I have noticed, that with merger mania in full swing, Wall Street is turning to increasingly disjointed and exotic valuation metrics in order to justify the high valuations, (by value investing standards) that are now being placed on many stocks. None of these is more prevalent and suspect than the EV/EBITDA ratio.
Once described by Warren Buffett's lieutenant, Charlie Munger as, "bulls**t" the now widespread use of the EBITDA figure leaves much to be desired. (The full Munger quote was: "Every time you see the word EBITDA, substitute it with the words ‘bulls*it earnings’!” -- Referring to the potential for earnings manipulation.")
The traditional P/E ratio's downfalls
The EV/EBITDA figure goes some way to mitigating the traditional P/E ratio's downfalls. It does so by including interest cost, tax, depreciation and amortization within the calculation, comparing this figure to the firm's enterprise value, reflecting the contribution of both a firm's shareholders and its lenders in arriving at a valuation.
However, there are problems to this approach. For example, the EV/EBITDA ratio is not as widely published as the P/E ratio, so comparisons across sector universes are hard to conduct, unless you have an army of analysts at your disposal.
Then there is the issue of how enterprise value is calculated. In particular, the value may not capture all debt outstanding, banks are a prime example and long-term leases many be excluded. Negative enterprise values can give some interesting results when trying to calculate EV multiples.
Of course, the P/E multiple is open to more manipulation than the EBITDA figure; there are multiple ways of calculating depreciation and amortization. What's more, the P/E multiple effectively penalizes companies for taking on leverage.
Nevertheless, the P/E multiple has been used to value the market for centuries and the sheer volume of data, both historical and current, available with regards to the P/E multiple gives it more weight when it comes to statistical analysis.
But the question is, does investing according to the EV/EBITDA ratio produce the same returns as P/E multiple investing? According to James O'Shaughnessy, writer of What Works On Wall Street, the EV/EBITDA multiple actually produces better returns.
O'Shaughnessy conducted two experiments, firstly he selected the market's 10% of stocks with the lowest P/E ratio and then, the market's 10% of stocks the lowest EV/EBITDA value. Returns were taken over the period 1964 to 2009. The stocks selected with the lowest P/E ratio produced a compounded annual return of 16.3%, while the EV/EBITDA selection produced a compounded annual return of 16.6% over the period. Surprisingly, both selections produced returns less volatile than the wider market during the period observed. It's also interesting to note that the stocks selected under performed during the market euphoria of the late 1990s and 2006/07. Over the same period the S&P 500 returned 1.3% per annum.
So then, it would appear that despite concerns over the EV/EBITDA's credibility, its use in valuing stocks in an attempt to outperform the market can be lucrative.
O'Shaughnessy's research shows that the EV/EBITDA multiple could be more effective for value investors searching for opportunities. However, the world's most famous value investor, Warren Buffett, believes that it should never be used.