-book updates, price changes, best bid and ask, spread, and timestamps that an automated workflow can use to assess executable conditions. See Polymarket’s Real-Time Data documentation. The key is not merely receiving a stream; it is refusing to act when the information is too old or internally inconsistent.

• Maximum position size: cap the number of shares or dollars that any one preset may open. The exit logic should be tested against that full capped size, not a small sample trade.
• Maximum spread: decline to place the normal exit order when the spread exceeds a defined threshold. Your escalation policy might pause, reduce size, widen the acceptable limit within a pre-set boundary, or require review.
• Minimum executable depth: require enough bid quantity at or above the limit for a minimum exit quantity. This prevents a single small quote from falsely suggesting that a full exit is available.
• Data staleness limit: if the latest market timestamp is older than your allowed interval, do not treat the best bid as current. Pause and wait for fresh data rather than acting on an old book.
• Timeout: define how long an order may remain active. At expiry, cancel it, reconcile its actual fills, and decide whether one bounded retry is permitted.
• Retry limit: specify the number of retries, the delay, and whether the minimum price may change. Unlimited retries can become uncontrolled chasing.
• Kill switch: halt new entries and cancel or stop managing orders when a critical condition occurs, such as repeated errors, stale data, an unexpected position mismatch, or a breached daily loss boundary.

Position limits are the first line of defense because liquidity is not independent of trade size. A position that is manageable at 20 shares may be difficult to unwind at 500. Set your maximum position from the loss you can tolerate if the stop fills worse than planned, or does not fill promptly—not from the most attractive quoted price. For the broader sizing framework, see bankroll management.

Translate the policy into no-code OddsFantasy presets

A preset should make the trading decision reproducible, not hide it behind automation. Start with the complete process in the no-code Polymarket trading workflow: market selection, entry conditions, stake limits, exit conditions, and review. In OddsFantasy, keep the entry preset and its linked exit preset as separate objects so each rule set remains auditable.

1. Create an entry preset. Define the markets, chart interval and timeframe, odds rule, spread rule, volume or liquidity filters, market-structure conditions, and stake per market. Save only conditions you intend the automation to enforce.
2. Create a protective exit preset and link it to the relevant entry preset. Enable the maximum-drawdown close, choose when checking begins before kickoff, and decide whether the same rule may continue live.
3. Choose the supported execution mode deliberately. OddsFantasy exposes Market (FAK), which takes available bids after confirmation, and Limit (GTC), which posts at the current ask and waits. FAK can reduce exposure immediately but may leave a residual position; GTC preserves price control but may remain unfilled.
4. Use the native spread-confirmation controls. Set the maximum spread and require the drawdown and spread conditions to hold for the full confirmation time. This avoids treating a single transient quote as a complete exit signal.
5. Add only the other exit modules the strategy needs, such as pre-match profit or time closes, live profit or time closes, or live momentum. More rules do not automatically make the strategy safer; overlapping conditions must still produce an outcome you can explain.
6. Document the safeguards that are operational checks rather than preset fields, including minimum executable depth, acceptable quote age, retry escalation, and the account-level kill switch. Keep manual oversight wherever the current product controls do not encode the full policy.
7. Save the linked presets in OddsFantasy Trade, review the summary, and test the workflow with deliberately small exposure before relying on it for a larger position.

The useful test is adversarial: ask what the preset does when the bid disappears, only 10% of the position fills, the stream pauses, or the market reopens at a lower price. If the answer is “the bot will probably get out,” the rules are not complete. If the answer specifies the trigger source, acceptable price range, residual handling, timeout, retry cap, and halt condition, the automation is at least auditable.

Preflight checklist before enabling automation

• I know whether the trigger uses best bid, best ask, last trade, or midpoint, and I can explain why.
• My exit limit defines the worst acceptable price; I am not treating the trigger as a guaranteed fill.
• My maximum position is small enough to be plausibly supported by the depth I inspect.
• I have set a maximum spread, minimum depth, and maximum data age.
• I have selected FAK or FOK intentionally and documented the treatment of any residual position.
• Every order has a timeout, bounded retry rule, and a reconciliation step for actual fills.
• My kill switch stops new exposure when data, order status, or risk limits are unreliable.
• I have reviewed the preset using small size and adverse scenarios, not just the ideal path.

Automation can make a risk process more consistent, but it cannot manufacture liquidity or guarantee an exit. Build the stop as a transparent trigger-and-execution policy, keep its limits conservative, and revise it when real fills show that its assumptions were too optimistic. When you are ready to turn defined rules into reusable no-code configurations, explore OddsFantasy Trade.