The AUM Illusion in Venture Capital

Recently, I had someone from the founding team of a startup ask me, “How much does the fund manage?” That was a fair question, but I believe knowing the dry powder or fund mandate of a venture firm matters more to a founder. By asking this, a founder knows if the venture firm has the capacity to meet the investment required by the startup or if the venture firm is likely to follow on their original investment (let’s discount the startup’s performance here).

Let’s rewind to the original question. The question simply refers to assets under management (AUM), but at what point did the amount of capital a venture firm manages become an indication of how well it invests?

I believe the metric is useful within the venture industry but more towards the limited partners (LPs) and the venture firm itself. The best use case I’ve found is using the metric to identify whether managing more capital improved or reduced a venture firm’s ability to generate returns.

In their 2005 paper titled Private Equity Performance, Returns, Persistence, and Capital Flows, Kaplan and Schoar studied the relationship between fund size and investment performance across venture capital and buyout funds. They found a concave relationship between fund size and returns, with the relationship being statistically significant especially for venture capital funds. ¹ Their findings indicate that while larger funds may initially generate more returns, the benefits of increasing fund size eventually diminishes, and performance can decline as funds become larger.

Here’s a simple hypothesis test.

H₀ : AUM is not positively associated with venture capital investment returns

Ha : Higher AUM is positively associated with venture capital investment returns

^{*A little caveat: The hypotheses above concern the relationship between AUM and venture capital returns, but the example below is a thought experiment rather than a statistical hypothesis test. Establishing such a relationship would require observations across multiple venture firms, ideally controlling for fund vintage, investment strategy, market conditions and other variables that could influence returns. A comparison between two hypothetical firms cannot establish causation or statistical significance. Unfortunately, two rows in a table do not constitute an econometric study.}

So, does managing more capital mean generating better returns?

Let’s take two hypothetical venture firms with different levels of AUM and investment performance as an example.

Based on the table, its evident that Firm A manages fives times more than Firm B. Look closer and you’ll notice that albeit managing more capital, Firm A has only distributed $0.30 for every dollar contributed by its LPs. On the contrary, Firm B returned $2.00 for every dollar contributed by its LPs despite managing a lower capital. This means that in absolute terms, Firm B has distributed $200 million to its investors, compared with Firm A’s $150 million. So, does this mean that Firm B is the better venture investor? Well, not necessarily. Here’s why.

First of all, DPI only measures realised distribution relative to paid-in capital. It does not tell us anything about the remaining unrealised value of a fund’s portfolio. Firm A could have other promising investments yet to be exited, while Firm B might have already realised most of its portfolio. For this hypothetical experiment, let’s assume these two firms were founded in the same year, and that the funds being compared share the same vintage and investment strategy, so differences in their DPI cannot simply be attributed to differences in fund maturity or investment focus.

^{Side note: a younger fund could report a lower DPI simply because its investment are yet to be matured.}

That’s why DPI should be analysed alongside Total Value to Paid-In Capital (TVPI), which includes both realised distributions and the remaining net asset value of a fund’s investments. Even then, TVPI depends partly on unrealised valuation that may not materialise at exit. The distinction matters because paper gains and cash distributions are two different things, especially when LPs are expecting the latter.

Regardless, the comparisons shows an important point whereby AUM provides information about the scale of a venture firm’s operations, but it does not tell us how effectively that capital has been deployed or returned to LPs. More importantly, the example does not allow us to reject the null hypothesis. It merely demonstrates why a positive relationship between AUM and investment returns should not be assumed.

So, is AUM is useful metric? Well, I think it is, so long as we are clear about what it answers. For LPs, it gives context on the scale of a firm’s operations and its fundraising history, and for the venture firm itself, it informs the capacity and resources needed to manage a growing portfolio. Regardless, I’d still be more cautious using it to establish a firm’s investment credentials, because the ability to raise capital is not the same as the ability to deploy it well or return it to LPs, and no single performance metric tells the full story either.

This brings me back to the question the founder raised. To be frank, it’s a fair question if you are trying to gauge the scale of the investor you are approaching, but I’d be more interested in how much deployable capital sits in the fund, the typical initial cheque size, the investment mandate and whether money is reserved for follow-ons. A firm managing $5B is not automatically better placed to back your startup than one managing $50M.

For LPs, they should be more interested in knowing whether a GP can keep generating attractive returns as its capital grows, and that is where AUM earns its place, not as a scorecard of investment success but as a prompt to ask whether that success holds at scale. Managing more capital is one achievement, generating returns on it is another, and the industry would do well not to confuse the two.

Reference

Kaplan, S.N. and Schoar, A. (2005) ‘Private equity performance: Returns, persistence, and capital flows’, The Journal of Finance, 60(4), pp. 1791–1823. https://doi.org/10.1111/j.1540-6261.2005.00780.x