The Long-Term PE
It’s a PE ratio, but not just any old PE ratio
Last week I boiled the way I pick shares down to three questions, the only three that count. Having spent most of the blog building up to that point I didn’t elaborate, leading Mike to comment:
It is useful to have a checklist but I think it is dangerously seductive to think that successful investing can be distilled into answering just three questions.
Quite right. So perhaps I should reveal that there is more to these three questions than meets the eye. Taking them in turn:
1. Is the company cheap?
I usually use the long-term price earnings ratio to determine how cheap a share is. That’s the average of five to nine years of earnings per share divided into the price (per share). I would use up to ten years, but my data only goes back nine. Companies with less than five years of earnings don’t have much of a record, so I don’t bother with them.
Earnings, profits, are important because they belong to the investor after the company has paid all its bills. We benefit directly, through the dividend, or indirectly through money retained by the company and invested to make more money. Our ownership entitles us to a share of the profit, which we’ll reap as a dividend or through a rising share price, assuming the company invests retained profits well.
The long-term average is important because the most recent year’s profit or analysts’ forecasts for the current year, the profit figures used in the price earnings ratios you see quoted in newspapers and on websites, may not represent the company’s true earning power – that is how much it might earn in a typical year. If you base your decision to buy a stock after a particularly good year, you’ll be disappointed when it fails to meet your expectations in subsequent years.
To help protect me from clever accountants fiddling the profit figure, I compare a company’s record of profits to its record of cash flows, the actual money coming in and leaving the company. If the latter is markedly inferior to the former, I forget about the company.
By comparing the price of the company to its average earnings, we find out how much we’re paying for a typical year’s profits, ten times earnings, for example, if a company’s price earnings ratio (price per share divided by earnings per share) is ten.
As it happens, ten times earnings is quite a good price to pay. Assuming the profit is your return, you can do the division the other way, divide the price per share into the earnings per share to calculate what fraction of the investment you would receive as profit, in theory, every year (it wouldn’t work out that smoothly). A company on a price earnings ratio of ten would yield 10%.
The earnings yield is a convenient figure because you can compare it to the returns you might get from money in the bank (interest) or invested in bonds (yield). Ten per cent is OK and although I might pay more, and therefore accept a lower return now, for a company I expected to earn much higher profits in the future, I’m not sure I’d ever bother with the hassle of investing in shares for less than 5% (a long-term PE ratio of about 20).
That’s what I mean by cheap. Typically, I like to pay up to ten times long-term average earnings. Sometimes I’ll pay up to sixteen times, but rarely more than that.
Does it work in practice? Dr Keith Anderson, now at the University of York back-tested an eight-year price earnings ratio on UK stocks and found that shares with the lowest PE’s returned 6% more a year than shares with the highest PE’s, almost twice the premium of a conventional PE ratio. James Montier, analyst at SG, tested long-term PE ratios globally with similar results (I don’t have his excellent book to hand, so I can’t give you the figures). Benjamin Graham, the granddaddy of value investing, inspired them both. He favoured the long-term price earnings ratio.
Of course, shares are often cheap for good reason – they’re stinkers. They’re going bust, or they’ll never earn the profits they have in the past. I try to protect myself from them, and boost my returns, with questions two and three, which I will have to explain later in the week as this blog has seriously run over. 
Footnotes:
- By the way, The Times’ Money Central has put us in its list of top twenty British personal finance blogs. Thanks Times Online Money Desk people!
35 Comments
Leave a comment
Comments
- Richard Beddard on Learning from Lauren
- Mark Carter on Learning from Lauren
- Richard Beddard on Ambivalent about French correction
- Mark Carter on Ambivalent about French correction
- Ken Kahura on Towards the perfect PE
- Richard Beddard on Games Workshop in two minutes
- Ethereal on Games Workshop in two minutes
- Monevator on Throwing the net wide open
- Richard Beddard on Throwing the net wide open
- Monevator on Throwing the net wide open
- Richard Beddard on Throwing the net wide open
- Monevator on Throwing the net wide open
- The cyclically-adjusted P/E ratio (PE10 or Shiller PE) on State of the market
- Philip O'Sullivan on Churchill China in 1 minute 53 seconds
- Richard Beddard on Churchill China in 1 minute 53 seconds
- Market Musings 1/5/2012 « Philip O'Sullivan's Market Musings on Churchill China in 1 minute 53 seconds
- Philip O'Sullivan on Churchill China in 1 minute 53 seconds
- Brad on Million Dollar Traders
- Stefan | Simple Value Investing on Pensions: peril or profit?
- Richard Beddard on PV Crystalox Solar in 1 minute 56 seconds
RB on Twitter
- @pdosullivan Thanks Philip. Ditto.
- Interesting thoughts from @mcturra2000 on magic formula investing http://t.co/JEq0Ak1j my reply: http://t.co/UHRnYZKR
- Just discovered there are two Mervyn Kings. This one hits the bullseye https://t.co/JmPZPIIp This one moves the target https://t.co/JmPZPIIp
- @smarkus Thanks
- Thinking of tackling Next L:NXT next. It scares me witless. @spbaines and @GeoffGannon wld have me rely on earning power. Soooo hard...
Latest posts
- French Connection in 2 minutes 9 seconds
- Restoration man
- Redefining French Connection
- Leases key to retailer’s financial position
- Ambivalent about French correction
- Three earnings yields
- Premier Foods in 2 minutes 11 seconds
- Throwing the net wide open
- Holders Technology in 2 minutes 8 seconds
- Churchill China in 1 minute 53 seconds
Sections
Companies
Archives
Blogroll
- Alphaville
- Barel Karsan
- Eurosharelab
- Expecting Value
- Gannon on Investing
- Mark Carter
- Monevator
- Musings on Markets
- My Investing Notebook
- Neonomic
- Oddball Stocks
- Peston's Picks
- Philip O'Sullivan
- Seth's Posterous
- The Value Perspective
- Turnkey Analyst
- UK Value Investor
- Value Stock Inquisition
- Valuhunteruk.com
- Wexboy
[...] the company scores a perfect nine. It’s cheap too, the share price is just eight times its average profits over the last ten years. Beneath lies a more complicated story, [...]
[...] and bad years, I use the average of the past ten years of earnings to calculate the 10 year PE. Academic research indicates it’s a better measure of value and by restricting the Thrifty 30 to companies with 10 year PE ratios of less than 20, I’m [...]
[...] valuation too, is suspect. The shares cost less than seven times the average of its last ten years of profits, but that figure is only relevant if Wolseley will achieve a similar level of profit in future. I [...]
[...] the mid-price of 472.5p, the market values the shares at 10 times earnings over the past ten years. It’s just about a bargain share. Were I to include the company in the Thrifty 30 model portfolio [...]
[...] is a borderline case. Take away the two good years of earnings before the restructuring and its eight-year average price earnings ratio is 12, above my admittedly arbitrary limit of 10 for a bargain [...]
[...] criteria, a Graham & Dodd price earnings ratio of less than 16. The Graham & Dodd PE is the long-term PE. It compares the price to its average earnings over the past ten [...]
[...] the share price has crashed back down to earth, I’m wondering if the reward, its ten year price earnings ratio is just six which makes the shares look very cheap, justifies the risk of buying shares in such a [...]
[...] shares cost twelve times average earnings over the last ten years, which looks cheap for a company with a record as consistent as RM’s. But RM looks more like a [...]
[...] its increasing indebtedness makes me a little leery of the share price, which is 15 times earnings averaged over the past ten years. Investors willing to pay that much must believe the investment will earn higher average profits in [...]
[...] has been restocking since it last commented in October. Short of a disaster tomorrow, though, on a 10 year PE of eight, Haynes still deserves its place in the Thrifty [...]
[...] shares are in bargain territory, they cost six times average earnings over the last ten years, and perhaps more importantly four tenths of book value. It’s so cheap, it’s quite close to [...]
[...] a glass processing business. Despite the dramatic recovery in its share price, the shares cost only eleven times average earnings over the last ten years, just outside bargain territory, and, especially as management describes a [...]
[...] 90% lower than their peak, and although they’ve doubled since, the fact that they cost less than five times average earnings suggests Armour might still be on the critical [...]
[...] compared to £211m for Johnson. And investors like Davis more. They’re paying fourteen times average earnings for the shares, while they’re not even prepared to pay an average year’s earnings for [...]
[...] shares are cheap too, costing five times average earnings over the last ten years, but nearer ten times this year’s earnings. Under the company’s previous management its profits [...]
[...] an average PE ratio is therefore a conservative measure. In comparison to the most recent PE ratio, or forecasts, it [...]
[...] The performance of its shares mirrors the trajectory of its profits, albeit less smoothly. They traded above 500p between 1999 and 2000. At the nadir of the bear market in 2009 the shares touched 20p. Today they stand just over 50p, at a bargain price of about four times average earnings over the last ten years. [...]
[...] shares cost 14 times average earnings over the last ten years, which isn’t obviously cheap for a UK company, the average is 15, and is quite pricey for one [...]
[...] shares are certainly cheap enough for the Thrifty 30, costing just nine times average earnings. If the collapse in Game’s share price is a reaction to the temporary effects of recession and a [...]
[...] a small amount of shares issued to reward executives. Since the company costs just over one times average (adjusted) earnings over the last ten years, statistically speaking Autologic is the perfect Thrifty 30 company, as long as the new managers [...]
[...] second value measure is the long-term price earnings ratio. Ultimately it’s profit that rewards investors and many investors relate price to profit using [...]
[...] to add M&S to the Thrifty 30 portfolio though. Although the shares are cheap, on 12 times average earnings per share over the last ten years, they’re not cheap enough to displace one of the existing shares or use more of the small pool of [...]
[...] circumstances and I stand by my conclusion last October. At ten times average earnings over the last ten years, and 1.1 times book value the shares look cheap. Castings’ F_Score is 6, and its long history of [...]
[...] and to identify cheap companies he used the granddaddy of ratios, price to book value. I favour the long-term price earnings ratio but although Argos and Homebase, the retailers owned by Home Retail, are long-established [...]
[...] In putting together the case for Home Retail recently, I relied on the company’s book value as well its PE, partly because it’s only been listed on the stock market for four years and so doesn’t have a long enough earnings record to compute the average of 6-10 years I favour. [...]
[...] of serious trouble and has not been earning good profits I might remove it once it achieves, say, a long-term PE of 10. If however, like XP it’s consistently profitable, say its return on total assets has [...]
[...] looks like its regaining some strength, and it looks cheap too. The shares cost just two times average earnings over the last ten years, and although they’ve recovered slightly, they’re still well under 10% of their peak [...]
[...] parts, which identify cheap and financially strong companies, can be outsourced to statistics like PE and PB ratios, and the F_Score. In comparison, understanding businesses is fuzzy and time [...]
[...] have more than doubled since the price bottomed in early 2009 and they cost eleven times earnings averaged over the last ten years, and 1.1 times book [...]
[...] With an F_Score of four out of nine it looks financially weak, and costing nearly fourteen times average earnings, it doesn’t look particularly cheap, so if I were a purist, I’d eject it from the [...]
[...] shares are cheap. If we value the company at ten times its average earnings over the last ten years, my usual benchmark for value, they would be worth 30p each. The current mid-price is just 8p. [...]
Does anybody maintain detailed (e.g., a spreadsheet, not just a graph) and publicly available historical data about UK 10PE, like Robert Shiller does for the S&P500?
http://www.econ.yale.edu/~shiller/data/ie_data.xls
Clearly, the values of historical UK stock index levels and of UK CPI/RPI are easily and publically available. In order to reuse Shiller’s spreadsheet to track the UK 10PE, the only missing thing is quarterly/annual earnings data for the corresponding stock index.
In your “State of the market” page (http://blog.iii.co.uk/state-of-the-market/) you mention Sharelockholmes.com and Sharescope databases as your data sources. Is there a freely available historical data source about UK stock indices earnings, such as this one for S&P500?
http://www2.standardandpoors.com/spf/xls/index/SP500EPSEST.XLS
The caption on your UK long term PE graph (http://blog.iii.co.uk/wp-content/uploads/2011/03/1103ukltpe.jpg) is ambiguous because it says “Average (median)…”. Do you mean the average or the median?
Note that the difference between the average and the median can be substantial because the average is capitalization weighted and the median is not.
Or do you perhaps mean “capitalization-weighted median”? The latter could be defined, for example , as the smallest 10yPE of a company that has the following property: the combined capitalization of this company and all companies with smaller 10yPE exceeds 50% of the capitalization of the whole market.
Just a comment on my silly remark above that “the average is capitalization weighted and the median is not.” Of course, one could also take the simple average of 10yPEs of all companies, but that might be the least meaningful of the four options: capitalization-weighted average (as in the Shiller’s 10yPE of S&P 500 linked to in the 2 March comment above), simple median, capitalization-weighted median (as defined in the comment above), and simple average.
Hi lostinmidlands:
“Of course, one could also take the simple average of 10yPEs of all companies, but that might be the least meaningful of the four options”
- rightly or wrongly, that is what I currently do.