Why uncorrelated strategies barely blow

The two forex swing configurations post annual blow rates around 0.4 percent, against 6 to 14.5 percent on the futures profiles. They are not sized smaller relative to their buffer in any dramatic way, and their deepest stretches still use a third to a half of the limit. The difference is structural: their strategies do not lose at the same time.

The mechanism

An account blows when one stretch of losses crosses the floor. When strategies correlate, their losing streaks stack into exactly that stretch. When they do not, one bleeds while another earns, and the combined equity line never digs the single deep hole that ends the account. The blow rate is a property of the worst combined stretch, and decorrelation attacks the worst stretch directly, which sizing alone cannot do.

Why this beats sizing down

Halving size on a correlated portfolio halves the depth of every stretch, good and bad, so payouts slow in proportion to the safety gained. Decorrelation is closer to a free lunch: the modeled forex pair keeps a high payout frequency, near the top of the roster, while holding the lowest blow rate. Smoothness pays twice, in survival and in cycle speed, as the payout frequency data shows from the other side.

Building it

Decorrelation has to be real, not cosmetic. Two momentum systems on correlated instruments are one bet wearing two names. The combinations that work mix logic types and instruments, mean reversion against momentum, gold against an index, the design constraints covered in single strategy vs portfolio and the instrument side in the instruments guide. The verification is the same Monte Carlo as everything else: resample the joint distribution and watch whether the worst stretches shrink, per the sizing methodology.

FAQ

Why do uncorrelated strategies blow less often?

Because an account blows on one deep combined stretch, and uncorrelated strategies do not stack their losing streaks into the same stretch. One earns while another bleeds, so the joint equity line avoids the single hole that breaches the floor.

Is decorrelation better than sizing down?

Often, yes. Sizing down buys safety at a proportional cost in payout speed. Decorrelation reduces the worst combined stretch while keeping production, which is why the modeled uncorrelated pair holds both the lowest blow rate and a top payout frequency.

How do I know if my strategies are actually uncorrelated?

Not by eyeballing two equity curves. Resample the joint trade distribution across Monte Carlo paths and compare the worst combined stretches against the single-strategy ones. If the combined tail does not shrink, the diversification is cosmetic.