Shares for long-term value

Brexit negotiations drag on and there may well be surprises along the way to the ultimate deal-or-no-deal conclusion. Nevertheless, there’s arguably little point in re-running such a thematic screen given the market should have now made a fairly full adjustment to the event.  Instead, this week I’m running a screen addressing a current bête noir for investors: high valuations.

Indeed, stocks look expensive across a broad range of measures and there are fears that rising interest rates (the risk-free yardstick stocks are valued against) could prick a valuation bubble. One of the most vocal critics of current valuations has been John Hussman, a highly-respected American academic turned fund manager who has called other market tops. It seems fitting that this screen takes inspiration from a key valuation tool he uses to assess the market as a whole.

Mr Hussman has been bearish for a number of years and he now calculates that US equity market valuations are about 175 per cent above their historic norms. His research also suggests low interest rates have little bearing on long-term (10 to 12 year) returns because for them to remain low, growth needs to disappoint, which takes away a central justification for high earnings-based valuations – dang!

What's the big idea?

The ratio Mr Hussman uses to evaluate long-term (10 to 12 year) market returns is a slightly doctored version of the price-to-sales ratio (a “nonfinancial market cap to corporate gross asset value” ratio). His use of the metric is based on analysis of its success at predicting long-term market returns since 1950 compared with many other popular valuation tools. For the intellectually intrepid, his weekly comment from 9 October 2017 provides a more in-depth explanation of the ratio and his outlook.

The logic Mr Hussman gives for price-to-sales being such an effective way to assess value is compelling. And, with a few of contortion, the same principles can arguably be applied to individual stock valuations (more of that in a bit). Put simply, Mr Hussman believes profit margins vary massively over the economic cycle, but also massively revert to mean. So sales are actually a better representation of long-term earnings potential than actual earnings themselves, even when long-term average “cyclically adjusted” earnings are used – a method made famous by Nobel Prize winning economist Robert Shiller.

Stock take

So, how can the concept of finding value based on the sources of earnings rather than earnings themselves be applied to individual shares? I already run several screens based on the price-to-sales ratio and whilst they often produce impressive results, stock selection tends to be skewed towards certain types of low margin, and cyclical companies.

Another issue with looking at a price to sales ratio when analysing individual stocks is that sales are not the real source of profit for some companies. Indeed, for real estate companies, investment companies, banks, and other financial companies, it is net assets that can be seen as the ultimate driver of the bottom line.

Squaring the circle

The way this week’s screen attempts to get around these problems is by simply comparing each shares’ current rating against its ultimate source of earning to its historic range. For investment trusts, property companies and financial companies, the screen looks at the price-to-tangible-book-value (PTangBV) ratio. For all other companies the screen looks at price-to-sales (PS).

To draw a comparison between the current share rating and its history I’ve used a statistical measure called the Z-Score (anyone who’s followed my outperforming and overlooked investment trust screen may be familiar with this). The Z-score assesses where a share’s current rating sits within a historic range, with a negative score suggesting a stock is cheap against its history and a positive score suggesting a stock is expensive. The Z-scores used by the screen are based on a minimum of six years’ worth of monthly valuation data and a maximum of 12 years. As this is a Z-score based on the value put on the Earning’s Ultimate Source from a company, I reckon a fittingly brazen name for the ratio is ZEUS – the same as the Greek God with a penchant for turning himself into a bull!

True, some companies change in nature during their lives, which means their present and the past will not always be comparable. This is a clear flaw in the approach. Nevertheless, hopefully a few interesting and possibly surprising ideas are present in the screen results, which are listed in the tables below. The first table shows the 50 cheapest stocks from the FTSE All Share based on their ZEUS ratio. The second table shows the 50 stocks that look most promising based on a combined ranking of cheap ZEUS ratio and strong three-month momentum. The idea behind combining momentum with the ratio is to reduce the potential for highlighting value traps.

Just cheap

Cheap and rising

Source: S&P CapitalIQ