In  earlier editions of his book, “What works on Wall Street“, James O’Shaughnessy desribed the Price to Sales Ratio as “The King of the Value Factors”.  Whilst subsequent editions of the book make the more reasonable claim that Price/Sales vies with other value measures, the measure retains solid quantitative support.

In this post we examine the qualitative rationale for using EV/Sales and some practical issues regarding its application.

Overview of EV/Sales

The essence of any value story is the future capacity of an investment to generate free cash flow.  One option for analysis is to extrapolate from current free cash flow, but this approach has two potential problems;

  • Firstly, current free cash flow may not be a good indicator of future free cash flow.  In the case of growth stocks, current free cash flow may be low or negative due to high upfront investment.  In the case of cyclical stocks, current free cash flow may significantly over or under represent sustainable cash flow; and
  • Secondly, to the extent that current cash flows are easily observable, it is likely that they are well understood and valued by the market.

Using sales as a proxy for future cash flow potentially helps with both these issues:

  • Firstly, sales are generally much more resilient than earnings or cash flow.  They should therefore, be easier to predict.

Work by Dechow and Schrand  (Earnings Quality, 2004) highlighted that on average, 85c of every dollar of sales persisted into the following year, compared to 71c for earnings and only 41c for free cash flow.

  • Secondly, sales can be used to normalise for earnings.  Either  by applying through the cycle margins (in the case of cyclical stocks) or target/terminal margins (in the case of growth stocks).    EV/Sales normalises for margin in the following way:

 EV/Sales = EV/EBIT * EBIT/Sales

  • Thirdly, anecdotal experience suggests that analysts generally focus much more on direct proxies of cash flow such as Earnings, rather than sales.  EV/Sales will therefore provide a valuation metric that is less available than PER or EV/EBIT, and therefore, potentially more informative.  (This anecdotal evidence has quantitative support, as demonstrated in What Works on Wall street.  However, like all quantitative evidence in finance it needs to be adopted with some caution.)

Application of EV/Sales 

EV/Sales will be a  useful measure when current sales is a better indicator of future cash flows than current earnings.  This will potentially be the case in the following situations:

Cyclical¹ Businesses: 

Where margins are meaningfully above or below a long term average , current earnings are not necessarily a good indicator of sustainable earnings.

EV/Sales is a useful tool for cyclical stocks because it is a shorthand method for understanding sustainable or through the cycle earnings.  It does this by normalising margins:

EV/Sales = EV/EBIT x EBIT/Sales.

In the case of Woolworths ($WOW), margins took a step up due to management decisions to stop investing in price.  This was great for short term earnings and the share price.  Despite the share price re-rating, earnings multiples still looked reasonable.  EV/Sales multiples, however, reached record levels (Green arrows).

When margins normalised due to increased competitive activity, the stock fell sharply.  The previously low earnings multiples had provided a false value signal.  In contrast, the sales multiple had suggested some caution was necessary.

Post the earnings and share price decline, earnings based multiples looked expensive.  The normalisation of sales multiples was a better signal for the basing of the share price.  (Green circles).

In this situaiton, the EV/Sales multiple did not provide a perfect sell and buy signal.  It did, however, contrast with the signal from earnings based multiples.  It was a good cross check to prevent too much optimism at the peak and too much pessimism at the trough.

We have examined the application of EV/Sales in relation to cyclical stocks in more detail here…

Growth Businesses:

Where current earnings are:

  • depressed due to heavy investment in growth; and/or
  • understated due to this growth potential;

there may be very little information available from earnings based multiples.

For example, in the case of software companies, many of which have little or no earnings, earnings multiples are either non-existent or at a level that is not meaningful.

However, for reasons discussed in more detail here, the market does value the sales of these businesses.  Sales multiples are usually a function of sales growth and the profitability of incremental sales.

This is one example of using sales to value growth businesses, we have examined the issue in more general detail here….

Consolidating Industries

Using EV/Sales for cyclical stocks is a way of normalising for a company’s sustainable margin.  An alternative consideration is where the margins are “normalised” by an acquiror.  In a takeover, the acquiror believes it can extract synergies from the target which enable it to increase margins to a level commensurate to its own margins (or some industry average).

The acquiree is buying sales and their potential future earnings, rather than buying current earnings.  Therefore, the sales multiples of both the acquiror and the acquiree might provide a better valuation comparison than earnings multiples.   Conversely, the EV/Sales multiple paid is a useful cross check about the sustainability of acquired earnings, as per the cyclical example above.

A recent example is the takeover battle for Bulletproof ($BPF).  With earnings struggling, the stock had been de-rated to 0.3x sales, compared to an “industry” average around 1x sales.  For Maquarie Telecom ($MAQ) there would be clear synergies in an acquisition.  Not surprisingly the MAQ bid at ~0.5x sales has at the time of writing been rejected and a superior bid received.

Earnings Quality

A final example where EV/Sales might be useful is where current earnigns are not believable due to earnings quality issues.  In this case EV/Sales provides a back up valuation signal.

For example CIMIC (Formerly Leighton) was beset with concerns regarding their accounting for revenue and profit recognition on a series of major projects.  Whilst earnings multiples looked cheap – these earnings could not be trusted.  Even though there were some concerns regarding the revenue recognition policy, the potential errors with respect to revenue were much smaller than the potential errors with respect to earnings.

Using EV/Sales in this case provided some comfort for investors that didn’t want to rely solely on cheap earnings multiples.

Using EV/Sales – General Comments

In any of these situations in which EV/Sales is applied, the following general comments are worthwhile remembering:

  • Absolute vs Relative:  Whilst all valuations are ultimately relative, the relativity varies.  Discount models provide an “absolute” valuation relative to a bond proxy,  whilst earnings based multiples highlight relative valuations within equities.  The later are often turned into “absolute” valuation by selecting an appropriate target multiple.  We can use a similar approach to develop a target EV/Sales multiple as follows:

Target EV/Sales multiple = Target EV/EBIT Multiple * Sustainable EBIT Margin

For example, if we think a stock should trade on a mid cycle EBIT multiple of 8x, and through the cycle margins are 5%, then an EV/Sales multiple of 0.4x  (= 8 * .05) would be an approximate target multiple.

  • Use the correct Enterprise Value (EV) – I’ll write about this in more detail in a separate post, but a key issue in any enterprise value multiple calculation is that the EV is calculated appropriately.  Simplistically this is just Market Cap. + Debt, but there are often other assets and liabilities that need to be taken into consideration.
  • Understand the approximation at work – Sales is a proxy variable for free cash flow.  Its persistence and lack of focus by the market give it some advantages as a valuation tool.  However, its important not to overstate these advantages.  As a derivative, sales is generally further from free cash flow than measures of profit or dividends.  Whilst the estimate of sales is likely more resilient, its estimation of potential free cash flow is not necessarily so.  The situations in which sales will be a better estimate of future free cash flow than current free cash flow need to be understood carefully and appropriate error bands applied.  Like all tools, EV/Sales will only ever represent an initial screen or a cross check – it will not of itself provide an investing answer.

Conclusion

Every valuation metric has advantages and disadvantages, which is why it is important to use a variety of valuation tools.  EV/Sales is a useful tool both because it provides an alternative proxy of future cash flows and is often overlooked by the market.

Naturally it will not be foolproof, but it provides a useful cross check to earnings based valuation measures and is particularly useful in situations where these measures are at their weakest.


Footnotes:
  1. Note that in this case the term “Cyclical” describes both those situations where external demand is cyclical and those where internal margins are “cyclical” as a result of temporary internal issues within a company.