Portfolio Builder by correlation
It makes the leap from "having one good strategy" to "managing a team of strategies". Instead of keeping the best one, the Builder automatically picks a basket of strategies that complement each other: when one is losing, another tends to be winning, so the group is steadier than any of them alone.
It opens from the Portfolio → Portfolio Builder (correlation)… menu. It works on the active strategies of a Portfolio you have already saved.
The idea in one sentence
Two strategies that win and lose at the same time (highly correlated) do not diversify: it is like owning the same one twice. Two that win and lose at different moments (weakly correlated) do: when one goes through a rough patch, the other cushions it.
The Builder measures that correlation across all of them and assembles the team with the best Return / Drawdown (profit against the worst fall) for the group as a whole.
How it does it, step by step
It runs a backtest of each one (1 contract) on its own symbol and timeframe. It is fast: one backtest per strategy, not an optimization — which is why it takes seconds, not minutes.
It computes each one's profit month by month and compares: do they rise and fall together or separately? That is the correlation matrix.
It starts from the best one (the seed) and adds, one at a time, whichever improves the group most, skipping those that are too correlated. It stops when nothing improves it any further.
It draws the combined equity curve and summarizes the team's metrics: profit, max drawdown, Return/DD and average correlation.
What you see on screen
Each row is one analyzed strategy, sorted by Return/DD (its individual "grade"). Columns:
- · Net P. $ / Max DD $ / Ret/DD — profit, worst fall and their ratio (the higher the better).
- · Avg corr. — how correlated it is with the rest (green = diversifies well, red = looks a lot like others).
- · Basket ✓ — whether it makes the final team.
- · Reason — why it is in or out (see below).
The top two are shown in gold (the highest individual grades) and the basket ones in green. Careful! They are not the same thing: see the box below.
A grid with every strategy in the rows and columns. Each cell colors its correlation: green = low (good), red = high. The basket ones carry a yellow border. Hover over a cell and it tells you the pair and its value.
The sum of the chosen strategies over time. Compare it in your head with how bumpy a single one would be: the team should climb straighter.
"The best ones" ≠ "the basket" (this matters)
Gold rows (#1 and #2) = the two with the highest individual grade. That is just the table's ordering.
Green rows ✓ = the ones that form the best combined team. #1 is almost always there (it is the seed), but #2 can be left out if it looks too much like #1.
That is what the Reason column is for: "Seed", "In basket: improves the group", "Dropped: correlated", "Does not improve the group"… The basket has no fixed size: it grows while adding one improves the group (up to the cap you set), so sometimes it is 2 and sometimes 5 or 7.
Three settings
- · Max strategies — cap on how many go in the basket (default 8).
- · Max correlation — it will not add one whose average correlation with the basket exceeds this value (default 0.7).
- · Min trades — it drops strategies with too few trades (not trustworthy).
Saving the basket
The "Save basket…" button creates a new Portfolio with only the chosen strategies. From there it is an ordinary portfolio: open it in Strategy Portfolio to re-test it, chart it or export it to TradeStation/MT5.
The full flow: Mass Test → Portfolio → Portfolio Builder → Save basket → export.
AniQuant can be tried free for 30 days, with every module and no card.