In this article I calculate the cyclically adjusted PE ratio (CAPE) for the FTSE 100, compare the current value to its historic average and forecast the index’s expected value in 2030.
Updating the FTSE 100’s CAPE ratio for 2020
Robert Shiller’s CAPE ratio is one of the best long-term valuation ratios we have. It’s much better than the standard PE ratio because the standard PE ratio compares price to earnings over a single year, and earnings over a single year can be volatile and highly misleading.
To get around that, CAPE uses ten-year average inflation-adjusted earnings, which strips out most of the volatility that comes with annual earnings. The result is a ratio which is much better at comparing current price to future earnings potential, and in the end, that’s what really matters.
I calculate CAPE using a yearly snapshot of the FTSE 100’s earnings, so with 2020 underway it’s about time I updated my CAPE calculations. Here’s how I do it:
- Go to the FT World Markets at a Glance report (PDF)
- Note the index’s price and PE
- Calculate CAPE
Unfortunately, the World Markets report is only available for today and not for historic dates, so you have to download it on the day you want your snapshot.
I downloaded the FT World Markets report on Jan 1st and this is what it said:
- FTSE 100 price = 7542
- FTSE 100 PE = 16.45
From there it’s easy to calculate the FTSE 100’s earnings for 2019:
FTSE 100 2019 earnings = 7542 / 16.45 = 458.5 index points
The next thing we need is CPI inflation for 2019, which we can get from the Office for National Statistics website.
According to the ONS, CPI inflation in 2019 was 1.7%. Inflation is currently indexed to 100 points in 2015, and with that information, we can calculate the FTSE 100’s ten-year inflation-adjusted earnings:
| Year | Nominal earnings | CPI index | Real earnings |
| 2010 | 486.0 | 90.1 | 581.7 |
| 2011 | 553.9 | 93.5 | 638.7 |
| 2012 | 512.0 | 95.9 | 575.4 |
| 2013 | 489.4 | 98.1 | 537.7 |
| 2014 | 432.8 | 99.6 | 468.5 |
| 2015 | 363.3 | 100.0 | 391.7 |
| 2016 | 211.9 | 101.0 | 226.2 |
| 2017 | 346.2 | 103.6 | 360.1 |
| 2018 | 600.7 | 106.0 | 610.9 |
| 2019 | 458.5 | 107.8 | 458.5 |
Table 1. FTSE 100 inflation-adjusted earnings
With these updated inflation-adjusted earnings we can calculate CAPE:
ten-year average inflation adjusted earnings = 485 index points
FTSE 100 CAPE = 7542 / 485 = 15.6
So the CAPE ratio at the start of the year was 15.6, but this doesn’t really tell us very much unless we compare it to CAPE’s long-term average.
If CAPE is currently below average, then it’s probably cheap, although how cheap depends on how far it is below average. If it’s above average then it’s expensive, again depending on how far above average it is. And if it’s close to average then it’s close to fair value. That’s the theory anyway.
Is the FTSE 100 expensive, cheap or somewhere in between?
Over the last 30 or so years, the FTSE 100’s CAPE ratio has averaged 18.4. The current value of 15.6 is obviously below that, so the FTSE 100 is cheap, right?
Not so fast.
The last 30 years include the dot-com bubble, which was the largest stock market bubble in history. The FTSE 100 CAPE ratio exceeded 30 at the peak of that bubble, and this materially increases the CAPE ratio’s long-term average over that period.
Looking at CAPE data for multiple indices globally over many decades, I think a ratio of 16 is, for now, a better estimate of the true long-term average.
Note: I couldn’t find the exact report which I used to come to that conclusion several years ago, but it was written by Meb Faber of Cambria Funds, a known CAPE expert, and here’s his collection of links covering everything you need to know about the CAPE ratio.
15.6 is only very slightly below that estimated long-term average of 16, so CAPE is telling us that the FTSE 100 is very slightly cheap relative to historic norms.
This is not exactly earth-shattering news and I said basically the same thing at the tail end of 2017 when the FTSE 100 was at 7,400.
However, I still think it’s important to let investors know that the FTSE 100 is in no way expensive by any normal measure. Yes, the index’s price is at near-record levels, but it’s also more or less where it was in 1999, more than 20 years ago.
So while 7,000 was bubble territory 20 years ago, thanks to inflation and real earnings growth it’s pretty close to fair value today.
And if you like to use yield as your measuring stick, then on Jan 1st 2020 the FTSE 100’s dividend yield was 4.4%, which is comfortably north of the 30-year average of 3.3%.
As with CAPE though, the average dividend yield may have been affected by the dot-com bubble’s sub-2% yield. We can fix that by looking at dividend yield since 2003 (the trough of the dot-com crash), where the average turns out to be 3.6%.
A dividend yield of 4.4% for the FTSE 100 is below the average yield of 3.6%, so on a dividend yield basis, the FTSE 100 also looks somewhat cheap.
In summary, both CAPE and the dividend yield suggest the FTSE 100 at 7,500 is probably below fair value, although not by a huge amount.
FTSE 100 CAPE valuation rainbow
I should probably explain the rainbow chart from the top of this article. Here it is again:
The chart is an attempt to show a range of reasonable values that CAPE could have had at any point in the past. It works like this:
The long-term average CAPE is 16 if you ignore the dot-com bubble, and history suggests that CAPE spends almost all of its time within a range from about half to double that average. In other words, a range from 8 to 32.
The chart shows the range of FTSE 100 values that a particular CAPE range would have produced, with red showing where the FTSE 100 would be if CAPE was 32 (i.e. very expensive) and green showing where the FTSE 100 would have been if CAPE was 8 (i.e. very cheap). The FTSE 100’s actual value is overlaid in black.
I find this a useful visualisation tool.
For example, in 1999 when CAPE was above 30, almost all the reasonable CAPE values lay below 30, so it was very likely (although not inevitable) that CAPE would fall and that the FTSE 100 would fall with it. And that’s exactly what happened.
In contrast, at the start of 2009, most of the likely range of values lay above the FTSE 100’s actual CAPE value at the time of 12, so the likely (but not guaranteed) direction of travel for CAPE and the FTSE 100 was upward. And again, that’s exactly what happened.
Today the FTSE 100 sits in the middle of that reasonable range, so the odds are about 50/50 that CAPE will increase or decrease in the years ahead (or it might even stay close to fair value).
So that’s where we are today, but what about the future?
10-Year FTSE 100 forecast
Stock market forecasts are of course a bad idea because they almost never come true.
Even so, I still think a sensible forecast can tell us something about the outcome we should expect, even if the actual outcome is unknowable.
For this forecast I’ll make three reasonable assumptions:
- Over the next ten years, inflation stays close to the post-2000 average of 2%
- Over the next ten years, real average earnings growth (where “real” means growth above inflation) stays close to its post-2000 average of 2.4%
- Dividends grow at the same rate as average earnings
- Ten years from now in 2030 the FTSE 100’s CAPE ratio will be at its long-term average of 16
Under those assumptions, by 2030 we’ll have:
- CPI inflation index at 131.6
- FTSE 100 cyclically adjusted earnings of 743
- FTSE 100 dividends of 505 index points
- FTSE 100 price of 11,884
- FTSE 100 dividend yield of 4.2%
So under a reasonable set of assumptions, I think we can reasonably expect to see the FTSE 100 reach about 12,000 by the end of this decade.
And the annualised return from this scenario, including capital gains and dividend income, would be 9.4% per year.
That’s slightly above the long-term average, partly because of a small tailwind from CAPE’s upward mean reversion to 16, but mostly because of the index’s above-average dividend yield.
One problem with this forecast is that it gives very precise figures, whereas in reality, we have no way of knowing what level the FTSE 100 will be at next week, let alone in 2030.
In reality, CAPE could be anywhere within a reasonable range of values from about 8 to 32. The extreme ends of that range suggest we could see:
- The FTSE 100 fall as low as 5,950 in 2030 (this would be a depression valuation)
- The FTSE 100 climb as high as 23,800 in 2030 (this would be a bubble valuation)
Both these outcomes are extremely unlikely, but not impossible, and which one is better depends on your point of view.
Retirees about to buy an annuity will want the bubble valuation to maximise their selling price, while new investors just starting out will want the depression valuation to minimise their purchase price.
Personally, I don’t like extremes so I’ll be hoping for something in the middle, and 12,000 in 2030 has a nice ring to it.
And in case you’re interested, here’s my FTSE 250 valuation and forecast.
“The implication is that expected returns over the next five or ten years are likely to be about average, i.e. 7% annualised, or perhaps slightly above”.
Where does 7% come from please?
Hi Jonathan, thanks for pointing that out. On second reading I didn’t like that bit so I’ve removed it because there’s a more detailed “forecast” later in the article.
I also updated the “forecast” to include dividend income, which is important given the FTSE 100 high yield. The annualised return from the “forecast” is 9.4%, which is at the top end of the high single digit long-term average.
If you want to know more about long-term stock market returns, read Triumph of the Optimists and/or the related Credit Suisse Global Investment Returns Yearbook. Here’s a summary of that report:
https://www.credit-suisse.com/media/assets/corporate/docs/about-us/media/media-release/2018/02/giry-summary-2018.pdf
It states that since 1900, UK equity real total returns were 5.5% annualised. So if we assume that applies in the future and that inflation stays at 2%, then we can expect nominal total returns of 7.5% over the long-term.
The 9.4% from my “forecast” is above that because of a slight tailwind from CAPE mean reversion and from the index’s above average 4.4% yield.
John
You have probably been asked before but does the same data exist to enable the same analysis on the S&P500/
Hi Stephen, Professor Robert Shiller kindly makes a lot of S&P 500 data available on his webpage:
http://www.econ.yale.edu/~shiller/data.htm
The direct link to his spreadsheet is:
http://www.econ.yale.edu/~shiller/data/ie_data.xls
A very quick summary is that the S&P 500’s CAPE is currently 31. It’s 100yr average is 17.5, giving an expected range of about 9 to 35.
So it’s at the top end of that range, implying that valuation mean reversion is likely to be a significant headwind in the coming decade.
Just out of interest, the actual range over the last 100yrs is 5 to 44, so at the very extremes (great depression and dot-com bubble) valuations can go beyond even half or double the long-term average.
Looking at the CAPE ratio in isolation does not make sense. Ultimately you are ignoring 2 key variables here :
– A PE or CAPE ratio is an earnings yield measured over different time periods. To judge whether stocks are overvalued, you have to compare it to prevailing interest rates. Comparing it to averages calculated in an era where risk free rates were much higher tells you nothing. Currently in all markets in the world, earnings yield from stocks is significantly higher than risk free bond rates. All you can conclude from that is that stocks will do better than bonds long term from here and the actual returns will depend significantly on the prevailing interest rates at the end point of the calculation period.
– You also have to probe the source of FTSE 100 returns, not just look at PE or CAPE values i.e ask why it is cheap and is this likely to change. Basically the FTSE100 is stuffed full of zero growth sectors with very low returns on invested capital and it “should” be cheap . The quality of the businesses has been the main reason for the very low returns on the FTSE 100 for the last 15 years in comparison to say the S&P 500 especially when you consider that it is measured in sterling which has depreciated significantly since then.
Hi Lemsip, those are reasonable arguments, but I tend disagree.
Equity valuations and inflation are loosely correlated, but they are correlated. So interest rates matter. But, CAPE does mean revert at least partly because interest rates mean revert. Interest rates don’t go off to 1000% or 5m%, they tend to be somewhere between zero and 20% or so, most of the time.
So, given the historical fact that CAPE has mean reverted in most countries over the last century or more, it’s reasonable to assume that it will continue to do so in the future, with the underlying assumption that interest rates will mean revert to, at some point. Perhaps they will and perhaps they won’t, but it’s still a reasonable assumption.
As for the quality of the FTSE 100, I would disagree that it’s full of low quality companies. Since 2000, the FTSE 100’s earnings have increased by about 4.5% per year, which isn’t exactly great but it isn’t zero either. The FTSE 250 has been faster growing, but the FTSE 100 hasn’t been a disaster, and the FTSE 100’s dividend has increased 140% since 1999 too.
The low returns of the last 20 years has little or nothing to do with the quality of the companies. It is almost entirely due to the high valuations in 1999, when CAPE was above 30 and everybody seemed to think the stock market could return 20% per year forever. But of course it didn’t, and CAPE mean reverted, which is exactly what I would expect it to do.
Having said that, it is important to remember that the future is unknowable and that CAPE is not a precision instrument.
All it does is provide a dim light into a very opaque future, but a dim light is still better than no light at all.
For other readers who are not familiar with CAPE, here’s one report on its predictive capabilities, limitations and possible uses:
The CAPE Ratio and Future Returns: A Note on Market Timing (pdf)
some of your statements above are easily disproved by data.
Since Feb 2005 ( i.e well after the dotcom crash) –
The FTSE 100 with dividends reinvested is up 150%
The FTSE All World ex UK is up 350% – i.e more than twice in the same time period. So the performance difference has nothing to do with 1999 overvaluation but points to a fundamental reason such as the quality of business results.
( Source – the trustnet charting tool)
Regarding interest rates mean-reverting as a statement of fact. Interest rates have been steadily falling for the last 300 years all around the world apart from an inflationary episode in the 70s and part of the 80s. There is no law of nature that determines an inevitability of them returning to previous levels in the timeframe that is of relevance to an investor. You can check this from multiple sources. You can argue that interests will be in a range of 0-20% but the actual values in that range are of great consequence to an investor.
Similarly CAPE values alone tell you nothing about the returns the businesses within a narrow index like the FTSE100.
To say that “low interest rates justify high valuations in stocks” is also to say “low interest rates justify low future returns in stocks.” If one wishes to protect overvalued prices, one also has to accept meager long-term returns. – Hussman
John, the CAPE and any PE ratio comparison between markets, would need to take into consideration the dividend differential.
Let’s say FTSE 100 companies pays 4% out of 5% earnings in dividend and S&P 2%. out of 6% earnings. So S&P 500 retain 3% more earnings per share.
If we use a Price per book multiplier of 4, which is something that S&P500 shows at this moment, then you need to add 12% to your 16% CAPE for FTSE 100.
So about 28% for US market is OK. 31% is not much higher, and reflects that earnings of US companies increased at a higher rate than earnings of UK companies.
Hi Eugen, I’m not sure I follow your argument correctly, but there has been research done on differing dividend payout ratios over time and they don’t really make much difference to CAPE values or its long-term average. E.g.:
http://www.philosophicaleconomics.com/2015/03/payout/ (very long article but goes into lots of detail on total return CAPE and how it doesn’t make much difference).
My position on all this is that CAPE mean reversion is real, and is likely to persist in the future.
However, it’s important to remember that although CAPE is a better indicator of future returns than most other indicators, it’s still very weak unless it’s at extremely high or low levels.
I read the article you sent me before. However, I do not think it is written well.
I am also not that keen on how much the price is relative to past earnings. I tend to be more face looking investor, so for me it is most important to estimate future free cashflows correctly, the return on capital and barriers to entry.
As Charlie Munger said, if you buy a lousy company earning 6% per annum on capital, even if you buy it a discount, you end up with more or less 6% per annum return. However if the company has 18% per annum return to capita, you end up with a hell of a result!
I never did like that Charlie Munger quote because it gets taken out of context so much.
Munger is talking about VERY long-term returns approximating return on capital, but over shorter periods, say anything up to 10 years, changes to valuation multiples (i.e. buying at a discount or overpaying) have more impact than ROCE in many and perhaps the majority of stocks.
Interesting and thought provoking article. My challenge comes in your assumptions for future returns. A 2.4% real growth rate seems fanciful to me. Looking at some of the top constituents by market cap:
HSBC – annual revenues have shrunk from $77bn in 2012 to $55bn in 2019
AstraZeneca – revenues peaked in 2011 at $33bn, now at $24bn
BP – $400bn revenue in 2014, now $287bn
BHP – $71bn (2011) ->$ 44bn
GlaxoSmithKline – $38bn today compared to a $46bn peak in 2008
Royal Dutch Shell (x2) – $493bn (2012) -> $371bn
British American Tobacco – an exception that is consistently growing so far (although the rate of growth has recently nosedived, vaping anyone?)
You get the idea – this is with the backdrop of a global expansion. It’s hard to imagine these companies would perform better through a recessionary environment.
You’ve also forecast dividends to grow on these companies? Last dividend increases were:
HSBC – 2016
AstraZeneca – 2011 (reduced in 2012)
BP – 2019
BHP – 2019 (but very severely cut during the 2016 commodity price sell off)
GlaxoSmithKline – 2014
Royal Dutch Shell (x2) – 2019 but still lower than 2015-2017 levels
British American Tobacco – 2020, a rare success story but is tobacco the real engine of growth for the future?!
You can even see this playing out in your CAPE rainbow chart. What was a consistent rising trend has flattened off and even begun to decline.
Hi BV, you make a very valid point. There are a lot of question marks about the quality or growth prospects of many large FTSE 100 constituents. And, perhaps you’re right that the historic real growth rate will not be achievable going forwards.
However, perhaps these companies will turn things around over the next ten or twenty years, or perhaps other smaller FTSE 100 constituents will grow enough to offset the lack of growth from these larger companies. Both scenarios are entirely possible.
For me the important points are 1) that the future is very uncertain and you just have to accept it, 2) CAPE has been reasonably informative in the past about future returns, especially from extreme valuation levels, and 3) that it probably will be in the future if it’s used with a significant pinch of sale.
You’re probably correct, John. The poster is making the mistake of conflating individual stocks with an index. One of the marvellous things about an index is that it adjusts itself to take account of individual stock underperformance. Witness M&S recently. Either way you look at it, the FTSE isn’t as bad as the American market, where the top 5 companies account for something like 18% of the entire market, and include a glorified ‘yellow pages’, a grocer, and a seller of ad space in self-published scrapbooks. All of whom have an imminent date with the monopolies regulators, as far as I can see.
Hi IWN, that tends to be my stance too. The whole point of an index is to capture broad economic progress rather than worrying about individual companies. CAPE and my valuation rainbows are based on the same idea. Look at the past, extrapolate that into the future with a pinch of salt, and forget about the ups and downs of individual index constituents.
As for the S&P 500, I like your characterisation of everybody’s favourite FANG stocks, but I have no idea what companies will be in the index in ten or 20 years so I don’t worry too much about what companies are in it today. I just go by CAPE, and CAPE says it’s pretty expensive and therefore has expected returns which are lower than the long-run average.
Agree. The American markets are expensive by pretty much ANY metric (price/book, p/e, cape, capitalisation to GDP, you name it, it’s all in the top 5% or worse historically). So anyone buying US stuff right now really shouldn’t be surprised if their returns over the next 10 years aren’t exactly stellar. Obviously, there’s no way to absolutely predict the future, but what we CAN do is make educated guesses, that have a more than random likelihood of happening. Strange the FTSE 100 is currently such good value – the rest of the world seems to have lost faith in Brits!
Hi John, thanks I enjoyed this article. Do you know where country by country long term average CAPE figures can be found? The US (ie S&P500) and UK figures are in a few places but I’m struggling to find this information for other countries. Thanks in advance for your help
Hi Stephen, StarCapital have a pretty nice international CAPE resource. It doesn’t seem to have historic values but it does list and rank multiple markets on CAPE, PE, PB etc.
https://www.starcapital.de/en/research/stock-market-valuation/
Excellent article and thoughtful follow-up discussion.
Thanks to all.
OK, 9th March. FTSE 100 is at the ‘depression 2030’ scenario level. In my book, that makes it a buy. Yielding 6% plus, Brexit in rear view mirror, what’s not to like?
As I type the FTSE 100 is hovering around 6,000, giving it a CAPE ratio of 12.4.
For context, my “slightly cheap” level is 14 and “cheap” is 12, so yes, the large-cap index is definitely at the sort of low valuation levels you don’t see very often.
And the dividend yield is now well over 5%, which is not to be sniffed at.
Obviously we have no ideas what will happen in the future, but that’s always true. Coronavirus hasn’t made the future more uncertain, it’s just made people more aware of that uncertainty.
My best guess today is that well see a bounce (after a 6%+ fall, easy one!) then more downward pressure. Shame the FTSE 250 hasn’t fallen quite so dramatically – as your articles show, over time, it grows faster than the 100.
Hi John. Does the current subdued price of the FTSE suggest your predictions need revising? Or are your end targets the same, but there’s just more profit along the way?
The latter. The earnings side of the CAPE ratio won’t be updated until next year, so for now the assumption is that a lower price today will produce higher expected profits tomorrow. And that has historically been the case. See 2003 or 2009 for good examples.
Wild times, my friend!
Thank you so much for an outstanding article.
Do you know where an up to date CAPE value for the FTSE is published.
Many thanks
Have a look at this page:
https://www.starcapital.de/en/research/stock-market-valuation/
It covers a range of countries and is updated monthly.
Hey John – would you fancy doing a quick update on the FTSE 100? It’s been way below the price in this article (although it’s bounced back a ways since then). It might be interesting to highlight whether it’s now a ‘no brainer’ investment or a ‘punt’… BTW, your previous FTSE article looks almost prescient in hindsight – bl**dy well done!
Thanks, IIN
I’ll probably do a mid-year market valuation update. Would be nice to revisit valuations at the market low from May, i.e. sub-5,000 for the FTSE 100. How have returns been since the index was at that attractively valued low point?
As for prescience, thanks but I’m not so sure. Looks more like luck to me…
Understood. Returns have been pretty good, to be honest. I picked some 100 up at sub 5000 (even though I had a full allocation to the 100, it seemed too good to miss). The 250 has been fun too – looking at it, a quick back of the envelope calc seems to suggest annualised returns of 13% / 11% inflation adjusted or so, which seems insanely high to me so I’ve probably gone wrong somewhere in the equation. BTW, I guess you’ve already seen the good old ‘1/CAPE’ trick; so if the CAPE right now is about 17, 1/17 = 5.9% after inflation returns. Don’t know how reliable it is – I guess it takes no account of earnings growth…
Hey John. Quick Q – do you have FTSE earnings for data earlier than 2010?
Thanx
IIN
Yes, but the quality of the data gets worse the further back you go, although it’s still good enough for a broad-stroke metric like CAPE.
Here’s a Google Sheet with data for the FTSE 100:
FTSE 100 CAPE data
Older earnings data are extracted from books, charts, websites etc, whereas newer data comes from the FT website.
Ta John – great stuff.
Lots of useful info in there. Did you create it?
IIN
Yes, cobbled together over a long period of time from a variety of data sources (sadly there’s no Robert Shiller in the UK).