FTSE 100 CAPE Valuation and Forecast for 2023

2022 was a year that many investors will want to forget.

Bitcoin fell by 65%, the FTSE AIM index fell 30% and even the mighty S&P 500 was down by 20%.

The FTSE 100, on the other hand, went up by 1% and, in the process, beat just about every other major stock market index on the planet.

In fact, I'm writing this in the middle of January 2023, and the FTSE 100 is now less than 1% away from setting a new all-time high.

For some investors the FTSE 100's outperformance will come as something of a shock, given the index's historical underperformance, so in this post I'm going to look at why the FTSE 100 performed so well last year and how that outperformance has affected its valuation and potential future returns.

Table of Contents

• Why the FTSE 100 crushed other indices in 2023
• Where does this outperformance leave the FTSE 100 in valuation terms?
• My FTSE 100 forecast for 2023
• My FTSE 100 forecast for the next ten years

Why the FTSE 100 crushed other indices in 2023

Here's a chart showing the FTSE 100 (in black) over the last year compared to the S&P 500 (in blue).

The FTSE 100 went from 7,385 at the start of the year to 7,452 at the end, for a capital gain of 1%. As I've already said, the S&P 500 fell about 20%, so this is a very significant degree of outperformance for the large-cap UK index.

If you look at the FTSE 100's sector composition then some of this outperformance begins to make sense.

Three of the FTSE 100's largest sectors are healthcare, energy and natural resources, which together make up more than a third of the index's value. Healthcare, energy and natural resource stocks generally did very well in 2022 because healthcare has become a post-pandemic priority and because supply chain issues and war have driven commodity prices sky high. As a result, the FTSE 100 also did very well.

For example, three of the FTSE 100's largest holdings are AstraZeneca (healthcare), Shell (energy) and Rio Tinto (natural resources). These companies currently make up 8.2%, 9.2% and 3% of the index respectively, for a total of 20.4%. That's 20% of the index in just three companies, which I suspect is more concentrated than most investors realise.

In 2022, AstraZeneca's share price went up 30%, Shell's went up 43% and Rio Tinto's went up 19%. Those are some truly exceptional results that few people expected at the beginning of the year.

So, here's the main reason why the FTSE 100 demolished just about every other major stock market index in 2022: It got lucky, by being in the right sectors at the right time.

One downside of this crushingly superior performance is that the FTSE 100's valuation hasn't fallen as much as the S&P 500's or the FTSE 250's. Given that fact, a sensible follow-on question is:

Where does this outperformance leave the FTSE 100 in valuation terms?

Most people use the price/earnings ratio (PE ratio) to value stocks and indices. This is generally a mistake. When we value a company or an index, we should be comparing its price against its intrinsic value, i.e. what the price should be if investors were economically rational (which we aren't). But last year's earnings (which is what we use in the PE ratio) are a terrible proxy for intrinsic value.

For example, if Unilever has a bad year and its earnings fall to zero, does that mean its intrinsic value also falls to zero? Of course it doesn't. If Unilever earns nothing in 2023 then it is still very likely to make a profit in 2024 and probably for many years after that, so valuing Unilever or the FTSE 100 using a single year's earnings is a very bad idea. 

A much better way to value companies and indices is to compare their current price to their average earnings over a number of years. The most popular of these price-to-average-earnings ratios is Robert Shiller's Cyclically Adjusted PE or CAPE ratio, which averages earnings over ten years and adjusts them for inflation.

For example, at the end of 2022 the FTSE 100's price was 7,452. It had a PE ratio of 17, so it's 2022 earnings came to 7452 / 17 = 438. That's 438 index points, not millions or billions of pounds.

If we do the same calculation for the last ten years and adjust for inflation, we get a cyclically adjusted earnings figure of 490. If we divide the FTSE 100's current price of 7,452 by 490, we end up with a cyclically adjusted PE ratio of 15.2.

Unfortunately, that doesn't really tell us anything. What we need to know is how that 15.2 figure compares to the FTSE 100's long-term average CAPE ratio. That will tell us whether 15.2 is above average (expensive), below average (cheap) or close to the average (fair).

I have FTSE 100 earnings data going back to 1987 and during that time, CAPE averaged 17.7. 15.2 is below 17.7, so according to the CAPE ratio, the FTSE 100's current valuation is slightly below average. In other words, according to the CAPE ratio, the FTSE 100 is cheap, despite massively outperforming the rest of the world in 2022.

However, to err on the side of caution, I usually assume that the average CAPE is 16 rather than 17.7. The reason is that I only have 35 years of data for the FTSE 100, and that period included the dot-com bubble, where stock valuations went to the moon. That bubble skews the mathematical average CAPE to a level that is, in my opinion, probably above the "true" longer-term average.

In practice this doesn't really make a lot of difference because 15.2 is below 17.7 and 16, so either way the FTSE 100 appears to be cheap.

But how cheap is it? After all, 15.2 is only 5% below 16, so that suggests it's only 5% below "fair value". Is that as good as it gets, or could it get a lot cheaper (or a lot more expensive)?

To put all of this into context, we need to look at the FTSE 100's historical track record to see how its price and CAPE ratio have varied over time.

A picture is worth a thousand words, so here's a chart showing the FTSE 100's price and CAPE ratio since 1987. 

If you haven't seen this type of chart before then the FTSE 100 is the black line and the rainbow tells you what CAPE was at the time and where it was in relation to its long-term average.

• Increasingly Red means increasingly expensive (CAPE rising to double its average)
• Yellow means the price is fair (CAPE is close to its average)
• Increasingly Green means increasingly cheap (CAPE falling to half its average)

The rainbow extends from half the average CAPE to double the average because decades of data across multiple indices tells us that is where CAPE spends almost all of its time.

This might make more sense if I replace the CAPE figures with some meaningful descriptions.

For example, in 1999 (the peak of the dot-com bubble), CAPE reached 32, double its long-term average, but it didn't go any higher. At the other extreme, in the 2009 credit crunch (which almost turned into a global depression), the FTSE 100's CAPE fell to 9.0, just slightly above half the average CAPE, but it didn't go any lower.

We can interpret this pattern in the following way:

• Investors are rarely so optimistic as to buy stocks at more than double their average CAPE and they are rarely so pessimistic as to sell stocks at less than half their average CAPE

With the FTSE 100's CAPE currently at 15.2, it isn't anywhere near the red "expensive zone" or the green "cheap zone". Instead, it sits squarely in the yellow "fair value zone", where CAPE falls between 14 and 20. In other words, with CAPE at 15.2, the FTSE 100 is currently very close to fair value.

So, despite beating the S&P 500 by about 20% in 2022, the FTSE 100 isn't even remotely expensive. That's good to know, but it would be better if we could use the FTSE 100's CAPE valuation to estimate the index's potential future returns, so let's do that now.

My FTSE 100 forecast for 2023

As a general rule, you should take all one-year stock market forecasts with a gigantic pinch of salt. That's because the stock market moves around in an almost random manner, so predicting where it's going to be next week, next month or next year is impossible.

On the other hand, we can talk about where the stock market is likely to be in the future, especially if we talk about the range of values it might take.

For example, stock market indices usually have a CAPE ratio that is somewhere between half and double its long-term average. For the FTSE 100, that means CAPE is likely to be within 8 and 32 by the end of 2023, which would put it somewhere between 3,900 and 15,700. The problem is that range is so wide as to be effectively useless; it's like saying the FTSE 100 will end 2023 somewhere between one and one million.

But perhaps we can do better. CAPE is a mean reverting ratio, which means it periodically returns to the long-run average. You can see this in the rainbow chart above, where CAPE went to extreme highs in the dot-com bubble and extreme lows in the credit crunch, but in both cases, it subsequently reverted back to fair value. If you imagine an elastic band connecting CAPE to its long-term average, you'll get the idea. The further CAPE is from its long-term average, the more powerfully it's attracted back to that average. 

Perhaps the most reliable assumption then, is that CAPE will be fairly close to its long-term average in the future. By fairly close I mean within my "fair value zone" (the yellow part of the CAPE rainbow), where CAPE is between 14 and 20. If we assume that the FTSE 100 is close to fair value at the end of 2023 then we can make the following one-year forecast:

• If the FTSE 100 is close to fair value at the end of 2023 it will be between 6,900 and 9,800

And in terms of forecast gains and losses:

• If the FTSE 100 rises to the upper end of fair value (9,800) it will increase by 32%
• If the FTSE 100 falls to the lower end of fair value (6,900) it will decline by 9%

A gain of 32% would be exceptional and unlikely, but a decline of 9% is entirely possible. Of course, the FTSE 100 could end the year outside that range, but the above range is a reasonable estimate of where the FTSE 100 "should" be by the end of 2023.

Let's move on to the longer-term ten-year forecast.

My FTSE 100 forecast for the next ten years

When we look at the longer-term future, we need to remember that earnings grow over time, partly because of inflation and partly because of real economic growth. This is true even over a single year, but the amount is so small and the uncertainties so high that it's barely worth considering in a one-year forecast. However, over ten years the FTSE 100's earnings and therefore its cyclically adjusted earnings are very likely to grow by a fairly large amount, so we need to take that into account.

As I said earlier, the FTSE 100's cyclically adjusted earnings today are equal to 490 index points. Back in 1987 they were at 259 index points and that's an annualised growth rate of 1.5% over those 35 years. Although the future remains uncertain, I think it's reasonable and conservative to assume that a similar growth rate will occur in the future.

That 1.5% growth rate excludes inflation, so we have to add inflation back in to come up with a nominal growth rate. I don't have a crystal ball, so I'm going to assume (perhaps optimistically) that UK inflation over the next ten years is close to the Bank of England's target of 2%.

That gives us an estimated nominal growth rate for the FTSE 100's cyclically adjusted earnings of 3.5% per year to 2032.

If that turns out to be correct (which of course it won't, but hopefully it won't be a million miles off) then the FTSE 100's cyclically adjusted earnings would grow from 490 today to 691 ten years from now.

As with the one-year forecast, it is reasonable to assume that the FTSE 100 will be somewhere in my fair value zone (where CAPE is between 14 and 20), and that gives the following forecast:

• If in 2032 the FTSE 100 is close to fair value it will be between 9,700 and 13,800 

In raw percentage terms, that means:

• If in 2032 the FTSE 100 is at the upper end of fair value (13,800) it will have increased by 85%
• If in 2032 the FTSE 100 is at the lower end of fair value (9,700) it will have increased by 30%

What about annualised total returns?

For that we need to include dividends and at its 2022 year-end price of 7,452, the FTSE 100 had a healthy dividend yield of 3.7%. If we assume that the dividend grows in line with cyclically adjusted earnings (3.5% per year), then we get the following forecasts for annualised total return:

• If in 2032 the FTSE 100 is at the upper end of fair value (13,800) it will have generated total returns of 9.6% per year
• If in 2032 the FTSE 100 rises to the lower end of fair value (9,700) it will have generated total returns of 6.4% per year

The FTSE 100's long-term average total return is about 7% per year, so the above total return estimates are broadly in line with historical norms. That is precisely what I'd expect to see, given that we're starting from a historically average valuation (so mean reversion headwinds and tailwinds aren't a major factor).

Of course, innumerable uncertainties will unfold between now and 2032, but I still think the above is a reasonable estimate of where the FTSE 100 is likely to be ten years from now.