What metric should you use to value a Private Equity firm? In a recent Motley Fool article Daniel Paterson claimed that Blackstone (BX) was cheap compared to Carlyle (CG) and KKR (KKR). He used "earnings" (I can only assume he meant price to earnings), and return on assets to base this claim. This is misleading.
Price to Earnings (PE) is not a good valuation metric for Private Equity firms. While PE is indeed a good way to value a company with relatively predictable earnings, Private Equity earnings are highly volatile and not subject to good forecasting.
PE to Growth isn't useable either because PE isn't so meaningful.
Return on Assets is flawed the same way. "Return" is earnings and that fluctuates.
Price to Sales isn't meaningful because the sales part isn't like a normal company either.
Dividend yield isn't a good measure because the payouts change with the volatile earnings. If you buy a Private Equity company because it has a high yield this year, you may be disappointed with a low payout next year, or vice versa. The different such firms don't have earnings that fluctuate in sync with one and other so neither do the yields and the comparison is faulty.
All the above include earnings in the equation, which makes them unusable to compare companies.
Price to Book. This is the only metric that makes any sense. How much capital do they have to do their leveraged buyout magic, and how dear is it.
Based on price to book, Carlyle is the cheapest at 1.24, Blackstone is second at 2.01, and KKR (Paterson's claim for least expensive) is last at 2.35. Investors also may want to look at Apollo Investment Corp (AINV) which has a price to book of 1.04.
If you had a reliable estimate of future cash flows, that would be best, but you don't have that with these types of firms.
Venture Capital firms have a similar valuation challenge. I wrote a recent article on Opko Health (OPK) which needs to be valued as a venture capital firm vs. a regular company. Many people have cried "foul" that the stock is too expensive. But they are stuck using valuation metrics that don't apply. In OPKO's case a discounted sum of expected values of future sales times two would be a good, simple valuation model because the company could be sold at two times sales in the future. Just add up possible future sales, apply a probability to them, multiply by two, and discount them back to today.
Whether its venture capital companies or leveraged buyout companies, be wary of tradition valuation methods, they probably don't apply.