After an extremely bumpy 2020 the FTSE 100 has recovered and is now close to record highs.
At the same time, the FTSE 100 cyclically adjusted PE ratio (CAPE) is close to its long-term average.
At first glance that seems like a paradox.
For many investors, all-time high prices are a sign that the market could be overvalued. On the other hand, historically average CAPE ratios are usually a sign that the market is fairly valued.
So which is correct?
Despite appearances this isn’t a paradox and in this article I’ll briefly explain why and then look at what CAPE is saying for both the FTSE 100 and FTSE 250.
Table of Contents
- Prices alone tell us nothing about value
- The ratio of price to intrinsic value is what drives future returns
- Cyclically adjusted earnings are a reasonably good measure of intrinsic value
- The FTSE 100’s historically average CAPE ratio suggests the index is close to fair value
- CAPE and the 2020 stock market crash and rebound
- CAPE suggests the FTSE 100’s total return over the next ten years will be about average
- Some thoughts on the FTSE 250 and its now record-breaking price level
- Closing thoughts: The pandemic could be a spanner in the works
Prices alone tell us nothing about value
To get the ball rolling, have a go at answering these questions:
- If I offered to sell you a car for £10,000 would you buy it?
- What if I came back tomorrow and offered you the same car for £20,000?
- Is the car cheap at £10,000 or expensive at £20,000?
Unless you’re telepathic you’ll have no idea whether the car is expensive or cheap at either price because you don’t know anything about the car.
It could be a brand new Lamborghini or a worthless bucket of rust, so if you’re trying to work out if something is expensive or cheap then price by itself is no help at all.
The stock market is exactly the same.
Just because the FTSE 100’s current price of 7,100 is a mere 10% below its all-time high, that tells you nothing about how attractively valued the FTSE 100 is or isn’t.
And just because the FTSE 250’s current price of 22,650 is about 2% higher than its pre-pandemic all-time high, that also tells you nothing about whether the FTSE 250 is expensive or cheap.
So let’s ignore the fact that the UK stock market is close to record high prices and let’s focus instead on how those prices compare to an estimate of each index’s intrinsic value.
The ratio of price to intrinsic value is what drives future returns
The intrinsic value of a dividend-paying investment is the present value of its future dividends, discounted by an appropriate interest rate.
If you were an omnipotent super-being then you would know exactly what dividends the FTSE 100 and FTSE 250 would pay over the rest of their remaining lifetimes.
It would then be trivial to calculate the annualised rate of return from those future dividends using a discounted dividend model.
Sadly you aren’t an omnipotent super-being and neither am I, so we can’t know exactly what dividends the FTSE 100 and FTSE 250 will pay. Since their future dividends are uncertain, so are their intrinsic values, and that’s why their prices jump up and down so much. Investors’ opinions about intrinsic value change as they digest new news every day.
But we can at least try to estimate the intrinsic value of these indices, and CAPE is one of the better ways to do that.
Cyclically adjusted earnings are a reasonably good measure of intrinsic value
By averaging earnings over ten years we get a much more stable earnings number. The idea is that it’s more closely related to intrinsic value, which should also be a stable number because long-term dividends at the index level are very stable.
The stability of intrinsic value is in contrast to stock market prices, which are excessively volatile because of volatile human emotions rather than volatile future dividends.
When prices are high relative to cyclically adjusted earnings we can assume that investors are paying too much for future dividends because they’re overly optimistic about the future.
And when prices are low relatively to cyclically adjusted earnings we can assume investors are selling too low because they’re overly pessimistic about future dividends.
When the ratio of current prices to cyclically adjusted earnings (the CAPE ratio) is close to its long-term average, we can assume that investors are buying and selling at prices which are at least reasonably close to fair or intrinsic value.
That’s the theory and in practice CAPE has proven to be a reasonably good guide to whether markets are expensive or cheap, and whether future returns are therefore likely to be above or below average.
The FTSE 100’s historically average CAPE ratio suggests the index is close to fair value
Looking at the FTSE 100, over the last 33 years its price has averaged 18-times its cyclically-adjusted earnings, with a high of 34 and a low of 10.
It would be entirely reasonable to use a CAPE of 18 as the “fair value” level for the FTSE 100, but I prefer to use a slightly more conservative figure of 16, based loosely on the 100-year average CAPE of the S&P 500.
I do that partly to err on the side of caution (33 years is not a lot of years to draw historic averages from) and partly because that 33-year period includes the dot com bubble which undoubtedly skewed the average upwards.
So, let’s just say the FTSE 100’s CAPE has averaged 16 over the very long-term and it’s ranged between 10 and 34 over the last 33 years.
With the FTSE 100 at 7,100 its CAPE ratio stands at 15.8, fractionally below that long-term average of 16.
Here’s what that looks like in chart form along with the FTSE 100’s CAPE over the last 30-odd years:
One of the main features of that chart is the spike in prices around 1998-2000. At that time if you wanted to invest in the FTSE 100 you had to pay up to 34-times its cyclically adjusted earnings, a record multiple that still stands today.
The chart also shows that returns from that record high CAPE level have been terrible, and in fact the FTSE 100 is barely any higher today than it was in the late 1990s.
The decline of CAPE from 34 in 1999 to 16 today has acted as a massive headwind for the FTSE 100’s price, more or less negating the gains from the index’s increasing earnings and intrinsic value.
If you want to visualise it, you can think of the yellow part of the rainbow as a reasonable approximation for intrinsic value. So in the late 1990s the FTSE 100 would have been at fair value at something like 3,500, not the 7,000 it almost reached.
With the FTSE 100 near fair value today we don’t face those same headwinds.
CAPE and the 2020 stock market crash and rebound
Here’s a close up of the FTSE 100’s CAPE through the crash and recovery of 2020 and 2021:
In March 2020 the FTSE 100 reached its pandemic low of 4,990, some 35% or so below where it had been just a few weeks before.
That wasn’t a pleasant experience, but it left the FTSE 100 with a CAPE ratio of just 11.7. That’s well below its estimated long-term average of 16 and even further below its actual 33-year average of 18.
The implication was that as long as the pandemic didn’t decimate corporate earnings for many years, the intrinsic value of the FTSE 100 (i.e. its long-term future dividend stream) would be relatively unharmed and the FTSE 100 at that level was very attractively priced.
If that was correct then above average returns from March 2020 were likely.
Although it’s only a very short time since the lows of March 2020, above average returns (capital gains of around 40%) are what we’ve had.
But that was then and this is now, so what sort of returns can we expect from the FTSE 100 over the next decade from its current price of 7,100?
CAPE suggests the FTSE 100’s total return over the next ten years will be about average
If the above heading is to prove correct then we need several things to come true by 2031:
- The FTSE 100’s cyclically adjusted earnings need to increase at more or less their 30-year average rate of 2.5% above inflation
- Inflation needs to stay close to the Bank of England’s 2% target
- CAPE needs to be close to its average of 16 in 2031
- Dividends need to be paid at a historically normal level relative to earnings (since total return includes dividends)
I think those are reasonably likely to happen, but we’re in the middle of a global pandemic so of course the degree of uncertainty is higher than usual.
Given those assumptions, we get an estimate for the FTSE 100’s cyclically adjusted earnings in 2031 of 703 (index points).
If that came true then we’d get the following scenarios:
- BUBBLE 2031:
- The FTSE 100’s CAPE is 32 and the index is at 22,500
- BENIGN 2031:
- The FTSE 100’s CAPE is 16 and the index is at 11,300
- DEPRESSION 2031:
- The FTSE 100’s CAPE is 8 and the index is at 5,600
So if the FTSE 100 is at depressed levels in 2031 we would have yet another decade of zero capital gains.
If the FTSE 100 is at reasonable valuations in 2031 I think it’s very likely to break through 10,000.
And if the FTSE 100 is in the middle of a bubble in 2031 (stock market bubbles can happen in the UK too!) then the FTSE 100 may even break through 20,000, as unlikely as that seems today.
Some thoughts on the FTSE 250 and its now record-breaking price level
The FTSE 250 has recovered from the 2020 crash even more strongly than the FTSE 100.
In fact, the FTSE 250 is now at record highs and as I type it’s at 22,650, 2% above its pre-pandemic high.
As with the FTSE 100, the FTSE 250’s price tells us nothing about how attractively valued it may or may not be. For that we need a valuation metric like CAPE.
The FTSE 250’s average CAPE since 1993 is 23.2. As with the FTSE 100 I think that’s probably above the actual longer-term average. The FTSE 250 CAPE has ranged between 13.5 and 31.5 over that period, so the spread of values has been smaller than for the FTSE 100, largely because the dot-com bubble affected the FTSE 250 much less than the FTSE 100.
To err on the side of caution I assume the FTSE 250’s average CAPE is 22. That’s materially higher than the FTSE 100’s average CAPE, but it may be justified by the FTSE 250’s consistently higher earnings growth.
Whereas the FTSE 100 has grown its cyclically adjusted earnings by 2.5% above inflation on average, the FTSE 250 has grown them by 4.5% above inflation, on average. That extra 2% growth rate makes a big difference over decades, which is why investors are willing to pay more for the FTSE 250.
Today the FTSE 250’s CAPE ratio is 25, around 13% above its estimated long-term average.
This suggests the FTSE 250 is trading slightly above fair value, but in reality 13% above average is nothing, as shown by the chart below:
To get into the red “bubble zone” and beyond, CAPE needs to be at least 75% above average, so being 13% above average is nothing and for practical purposes the FTSE 250 is at fair value according to CAPE.
As for future returns, given that the FTSE 250’s CAPE is close to average those future returns are also likely to be close to average.
Here are some assumptions for estimating where the FTSE 250 could be in ten years:
- FTSE 250 cyclically adjusted earnings grow by their historic rate of 4.5% above inflation
- Inflation stays at the Bank of England’s 2% target
- The FTSE 250 CAPE is at the historically average level of 22 in 2031
Given those assumptions we would have FTSE 250 cyclically adjusted earnings of 1,700 index points and the following related scenarios:
- BUBBLE 2031:
- The FTSE 250’s CAPE is 44 and the index is at 74,600
- BENIGN 2031:
- The FTSE 250’s CAPE is 22 and the index is at 37,300
- DEPRESSION 2031:
- The FTSE 250’s CAPE is 11 and the index is at 18,700
As with the FTSE 100, a depression in 2031 would leave us with zero capital gains for the FTSE 250 over the next ten years (actually a slight loss).
More positively, healthy capital gains seem to be a reasonable central estimate, with 40,000 a not overly optimistic possibility under benign conditions.
Finally, if the UK market is as bubbly in 2031 as the US market is today, we could see the FTSE 250 exceed 70,000 in the early part of the next decade.
And speaking of the S&P 500, I hope to do a review of that index and how a variety of factors have pushed its CAPE ratio close to and potentially above the levels seen in the dot-com bubble.
Closing thoughts: The pandemic could be a spanner in the works
I haven’t mentioned the pandemic in these valuations and that’s because the outcome of the pandemic and its long-term economic impacts are impossible to even guess at, at least for now.
If we see no extremely bad variants and if vaccines get rolled out across the globe faster and cheaper over the next year or two then perhaps as a species we and our economy may have got off lightly.
But if new variants appear that are significantly more transmissible, more resistant to vaccines or more deadly, then all bets are off.
So for now I’m ignoring the economic impact of the pandemic. But if there are any obvious lasting impacts which are material at the index level, then of course I’ll incorporate them into future market valuations if it makes sense to do so.