Contents · 36 / 58
Test formulas in Alpha Lab
Updated 31 October 2026 · 3 min read · 14 Contents
Plain-language meaning
Alpha Lab evaluates a formula you type, such as Rsi(14) < 30 and Close > Sma(200), bar by bar over one listing's stored daily price history and shows where its value holds or changes. It is a single-listing expression tester, not a cross-sectional factor ranking.
Why it is useful #
It lets a user see exactly when a rule-based condition held in one listing's history, with the formula fully visible, before turning it into a screen, alert or backtest.
Where it appears and the workflow #
Open Alpha Lab (/AlphaLab; /Analytics/AlphaLab and /Quant/AlphaLab open the same page). Choose the exchange and symbol, set the lookback slider (60 to 2,000 bars, default 400), then type a formula or start from a preset or one of your saved formulas, and select Evaluate. The result plots the formula's value for every bar beside the price.
Calculation or source #
Each stored daily bar is evaluated with the formula's own functions, such as indicator functions (RSI, SMA) and price fields (Close). The value for a bar is exactly what the formula returns: a number, or true or false for a condition. A bar where any input cannot yet be computed returns null. Optional US short-volume and options fields read supplemental stored datasets only where display rights permit.
Prerequisites, inputs, period and unit #
Prerequisites are a listing with stored daily history, a valid formula and the specialised-research entitlement for Evaluate. The lookback is measured in bars (completed sessions). The unit is whatever the formula produces: a true/false condition, an indicator level or a price-based ratio.
Worked example #
Illustrative only: over 400 bars, Rsi(14) < 30 and Close > Sma(200) may be true on 9 bars. The first 199 bars cannot compute Sma(200), so they are null, not false, and are not counted as misses.
What high and low mean #
For a numeric formula, high and low follow the formula's own definition. For a condition, true means every part held on that bar. More true bars do not make a formula more useful.
Positive, negative and zero #
A numeric result can be positive, negative or a genuine computed zero. False is an evaluated result; null means the bar was not evaluated.
Limitations and common mistakes #
Trying many formulas on one listing invites curve fitting, and a historical match count is not a forecast. Alpha Lab applies no costs, sizing or exits; use Backtests for those assumptions.
Market-specific differences #
Any persisted equity venue in the exchange list can be used, with prices in the venue's native currency. Supplemental short-volume and options fields cover US listings only, and only where rights permit.
Suggested next steps #
Evaluate a preset on one listing and note the first non-null bar and the number of true bars. Change one threshold, record how many bars changed, and explain why that alone is not evidence the formula works.
Educational use only #
This guide explains descriptive research evidence. It is not investment advice, a price prediction, a recommendation, a claim of predictive accuracy, or an instruction to buy, sell, rebalance or place an order.