Risk
disclosure.

StrategyLab is a tool for testing ideas. Every number it shows is the output of a simulation. This page explains what that means, what a simulation leaves out, and how to keep its results in proportion.

Everything here is hypothetical

No trade shown in StrategyLab was ever placed. Backtests apply a rule to past prices. Replay shows the same backtest in order. Live paper trading applies a rule to current prices with paper money. None of them involves a real order, a real counterparty or real money, and none of them is a record of anyone’s actual trading, including ours.

Why hypothetical results differ from real ones

• They are designed with hindsight. You choose a strategy and its settings after seeing how the market behaved. Real trading has no such benefit. The more settings you try, the more likely the best one only fits the past by chance.
• Fills are assumed. The simulation fills orders at the next candle’s opening price, adjusted by the slippage percentage you set. Real fills depend on the order book at that moment, the size of your order and how fast the market is moving. They can be much worse, especially in thin markets or sharp moves.
• Costs are simplified. One fee percentage and one slippage percentage stand in for exchange fees, spreads, funding and borrowing costs, withdrawal fees and taxes. Short positions in particular are modeled simply, with no borrowing cost and no forced liquidation short of the account reaching zero.
• Liquidity is assumed to be unlimited. The simulation never fails to fill and never moves the market.
• Nobody panics in a simulation. A backtest follows its rules through every drawdown. People often do not. The psychological pressure of real losses is not simulated.
• Data has limits. Prices come from one exchange’s public data and can contain gaps, errors or later revisions. Results on another venue would differ.
• The past is one sample. A test covers one stretch of history. Markets change character, and a rule that suited one period can fail in the next.

How to read StrategyLab’s numbers

• Treat every result as a hypothesis to keep testing, not as evidence of what will happen.
• Compare with buy & hold, which the tool shows in the first sentence of every result.
• Take the warnings seriously. “Only 9 trades” means the result could easily be luck.
• Use the out-of-sample split and the parameter sweep to look for overfitting before you believe a result.
• Look at the max drawdown and ask whether you could have sat through it.

Risk of loss

Trading and investing involve substantial risk of loss, including the loss of everything you put in. Crypto assets are especially volatile and can lose most of their value quickly. Strategies that use short positions or leverage can lose more than was invested. Do not trade with money you cannot afford to lose.

No advice, no solicitation

Agent Creative, which operates StrategyLab, is not a registered investment adviser, broker or dealer. Nothing on this site is a recommendation or an offer to buy or sell any asset. The example strategies and their results are illustrations of how the tool works. They are not suggestions. Consider speaking to a licensed professional about your own situation before making financial decisions.

Shared results and exports

Share links and CSV exports carry the disclaimer above. If you pass a simulated result to someone else, pass the disclaimer with it, and do not present it as a record of real trading.

Questions

If anything here is unclear, ask through the Contact page.