Home › Documentation › Validation & robustness
Module

Multi-Timeframe Analysis

It re-tests one or several Portfolio strategies across many timeframes at once (say from 1 to 240 minutes) to answer the key robustness question: does the strategy win across a wide band of timeframes, or only on one isolated peak? A strategy that works over a broad range of timeframes is far more trustworthy than one that only shines on an exact one. You launch it with the ▤ MULTI TF ANALYSIS button in the Portfolio window, on the strategies you have ticked.

The parameters dialog

Every strategy is tested on a common symbol (homogeneous and fast: each timeframe is aggregated only once). You pick the symbol, the range of minutes with its step and, optionally, add Daily and Weekly timeframe rows.

// Range of timeframes to analyze
Symbol: ES  ·  Minutes from 1 to 240 (step 1)  ·  ☑ Daily  ☑ Weekly

Daily aggregates the bars by calendar day; Weekly by week (anchored to Monday). They show up as two extra rows at the end of the grid, sortable like any other. They are not drawn in the Curve view (its axis is in minutes); they are in the grid and in the Heatmap.

Results grid

Rows = each timeframe; per strategy, a block of four columns No. Trades · Net P. · P.F. · Win%. The headers read E1, E2… with a legend giving the full names.

Sort by column

Clicking any header sorts by that value (ascending/descending, with a ▲▼ indicator). It finds each strategy's best timeframe at a glance.

Highlighted cell

The selected cell is highlighted in bright yellow. Navigable with the mouse and with the arrow keys.

Double click → Performance

Double clicking a cell recomputes that (strategy, timeframe) pair and opens its full Strategy Performance Report.

N/A cells

A timeframe smaller than the native one (or not a multiple of it) is invalid and shows as —; an infinite PF shows as ∞.

⚡ Multithreaded computation: the analysis spreads the timeframes across every CPU core (with dynamic balancing, since the small timeframes weigh far more). With a live progress bar and a cancel button. Hundreds of backtests in seconds.

Graphical comparison

The GRAPHICAL COMPARISON button opens a window with two views switched by a tab at the bottom (classic style): Curve and Heatmap. At the top, a metric selector (Net P. / P.F. / Win% / No. Trades) and a checkbox per strategy («Strategy N», in its curve's color and with a yellow hint carrying its full code) to turn them on or off.

📈

Curve view

  • →X axis = timeframe (min), Y axis = metric; one line per strategy.
  • →It shows the band of timeframes where it works and where its optimum sits. A zero line for Net P.
  • →The line breaks at timeframes with no trades (it does not invent continuity).
  • → A crosshair reading out the timeframe and each strategy's value.
🔥

Heatmap view

  • → A colored grid: rows = timeframes, columns = strategies.
  • → Color by value with a neutral center set by the metric (Net P. at 0, P.F. at 1).
  • →Ideal for spotting the hot zones when you have many strategies and timeframes.
low → high
💡
How to read it

Look for strategies whose metric stays good across a wide range of timeframes (a green plateau in the heatmap, a high flat curve). Be suspicious of the ones that only win on one specific timeframe and collapse on its neighbors: that is usually overfitting to the timeframe.

Try it yourself

AniQuant can be tried free for 30 days, with every module and no card.

← Previous
Overfitting detection (PBO and DSR)
Next →
360 Analysis
More in Validation & robustness