forex-scalping.com

Running one strategy across several prop accounts, the correlation risk people ignore

some people run multiple evaluations at once and use a copier to place the same trades on each. I tried a smaller version of it and the main thing I noticed is that it isn't diversification. It's the same trade several times.

If your trade loses, it loses on every account at the same moment. If you have three accounts at the same size and a bad day happens, you don't have three independent chances, you have three failures. The only thing you've spread is the cost.

It also creates practical problems. Different firms have different rules and different server times, so the same trade might break a rule on one account and not on another. Fills are slightly different on each, so results drift. And the copier itself is a point of failure, since a delay can leave you with different positions on different accounts.

If I use multiple accounts, I keep the same discipline as for a single one, and I don't increase my total risk just because the cost per attempt is lower. I think about the worst day happening on all accounts at once, and whether I'm fine with that.

To put numbers on the worst day idea without inventing anything, I take my own largest daily loss from my journal and multiply it by the number of accounts. If that total would breach the daily limit on one of them or make me feel sick, the setup is too big. It is a very simple test, but it made me cut back from what I was planning. Seeing the combined figure written down is a lot more sobering than thinking of each account separately.

The server time point caught me out in a practical way. One firm reset its daily loss at a different hour from the other, so a trade that was fine for one account counted toward the next day on another. If you run several accounts, write down each firm's reset time and news rules in one place. It is boring admin, but mixing them up is the kind of mistake that breaks a rule on one account while you think everything is fine.