Building a Portable Track Record

A researcher who spent three years generating a strong signal that made the firm real money often discovers, when interviewing elsewhere, that they can't say much about it: the code belongs to the firm, the exact P&L numbers are confidential, and the new employer has no independent way to verify a claim like "my strategy returned 18% annualized" beyond taking the candidate's word for it. A track record that exists only inside one firm's private systems is, functionally, not very portable, and building one that survives a job change takes deliberate effort well before a move is on the table.

The core tension is that the specifics that make a track record credible, exact returns, exact positions, the actual code, are usually exactly what a firm won't let a departing employee take or disclose. What tends to survive that constraint is a level up in abstraction: documented general approach and methodology (without proprietary specifics), verifiable outcomes described in relative or qualitative terms that a former manager can confirm in a reference call, and any work that was done independently of the firm's proprietary systems, personal research projects, open-source contributions, competition results, or published work, which belongs entirely to the individual and travels cleanly.

A practical habit is to keep a private, ongoing log of accomplishments as they happen, in terms specific enough to be useful later but abstract enough to be shareable: "built a signal that became one of the top-three contributors to desk P&L in its category" rather than the literal number, which usually can't be repeated outside the firm. Strong references from former managers and colleagues end up doing much of the work that a literal track record would otherwise do, since they can vouch for specifics a resume can't state directly.

What this means in practice

The time to start building a portable record is the first year on the job, not the year a move starts to look likely, a habit of documentation kept consistently over years produces something usable, while a rushed reconstruction right before interviewing rarely does, and often runs into the exact confidentiality restrictions the habit was meant to work around.