Last Updated February 24, 2014
In recent posts I’ve looked at a variety of ways to track down and compare defensive shares. I’ve covered looking for consistent profits and dividends, consistent growth, high rates of growth and low debt. But finding defensive shares is only half the battle; the other half is having a reliable way to value them.
The problem is it’s usually not obvious whether a company’s shares are expensive or cheap, and therefore it’s not obvious whether any given company is a buy or a sell.
This lack of obvious bargains is partly down to market efficiency. If a company is obviously cheap then investors will start buying, which will push up the price until the obvious cheapness goes away.
On the other hand it’s also down to the tools that most investors use to value companies.
Investor are only as good as their tools, and most investors use very blunt tools
The de facto standard valuation tool is the Price to Earnings ratio (PE). It compares the current share price to the company’s earnings per share last year.
In principle it makes a lot of sense because earnings provide the basis for most of the returns that investors get. They’re the starting point for future dividends and they’re also the source of much internal investment within companies. That internal investment goes into things like new product research or new factories, which drive future earnings growth and capital gains.
So the lower the PE ratio the higher the earnings yield, and the more earnings you’re getting per pound invested; so far so good.
But earnings are often volatile, even in defensive companies.
A company with shares at 100p and earnings of 10p has a PE ratio of 10. But if the earnings drop to 5p next year the PE will change to 20. If the average PE in the market is 15 (which is a reasonable ballpark figure most of the time) then a PE of 10 looks cheap, while a PE of 20 looks expensive.
And yet, fundamentally, almost nothing may have changed in a company whose earnings have gone from 10p to 5p. A 50% drop sounds like a lot, but there could be many reasonable and legitimate reasons for it, and it may have no bearing whatsoever on how much the company is earning 5 years from now.
In this case the standard PE would give you a faulty signal. It would tell you that the company had become more expensive when in fact it hadn’t. The company’s long-term fundamental or intrinsic value could well be exactly the same as it was before (note that here I’m assuming the true or intrinsic value of the company is the discounted value of all the cash it will ever return to shareholders, which may or may not have anything to do with this year’s earnings).
The same argument applies if the company’s earnings spiked from 10p to 20p.
The PE, with the shares at 100p, would drop to 5, making the company look incredibly cheap. But if those high earnings of 20p were a one off, and the company’s ability to generate profits across the business cycle hadn’t fundamentally changed, then the PE ratio would be giving you another faulty signal.
Most standard valuation ratios are too focused on the short-term
As a long-term investor rather than a speculative trader, I’m looking to own companies for at least 5 years. I may own them for a shorter period of time or a longer one, but 5 years is my default expectation.
The problem with the standard PE is that it only tells me about how a company’s share price compares to its earnings in the current year. As such it tells me almost nothing about where the share price will be 5 or 10 years ahead.
Of course there is no way to know what the share price will be tomorrow, let alone 5 years from now, but there must be a better way of predicting whether the shares are likely to go up or down over the next 5 years than by looking at last year’s earnings. And there is.
A better long-term valuation tool is the Cyclically Adjusted PE ratio
Professor Robert Shiller popularised the idea of comparing current prices to long-term average earnings. Specifically, he used 10 year inflation adjusted earnings in his Cyclically Adjusted PE (CAPE). By applying this measure to the S&P 500 index, Shiller was able to show that the US market in 2000 was clearly expensive relative to its historic norm.
The idea of using average earnings instead of just last year’s earnings is simple, and dates back at least to Ben Graham’s PE10 in the 1930’s (PE10 is the same as CAPE except the earnings are not adjusted for inflation).
By averaging out earnings over a number of years we can reduce volatility in the E part of the ratio.
That’s because as each year goes by 9 out of the 10 years remain the same and we simply replace the earnings from 10 years ago with the latest earnings. Year to year earnings fluctuations then have little impact on the overall average.
So by averaging across a number of years we have a much steadier, more stable E number which doesn’t change much year to year. The only part of the CAPE ratio that bounces around is the P part. Because E is relatively stable over many years a high P today will still be a high P relative to next year’s E, and it will probably be high relative to whatever E is 5 years from now.
So CAPE gives us a much clearer signal of cheapness and expensiveness. A low CAPE value tells us that the price is low today, not to today’s earnings but to the earnings that the company will likely produce for a number of years.
This means it can be used to make far more effective buy and sell decisions.
Using CAPE to get high returns with low risk
Professor Shiller found that over the past century the S&P 500’s CAPE had averaged around 15. When CAPE was higher than average it tended to fall in the years ahead, partly through earnings growth but primarily because the index fell. When CAPE was lower than average it tended to expand in the years ahead, not by shrinking earnings but through a rapidly increasing index.
Most of the time this mean reversion was clouded by the random movements of the market. If CAPE was only slightly high then it would be almost impossible to tell if the market was going to go up or down in the next few years. But when CAPE was close to extreme values, perhaps as low as half or as high as double its long-run average, it almost always mean reverts aggressively, taking the market up or down with it.
This means that if you can buy an index like the S&P 500 or the FTSE 100 when their CAPE values are very low – in other words when other investors are running for the exit and there is blood in the streets – you have a good chance of getting high rates of return with little risk:
- High returns from high dividend yields – Dividend yields make up about a third of the total return from the stock market. When markets have a low CAPE, their yields are usually high, which means more income from dividends, which boost future returns.
- High returns from expanding CAPE ratios – If the market’s average CAPE is 15 and the current ratio is 10, there’s a good chance that you’ll see a 50% increase in shares simply from the CAPE ratio expanding back to normal levels. Think of CAPE like a spring – the more you compress it the more it wants to bounce back to its ‘default’ value. Conversely, if you start with a CAPE of 30 (as we did in the year 2000) you’re likely to see price falls of up to 50% as the CAPE ‘spring’ mean reverts.
- Low risk as CAPE has little room to fall further – Market valuations mean revert, so the lower the valuation ratios get (CAPE in this case) the more likely they are to go up (and create high returns) and the less likely they are to go down. In other words, low valuations create a potentially low risk, high return opportunity.
Applying CAPE (or PE10) to defensive companies
You may be wondering what CAPE has to do with valuing defensive shares. After all, CAPE is a tool used to value market indices, not individual companies. That’s true, and there’s a good reason why CAPE isn’t used to value individual companies.
With an index you’re starting off from a steady base. Indices like the FTSE 100 always make profits and pay dividends, and this means that most of the time the past looks like the future, so taking an average of the past 10 year’s earnings is a good way of predicting the next 10 year’s earnings.
This in turn means that a low CAPE today is an indicator that the current price is low relative to future earnings and not just past earnings.
But most companies don’t make profits every year in a decade. They make losses, they cut or suspend dividends, they get taken over and they frequently go bust.
This volatility and uncertainty makes CAPE less appropriate for individual companies than it is for markets. If a company doesn’t have a strong position within its market then what it earned in the last 10 years ago may be of little relevance to what it may earn in the next 10 years (assuming it can even survive the next 10 years).
But defensive companies are different. They’re strong, they produce reliable dividends and, for the most part, can be expected to produce profitable dividend growth for many years if not decades to come.
Many defensive companies are so reliable that their financial results can look surprisingly similar to those of an index, and that’s why CAPE can be a viable and powerful tool for valuing them.
When it comes to actually valuing companies with CAPE, I prefer to use Ben Graham’s PE10 instead, which doesn’t involve adjusting the earnings for inflation. In theory inflation adjusted earnings are better, but for practical purposes PE10 is just as good, and is that bit easier to put together.
The process of valuing a company using PE10 is simplicity itself. You just calculate the average earnings per share over the last 10 years (I prefer adjusted earnings rather than basic earnings per share) and work out the ratio between the current price and that earnings figure. A lower PE10 is of course better, all else being equal (which it never is; a point I’ll come back to later).
Applying the same thinking to dividends
Defensive investors are usually interested in dividends and the dividend yield of their investments, but the standard dividend yield suffers from much the same problems as the standard PE ratio. Perhaps not quite to the same extent, as dividends tend to be more stable and progressive than earnings, but the basic problems are still there.
If a company cuts its dividend the dividend yield will fall. In most cases a dividend cut leads investors to sell, because the lower dividend leads to a lower dividend yield, which for many is a sell signal.
But time and time again this has proven to be a mistake. When a defensive company cuts its dividend it’s usually only a matter of time before the dividend can be rebuilt. It may take several years to recover, but to a long-term investor a period of several years should be nothing.
Eventually investors realise this and the share price recovers, usually long before the dividend does, and those who sold out after the cut are left with losses that they could easily have avoided.
So rather than just comparing the company’s current share price to its latest dividend, I prefer to compare the share price to the average dividend paid over the last decade. The result is the PD10 ratio, which sits alongside the PE10 ratio as a core valuation metric for defensive stocks.
Once again, the ratio’s robustness and stability are key. A dividend cut today will have little effect on the PD10 ratio, so if a company was cheap according to PD10 before the cut, it’s likely to still look cheap after the cut. This is in contrast to the standard dividend yield which will tell you that your shares are expensive if the dividend is cut.
By using both PE10 and PD10 you’ll have an incredibly robust, stable and usable means of valuing defensive companies.
Year to year volatility in earnings and dividends will no longer give you wild and unpredictable buy or sell signals. You’ll be able to take a much more long-term view of the companies that you own, and you’ll be at least partially insulated from the unpredictable ups and downs of short-term events.
Combining defensive investing and value investing
Of course no two companies are the same, and two companies that have the same PE10 ratio are unlikely to be equally attractive. For example, a company that has been growing at 10% every year like clockwork is going to be worth a substantial premium over a company that hasn’t grown at all.
To take account of the varying defensiveness of different companies it’s important to combine these valuation ratios with the other metrics; specifically, the ones I’ve recently covered that measure a company’s growth rate and growth quality and debt levels.