In this post, I’m going to do a valuation and projection for the S&P 500 using Robert Shiller’s CAPE ratio.
I think this is worth doing on a regular basis because it gives you a good idea of where a particular market is in its valuation cycle and what sort of returns it’s reasonable to expect over the next few years.
Valuation: At 2,390 the S&P 500 is very expensive
Okay, so the S&P 500 is very expensive, but what does that actually mean?
It means that at 2,390 points, the S&P 500’s price is 27.2-times its average inflation-adjusted earnings over the last ten years (using annual data).
In other words, its CAPE ratio (an improved version of the standard PE ratio) is 27.2.
Over the last century, the S&P 500’s CAPE ratio has varied between 5 and 44, with an average value of 16.9.
So at (about) 27, the current S&P 500 CAPE ratio is 59% higher than its average of (about) 17, which I think is a lot.
Another way to think about this is to look at how often the S&P 500 has been cheaper (i.e. had a lower CAPE ratio) than it is today.
In fact, the S&P 500 has been cheaper than it is today 93% of the time over the last century.
The only time the index has been more expensive than today was:
- In 1929 at the peak of the roaring ’20s, just months before the Great Depression
- In the dot-com bubble between 1996 and 2002, before declining almost 50% in the subsequent bear market
- In 2007 at the peak of the credit bubble, just before a massive 55% decline as the credit crunch exploded
If you’re interested in how the S&P 500’s CAPE has varied relative to its historic average, below is my ever-popular “rainbow chart” which shows exactly that:
This is what the chart shows:
- Black line: This is the S&P’s price
- Red band: This is where the S&P 500 would have been if it was about twice as expensive as its historic average (i.e. CAPE close to 34 rather than 17)
- Yellow band: This is where the S&P 500 would have been if was about as expensive as its historic average (i.e. CAPE close to 17)
- Green band: This is where the S&P 500 would have been if it was about half as expensive as its historic average (i.e. CAPE close to 9.5 rather than 17)
CAPE ratios higher than double or less than half the historic average are classed as bubbles and depressions respectively.
As you can see, the S&P 500 is currently on its way towards bubble territory, with a long way to fall if investor sentiment turns negative.
What about the Global Financial Crisis?
Just hold your horses one dang minute, I hear you shout. What about the Global Financial Crisis?
Surely it negatively impacted average earnings over the last decade? That would reduce the “E” part of CAPE which would in turn push the CAPE ratio higher.
So perhaps the S&P 500 isn’t really expensive? Perhaps it’s just an anomaly caused by the way CAPE is calculated combined with extremely low earnings during the financial crisis?
This is a legitimate point, so let’s have a look at a chart of the S&P 500’s earnings over the last decade, using Robert Shiller’s excellent data:
As you can see, the financial crisis had a massive impact on earnings which definitely has had an impact on the CAPE ratio.
To get around this problem we can just pretend that the financial crisis never happened.
To do that I’ll just edit the earnings data so the peak earnings value of 100 in 2007 is carried right through to 2013, which is when earnings finally moved back above the 100 level.
Here’s what the earnings would have looked like in this crisis-free world:
What difference does this implausibly optimistic view of the past make to my very negative view of the S&P 500?
The answer is, not a lot.
The S&P 500’s 100-year CAPE average declines from 16.9 to 16.6, hardly an earth-shattering difference.
Somewhat more impacted is its current CAPE value, which falls from 27 to 24.
If we use this more cautious CAPE ratio then the S&P 500 is now 45% more expensive than average rather than the original 59%.
And with CAPE at 24, the S&P 500 has been cheaper than it is today a mere 87% of the time rather than 93% of the time.
Another way to think about this is to think of a 100-sided die. If you rolled an 87, do you think the next roll would be higher or lower? The same idea applies to the stock market.
Okay, so the S&P 500 is expensive with or without the financial crisis.
Let’s take that optimistic CAPE of 24 and use it to make a projection of where the S&P 500 will be in one year’s time.
Projection: The S&P 500 is going down
I used to call these short-term projections forecasts, but that’s wrong because:
- A forecast is a description of a future which you think is actually going to happen, or which you think is most likely to happen.
- A projection is a description of a future based on a set of assumptions, which may or may not be likely to happen. In other words, it’s a “what if” analysis.
I have absolutely no idea what the most likely short-term outcome is for the economy or the stock market, so I don’t make forecasts.
However, anybody can make assumptions about the future, so anybody can make projections.
What I’m after is a projection which gives a good indicator of the headwinds or tailwinds the market’s current valuation is creating.
So here are my assumptions, which I think are reasonable but which may or may not actually happen:
- Earnings will grow: The S&P 500’s cyclically adjusted earnings will increase by 5% in the next year, which is historically average
- The market will mean revert: The S&P 500’s CAPE ratio will close the gap between its current value (24) and its average value (16.6) by a half over the next year
Now we can make some projections using those assumptions:
- Earnings will grow: The S&P 500’s cyclically adjusted earnings will grow by 5% from 99 points in 2017 to 104 points in 2018
- The market will mean revert: CAPE will close the gap to its average value by half, moving from 24 to 20.3
With 2018 cyclically adjusted earnings at 104 points and the CAPE ratio at 20.3 we can now make a projection of the S&P 500’s value for May 2018:
- S&P 500 projection for May 2018 = 2,110
That’s a projected decline of 12%, which is not insignificant.
Remember, this is not a forecast of what will happen, but rather an indicator of the valuation headwinds the S&P 500 is currently facing. If you want to forecast the future by working out what’s going to happen with Trump, China, North Korea and the combined sentiment of millions of investors then good luck!
As a final point, I’ll mention the index’s projected “fair value” price, which is the price it would have a year from now if its CAPE ratio fell back to the historic average of 16.6 (again, using the optimistic assumption that the financial crisis never happened).
This simply involves multiplying the 2018 cyclically adjusted earnings of 104 by the fair value CAPE of 16.6, giving:
- S&P 500 fair value projection for May 2018 = 1,730
That’s a decline of 28% to fair value, and while such a decline isn’t guaranteed to happen, it is a very plausible scenario.
So of course I’m bearish about the S&P 500, but that doesn’t mean I wouldn’t invest in it. It just means I would be cautious and mentally prepared for the next inevitable bear market.
Another very interesting article and really like that CAPE diagram. Everybody tells me its impossible to time the market but I cant help but be doubtful about investing a lump sump of money when the trend is so high above the mean. Just wondering what your thoughts would be about how best to invest a lump sum 7 years into the existing bull market and given the data from your analysis?
One idea could be to invest in a fund with a high proportion of bonds to equities – i.e.: 40/60 or 30/70 equities bonds and adjust up the equities portion as time goes on and valuations become cheaper and thus reduce the risk of a sequence of low returns. What criteria would you use for your portfolio of income equities if the UK market was showing a similar trend to the SP500?
Would appreciate your thoughts.
Thanks
Hi Barry, thanks.
“how best to invest a lump sum 7 years into the existing bull market” – Sorry Barry but I can’t really get into details with this one as I’m not a financial advisor. However, I can talk more generally.
In terms of your idea of tilting a portfolio towards bonds rather than stocks, I think that’s a valid approach, but I don’t think there’s much evidence about good ways to do it. Years ago I dreamed up a mechanical countercyclical CAPE-based asset allocation system which would move a portfolio into cash as the market became more expensive and into stocks as it became cheaper. It worked well in back-testing with market-equalling returns and much lower risk, but I never really used it in practice.
If I was going to invest in the US today then my approach would depend on whether I was investing actively or passively.
As a passive investor I would probably just invest 50/50 in global stocks/bonds, rebalance once each year and not worry about market valuations.
As an active investor I would do exactly what I do in the UK, which is to look for good companies at good prices regardless of the market’s valuation. Even if the market is expensive I would expect there to be some good quality but out of favour companies, and more than enough to build a portfolio from.
Although the market’s valuation wouldn’t affect my investment approach, it is still very useful to know because it tells you a lot about what the future may hold, which is useful from a psychological point of view. Both in terms of not getting too excited during booms and not panic selling during busts.
John
Interesting set of statistics.
CAPE aside John, and Robert is a bit of a smug character in himself, there is always a bull element amongst the bears. A lot of the lift in the CAPE seems sector driven, with oil and commodity stocks rebounding rapidly in the last 12 months and banks more recently (I don’t like banks).. Some areas, or more specifically some stocks are either just undervalued, or exposed to growth areas that are to some extent due to structural change (such as growing online payments or the adoption of cloud services over in-house computing systems). These are just two examples and I guess there are more which sets them aside from the expense of the market index..
To compensate for CAPE at these levels I guess you have to be more selective and that’s exactly what you do best John. For the US market, I think this selection is capable of outperfoming still in a threatened overall market :-
Microsoft – P/E of 20 — is growing cloud at a significant rate (P/E was 59 at the peak of the early 2000)
Apple – P/E of 13 — not quite the bargain it was at $89 but still has some significant potential upside
Disney – P/E 17 — growing EPS at a great rate across films/online and studios,could be bought by Apple
PayPal — P/E 22 — just bust it’s numbers again – hit almost 100 billion transactions – still in it’s infancy
Novo Nordisc – P/E 16 — In a world of growing diabetics this is a giant in it’s field and undervalued
Skyworks – cash rich ($1.35Bn) / no debt – has the most complex analogue components in smart phones and the connected world (Internet of things). With the switch over to 4G, Skyworks supplies 4X the content in the smart phone compared to 3G — a small goldmine I think.
Oh well – time will tell as they say.
LR — Good look, careful not to get wrapped up in CAPE Fear Mr De Niro – lol !!
Hi LR
I don’t really know much about those specific companies, but you’re right that there will always be pockets of cheapness even in the most expensive market. In the dot-com bubble it was the “old economy” boring stocks like consumer staples! How could anybody want to invest in soap when you could invest in Web Vans?!?
The more things change the more they stay the same, as someone once said.
Not sure what’s cheap in the US, but in the UK it’s UK-focused retailers mostly, plus a few other bits and pieces. So of course that’s where I’ve been investing recently. As you say, time will tell if that’s a good idea or not.
And Cape Fear, very funny. I’ll have to steal that from you when the bubble’s about to pop!
“That’s a decline of 28% to fair value,”
What’s the record trough? Down to 50% of fair value? 20%? …. lower?
The biggest peak-to-trough decline ever was in the great depression. Down 84% between September 1929 and July 1932, which of course was much worse because lots of people invested with margin (i.e. borrowed to invest) back then, so they lost much more than everything, including the shirt on their back.
Then in the early 70s there was a 43% decline, although high inflation made a 57% decline in real terms. In real terms the peak of this 70s stagflation decline was in 1982 with a 63% real decline from the highs of 1968.
Then of course we have the dot-com crash with its 44% decline and the credit crunch managed 51%.
So 50% or thereabouts seems fairly normal for a massive bear market, with only the great depression being much, much worse.
Personally I expect to see another one or two 50% declines in my lifetime, but hopefully no more 84% declines!
Nice piece on S&P 500’s CAPE.
I think we need to consider two other asset class (bonds and property), as well as equity. U.S. property is still cheaper than it was in 2007. (Here is the Case-Schiller Index: http://i.imgur.com/w1OHB1V.png )
The recent rally in the S&P 500 corresponds to the sell-off on bonds. But, U.S. bond prices looks historically low.
Then, there is the foreign money from the Far-East and Middle-East nations seeking assets in the West.
So, should we consider two other giant asset class in property and bonds, before evaluating if the stock market looks overvalue?
I feel if U.S. 10-year bonds hit 3%-3.5%, then U.S. stock market could hit 3,000 on the S&P 500. And then, it would sell off because equity dividends yields are too low.
Hi Walter. I tend to look at stock market valuations in isolation from other asset classes. From the various papers I’ve read over the years it seems that controlling for other factors like interest rates, property prices, bond prices and so on, doesn’t really improve the predictive power of CAPE.
In other words, I think CAPE works quite well as a long-term valuation metric, and doing lots of legwork to calculate other things doesn’t add much value, so I don’t bother.
I’m sure there are some very clever people somewhere that can look at bonds and property and the fed rate and who knows what else, and come to a better prognosis for the market than one derived solely from CAPE. Good luck to them. But I can’t do that sort of analysis with any rigour so I’m sticking with my plain vanilla CAPE analysis.
The difference between the valuation of the US market (CAPE, P/B, etc.) and of other markets is bigger than their economies account for. Long Europe!
Hi Gregory, yes that’s what CAPE implies. You do have to have the stomach to invest in out of favour markets though!
No pain no gain.