Some Thoughts On Rolling Funds

An intro post from Angel List’s website

I’ve heard of a handful of angel investors raising Rolling Funds as an alternative to a “micro fund.”

I wanted to lay out a handful of incomplete gut reactions on the concept:

(1) What is a Rolling Fund?

My understanding is that a Rolling Fund is actually a series of funds that get set up once per time period (example: once per quarter). Investors can choose to invest a set amount per time period, and essentially that just means they contribute a pre-set amount to each of these time-stamped funds.

Example: “I’d like to invest $25k/quarter ($100k per year) into all deals that “XYZ Angel Investor” commits to.

It’s a really nice way for an investor to think about their commitment. Instead of thinking about making a $300k-$400k commitment that would be drawn over 3–5 years between new deals and follow ons, it’s a way to toggle up and down across vintages, and be able to shut off the spigot if liquidity gets tight later. It also makes sure that the pace of capital calls is even so that there aren’t any big surprises at the wrong time.

If I were considering investing in one of these (which I’ve decided I’m not), the way I would size my commitment is to 1/7th of the total allocation I actually wanted. For example, if I wanted to have $250k in a fund, I’d do $250k/7 = ~$35k/year. This way, by the time year 8 comes along, I’ll have started recycling funds initially deployed in year 1. If I were to become wealthier or receive realizations earlier than expected, I could always toggle that amount up later on.

This way, I wouldn’t be afraid of running out of dry powder.

These rolling funds can be great solutions for individual investors who can’t hit normal fund minimums. On top of that, it seems like the way they are being set up allows for general solicitation, which is a way more open and transparent process for providing fund returns.

But here are some concerns I have:

(1) Trying to deploy X/time period is really hard. Accelerators do it, and I’ve never understood how. I’ve written about this before, but for me — deals often come in waves. Some quarters we do 3–5, others we do none.

I often joke that the only thing scarier than writing 3 checks in a month is writing none for 6 months. Deal flow doesn’t come periodically. For example, I’d want to contribute less to deals in Q3 and Q4 than I would in Q1 and Q2, generally…

(2) Carry: The carry of these Rolling Funds favors the GP. Normally, in a fund, the bad deals wipe out gains from the good deals. This means the GP doesn’t take carry on a homerun deal if all the other deals flop, and the LPs lose money.

But in the case of a rolling fund, you could invest $100k, lose 30% of your money, and the GP could still get carry. How? The fact that each vintage is its own distinct entity means that the carry on each deal is not cross-collateralized. And therefore if one deal flops, it won’t wipe out part of the carry from the next winner.

(3) Reporting: Every single quarterly vintage will have a different makeup of investors (likely). Since different people can subscribe at different times, and have the flexibility to opt-in and out and toggle their commitments up and down. If they couldn’t, it would defeat a lot of the flexibility Rolling Funds are purposed for.

This means the combinations of exposures various investors have to a portfolio could reach infinite! The only three solutions are:

  • Draft the same report to everyone, but this means giving confidential info on startups some people have no exposure to
  • Draft no reports to anyone, creating an opaque process
  • Come up with some custom solution to draft reports to each person based on individually customized vintages, which seems hard w/o software. Maybe AL will build this.

(4) Allocations: It’s really hard for VCs of any size or merit to get exactly their perfect allocation each time. Even the best firms like First Round or Floodgate don’t do this perfectly. So odds are, there is going to be either too much demand for deals, or too much capital, leaving excess cash in the vintage. What happens to that cash/

(5) Pro-Rata Decisions: Often, pro-rata decisions are made in part to support an initial check into a business. Bridge rounds etc. are other examples of new investments made, to support past ones. But if new investors are part of a newer vintage, with no exposure to a deal already, there may be a misalignment in investment decision making.

*******

In any case, there are counters to each of these concerns. But they all feel real to me, and I’m not comfortable with the concept until some of them have been ironed out.

Luckily the people raising them are seem high quality, and have a chance to do so!