Closing Out The Fund

The art and science of ending the investing period of venture capital funds

We are in the final stages of closing out new investments for Compound I, our 3rd fund in the management company, but first official Compound fund. The concept of closing out a fund in terms of new investments is something that isn't talked about much but in my opinion is an art as much of a science.

Understanding Our Model

Much has been written about portfolio modeling, but the math is effectively interconnected variables of # of companies, desired % ownership, and initial:follow-on ratio (Compound I is 1:1), all of which influence check size and capital deployment. The last part of this equation is pace, an area where we have "zagged" a bit away from the market, investing over ~5 years (we aim for ~4 years but were slow to deploy), where the market has instead seen deployment periods compress from ~3 years to 12-24 months.

Our goal at Compound as an emerging/growing institutional investing firm is to be as true to our model in year 5 as we said we would be at year 1. Investors can take bad news, but they don't enjoy surprises.

Because of this, we use a very granular model I built about 4 years ago that projects forward every portfolio company's future in order to model follow-on scenarios, outcomes, and future fund performance. Many would call this a fool's errand, but it helps guide how we think about capital deployment in real-time. The inputs to the model can change rapidly based on the economic environment (for example, when the model was originally built in 2017, Series A and B valuations and thus our pro rata checks were...quite different).

We continually update the model as we deploy more capital as well as when more capital is returned to the fund and we get better resolution on recycling capital (we plan to invest ~105%-110% of raised capital in Compound I).

The dynamics surrounding incorporating our crypto portfolio within the fund also throws a wrench into things as follow-on dynamics are different and our check sizes have more variance. This has historically been the most fuzzy part of our model, both for us as well as our LPs.

Internally, we initially applied a blanket future multiple on the crypto portfolio of an “LP acceptable” but conservative 4x (today we heavily discount mark to market and use a 30 day trailing average price for valuation). Back in 2017, many LPs were dubious that our crypto portfolio would be able to return the fund, however this has proven likely to be true and our return on the crypto portion of our fund will likely end up many multiples of the modeled 4x in value.

As you enter into the final quarters of a fund's new investments however, the model begins to look slightly more malleable and optimizations occur. Psychology and the art of capital deployment starts to settle in.

On Psychology

Closing out a fund in theory shouldn't be an art if you stick to your fund model. Despite this, after many conversations with other GPs, varying types of psychology take over.

Because of our longer investment period with Compound I (a $50M fund), after ~5 years, we have significant insight into the potential performance of this fund.

Currently the portfolio looks like this:

  • 28 Active Core Companies of which ~50% are Series A+, and 2 companies make up >15% of the fund's total capital

  • 4 Exits (cumulatively these have returned a meaningful portion of the current fund)

  • 3 Dissolutions (representing ~$1.6M)

  • A crypto investment portfolio of 18 investments

As you can see, our longer deployment period has given us good resolution as we close out new investments in the fund, relative to many peers that have only 2 years of data. IMO this makes it significantly easier to head into our last ~6 months of planned deployment and figure out whether or not we want to continue to plan, or reduce the # of portfolio companies and increase planned reserves to deploy into "more clear" winners, effectively running expected value calculations for each incremental dollar deployed.

In summary, we have high-confidence data on which companies are material holdings for the fund, how much capital we will be able to recycle, as well as some dissolutions that allow us to have certainty no further capital will be required to support those companies.

Other GPs I've spoken to with shorter deployment periods have a different psychology. Some just trust the process/model, reach their desired # of companies, and move on (having a very stable LP base can help this). Others look at a collection of Seed/A companies (all they've now known for at most ~24 months), of which maybe a few have early product-market fit, and think "well...it might be nice to get a few more shots on goal". Psychology and conviction, art versus science.

I see the value in both approaches but despite my confidence in our investment process and future returns, I think I would have far higher anxiety in the latter example. This is one of many reasons why we don't intend to go to a 2 year planned investment period despite constant market pressures.

On Moving Forward

I will have much more to write about this in due time, but closing out a fund's new investments also allows you to reflect upon your past few years of partnering with new founders and optimizing that process for our future relationships with those teams, as well as an entirely new set of founders we plan to partner with.

Some of this happens in monthly internal reviews, a portion of this happens during fundraising, and a smaller amount happens in the daily rollercoaster of confidence and anxiety in a world with sparse feedback loops and seemingly lots of false positive signal.

Despite all of this, closing out the first chapter of a fund's portfolio allows you to measure how closely you did what you said you were going to do, while also forcing you to think about what you should be doing better or what you want to change due to decaying benefits.

Our job for our existing funds is nowhere near done (I spent last week literally tweaking a similar granular model for fund 2, 8 years in), but in a job with a continuous stream of "jobs to be done" and commitments to fulfill, it's important to take a step back and reflect for a moment, if only to take a breath.

This is in some ways the beauty (and pain) of venture capital. We raise 10 year funds, invest them over a time period, work with founders to create generational companies, and then wipe the slate clean and start from 0 with the next fund. You're only proven until you're once again, unproven.