Last month The Wall Street Journal reported that the Securities and Exchange Commission (SEC) was looking into how U.S. mutual funds price shares of private technology companies. The article cited research noting that on 12 occasions in the past two years, the private companies worth over $1 billion were valued differently by different mutual funds on the same date. Uber, for example, was valued by Blackrock about 12 percent higher than Fidelity valued it – a swing of approximately $6 billion! This news came on the heels of Fidelity’s markdown of several well-known unicorns (and subsequent markup a month later) and the announcement that Square’s IPO price range would be $11 to $14 (ultimately shares were priced at $9), lowering their valuation by 42 percent and effectively declaring its highly-anticipated IPO would be a down round. The news also came with many headlines that used words like bubble, burst, deflate, pop, and erupt.
The action by the SEC is perfectly understandable. The agency is charged with, among other things, ensuring that capital markets operate in an environment in which all participants have access to all the information they need to make informed decisions. If a public mutual fund owns shares in a private company, it follows that fund investors ought to know what it’s worth.
There is only one problem: with private companies – particularly early-stage private companies – it is often not possible to determine fair value with very much precision. Moreover, efforts by well-meaning rule-making bodies – made of individuals who are convinced that truth can be revealed with enough process, arithmetic, and rigorous methodology – may be misguided. The valuation methodologies employed by private equity funds – including those at Revolution – are well-established now and provide a familiar framework for valuing private companies. But there is probably a better way.
The challenge is quantifying “fair value.” Just what is a share of some company actually worth? The experts tell us that “fair value” of any asset is the price that at which a willing seller and a willing buyer transact for that asset, assuming that both have all material information and neither has any unfair advantage over the other. In the stock market, the “fair value” of a public company is calculated thousands of times a day – with each trade – as cash changes hands and shares moves from sellers to buyers. So far, so good.
But with private companies, that construct breaks downs. Or at least, it doesn’t happen very often. Months, or even years, may pass between transactions where a willing buyer and a willing seller meet to transact company shares. And experience tells us that the value of a private company at one moment in time is unlikely to have the same value months or years later. So how should they be valued? One approach would be to “go back to the future.”
In the old days (that is, before the 2008 financial crisis), venture capital fund managers understood those limitations and conservatively valued their portfolio companies at either (i) their cost or (ii) the value set by a subsequent arms-length transaction for the company shares (usually a financing for the company at a later date). The subsequent financing would establish a new value – higher or lower – and investors understood that those arms-length transactions represented the most accurate determinations of fair value.
And there was a natural conservatism to that approach. After all, if the portfolio company was struggling, it would return for private financing more frequently, which had the effect of resetting fair value frequently as well. And if the company was prospering and didn’t need additional capital, then portfolio managers were left with the happy problem of potentially understating their portfolio’s value.
But in the past half-dozen years that traditional approach has been abandoned. Auditors and now regulators have replaced it with more elaborate mathematical exercises that are less helpful in understanding the value of a small private company. This is called “mark-to-market” accounting, and one may reasonably question whether the precision implicit in that form of accounting is justified. That’s because for most of the time there is no market to mark to. Trying to mathematically nail down a private company’s valuation may be the “triumph of science over art.”
Mark-to-market accounting requires that fund managers apply various methodologies to private companies to determine fair value. Those methodologies include a comparison of the private company’s revenue and profits with those from similar public companies and the application of the respective revenue or profit multiple to the private company. However, often early-stage venture capital companies have no revenue or profits; for a biotech company revenues could be more than a decade away. And if the multiples are forward-looking (i.e., multiples of next year’s performance), the arithmetic is even more suspect – and less susceptible to actual auditing rigor.
Another common methodology for valuing private companies is discounted cash flow. That approach requires a forecast of the cash flows (both negative and positive) for the private company for a decade or longer, at which point it is assumed that the business can be sold for some multiple of that tenth-year cash flow. That stream of cash (including the ultimate sale) is then discounted back to a dollar value today, accounting for time and risk involved. But again, the precision implied from the arithmetic may be suspect. Indeed, because cash flows are typically negative for small private companies in the early years, more than 100% of the “fair value” is calculated from the imaginary sale of the business more than ten years into the future.
The simple truth is that the most reliable conclusions about “fair value” require a liquid market. By definition early–stage private companies don’t operate in liquid markets. And trying to apply elegant mathematics to those early-stage private companies does not remove the need for liquid markets to provide the data to make the mathematics reliable. It may be time to revisit the more traditional valuation methodology: re-setting valuations only when there is an actual transaction where cash changes hands.