A Few Thoughts on Valuing Investments in Startups

Alternative Asset Managers


Concurrent with Madeleine Harrigan’s post last week about IPOs being the new private equity downround, the financial reporting group at Mercer Capital published an interview with the head of the group, Travis Harms, on the difficulties mutual funds face in valuing level 3 assets (think Square). The following is an excerpt from that interview.

With respect to portfolio valuation, who are your clients and what services do you provide?

In our portfolio valuation practice, clients cover the spectrum from debt-focused funds, to hedge funds, traditional private equity funds, venture funds, and sector-focused credit and equity funds. Despite the diversity of strategies, what they all have in common is the need to develop reliable, defensible fair value marks for hard-to-value assets in a real-time reporting cycle.

That reporting cycle varies by client – we mark some assets on a monthly basis, while we look at others annually. The frequency with which we mark assets is generally a function of the fund manager’s ability to develop interim marks on their own – do they have the requisite expertise and staffing to develop and document reasonable interim marks?

Now, of course, the fund manager has the expertise to value assets. However, the fund manager’s valuation objective is to determine “intrinsic” or “investment” value, which may well differ from the prevailing market consensus. That is not the objective of fair value reporting, though. Fair value is not the fund manager’s price target based on his investment thesis. It is a particularly defined standard: fair value is the price that would be received to sell an asset in an orderly transaction between market participants at the measurement date. Developing and documenting the corresponding market participant inputs can be time-consuming and requires a different perspective than the fund manager is accustomed to using.

Sometimes we are developing our own independent estimates of fair value from scratch; other times we are examining the fund manager’s own estimates for the purpose of providing positive assurance that the marks are reasonable. Regardless of the scope of our work, documenting, presenting, and defending our conclusions to auditors and, potentially, regulators is always part of our job.

Looking to the VC markets a bit, you have commented on the Unicorn phenomenon and suggested that from a valuation perspective, “What’s obvious isn’t real, what’s real isn’t obvious.” What do you mean?

What we mean by that is that – while the headline valuation ascribed to a company following a fundraising round is obvious (price x fully-diluted shares outstanding), that is not the real value of the company. What is less obvious, but considerably more real, is that the price per share in the most recent round reflects all of the rights and economic attributes of that share class. Those rights and attributes are not the same for all of the other shares included in the fully diluted share count. It’s like applying the per-pound price for filet mignon to the entire cow – you can’t do that because the cow includes a lot of other stuff that is not filet. In the same way, the “obvious” pxq calculation overstates the value of an early-stage company. Now, no doubt the values of many “unicorns” are substantial, even when calculated correctly – but the real values are not nearly as obvious as the often breathless headlines would suggest.

Last week, a Wall Street Journal article elaborated on some of the difficulties that mutual funds face in valuing their investments in startups. Based on your experience with providing periodic fair value marks for VC funds, what are some of the elements that go into valuing such investments? What are some of the pain points?

Valuing startup investments, including “unicorns” such as those mentioned in the Wall Street Journal article, requires developing a thorough understanding of the economics of the most recent funding round, which provides a market-based “anchor” for valuation at subsequent measurement dates. What we find most effective is to build our valuation model so that it corroborates the “anchor” value as of the date of the most recent external funding round. Once our model is appropriately calibrated, we can then develop appropriate market participant model inputs for the measurement date that take into account changes in markets, company performance, and prospects for future exit with regard to timing, amount, and form.

Valuing these investments is particularly challenging given the illiquidity of the securities. When observable transactions occur only sporadically or at long intervals, it can be difficult to assess how changes in the market and company prospects will influence value. The longer the holding period – in other words, as you move from days to months to years – the greater the uncertainty regarding reasonable inputs and the wider the range of potential outcomes. Things become even more difficult when you layer in the unique features of many of these securities, such as liquidation preference, conversion, participation, and other rights and features.

Finally, the WSJ article discusses the fact that there is variation, sometimes substantial, in the valuation marks provided by different investors in the same company. Is that troublesome?

Is it troubling that different reasonably informed investment professionals come to different good faith estimates of the fair value of the securities we’ve been discussing? No. As we mentioned previously, illiquidity necessarily increases uncertainty. This is a phenomenon that you can observe even in securities that trade in markets – the less liquid and shallower the market, the wider the bid-ask spread will be. Even if you follow a rigorous calibration process like we outlined earlier, there is uncertainty about inputs. For example, you may know – on the basis of an observed market transaction – that a company’s value was $40 at a particular date, but what you still cannot directly observe is whether that was 8 times 5 or 10 times 4. Those unobservable inputs will necessarily breed good faith differences of opinion as the $40 value becomes stale with the passage of time. That is not to say that anything goes – there is a range of reasonable conclusions. But no, different estimates of fair value for these securities are not in themselves troubling.

A different question, whether it is troubling – given this valuation uncertainty – that an open-end mutual fund owns such securities is for the regulators to decide. It may be that the fair value estimates are reasonable, and reasonably different, but those differences are simply not tolerable from a regulatory standpoint. That, however, is ultimately not a valuation question.

Related Links


Mercer Capital’s RIA Valuation Insights Blog

The RIA Valuation Insights Blog presents a weekly update on issues important to the Asset Management Industry.