Under the rule of thumb

For something so simple, the Price to Earnings Growth (PEG) ratio has to be one of the most misunderstood of all financial ‘metrics’.

The PEG ratio attempts to take the basic single period P/E ratio further – bringing in time and growth.  This makes sense as an objective, but the ratio is not based on much in the way of maths that would suggest it should provide a reliable, widely applicable or consistent result.

The PEG ratio is a rule of thumb.  It has, however, taken on a far higher status among some investors.

Standing on the toe caps of giants

The application of PEG as a tool first rose to prominence following its inclusion in Peter Lynch’s book ‘One Up on Wall Street’, published in 1989.  In the UK, the idea was given a real drive forward in the 1990s with Jim Slater’s book ‘Zulu Principle’ and with the associated REFs investment guides.

While PEG is now seemingly applied with abandon, looking back what these investment masters actually proposed was far more restrictive and limited than today’s application, with both Lynch and Slater giving guidance on what constituted a PEG and when it was appropriate.

Lynch, it must be said, only wrote a few summary paragraphs on the subject in his book.  Slater laboured the point on definitions and the insight that could be gained from correctly calculated PEGs.  His involvement with REFs, a commercial source of correctly calculated PEGs, might point to why he was quite so keen for everyone to get the right data.

The basic message was that a PEG of 1 suggested that a company’s shares were fairly valued.  Slater was more specific with his suggestion that 0.75 was the number to be looking for, but it has been the PEG of 1 that has stuck fast.

It is unclear whether Lynch and Slater came to the PEG of 1 figure by trial and error in practice, by random chance seeking a figure that sounded good for punters, or after much toying with the valuation formulae of financial academe.

What is clear is that much has changed since they were writing and it is worth considering whether the PEG is still a valid tool, just a badge of intellectual laziness or perhaps simply another useful mechanism for analysts to try to mislead investors.  It is also clear for anyone who bothered to read more than a few lines in either book that both Lynch and Slater understood that the PEG of 1 was not a fundamental principle or something that would hold across a wide range of scenarios.

What’s behind it, not who is behind it

Investopedia claims that the PEG of 1 is theoretically underpinned by the Sum of Perpetuities Model.  If the company does not pay any dividends and has a risk adjusted discount rate of 10%, then the theoretically correct PEG generated by the Sum of the Perpetuities model is 1.  Some – it seems quite a few – might be satisfied with this.

Looking at the equation for the Sum of Perpetuities below, it is evident that other combinations could lead to a PEG of 1, but also evident that most combinations do not. So the PEG of 1 is inaccurate most of the time even within the confines of the assumptions that drive the Sum of Perpetuities Model.  Perhaps not a great start, but we are only looking to the PEG as a guide, a rule of thumb, not as a means to derive a precise valuation.

(Where P=price, E=earnings, G=constant growth rate, D= dividend payout, and K=cost of capital)

For anyone recoiling at the horror of such an equation, with memories of schooldays and lines on charts following weird and wonderful paths, fear not. No further mathematical analysis is required because your natural reaction has already provided the answer to the question about how reliable PEG is going to be as assumptions change.

To step away from the strictures of the Sum of Perpetuities model, we can calculate what the correct theoretical PEG is for a basic two-stage DCF.  We are going to look at an assumed ‘generic’ company (growing at a higher rate for five growth years, then growing at a lower mature rate of 8%) with a dividend payout rate in the growth period of 20%, rising to 50% in the later times.  We can then apply different costs of capital.

The chart below shows the theoretical PEGs (y-axis) for scenarios with differing growth rates (x-axis) in the first five-year growth phase with a 12% cost of capital. Looking at a risk-free rate of around 4% and equity risk premium of 7%, this is not unreasonable and not completely out of line with much of the 1970s and 1980s (we are talking in broad terms here!) when Lynch and Slater were investing and writing.

But this is only a starting point – admittedly a carefully chosen one. As we can see when we add in PEGs for costs of capital of 11% and 14%, PEGs can only be a rough guide. A lower cost of capital means a higher PEG is justified – but the impact is to ‘lift’ the line, not radically change its shape or slope.

We can then add in another major factor – the length of the growth period.  As can be seen from the diagram below, the ability to sustain premium growth for more than few years has a considerable impact on the valuation, with a clear suggestion that paying more than 1x growth has merit when growth is both strong and sustained.  This is perhaps no great surprise now, but it does give permission to let go of the 1x PEG fixation.

Hammer, not a Manchester screwdriver

So it would appear that PEGs can’t be reliably compared across companies with different growth rates, growth profiles and different costs of capital. This shouldn’t come as a great surprise – the association of wavy lines and complex formulae was enough to tell us that.

Another reason why it might not be a surprise is that this is exactly what both Lynch and Slater wrote, only a few pages on from delivering their headline insights.  Unfortunately, once again the attractions of the simple rule of thumb override the attractions of actually understanding.

To compare PEGs between companies requires similar growth profiles (and cost of capital etc.), and once you have established that, then you might just as well compare P/Es or go into what might be described as real detail about the nature of the earnings.

We should not, however, forget where PEG came from and how it was originally used. PEG is more of a filter to identify potential value. It is also a discipline based on significant and profitable experience, so it would be wrong to dismiss it as irrelevant to the modern stock market.

We just have to understand that a company with 25% annualised earnings growth over five years and with shares standing on a P/E of 25 could well be considerably better value than one with 10% earnings growth on a P/E of 10 – something that has use in itself.

A final thought is that it is perhaps a shame, and ironic, that overly strict adherence to a rule of thumb from leading growth investors may have stopped later investors who cast themselves in the mould of the greats from investing in truly undervalued growth opportunities.

Ian Robertson

irobertson@progressive-research.com