The Art and Math of Follow-on Decisions

Disclaimer: This is a VC math nerd post… and it isn’t perfect. Feedback welcome! :)

Portfolio construction among GPs and LPs is getting a lot more attention as capital markets shift (endowment taxes, mega-funds, small seed funds, etc.). In general, less scrutiny is paid to optimal construction when capital flows into a sector — and much more when capital becomes scarce.

The hard part is this:

While VC can be statistically modeled pretty well at the asset class level, investment decisions are made dynamically and sequentially within the constraints of a time-bound fund. It’s very hard to have a singular framework at the right level of abstraction that captures all the variables...but I’m going to try!

For us, the core question at every investment decision point is:

“What size exit is required — before and after this decision — for this company to return the fund ("RTF")?

Obviously, more ownership up front means a smaller required outcome. Less ownership means a larger one. But — as companies and portfolios develop — we get more information. That means we can, in theory, make more informed decisions about whether to follow on into Company X versus reserving capital for a brand-new Company Y. This dynamic becomes especially important for very early-stage funds like ours, where follow-on opportunities often arise while the fund is still actively investing.

Here’s a scenario:

You invest in the pre-seed of Company X. Nine months later, they raise a great seed round. At that moment, you have two options:

1. Take the dilution: Accept a lower ownership stake, and recognize that Company X now needs to achieve a larger outcome to RTF. (We model expected dilution initially, but sometimes you can "buy back" some of it if things are going better than expected.)

2. Buy back dilution by doing pro rata: Maintain your original RTF math, including initial modeled dilution, but at the cost of deploying capital that could have funded a new initial investment elsewhere.

Neither answer is perfect. The real trade-off is between consolidating conviction or creating optionality. Follow-on = Consolidate conviction; New investment = Create optionality.

The decision depends on your evolving belief in Company X relative to the opportunity cost of a future Company Y you haven't even met yet.

In the end, this is more art than science. You should have a framework. You should run the math. But you also have to trust your gut. Relationships are where outlier returns come from, the math can just serve as a guide.

Also — I really like round numbers, and sometimes that factors in too. :)

Would love thoughts, critiques, or other frameworks people use here!