Every year, Sand Hill Angels invests in number of start-ups. We know how much money we’ve invested (several million dollars a year) and what we’ve received in return (securities of various types). But, we have not, until now, tried to determine what those shares are worth.

One approach is to declare the problem vacuous. Our portfolio consists mostly of companies whose shares will be illiquid for a long time to come. So, whether the shares are (in some theoretical sense) worth one dollar or one million dollars, there is no way to exchange the shares for dollars. And companies that look great often fall apart, companies triumph after appearing all-but-dead, and markets shift quickly.

But, I believe it is important to value these companies precisely because they are so illiquid. Valuation, even if far-from-perfect, provides useful feedback to investors about their portfolios.

So, how can we value a portfolio of illiquid start-ups?

The first step is to eliminate most of the traditional methods of valuing companies. For example, the discounted cash flow method estimates value by looking at the cash flow likely to be generated by the company over the coming years and discounting for the interest rate you could obtain by investing in government bonds. That’s a reasonable approach for a large, slow-growing company that generates consistent returns — but completely impractical for a start-up that is expected to be unprofitable for years to come. Another method is to determine the enterprise value-to-revenue multiple for public companies in the same industry, and then apply that multiple to the revenue of the portfolio company. But, even assuming the start-up has revenue, it’s ludicrous to assign the same multiple that would be applied to companies millions of times larger. A third approach is to look at transactions (such as acquisitions or investments) involving comparable start-ups. But, it is difficult to find comparable companies — and even more difficult to obtain transaction data. Instead, we need a simple method that can be consistently applied across the portfolio..

Start-up investors generally hold a combination of the following three kinds of securities:

1. Stock
2. Warrants
3. Convertible debt

We have decided to value stock at the value established in the last financing round. For example, if the company recently issued 1,000,000 shares at $1 per share, then shares are worth $1 per share. Of course, since the point at which stock was last issued, the value may have increased (because the company has made progress, or because the market has strengthened) or decreased (because the company is in trouble, or because the market has weakened), but it’s difficult to determine appropriate adjustments. Liquidation preferences also affect payouts, but the full capitalization table with all preferences is often too difficult to obtain and model.

Warrants (the right to purchase stock at a pre-agreed “strike price”) are valued at their “intrinsic value”, i.e., the difference between the stock price (as determined above) and the strike price. If the warrant is underwater (i.e., the strike price is higher than the stock price), then the warrant is considered worthless. There is no way to apply more sophisticated options models (like Black-Scholes) because those models require estimates of volatility to determine the “time value” of the option.

Finally, for convertible debt, we have concluded that the appropriate valuation is the amount owed by the company, i.e., the principal amount loaned to the company plus any interest accrued. In general, the debt converts into equity when a future equity financing occurs. Therefore, if the debt hasn’t converted, no equity financing has occurred, and it is impossible to value the shares which will be received when the debt does eventually convert. More sophisticated approaches would take into account whether the debt is senior, the company’s cash position, and so forth — but that information is often hard to obtain.

A recurring theme is that there is a lot of information that is available in principle (i.e., does not require making predictions), but which is difficult for investors to obtain. I hope that tools like eShares or captable.io will make it easier for investors to value their investments in the near future. And I look forward to your suggestions!