A hiding to (Con)naught
Posted on November 28, 2007 by Richard Beddard
Filed Under Companies, Investing |
Not being a momentum investor, I can’t offer an informed view on how long that might be and, in the very short term. (i.e. this month) it’s on the way down.
But with my value investing goggles on I can offer a new perspective. Inspired by the work of James Montier and Keith Anderson, I calculated the long-term price earnings ratios of companies listed on the London Stock Exchange. Five hundred and twenty four were profitable in eight of the last nine years.
Otherwise known as Graham and Dodd PE’s these ratios use earnings per share averaged over at least five years, and preferably ten. The PE ratios you see on websites and in the financial pages use only one year (usually last year’s, or a forecast).
Benjamin Graham, the father of value investing, and his partner preferred long-term PE ratios to avoid being swayed by a few years of exceptionally good (or bad) results. They recognised that profits are, generally, cyclical, or to put it another way exceptional profit growth rarely persists for a decade.
Although their insight predates the Second World War it’s never really caught on, a testimony perhaps to the short-term mentality of investors, or our innate optimism that every growth share is potentially the next Microsoft.
But James Montier’s research on global stocks shows that companies with the lowest ten-year PEs (the bottom 10%) returned 21% per annum between 1980 and 2005. Companies with the highest PEs (the top 10%) returned -4%. Keith Anderson uses long-term earnings averages as well, but makes other adjustments to the PE. His back testing on UK stocks with extremely low PE ratios returned 39% a year.
That doesn’t mean that every share on a high PE ratio is set for a fall, but in the medium to long-term the odds favour low long-term PEs.
Glancing through the list, I noticed Connaught. It’s in 526th postion with a long-term PE ratio of 89. It’s not just in the top decile. It’s in the top three. Yikes! The business is going great guns, but what happens when there’s a hiccough?
At 365p, there’s a lot of air between it and a price based on a sensible valuation.
Comments
4 Responses to “A hiding to (Con)naught”
Leave a Reply
Hi Richard,
Thanks for writing this article and highlighting the fairly amazing result from Montier’s book about the average earnings PE ratio. You made me go back and read that chapter carefully again.
The fact that the Montier results are so volatile (in the long-only variation) suggests that there might be a danger for Keith Anderson, as he is only choosing a few shares each year (he might choose the wrong ones). I would have thought that you would get safer performance if you selected a larger subset of low average PE shares. Assuming that there really is a high correlation with out-performance, then you could expect better performance through this diversification.
I am confused by the following which you might be able to clarify. G & D were emphasizing using average earnings based PE to eliminate the effects of the 10 year cycle. This suggests an underlying long term value strategy. However both Montier and Anderson only invest in shares for a short term (1 year) period. It seems they are using the same idea (average earnings) but expecting a different effect (short term upward momentum in share price).
They might get better results by actually timing the investment to select shares which have a lower forward PE than the long term average PE. I imagine getting hold of historical forward PE’s will be very difficult!
Hi Robin
I must admit it had occurred to me that testing the long term pe on portfolios rebalanced every year is unlikely to bring the best out of the value approach. I suppose you have to use some time frame and a year is ’standard’. I know Keith doesn’t advocate trading annually in practice.
Although James Montier’s book also uses 12 month rebalancing to demonstrate the supremacy of long-term (Graham & Dodd) PE ratios, in a later chapter (31) he does demonstrate that a simple value strategy based on normal 1 year trailing PEs generates 3% excess return (over the market). If you hold, though, you get 8-10% in years 3,4, and 5.
I don’t know if either of them have tested holding companies with low long-term pe ratios over the long-term. It seems like an obvious test to me… I haven’t got to the end of Montier’s book yet though!
Hi Richard,
Ken Fisher shows, in his book, that there is no correlation between PE’s and out-performance. I do not know what to think anymore.
Hi Robin,
I know what you mean! I take some comfort from the fact that Keith’s results and James Montier’s were independent of each other and in different markets (Keith=UK, James=Global).
Also (from memory) Ken was talking about 1 year trailing PE’s. James Montier apparently got 3% outperformance using 1 year trailing PEs and Keith something similar, maybe slightly less. The point is with spreads, charges and taxes that’s going to be a pretty miserly advantage. I think both James and Keith say that. Keith’s certainly of the opinion there’s no point in following a strict low trailing PE strategy.
So maybe Ken’s right if you frame value as just buying shares in companies with low trailing one year pe ratios. I don’t though. I’ve always added other things into the calculation (low debt primarily, and no obvious reason why the business is likely to go down the pan). Now I think the long-term pe is an important factor too - in fact as a variation on that theme I always preferred shares that were on a lower PE ratio than in previous years, which is saying a similar thing.