A portfolio run trades several instruments on one account. It is not several runs added together, and the difference is the point: separate runs never compete for margin and never draw down at the same time, so adding them up flatters the basket.
Adding a leg
The Portfolio section of the settings panel holds the legs. The run's own instrument is the first leg. Each row you add is another.
| Field | Left empty |
|---|---|
| Instrument | Required. This is the leg |
| Timeframe | The run's own timeframe |
| Lot size | The run's own lot size |
| Strategy | The run's own strategy |
A leg that names only a symbol is that symbol traded exactly as the run is otherwise configured, which is what "the same idea across a basket" looks like and should not require repeating five values. Giving a leg its own strategy is how you test two ideas competing for the same money.
Each leg can also carry its own cost schedule. Absent, it uses the run's schedule when there is one and the run's flat commission and swap otherwise. See Costs and Fills.
One account
The legs share one starting balance, one pool of free margin and one equity curve that all of them move. Each starts with an equal share of the balance.
Two consequences fall out of that, and they are the reason to run this way:
- A leg can be refused an entry because another leg is already holding the margin. The leg result counts those refusals, so a basket that is too large for its balance says so in a number rather than in a quietly worse return.
- A leg stops when the account is wiped out, not when its own share is. A leg deep in the red inside a solvent portfolio has not been stopped out by anybody, and a solvent leg inside a wiped out portfolio has.
Reading the result
The headline return, drawdown and equity curve are the portfolio's. Under them the result carries what only exists because the legs ran together.
Per leg: its own trades, its own curve and its own statistics, so you can see which leg carried the result and which one paid for it. A leg's curve is its contribution to the account, starting at its share of the balance and moving only with its own fills, so the legs add up to the portfolio curve.
Correlation: the Pearson correlation of the legs' per bar returns, on the portfolio's own time grid. A pair that never has a bar in common reads 0.0.
Drawdown, twice: the account's deepest fall, and the legs' deepest falls added together. The gap between them is the whole argument for the basket. Two strategies that fall in different weeks give a portfolio drawdown well under their sum. Two that fall together give one close to it, and that is a fact about the basket that neither run shows on its own. The percentage of the summed drawdown that the portfolio avoided is reported as the diversification benefit.
A high diversification benefit on one period is not a diversified basket. Two instruments that share a driver, two majors against the same quote currency for instance, will separate for months and then fall together on the day it matters. Test the basket across regimes, the same way you would test one strategy.
What it costs
Every leg is a full simulation. A four leg portfolio is roughly four times the work of the same period on one instrument, and the legs advance together bar by bar rather than one after another, so there is no shortcut in the middle.