Risk disclosure

Futures trading involves substantial risk of loss and is not suitable for every investor. Futures are leveraged. The margin you post is a small fraction of the value of the contract you control, so a small price move can produce a large loss, and losses can exceed the money in your account. If they do, you may owe your broker the difference.

• Markets can gap past your stop: overnight, on news, at the open or when liquidity disappears. A stop order does not guarantee the price you get.
• Your broker can raise margin requirements or close your positions at a loss without asking you first.
• Micro contracts are smaller, but they are still leveraged.
• Prop-firm evaluation fees, resets, platform and data fees can all be lost in full, whether or not you pass.

Only trade with money you can afford to lose. Think about whether trading suits your finances, experience and goals, and talk to a licensed professional if you are unsure.

Nothing in QuantReplay guarantees, promises or predicts that you will make money, earn any particular return, pass a prop-firm evaluation, get a funded account or receive a payout. Many people who trade futures lose money, and many evaluations are failed. Plan for the possibility that you will lose.

Results shown on our website, in examples, templates and demos, are illustrations of how the product works. They are not typical results and say nothing about what you will achieve.

Some charts and backtests on our website use real historical CME futures data, such as a past ES or NQ session. Real data doesn’t make the results real: those backtests are hypothetical, and the sessions shown were chosen to illustrate the product, not because they are typical.

Almost every number QuantReplay shows comes from a simulation, not from real trading. That includes backtests and their reports, the results of changing a bot’s parameters, pass odds and other Monte Carlo figures, replay sessions and their simulated account, Paper deployments, anything run on a simulated feed, the charts and example results on our website (including those built from real historical CME data), and the figures Quinn reports. Past results, real or simulated, do not predict future results.

Our backtest engine makes simplifying assumptions. Real trading will differ, often for the worse:

• It works from 1-minute bars, so it can’t know the order of prices inside a bar. When a stop and a target fall in the same bar, it assumes the stop filled first, but real fills can differ either way.
• Market orders fill at the next bar’s open with a fixed slippage (1 tick by default), and limit orders fill as soon as price touches them. Real slippage can be much larger, especially in fast markets, a real limit order can go unfilled when price only touches it, and real orders can fill partly or not at all.
• It doesn’t model your place in the order queue, the market impact of your orders, or delays between your bot, the bridge and the exchange.
• Commissions and fees are estimates and may not match your broker’s or your firm’s.
• Historical series join contracts across expirations (continuous front-month, rolled by calendar). Prices around a roll can differ from the contract you would actually trade.
• Tuning a strategy until its backtest looks good, including asking Quinn to improve results, tends to fit past noise (overfitting). A strategy that looked great on history can fail at once in live markets.
• A small number of trades or trading days makes every statistic unreliable.

Pass odds resample a backtest’s past trading days through an evaluation’s rules thousands of times. They show how past results interact with the rules, not what will happen next. They assume future days look like past ones, and the rule templates follow common evaluation structures rather than any firm’s exact, current rules.

Replay lets you trade history with a simulated account. Trading in replay carries none of the pressure of real money, and you may trade differently when your own money is at stake.

QuantReplay uses either synthetic sample data for ES and NQ, which is generated rather than real and is labeled sample data wherever it is used, or historical CME futures data licensed from Databento. Anything computed from sample data says nothing about real markets.

Historical data from any source can contain errors, gaps and bad prints. Live data can lag, freeze or drop out. A live Databento feed for deployments is built but not yet tested against the live service, and Paper deployments can run on a simulated feed that replays history (always labeled Simulated feed).

The live chart on our home page is TradingView’s widget showing index CFD prices (FOREX.com’s S&P 500 and Nasdaq 100), not the CME futures that bots trade. Its prices and moves can differ from the futures.

A bot does exactly what its code says, which is not always what you meant. Once deployed, it can trade quickly and repeatedly without you.

• Bots can contain bugs, including code written or changed by Quinn.
• Today the runner that keeps bots going runs on the computer where QuantReplay is installed. If that computer sleeps, restarts, loses power or loses its connection, bots stop receiving data and sending orders.
• Orders can be delayed, rejected or filled at a different price or size than expected. An order whose request times out is marked unconfirmed: it may or may not have reached your broker.
• Webhook bridges (TradersPost, PickMyTrade, custom webhooks) don’t report fills back. Positions on those routes are labeled Estimated and can differ from your broker’s records.
• On routes where no stop is held at the broker, an outage of the runner, your computer or the network can leave a position unprotected.
• When the data feed goes stale, new entries pause, but positions already open stay exposed to the market.
• Stopping a deployment doesn’t close its open position. Use Flatten, or close the position at your broker.
• Native Tradovate and ProjectX connections are experimental: built from the brokers’ documentation, not yet tested against live accounts, and off unless the server enables them.

Watch your positions while bots run. Keep your broker’s platform open and know how to flatten and cancel orders there yourself.

The guard checks each entry against the limits you set (maximum contracts, daily loss, trailing drawdown, trading window and flatten time) and blocks entries whose worst-case stop-out would cross one. If a fill or a fast market trips a limit anyway, it flattens the position and locks the deployment until the next session.

It can still fail to protect you:

• It can act only on the data and fills it receives. On Estimated routes it works from positions that may be wrong.
• It can’t prevent slippage, gaps or broker delays that carry a loss past a limit before it can act.
• The limits you enter can be wrong, or differ from your firm’s actual rules.
• The guard is software and can have bugs.

Your broker’s and your firm’s own risk systems decide what actually happens to your account.

Quinn runs on Anthropic’s Claude models. Like any AI, it can be confidently wrong. It can write code with bugs, misunderstand your idea, misread or misreport results, explain things incorrectly, and state things about markets or firm rules that aren’t true. When it turns your replay trades into a bot, the patterns it finds in a small sample may be coincidence.

Quinn’s replies are not advice. You are responsible for reviewing every line of code and every result before you rely on it, and for testing on Paper before you trade real money with it.

Each firm decides its own rules: drawdowns, daily limits, consistency rules, news restrictions, scaling and payouts. Firms can prohibit or restrict automated trading, copy trading, third-party bridges or API access, and some require orders to come from your own device. Breaking a rule can cost you the account, its fees and any payouts. Many evaluation and “funded” accounts are simulated accounts, and payouts depend on the firm’s terms.

The firm summaries, rule templates and compliance checks in QuantReplay are aids. They can be out of date or wrong. Confirm your firm’s current rules, including its position on automation, directly with the firm before you deploy. QuantReplay is not affiliated with or endorsed by any prop firm, broker or bridge unless we say so.

QuantReplay is software. Nothing it shows, including templates, examples, pass odds, firm summaries, Quinn’s replies and anything on our website, is investment, financial, tax, legal or accounting advice, or a recommendation or solicitation to buy or sell any contract or to use any strategy. None of it is tailored to your circumstances. Taxes on your trading are your responsibility.

QuantReplay is not a broker, futures commission merchant, introducing broker, commodity trading advisor, commodity pool operator, investment adviser or exchange, and it is not registered with the CFTC, the NFA or any other regulator in any of those roles. [Counsel to confirm the registration position and any exemption relied on.]

• It never holds, receives or manages your money. Your funds stay with your broker or firm.
• It has no discretion over your accounts. Orders are sent only by bots you deploy, to accounts you connect, under limits you set.
• It doesn’t trade for you, pool funds or manage accounts for anyone.

You decide whether to trade, what to trade and how much to risk. You are responsible for every order your bots send, for monitoring them, for following your broker’s and firm’s rules, and for making sure trading futures and using automation is legal where you live.

By deploying a bot you confirm that you have read and understood this disclosure. Our terms of service (draft) set out the rest of the agreement.