MIZUHO SECURITIES USA INC. | US EQUITY RESEARCH
Summary
Valuation is perhaps the most difficult and delicate art practiced by analysts and investors. We seek, in a rigorous way, to assess current and project future stock prices using methods that purport, but often fail, to encompass all of the aspects of the enterprise: cap structure, growth, margin, FCF productivity and the 'quality' of the business. In this report, we check the academic literature for help in finding a rigorous, definitive approach and, finding none, revert to arguing for a blended metric of EV, PEGs and FCF multiples.
Key Points
It shouldn’t be a surprise that the commonly used valuation methods yield strikingly different valuations, conclusions and implications for stock prices and, therefore, our ratings. Indeed, the most useful thing we found in the academic literature is that even people who study this issue for a living still can't derive a robust valuation model, i.e., one with a standard deviation less than about 25%. Random noise still accounts for a large part of equity valuation. Investors and analysts, however, should take heart: it is our work, coupled with exogenous whacks and boosts, that generates much of this random noise. In large part, we too are trying to reduce that random noise and improve the fit of the model by forcing multiple methods together to take into account not only the impact of the cap structure (EV/EBITDA less NCI), growth (PEG ratio) and FCF (FCF yields), but also the long-term value of the enterprise (DCF, where positive). We draw the following conclusions:
First, even a flawed EV/EBITDA less NCI still yields useful information about the enterprise and, when the EBITDA part is carefully defined across firms, a comparable metric. But, we show that the companies will often have the highest EV multiples for no reason other than that they have lots of debt. Second, PEG ratios are dynamic (growth-based) and address the use of FCF to buy back stock (EBITDA doesn’t take into account higher interest). Third, FCF metrics essentially do the same thing as PEGs, but FCF moves around a whole lot more than earnings, which are more often the basis for guidance. That volatility makes FCF-based metrics less reliable, much to our chagrin. Finally, a DCF example shows that if we were to be really rigorous and value companies on the basis of DCFs, some might have negative implied stock prices because of their significant debt burdens. Therefore, we will blend our valuation methods to come up with a more balanced approach to valuing hospital stocks. And then when we have a positive DCF under ‘reasonable’ assumptions, we’ll compare against that as a test for rigor.