news article background image
2017-11-16Macro `n Cheese

Long only

We like to think of Roger as our very own “Éminence grise”, not because we think he’s very influential (God forbid!), but rather because he is such a perfect representative of the dismal science. In this week’s piece on fundamental stock market valuations we see him make yet another triste figure.

There is always a fair share of discussion on equity market valuations, but my impression is that those discussions have intensified during autumn. If nothing else FT’s celebrated Chief Economics Commentator, Martin Wolf, recently wrote a column on the subject and I thought it was an interesting read indeed, despite the argument of an inevitable (?) rise in real interest rates being somewhat overlooked. For that reason, and out of curiosity, I wanted to have my own look at equity valuations. No matter how I turn, it seems hard to escape a view of lofty valuations.

The CAPEd crusader

To start off, let’s look at a couple of oft quoted measures for long-term valuations: The Shiller Cyclically Adjusted P/E-ratio (CAPE), the Siegel-Bianco total return CAPE, the Hussman GVA-indicator and the “Warren Buffet-indicator”. The Siegel-Bianco total return CAPE tries to take into consideration the structural (mainly tax-induced) changes in dividend policies that has affected the original ‘net’ calculations (this is also a point that Shiller later addressed1), but the validity, reliability and effectiveness of such (and other) alterations are hotly debated.

1 Macrobond Document
Figure 1: The cyclically adjusted p/e-ratios are disturbingly high.
Note:The series used in the above calculations are only available to those who have access to the MSCI-data base (and computable for a number of other countries). The Shiller CAPE is nonetheless available for all Macrobond users.

To be sure, the cyclically adjusted p/e-ratios have proved efficient in forecasting the subsequent 10-25 year returns, and studying the full sample Shiller CAPE-ratio we can conclude that valuations have only been higher during the turn of the millennia (during the IT-bubble) and in sept 1929. So, according to the most used CAPE-ratios, stock markets are richly valued, and this suggests some type of upcoming correction.

Fundamentally yours

What about other popular measures of equity valuations? I like measures that use economic logic (and data) which is why I have put together a few of those that I think make good sense; the Warren Buffett-indicator (applause) and a, perhaps, more evolved variant of Buffett’s measure, developed by Hussman Strategic Advisors. I have also put in Tobin’s Q for good measure.

Macrobond Document
Figure 2: Fundamental measures are cause for concern.
Note: Tobin’s Q is probably the preferred measure from a theoretical standpoint, but as companies generate an increasing amount of income from abroad this measure becomes skewed (as the denominator does not take into account foreign assets).

As can be seen, these metrics demonstrate much the same patterns that the CAPE-ratios did. Importantly, and as with the CAPE-ratios above, on the subject of timing the metrics computed above remain completely silent.

That said, concerns surrounding asset prices are probably starting to feed into the minds of policy makers as well. The latest FOMC-minutes (among others) stated that a couple of “[…] participants expressed concern that the persistence of highly accommodative financial conditions could, over time, pose risks to financial stability”. The perceived risk of “irrational exuberance” in asset prices is probably one important reason to why the FED seems less inclined to take a cue from a string of continued low inflation outcomes.

When all is said and done, are there really no valid economic arguments for higher stock markets? But of course, this is economics after all!

A key objection to all the above metrics is, for all their previous successes, that they have no forward-looking element to them. While they have worked well in the past, what if this time really is different?

Beware the snake oil

Many pundits as well as academics are putting forward low discount rates as a possible answer. This argument relates to the fundamental asset price model where an asset’s price is determined by the present value of all (discounted) future cash flows.
But a lower discount rate is an utterly insufficient argument. The discount rate has two parts; the risk-free interest rate and the equity risk premia. If your equity salesperson refers to a drop in the risk-free interest rate, this should be balanced by an equivalent drop in the growth of cash flow, leaving the result of a traditional NPV-calculus unaffected. That argument is snake oil.

Then again, should he/she allude to a drop in the equity risk premia (hence lower discount rate relative to the growth of cash flows), it could indeed serve as an argument for structurally higher stock prices. Who knows, it might even put an end to the “equity premium puzzle” that has haunted academics and gratified equity investors ever since (at least) the 19th century. Unfortunately, I have so far found little discussion on these important distinctions and their implications (reading suggestions are welcome). This argument is, perhaps, Chinese snake oil.

Sure enough, just as recently as a few days ago, in a FRB SF Economic Letter2, the discount rate was put forward as an explanation again. Albeit sketchily discussed (i.e., assumed away), the proposed model fails to directly address how investors, in aggregate, could expect the risk-free interest rate to remain low, without making equivalent reductions in aggregate future cash flows (GDP). Tellingly, the author used different sets of data (and methodology) for interest rates and potential GDP-growth3, which not only provide a better fit and higher R2, it also provided us with an opportunity to discuss differing views of the stock market.

3The model uses Laubach-Williams (LW) measure of r* while it uses CBO:s measure of potential GDP, despite LW also providing an estimate of potential GDP (y*).

Now, I would assume that the author’s main intention was simply to exploit the CBO’s forecasts thinking that there should be no qualitative differences between estimates of potential GDP; as a swift glance on the graph below (upper pane) also would indicate. However, and as the lower pane in the graph shows, the deviation between the two potential GDP-estimates increase first gradually (from 2000) and then suddenly (around 2008). Which is why chances are that the model inadvertently picks up something ‘else’ in that shift (remember that the model’s explanatory power and test statistics improve with the use of CBO’s measure of potential GDP).

Macrobond Document
Figure 3

Admittedly, we cannot say for certain what this ‘else’ is. It could be a substitute for an expectation of stronger future cash flows, which would eventually lead the (risk-free) interest rate higher and, hence, eventually push down the CAPE-ratio to historically normal levels again (around 15). However, and more interestingly, the ‘else’ might also reflect a drop in the equity risk premia over the last 10 years or so, in which case the CAPE-ratio could very well remain on elevated levels for the foreseeable future4.

4To be clear, this hypothesis could be tested in a number of statistically more satisfying ways, but I think that would not have provided us with such an interesting backdrop.

If this latter view proves to be correct, the projections made may not be so far off after all. In the graph below, and to give you an idea, I provide an artless version of the more advanced statistical maneuverings in the original.

Macrobond Document
Figure 4

As can be seen, the model captures current elevated levels of the CAPE quite well. At present, the predicted CAPE-ratio is approximately 8% below the actual CAPE-ratio, while during the IT-bubble it was some 43% below the actual CAPE-ratio. To many this will probably evoke some hope that even a macro founded view of stock markets might not be all snake oil and that perhaps this time really is different. Yes, perhaps. At least this (somewhat haphazard) model does not suggest that a crashing bubble is imminent.

No matter how you turn, stock market valuations are high

Then again, when looking ahead, the model still sends a sobering message as it seems the CAPE-ratio must nonetheless come down almost 15% from current levels. This would, no matter how you turn it, imply a very poor stock market performance compared to recent experiences.

In conclusion, most fundamentals seem to indicate that stock markets are richly valued. However, the metrics normally employed are inherently backward-looking. In particular, they overlook the possibility for structurally higher CAPE-ratios from a conceivable resolution of the equity premium puzzle (i.e., from a lower equity risk premia). In addition, and importantly, they say nothing about the timing of any correction.

Get these posts delivered to your inbox

If you like what you read, why not subscribe to have these analyses delivered to your inbox every Friday?

Sign up


We don’t usually have views and opinions about economic and financial states of affairs, (not ones that we express publicly as a company, anyway). We do believe, however, that people can and do appreciate a variety of perspectives. What you’ve just read is the perspective of our resident chief economist. While we think he’s very smart, Macrobond Financial does not expressly endorse the views he presents here. And, as the old adage goes, you shouldn’t believe everything you read (not without finding the data, performing a few analyses and presenting it in a nice chart). We want to make it clear that we are not offering this information as investment advice. That being said, if you have the application you can easily check everything that’s mentioned here, and decide for yourself. If you don’t have the application, now you have a great reason to get it.