Hidden Common Factor Exposure Across Sleeves

A multi-strategy fund runs five sleeves: a value equity book, a merger-arbitrage book, a credit long-short book, a quant momentum book, and a macro book. On paper this looks well diversified — five different asset classes, five different research processes, five different teams. Then a single quarter arrives where small, cheap, financially stressed companies underperform sharply, and all five sleeves lose money together, because each one, through its own separate logic, happened to be tilted toward exactly that kind of company.

That is hidden common factor exposure: several strategies that look unrelated on the surface but share an underlying statistical driver — a "factor" — that none of them explicitly targets but all of them are quietly exposed to.

Why it hides so well

Each sleeve's own risk report is built around the risks that sleeve's own team thinks about. The merger-arb desk monitors deal-break risk, not size factor exposure. The credit desk monitors default risk, not whether its issuers happen to be small-cap. None of these reports is wrong on its own terms — they simply were never designed to answer the question "does this book, incidentally, lean toward small or distressed names?" Only a firm-wide factor model, run across every position regardless of which sleeve it sits in, can answer that question, because the exposure is a side effect of each strategy's normal selection criteria, not something any team deliberately chose.

Worked example

A firm runs a factor decomposition across its five sleeves and finds that, netted together, the portfolio has an aggregate "small-and-distressed" factor tilt equivalent to being short $150 million of large, financially healthy companies and long $150 million of small, financially stressed ones — even though no single sleeve holds a position anywhere near that size, and no sleeve's mandate mentions size or credit quality at all. When that factor sells off sharply, as it periodically does during flights to quality, all five sleeves post losses in the same week, and the fund's leadership discovers its "five uncorrelated strategies" were, in this one dimension, really one strategy wearing five costumes.

What this means in practice

The fix is a periodic, firm-wide factor risk report that looks straight through strategy labels to the underlying exposures — size, value, momentum, quality, credit spread, and so on — computed across every position in every sleeve simultaneously. When a common factor tilt is found, the firm can choose to hedge it out with a targeted offsetting position, cap it as a limit that no combination of sleeves is allowed to breach, or simply accept it as a known, sized bet rather than an accidental one.