More on current CAPE

[MichaelB]

An interesting blog post by Meb Faber on CAPE ratios, cleverly titled "Don't Use the CAPE Ratio". What makes this article different is two things: the focus on other countries, and some good links, including one to Star Capital, which provides detailed CAPE ratio data for other countries.

Don't Use the CAPE Ratio | Meb Faber Research - Stock Market and Investing Blog

 

[Gone4Good]

While the findings are not surprising it is compelling how clearly valuation measures like CAPE predict future returns not just in the U.S. but also in 8 other countries around the world.

 

[eta2020]

But what use is this to me... besides telling me to put new money into VXUS and not into VTI?

If I have lets say 4 000 000 bucks in VTI with 2 500 000 of market gain should I sell it and pay 500-600k dollars in taxes and wait for Bear Market? I think not.

 

[samclem]

Thanks, an interesting read.
- Some analysts make the case that relatively recent changes in the US market and in accounting rules make current US PE10 numbers non-comparable to the historic numbers. If that's the case, that the US historic CAPE can't be compared to the present US CAPE, I wonder if it is reasonable to compare CAPEs across national borders/economies? Seems like the baseline differences could be very significant, posing the same problem.

- There have been several good cases made for US investors to consider reducing equity exposure when US valuations are high, with concomitant higher exposures to bonds (or, typically govt bonds when US equity valuations get especially high). Maybe there's some additional fine-tuning that can be done with foreign equities, too.

For me, I plan to keep my US-Foreign equities ratio fairly static (but rebalancing should still result in purchases of the "cheaper" ones, selling of expensive ones, on a relative basis). I do/will reduce total equity exposure when US stock CAPE is high, increasing allocations when US CAPE is low. But only a bit--more of a "tilt" than a big bet. And history indicates that the changes might be required at about 10-20 year intervals, so hardly day-trading. Still--"market timing."

 

[eta2020]

What if inflation and interest rates will stay low for decade or decades? I can google up many articles expecting this.

 

[MichaelB]

But what use is this to me... besides telling me to put new money into VXUS and not into VTI?

It's not a suggestion that you do anything - only a link to a CAPE discussion that includes international markets, for those interested.

 

[photoguy]

Well one can always invest in Faber's GVAL etf. (Although I'm quite a fan of CAPE I wouldn't invest in this specific etf myself for various reasons)

Thanks, an interesting read.
- Some analysts make the case that relatively recent changes in the US market and in accounting rules make current US PE10 numbers non-comparable to the historic numbers. If that's the case, that the US historic CAPE can't be compared to the present US CAPE, I wonder if it is reasonable to compare CAPEs across national borders/economies? Seems like the baseline differences could be very significant, posing the same problem.

The last time I looked at this, Faber was looking at countries like greece, russia with CAPE values around 10 or less. Way below the US so probably slight differences in how earnings are computed won't make much of a difference.

 

[Totoro]

I wonder if it is reasonable to compare CAPEs across national borders/economies? Seems like the baseline differences could be very significant, posing the same problem.

I've read quite a few stories about investors discussing different accounting rules across Europe (IFRS) and the US (GAAP) specifically resulting in very different numbers.

The example I remember best are telcos. Apparently depreciation and revenue recognition works so differently that you can have book value and earnings varying by as much as 30% depending on how one interprets the accounting rules.

That was for individual companies but I can imagine it goes for whole countries as well.

It was a while back (early 2000s) that I last read about it, so situation may have changed.

 

[Totoro]

The last time I looked at this, Faber was looking at countries like greece, russia with CAPE values around 10 or less. Way below the US so probably slight differences in how earnings are computed won't make much of a difference.

Russia specifically is low priced because of political risk.

Gazprom is my favorite example since the Ukraine crisis:

• P/Cashflow: around 2 (S&P 500 = 12, XOM = 12)
• P/B: around 0.4 (XOM = 2.1)
• P/Sales: 0.7 (XOM = 1.4)

A steal right? Until Putin steals it from you indeed ..

 

[Gone4Good]

I've read quite a few stories about investors discussing different accounting rules across Europe (IFRS) and the US (GAAP) specifically resulting in very different numbers.

Yes accounting rules are different from place to place. Earnings in Hong Kong will not be directly comparable to earnings in the U.S. or Germany without making substantial adjustments. But accounting rules within a country are the same.

What that means is that you can't compare CAPE values from one country to another. A CAPE of 20x in the U.S. doesn't mean the same thing as a 20x value in Europe.

But you certainly can compare them within a country. And what you find is the same strong relationship where low CAPEs precede higher future returns and high CAPEs precede low future returns.

From the graph attached above we can see that a 15x multiple is the dividing line for many countries between double digit future returns and single digit returns. For Switzerland that same line is around 20x and in Hong Kong it's more like 10x. It's then easy enough to use these bands to make rough assessments about each market's relative over or under value position.

 

[Gone4Good]

But what use is this to me... besides telling me to put new money into VXUS and not into VTI?

If I have lets say 4 000 000 bucks in VTI with 2 500 000 of market gain should I sell it and pay 500-600k dollars in taxes and wait for Bear Market? I think not.

Some of us like to make hay while the sun is still shining. Measures like CAPE help tell us the hour of the day, which is now closer to dusk than dawn.

 

[samclem]

From the graph attached above we can see that a 15x multiple is the dividing line for many countries between double digit future returns and single digit returns. For Switzerland that same line is around 20x and in Hong Kong it's more like 10x. It's then easy enough to use these bands to make rough assessments about each market's relative over or under value position.

I understand the rationale, and can see that might work. But anyone going this route should understand they are using an extrapolation (cross-economy appraisal/handicapping) from a point that is none-too-steady to begin with (using present PEs relative to historic ones to determine likely future returns in single national markets). It's a bit like putting a step ladder on a thick mattress, then climbing it with a pole saw in hand to cut a much higher branch. High adventure. But, it could turn out fine.

 

[Gone4Good]

I understand the rationale, and can see that might work. But anyone going this route should understand they are using an extrapolation (cross-economy appraisal/handicapping) from a point that is none-too-steady to begin with (using present PEs relative to historic ones to determine likely future returns in single national markets). It's a bit like putting a step ladder on a thick mattress, then climbing it with a pole saw in hand to cut a much higher branch. High adventure. But, it could turn out fine.

I'm completely sure I don't follow.

All they're doing is using a Price Earnings multiple in each country to asses whether those equity markets are rich or cheap. It seems like a very straightforward thing to do.

 

[samclem]

I'm completely sure I don't follow.

All they're doing is using a Price Earnings multiple in each country to asses whether those equity markets are rich or cheap. It seems like a very straightforward thing to do.

Yes, the math is straightforward. But unless we have a firm understanding of the underlying mechanics and the likelihood they'll continue (preferably based on considerable historical data), then I'm not sure how valuable the whole exercise would be.
Example: Suppose we have good, reliable data since 1970 on PE10s for lots of countries, and the subsequent equity returns. Lets take 4 for example (data entirely notional):
Mexico: Present PE10 = 15, historical real expected annual total return for next 15 years at this PE10: 6%
United States: Present PE10: 30, historical real expected annual total return for next 15 years at this PE10: 3%
Germany: Present PE10: 25, historical real expected annual total return for next 15 years at this PE10: 2%
Greece: Present PE10: 15, historical real expected annual total return for next 15 years at this PE10: 8%

What do we do with this data? Go overweight into Greek equities? There are lots of factors that explain why the 47 year historical returns of Greek equities at the present PE10s will not be a good guide to the future. A better guide would likely be what investors are willing to pay for these stocks today, relative to the stocks in other economies. That (in theory) has all the factors of the present environment and risks of the future environment baked into it. And we can do that easily using a cap-weighted basket of securities. Sure, it's not perfect, but (IMO) given the changing dynamics of international markets (incl currency fluctuations, politics, etc), I'd feel more secure with that, in the case of most countries, than an estimate of future returns based on their past. That goes especially for more volatile economies/countries, less so for more stable ones.

 

[haha]

A better guide would likely be what investors are willing to pay for these stocks today, relative to the stocks in other economies. That (in theory) has all the factors of the present environment and risks of the future environment baked into it. And we can do that easily using a cap-weighted basket of securities. Sure, it's not perfect, but (IMO) given the changing dynamics of international markets (incl currency fluctuations, politics, etc), I'd feel more secure with that, in the case of most countries, than an estimate of future returns based on their past. That goes especially for more volatile economies/countries, less so for more stable ones.

This is exactly the inverse of a value philosophy. Some like chocolate, some strawberry.

I am not surprised at the tenor of many responses here. People mostly like action, and an extreme value proposition is like fishing in a lake with a few big fish. The fish are big, but there are only a few of them, and no little fish to keep us entertained.

Ha

 

[MichaelB]

What do we do with this data?

No one is suggesting this is intended to be actionable. Personally, I find it is more helpful to follow (and track) the same advisors (and data) over time, and the link in the OP leads to recent and updated opinions and data that some members read and track regularly. It is clear to me that not everyone finds value in PE10 and CAPE data, but some do, and this thread is for them.

There aren't many good, regular sources of CAPE data for other countries, which is why I posted it here. The relative valuation between the countries is not a surprise, and I would be shocked if someone here jumped at this as a signal to make a radical shift in one's allocation model. Most of the data, and opinion, we get on market valuations originate from the sell side, and I thought this would be a nice counter view.

 

[CaliforniaMan]

I have been trying to make heads or tails of CAPE for a while now, and for myself I can't figure it out. For one thing, CAPE is somewhat of a predictor of CAPE 20 years in the future. We have these nice approx 60 year quasi-periodic cycles in the data. Is much of the predictability we see an artifact of this? And when I plot out CAPE vs. future returns for different 20 year timer periods (each which require 40 years data) there are vastly different plots. And some very smooth...but... the data are so auto-correlated, there is so little independent data, the plots look smooth but it is mostly artifact. And looking at the recent period, say 1980-2006, it looks more like mid point of 25 rather than 15. And what about 2009, why did we get only to 13 instead of around 8 like in 1974? Is 13 the new 8? Or was the government response to the crisis just so much better in 2009 than it was in 1974, and if so, will it continue to be so? I just can't figure out what I can do with CAPE or how to use it. Obviously (to me anyway) you cannot compare the CAPE of for example the U.S. with Russia or Greece or emerging economies, but can you compare the U.S. with Germany? Yes, stocks are sometimes too high (relative to the future) or too low (again relative to that unknown future agaom), but how does CAPE help. Just looking at the data it seems to me (and really I am really trying not to fool myself) that a CAPE of 25 right now is fairly valued.

 

[samclem]

It is clear to me that not everyone finds value in PE10 and CAPE data, but some do, and this thread is for them.

I'm among those who find value in PE10, at least as a guide to likely future returns within a single, relatively stable economy. And, it's useful to see the tools and models being used by others.

 

[samclem]

Yes, stocks are sometimes too high (relative to the future) or too low (again relative to that unknown future agaom), but how does CAPE help.

It is surely a highly imprecise tool with unpredictable timing, and that's why I'd never go "all in" or "all out" based on it. But it does seem to have a good record, across several markets and many time periods, of improving risk-adjusted returns, in general.

PE10 doesn't tell us when the market will turn, but it can apparently indicate that the chances of continuation are changing. That's still useful.

Some pretty good discussion here and here.

 

[papadad111]

Global CAPE 10 is interesting. Good topic.

Isn't the historical market data within firecalc solely based on USA equity returns. That's always been a glaring gap in firecalc in my opinion.
Expansion to include behaviors of the top global equity markets would more realistically model today's investors' AA.

And then, somehow add in currency exchange risk...

 

[kramer]

Great detailed article on CAPE here by Larry Swedroe:

Swedroe: CAPE 10 Ratio In Need Of Context | ETF.com

He covers how the accounting rules have changed in the US, idiosyncracies and history of CAPE, etc. A very well written and detailed article.

 

[Gone4Good]

There are lots of factors that explain why the 47 year historical returns of Greek equities at the present PE10s will not be a good guide to the future.

That's the nature of value. There's always lots of good reasons not to own cheap securities. That's why they're cheap. And lots of good (or at least somewhat plausible) reasons to own expensive securities. That's why they're expensive.

Whether or not anyone finds any of this useful is mostly a matter of investment style, temperament and discipline.

 

[Gone4Good]

Great detailed article on CAPE here by Larry Swedroe:

Swedroe: CAPE 10 Ratio In Need Of Context | ETF.com

He covers how the accounting rules have changed in the US, idiosyncracies and history of CAPE, etc. A very well written and detailed article.

Not a fan. Most of his adjustments are nonsense. A longer critique of this Swedroe piece is here

 

[Boho]

I'm trying to shorten a short list of countries to invest in and I found this advice from Burton G. Malkiel: " 'U.S. investors need to overcome their home-country bias,' noting that the CAPE ratio of several foreign markets is below historical averages (unlike in the U.S.)."

In researching more on CAPE, I found this thread. The title of this thread and of the linked article scared away from CAPE, but I read the thread and the article and the first chart in the article. It sounds like CAPE works. If you sell at the worst time, then using CAPE is worse than if you don't use valuation indicators like CAPE, but overall, using CAPE will make you more money, right?

I've reduced this thread to:

What that means is that you can't compare CAPE values from one country to another. A CAPE of 20x in the U.S. doesn't mean the same thing as a 20x value in Europe.

But you certainly can compare them within a country. And what you find is the same strong relationship where low CAPEs precede higher future returns and high CAPEs precede low future returns.

I'm not sure now though. Low CAPE means cheaper, higher value, right?

 

[bmcgonig]

I'm trying to shorten a short list of countries to invest in and I found this advice from Burton G. Malkiel: " 'U.S. investors need to overcome their home-country bias,' noting that the CAPE ratio of several foreign markets is below historical averages (unlike in the U.S.)."

In researching more on CAPE, I found this thread. The title of this thread and of the linked article scared away from CAPE, but I read the thread and the article and the first chart in the article. It sounds like CAPE works. If you sell at the worst time, then using CAPE is worse than if you don't use valuation indicators like CAPE, but overall, using CAPE will make you more money, right?

I've reduced this thread to:



I'm not sure now though. Low CAPE means cheaper, higher value, right?

Mebane Faber runs an ETF that invests in the countries with the lowest CAPEs. Might be interesting to you too see what he does

http://www.cambriafunds.com/gval.aspx