Strategy Health
Updated Nov 13, 2026
Strategy Health checks two things about each of your bots: whether its latest trades are in line with its own earlier ones, and whether it trades the way a backtest of its own settings says it should. Whispertrades runs both for you, free, with nothing to set up.
Where to find it
A bot's Strategy Health tab holds both for that bot, under its status: Recent Performance vs Historical, then Live vs Backtest.
Strategy Health in the main menu lists your enabled live bots together, those that need attention first, with their totals above; Disabled or All in its filters shows the rest. Paper bots are not listed there.
Each bot's status is the worse of its two readings: Needs attention for Drift or Far behind, then Watch, Healthy or Not enough data. The line under it says which reading set it.
The Strategy health module, which you can add to your dashboard, shows the same list in brief.
Each comparison has its own Turn off, on its line in the bar at the top of a bot's tab. A bot with one off shows Off in its place on the Strategy Health page and the dashboard; with both off it leaves them.
Recent Performance vs Historical
Recent Performance vs Historical (Recent vs History in the iPhone app) reads a backtest of the bot's current settings, which Whispertrades runs for you, free, from up to ten years before the bot went live and extends every night: its last year of trades against all its earlier ones. It shows whether the strategy itself is still working; Live vs Backtest, below it, shows whether your bot trades like its backtest. Every figure is P/L %: a trade's profit as a percent of the premium at entry, as in your positions table.
Its label, beside the heading: In line; Watch, when the backtest's recent trades are running below its history by more than usual, which many strategies pass through by chance; Drift, when they have stayed below it long enough that chance rarely explains it; or Not enough data.
Stretches of the backtest's history this low or lower says how often stretches of the backtest's history, reshuffled, averaged as low as its recent one. A small figure means the recent stretch is unusual for this strategy, not why. A colored border on each recent figure says the same at a glance: amber when fewer than 1 in 4 of those stretches were as low, red when fewer than 1 in 40 were. On the Strategy Health page, the bot's figure takes the same color.
Your bot's own trades are shown beside the backtest's over the same months, or its trades so far for a newer bot.
Windows: the last 2 years, 1 year, 6 months and 3 months show side by side, each with its average and an arrow for how far it sits above or below the trades before it. The label always reads the last year; a window with fewer than 20 trades shows its figures without color.
On a backtest: every finished backtest shows the same comparison over its own last 2 years, 1 year, 6 months and 3 months, each with its own label. Each window's trend, how many points of P/L % it sits above or below every earlier trade, is also an optional column on the Backtests list (Columns), in Backtest Comparison and the list export, and a sweep heatmap metric, and the 1 year and 6 month trends show at the top of the backtest's page.
After a settings change that matters (an entry, exit, adjustment or variable setting, the type or symbol, or switching between paper and live), it reads again once a backtest of the new settings is built, within a week. The name, notes, sizing and entry and exit speed don't count.
A bot whose settings can't be backtested is read against its own earlier trades once it has 80 closed trades under its current settings. A trade counts three trading days after it closes, and manual and API entries don't count.
It updates once a night. On the Strategy Health page, each bot's row shows the same comparison as a small chart of the backtest's line against its history. In the iPhone app it's on a bot's Performance screen, without the chart.
Its Log tab notes when the bot enters Drift and when it leaves it. To also get these by email, text or push, turn them on under Settings > Notifications > Bot activity.
Live vs Backtest
Live vs Backtest backtests the bot's settings over the days it has traded, scaled to its own position sizes, and shows where live trading and the backtest part ways. It is on for every bot and rebuilt every night, starting the morning after the bot's first closed trade.
Live P/L, Backtest P/L and Live vs Backtest difference: what the bot made after fees, what the backtest made over the same days, and live minus backtest. The percentage is of your account's value when the comparison begins; a paper bot shows dollars only.
Where live differs from the backtest splits it into strike selection, entry timing, entry and exit fills, exit timing, unfilled profit targets, fees, missed and extra entries, and whatever could not be split, shown as Unexplained. Click a part to list the trades behind it; on the Strategy Health page, click one to list the bots behind it. Largest difference first puts the trades that moved the total most at the top, and a trade under $5 from its backtest names no main cause. The backtest fills the way your own backtests do, at what real Whispertrades orders of the same kind filled at.
Missed and extra entries are trades the backtest took that your bot did not, and the reverse. A missed entry shows the reason your bot logged for skipping it, when there is one.
Not counted: manual and API entries, and backtest trades from days your bot was disabled or a bot variable kept it out, are listed but not counted. Settings a backtest can't fully reproduce are named on the tab.
What it covers: the trades since the bot's settings last changed in a way that matters, the same changes that restart Recent Performance vs Historical. Set a start date, at the top of the tab, starts both comparisons on a day you choose, even before a settings change you've already made. A change you make after saving it starts them again at that change.
Refresh rebuilds a bot's comparison right away, once a day per bot.
Colors: a bot's Difference % turns amber or red when it trails its backtest further than chance alone usually would (fewer than 1 time in 4, or 1 in 40), and stays gray until 20 of its trades are compared.
Far behind, beside the heading, marks a bot whose Difference % has turned red, until it is back to green.
Its Log tab notes when the bot becomes Far behind and when it is no longer. To also get these by email, text or push, turn them on under Settings > Notifications > Bot activity.
Live vs Backtest is on the website only.