Strategy Lab
Strategy Lab — test the entry rule
Change the entry rule below and the Lab re-runs the signal counts on the same real daily closes the backtest used. Every change recomputes instantly, in your browser, from the data file — no server, no estimates.
Entry rule
Bands
10–50, default 20
1–3, default 2.0
Regime
Default: the day’s low touches or crosses below the lower band.
Default: close above the 200-day SMA. Off counts every touch as a signal.
Strike proxy
5–15%, default 10%. Sets the “staying above” threshold below.
Universe
Only tickers with price history on file are offered. BMNR’s history starts 2025-06-05.
Range
The rule is always evaluated on the full history (bands and averages need the earlier bars); only the signal dates are filtered to this range.
Loading…
Loading the Lab data…
How it is computed
- Signal. A day counts as a signal when the touch price (the day’s low by default, or the day’s close if you switch it) touches or crosses below the lower Bollinger Band — the N-day mean minus the multiplier times the population standard deviation — and the day’s close is above the regime SMA. This is the same definition as the live screener and the backtest.
- Touches vs signals. A touch is a band-touch day, whether or not it passes the regime filter. The table shows both, so you can see how much the regime filter actually cuts.
- % touches bullish. The share of touch days whose close was above the reference SMA (the selected length; the 200-day when the filter is off). It is measured over touches, not signals: with the filter on, every signal is bullish by definition, so a percentage over signals would read 100% and mean nothing.
- Avg days between signals. The mean calendar-day gap between consecutive signal dates.
- Staying above −OTM%. The share of signals whose lowest close over the next 45 trading days stayed at or above minus the OTM% you set. Signals too recent for a full 45-day window are excluded, not assumed. It is a proxy for price staying above the short-strike region, not assignment — and no premium is involved.
- Date range. Bands and averages are always computed on the full history, then signal dates are filtered to the range you pick. Narrowing the range never changes the rule itself.
- Data. Daily closes and lows from yfinance, the same source as the backtest closes. SPY/QQQ/IWM/DIA are the exact bars the backtest used; NFLX/BMNR are yfinance daily bars to the same end date. Closes are not CBOE data. BMNR only has 364 trading days (from 2025-06-05), so its 200-day average is evaluable on just 165 days — treat its counts as thin. The 5-year tickers have 1,056 evaluable days with the 200-day average.
- What is not claimed. No premium collected, no put profit or loss, no trade win rate, and no annualised return. A premium backtest needs recorded option prices, which only began accumulating recently.
Scenarios — what happened after past events
Base rates from history, not forecasts. The last N times an event happened, this is what followed on average over 7/14/30 trading days. Past behaviour does not predict future outcomes. Recomputed weekly (Mondays) from 10 years of Yahoo history.
Loading scenarios…
Education only. Past price behaviour after a signal does not predict future outcomes.