Is this ratio the Holy Grail for investors?

The promise of a shortcut to riches is always attractive. No sane investor wants to grind through every page of a set of financial statements to hunt down a bargain share if a single ratio will do the job instead.

Enter James O'Shaughnessy, an American money manager who has just released an updated version of his 1996 book (described as "something of a bible" by John Reese on Seekingalpha.com), What Works On Wall Street. He claims that one number, the intimidating sounding enterprise value (EV) to earnings before interest, tax, depreciation and amortisation (EBITDA), works much better than the rest. So has he uncovered investing's Holy Grail?

What's wrong with a p/e ratio?

The easiest way to understand EV/EBITDA is as an alternative to the commonly used price/earnings ratio (p/e). A p/e ratio compares the firm's current market capitalisation (the number of shares issued multiplied by the current share price) to one year's earnings. If a firm has a market capitalisation of £3m and the latest earnings figure is £500,000, the p/e is six (3/0.5).

Subscribe to MoneyWeek

Subscribe to MoneyWeek today and get your first six magazine issues absolutely FREE

Get 6 issues free

Sign up to Money Morning

Don't miss the latest investment and personal finances news, market analysis, plus money-saving tips with our free twice-daily newsletter

Don't miss the latest investment and personal finances news, market analysis, plus money-saving tips with our free twice-daily newsletter

Sign up

The lower the result, the better (assuming you are after a cheap share). That's because a low number suggests that the market is pricing the share at a low multiple of current earnings. So if earnings pick up in the future, the share price, and overall market capitalisation, should also rise quickly.

Sounds simple. But fans of EV/EBITDA point to several flaws in the p/e ratio. The biggest is the number used for earnings. This is the profit after tax figure from the bottom of a profit and loss account. As an accurate gauge of a firm's true operational performance, this suffers two key drawbacks.

MoneyWeek videos

The EV/EBITDA ratio

Tim Bennett explains the EV/EBITDA ratio, used to tell whether a share is cheap.

Watch all of Tim's videos here

Firstly, distracting financing costs (typically interest on loans and overdrafts) have been deducted to arrive at the final number. So too has a tax charge that is calculated using obscure tax rules.

Secondly, the p/e is stated after deducting two highly subjective costs that reflect the directors' best estimate of how long the firm's assets will last before being worn out depreciation and amortisation.

Critics argue that all these items obscure the firm's true performance, and mean that the p/e ratio as a valuation tool simply doesn't sufficiently reflect the lifeblood of any firm cash flow.

How does EV/EBITDA work?

In a nutshell, EV/EBITDA tries to correct this flaw. It does so by replacing earnings (after interest, tax, depreciation and amortisation) with EBITDA earnings before these four costs (what I like to call "earnings before interest, tax and dodgy accounting").

Better still, it replaces the market capitalisation on the other side of the ratio with a firm's enterprise value (EV). The two key components of EV are the market value of debt as well as the firm's market capitalisation. In other words, EV reflects the contribution of both a firm's shareholders and its lenders in arriving at a valuation.

So if the firm we mentioned earlier has £1m of outstanding debt on top of its £3m market capitalisation, its EV would be £4m (£3m + £1m). If interest, tax, depreciation and amortisation charged against earnings were £200,000 in total, EBITDA would be £700,000 (£500,000 + £200,000). Combine the two and the EV/EBITDA ratio comes out at 4m/0.7, or around 5.7. As with a p/e ratio, the lower the better.

EV/EBITDA is a magic number

So does it work? O'Shaughnessy looks at the performance of the top 10% of US stocks chosen using various popular valuation ratios between 1964 and 2009. His findings are clear. The average annual compound return when stocks were selected using EV/EBITDA was 16.58%.

This compares to 16.25% using the p/e ratio, 14.53% using the price-to-book ratio and 13.3% using the dividend yield. His findings are backed up elsewhere. As the Daily Wealth's Brett Eversole notes, a study by Drexel University's Wesley Gray and John Vogel reveals that picking the cheapest 25% of stocks using EV/EBITDA between 1971 and 2010 returned 17.66% a year versus 15.23% for the p/e ratio.

Only fools rush in

So is the quest for a foolproof valuation ratio over? Afraid not. EV/EBITDA is a very useful ratio but it isn't foolproof nothing is. The biggest catch, as O'Shaughnessy notes, is the volatility of returns otherwise known as risk. Hence getting your timing right (ie, when you decide to buy or sell shares) matters more with this ratio than the rest.

To test that, O'Shaughnessy looks at factors such as the biggest annual declines, the percentage of years in which returns were positive (remember the figures quoted above are averages), and the consistency of returns. You can bash this into a number that combines returns and risk (known as Sharpe and Sortino ratios).

We won't dwell on the details here because it's the conclusion that matters EV/EBITDA may offer the best returns, but the cheapest 10% of stocks selected on this basis also carry the highest risk compared to selecting them using another ratio. In the case of, say, EV/EBITDA versus dividend yield or the price-to-book ratio, the difference is large.

What this means is you can't trust EV/EBITDA to deliver in all investing climates and you run a much greater timing risk than with some other ratios.

There's safety in numbers

So what to do? O'Shaughnessy recommends screening stocks using a range of valuation metrics to get the maximum average return with the fewest nasty single-year surprises along the way.

Our advice? Don't abandon key ratios such as the p/e ratio in favour of EV/EBITDA the latter isn't always reliable as a guide to value and it's less widely published and harder to calculate. However, if you've never considered using it before, O'Shaughnessy's work suggests it's well worth adding to your stock screen. For more, see my EV/EBITDA video.