Roll or cash out.

The letter is standardized by now. A portfolio company — usually the fund's best — is moving into a continuation vehicle. The price has been set through a process, a fairness opinion is attached, and every limited partner faces the same election with the same deadline, measured in weeks: take the cash, or roll the interest into the new vehicle. The paperwork frames it as a liquidity option. What it actually asks the LP to do is underwrite an investment.

The two paths are less symmetrical than they look. Cashing out means accepting the process price — whatever the competitive bidding and the opinion produced — and ending the position there, with certainty and a wire date. Rolling means staying invested in the same company, run by the same manager, inside a new partnership with a fresh clock; under current institutional guidance, LPs should be offered a status-quo option that carries their original economics forward, alongside whatever new terms the vehicle sets. One path closes the book. The other opens a new one, on a business the LP has now watched for years.

What the LP actually knows

The election is structurally tilted, and everyone involved knows it: the general partner understands the asset more completely than any outsider ever will, including the LPs deciding whether to stay. But the tilt runs the other way too, and it's the underused half. Over a holding period the LP has accumulated something no new investor in the continuation vehicle possesses — years of reporting from this manager, the record of how candid the marks were, how bad quarters were communicated, how projections aged. The roll decision is one of the few moments in private markets where that accumulated file can be acted on directly.

The GP knows the company better than the LP ever will. The LP knows the GP better than any new investor ever could.

Plenty of LPs will take the cash for reasons that have nothing to do with the asset — allocation limits, liquidity needs, a policy of not rolling. Those are legitimate. The failure mode is answering on autopilot in either direction: rolling out of inertia without re-underwriting the new terms, or selling out of habit and surrendering the one company the manager fought to keep — which is itself information, since a continuation vehicle is a general partner showing you, with its own money and reputation, which asset it ranks first.

The deadline makes the election feel administrative. It isn't. It is the rare moment a private-markets investor is handed a genuine decision in the middle of a hold, with more first-hand history about the manager than any buyer in the process. The quality of the answer depends less on the deal documents than on the file the LP has been keeping all along — which is, among other things, an argument for keeping the file.