Unfortunately, most of the investment media focus on the daily movements of the FTSE 100. While this makes a nice soundbite at the end of a news feed, it has about as much use for investors as a chocolate fireguard. Instead, investors should focus on where we are now in the valuation cycle.
Smart investors have long known that the current PE ratio is a poor indicator of value. Instead, it is better to “cyclically adjust” the earnings by averaging them over a decade or more. This idea goes back at least to the 1930s and may go back many centuries more.
A simple but powerful idea
The idea is simple. Over time the price of an asset (whether it’s a share, a house or something else) will move up and down depending on demand from investors. But prices are anchored to future incomes, whether dividends from shares, or rental income or mortgage costs on a house.
Data is available on the S&P 500 index going back over a century, and it shows that the American index’s current price has averaged about 16 times the cyclically adjusted earnings of its constituent companies.
When demand is low from investors, such as during recessions or depressions, valuations fall. At their lowest points, they can fall below half their long-run average. In contrast, when investors are jubilant and expect trees (and stock prices) to grow to the moon, they pay a far higher price, perhaps two or three times the long-run average.
Although we don’t have data for the UK going back a century, we do have a quarter century of data which I will gladly share with you in the chart below:
FTSE 100 Valuations – 1988 to 2013
The main points to note are:
- The black line is the FTSE 100 using monthly data
- The scale on the left is logarithmic, so each step upwards doubles in value. This means steady growth produces a straight line on the chart rather than an exponential curve upwards. It also means that a 10% move in the value of the FTSE 100 in 1988 will be the same size as a 10% move in 2008 (for example)
- The green bands represent CAPE values (Cyclically Adjusted PE), which are the basis for the valuations
- The UK average CAPE is assumed to be 16
- The steady upward slope of the green bands shows how the UK and global corporate earnings have grown through this period (in nominal terms, although they have grown after adjusting for inflation too)
The valuation bands run from half the assumed average of 16 right up to about double that value. This range encapsulates the vast majority of peaks and troughs, both in the US and the UK. Some markets are different and have different average valuations (Japan’s median CAPE of the past 30 years is more than 40!), but the same principle applies.
FTSE 100 still “slightly cheap”
The current value of the FTSE 100 in the chart above is 6,621. This is close to the all-time high in terms of index points, but in terms of valuation, it is below average. That means future returns may be slightly above normal (across a range of time periods), but not by much, and normal means something in the region of inflation plus 5 percent.
Where to next?
Sadly my crystal ball is broken, and particularly so at the moment. With valuations at middling, almost “normal” levels, it is impossible to say what will happen. But that’s okay because most of the time, that’s true. The market is only mildly predictable at extreme valuations and isn’t predictable at all at other times.
But if I must don my prognosticator’s hat, then the chart implies that anything under 5,000 is an absolute bargain, up to 8,000 is reasonable, while 10,000 is pushing it a bit. If we get to 14,000 in the next few years, just remember that it’s a long way down from the top…
Hi, very interesting article, thanks.
I’m a little confused by the graph. The values for the FTSE on the graph look to be correct until the period between 2003-2008. The ramp up here in reality was a lot sharper than what’s on your graph, I think? The peak on your graph in 2008 looks to be just over 4000, whereas in reality it was around 6700 – basically the same level as the peak in 2000. The slump in 2009 on your graph looks like it got down to almost 2000, the reality was more like 3800.
Am I missing something to do with the logarithmic scale? I understand that the scale is logarithmic, but the values are meant to match the values of the FTSE 100 right?
Cheers
Hi Bob, sorry about the confusion.
The image is a little small perhaps. Here’s a larger version. I’ve also added horizontal lines to help make things clearer, but it’s still not perfect, but is a bit better I think.
As for the data, it’s fine. The data is taken at Jan 1st, or thereabouts, so it doesn’t represent the high or low for a year. So for example, in 2000 the FTSE 100 is shown at 6,930, while in 2008 it is at 6,414. From the larger image that looks about right, so it’s also right in the small image, just harder to see.
In 2009 the index is at 3,500, which also seems to be about right. If you grab a ruler or sheet of paper and make a horizontal line where the 2009 data point is, then it does line up on the vertical axis (FTSE 100 values) slightly below 4,000 in 2009, which, given the logarithmic (base 2) nature of that axis, is what I’d expect.
I hope that’s a bit clearer? Let me know if it isn’t and I’ll whizz up an even clearer graph if possible!
John
Hi John,
How strange, you’re right. I lined up some paper and it does indeed do as you say. The human eye is a funny (and inaccurate) thing.
Having held a FTSE tracker since June 2009, the recent ramp up has me wondering whether I should think about selling. I don’t really have a need to sell in order to buy something else, so for the moment I’m leaving it as is. With dividends, it’s returned something like 65% over the last 4 years which I’m not complaining about, and the yield on the cost I paid is almost 5%.
At some valuation it will have used up most of its margin of safety, and its likelihood to extend further into positive territory will be reduced. It’s just a matter of whether I hold it for a 5ish year period, or sit on it for a long time and let the dividends roll in. All I know for sure now is that if there was another plummet like we saw up to 2009, I’d park some more money in and forget about it for a while.
Thanks again!
Hi,
You mentioned that the UK average CAPE is around 16, but just from eyeballing your chart it looks to be higher, it looks as though it has only been below 16 for 5 of the last 25 years. Could you explain why? How do the bands work also, the darkest band corresponds to CAPE 18, so is that really CAPE 20 to 22?
thanks
Hi jbox, you’re right, the average UK CAPE from my data is about 18 over the past 25 years. I usually say it’s around 16 because the true average (I don’t have any data going back before 1988) is likely to be lower than 18. The S&P 500 average over a century is just over 16, so I take 16 as a simple estimate of a more accurate long-run average. For example, since 2002 (after the dot com bubble) the average has actually been exactly 16.
Of course 16 is likely to be wrong, but whether it’s 14, 16 or 18 doesn’t really make that much difference. What matters is being able to see when markets are hugely out of whack and CAPE is 10 or 30.
As for how the bands work, the design isn’t perfect and is limited by my expertise (or lack of) with MS Excel. The lower white area is CAPE 0-8. The first green band is CAPE 8-10, the next one is CAPE 10-12, and so on. So the middle, darkest green band covers valuations from CAPE 14 to 18. The very top border between green and white is CAPE 30, and the top white area goes from 30 to infinity and beyond.
Hopefully that makes a bit more sense now?
perfect, thanks