Multi-Coin Scanner
Runs one strategy across many coins with a walk-forward split: the in-sample segment is where an edge would have been discovered, the out-of-sample segment is scored blind. Results are ranked by OOS Sharpe and summarized by how many coins stay profitable — a formula that only works on one cherry-picked chart is noise. Note the coin list itself is today's survivors, so historical results carry survivorship bias.
Signals are computed once per coin over the full window and sliced at the split date; the OOS simulation starts flat with full warmup history, exactly like an account funded on that day.
Scan the MA Cross 5/20 strategy across every coin
One chart can flatter any strategy, so this scan runs the MA Cross 5/20 rule — long on the 5/20 cross up, flat on the cross down — across the whole coin universe in one pass. Each coin's history is split walk-forward: the in-sample segment is where the edge would have been discovered, the out-of-sample segment is scored blind.
The ranking defaults to out-of-sample Sharpe, and the summary shows how many coins stayed profitable after the split. A trend-following rule like this one lives or dies by regime: the scan makes visible whether the recent trend paid for the whipsaws across the market, or only on a lucky few charts.
How it works
- Pick an interval, the coin universe (majors or all) and the walk-forward split.
- The strategy formula loads into the editor, fully editable.
- Set fees and slippage, then scan — every coin is backtested with the same rules.
- Rank by out-of-sample Sharpe, return, drawdown or win rate.
Frequently asked questions
What does the walk-forward split in this MA cross scan mean?
Each coin's candles are divided into an in-sample segment (typically 70%) and a held-out out-of-sample segment. Metrics marked OOS — the default ranking — come only from the held-out data, i.e. performance the strategy was never tuned on.
Why scan the MA 5/20 cross across many coins instead of one?
A crossover that only works on one cherry-picked chart is usually curve-fitting. If the rule earns positive out-of-sample Sharpe on a broad set of coins, the edge is more likely the market's trend behavior than a lucky parameter choice — and the coins where it fails tell you where regimes differ.
More multi-coin scanner pages
MACD Cross · RSI 30/70 · Bollinger Breakout · KDJ Cross · Donchian 20/10
Related tools
Moving Averages in the Indicator Lab · MA Cross Signals in the Indicator Lab