What happens when your fund runs out of follow-on capital

A named failure pattern

Fred Wilson has been direct about this for a long time. Writing about reserves, he identifies a characteristic mistake among new managers: they do not sufficiently reserve for follow-on investments, and consequently run out of money and cannot participate in follow-on rounds. He puts the difficulty plainly, noting that what happens to a portfolio after it has been selected is the other half of the job, and the harder half to learn.

It is worth being precise about why this is a failure rather than a disappointment. A fund that cannot follow into its best company has not simply missed an opportunity. It has concentrated its returns into a position it is now unable to defend, at exactly the moment the position became worth defending.

Three states, only one of which is loud

Running short does not arrive as a single event. The platform distinguishes three situations, and they call for different responses.

• Follow-on deployment has passed the planned reserve

  The fund is still inside its total capital, but it has spent more on follow-on than it set aside. Everything from here comes out of capital earmarked for something else.

• Total deployment has passed fund size

  The arithmetic no longer works at all. This usually indicates a recycling arrangement in play, or a data problem worth resolving before reading anything else.

• The reserve is intact but the portfolio needs more than remains

  Nothing has been overspent. The companies still to raise simply need more than the reserve left can cover. This is the state that arrives quietly and the one worth catching early.

The third is the one that matters most and shows up least. Nothing has gone wrong on any ledger. The fund is inside its reserve and inside its total capital. It is simply that the companies still ahead of it will need more than what is left, and that becomes visible only if someone is looking forward rather than at the balance.

What it costs beyond the money

The direct cost is dilution in the companies the fund most wanted to hold. The indirect cost is information. An existing investor declining to take its pro-rata is read by everyone else in the round, and it is rarely read as a capital constraint even when that is exactly what it is. A new lead has to decide whether the fund knows something, and the fund is in the awkward position of explaining that it does not.

There is also a compounding version. A fund short of reserves late in its life faces every subsequent decision with the same constraint binding, and stops being able to distinguish between companies it is declining on merit and companies it is declining on arithmetic. The portfolio ends up shaped by the capital that remained rather than by any judgment about the companies.

What can be done, in rough order of preference

Recycling comes first, because it is the only option that adds capital without asking anyone for anything. Proceeds from early exits are redeployed rather than distributed, which is why partnership agreements carry recycling provisions at all. Wilson describes his own firm capping this at somewhere between a quarter and a third of fund size, which is a reminder that recycling is bounded rather than a general solution.

Beyond that the options get progressively less comfortable: a special-purpose vehicle alongside the fund, which asks LPs for capital outside their existing commitment and takes time the round may not allow; or deliberate non-participation, choosing which positions to let dilute rather than discovering it one round at a time.

Choosing deliberately is much better than it sounds. A fund that has decided which three companies its remaining reserve is for is in a materially stronger position than one holding the same capital against an undifferentiated portfolio, even though the balance is identical.

The discipline that avoids it

Wilson describes his firm modelling the financing needs of its portfolio companies and holding enough reserve to fund them with high confidence, revisited continuously rather than set once at construction. The specific technique matters less than the cadence: the reserve is a live position that has to be re-checked against a portfolio that keeps changing, not a number chosen in year one and consulted in year five.

That is the whole lesson of this page. Nearly every fund that runs out of follow-on capital had a reserve percentage it was happy with at the start.

Where this shows up in the platform

Two places. The fund view tracks reserve deployment against the reserve planned, so the first two states above are visible as they develop rather than after the fact. And when the Scenario Engine evaluates a specific follow-on, reserves are one of the constraints it can name as binding, which is the same condition arriving at the level of a single decision.

A binding reserve constraint on a company the evidence otherwise supports is the clearest form this problem takes: the module is saying the case is there and the capacity is not.

Sources

• ReservesFred Wilson, AVC, January 2017Names under-reserving as a characteristic emerging-manager failure, and describes the modelling discipline used to avoid it.