New Sunday Times Feature - Diary of a Private Investor

Where did I guffaw at anybody?!

Again, I think it's entirely reasonable to refer to the past when determining your current expectations.

CAPE is simply the price of a stock divided by the average of 10-years previous earnings of that company, adjusted for inflation. It has nothing to do with reversion to mean.

I absolutely agree with you on one point though - a large part of investing is behavioural.

A lot of people invest in equities through their pensions while carrying a mortgage. That's leveraged investing but most people don't see it that way for some reason.
 

Incorrect; it has everything to do with reversion to mean. The ratio is useless unless compared to something. We can’t say whether it’s high or low without assuming mean reversion. Surely you get that?

It is also oversimplistic to look at the mortgage/pension piece in that manner. It’s clouded by the fact that an individual’s pension contributions operate within a “use it or lose it” framework. If I don’t contribute for 2017 before 31 October 2018, I never can.
 
I know Colm's post is partly tongue in cheek, but just because you "need" 6,7,8,9%, does not mean equities will fulfill your needs. With bonds yielding 1% and many people "needing" far more than that, I have to strongly suspect that equities (like bonds) are currently over priced and in the end of the day, for someone investing their ARF now, the return they will get is decided by the price that they pay for their equities in the second hand market.
 
Duke of Marmalade said:
I know Colm's post is partly tongue in cheek,
Click to expand...
Not really, Duke. My long-term financial plan is based on that assumption. It's worth adding the later sentence from the article, however:
I rely particularly on the latter. My investment philosophy is to only invest (and only to stay invested) in shares/bonds (I have bought one particular bond in the past that qualified under these criteria - I may write about it sometime) that I believe will deliver that return over the next 5, 10, 20 years. I am confident that all the stocks in my portfolio, a large proportion of which have been mentioned in the diary by now, will deliver the required 6% a year (I do have a small margin of safety in that I'm budgeting on getting slightly less than 6%).
 
Okay, Colm, your dividend argument in respect to FTSE at least is persuasive. As to history being a precedent, we have never before seen negative yields or indeed yields of 1% on long term bonds, so we have no reliable historical guide as to where these astronomical asset prices will go from here.
Gordon Gekko said:
Incorrect; it has everything to do with reversion to mean.
Click to expand...
I love these semantic debates. In this case I must adjudicate in favour of Sarenco. As I understand it, CAPE tries to measure current market prices by reference to an average of recent earnings over a period of years. This seems entirely valid and closer to reality than using the average earnings over the last 12 months. For example, if a company reports a loss in a year, its price doesn't go to zero, instead investors give credit for past earnings. Maybe GG is arguing that this shows that investors demonstrate a certain reversion to mean in their valuations but if so that does not invalidate the metric.
 
Last edited:
Duke, perhaps you both had a late night, but for the avoidance of doubt, virtually ALL measurements like this assume mean reversion!

Otherwise they’re meaningless numbers (e.g. 32). What is 32? Oh, it’s high is it? Based on what? Oh, I see, mean reversion!
 
Ahh Gordon are you sure you haven't had a wee dram after breakfast:rolleyes:
CAPE is a measurement. It makes no assumptions at all. It is motivated by the plausible premise that market participants give some weight to recent past years' earnings rather than dismissing all statistics other than those pertaining to the last twelve months or indeed their forecast of the next twelve months. I thought the example of the company which made losses in the last year but still has a market valuation would convince, but possibly that dram wasn't so wee after all.
 
Last edited:
It’s a measurement that produces a number (e.g. 32).

The number, e.g. 32, is only relevant if it can be compared to something else.

Otherwise, it’s about as useful as a chocolate teapot.

The number, e.g. 32, is useless unless we make certain assumptions around mean reversion.
 
So you would have had the same objection if he had used straight P/E rather than CAPE?
I thought your objection was to the averaging process involved in CAPE. Misunderstanding then, the pitfalls of blogging.
You would then level the mean reversion criticism against anyone who says that a current figure is historically high or historically low. It's a fairly tautological criticism which gave first impressions of having scientific substance.
 
Not really; I’m merely pointing out the inconsistency of highlighting measurements such as CAPE whilst frequently being critical of anyone who assesses future market performance based on its past performance.

It’s a little like trying to have one’s cake and eat it.
 
As I understand it, CAPE measures the P/E ratio, adjusting the "E" for cyclicality. On this measure, the US stock market is very expensive compared to the average cyclically adjusted P/E ratio in the past. Isn't this what Duke is saying? I've no problems with this. I don't invest in the market. I invest in particular stocks that I think are not overvalued.
 
I could add that I make "negative" investments in stocks that I think are overvalued, e.g. Tesla (still!)
 
@Gordon Gekko

Is your argument that a statement to the effect that a particular price earnings ratio is above (or below) its long-run average necessarily reflects an implicit assumption that the ratio will inevitably revert to that long-run average? My apologies but I thought you were talking about the CAPE methodology itself (the phrase "built on" threw me).

If that is your point (and please do correct me if I have misrepresented your argument), I don't think it's correct - there could be very good reasons for a secular shift in the ratio, it doesn't necessarily have to revert to mean.

Again, I think it is entirely reasonable to refer to history when framing your expectations regarding future asset returns and I have said so repeatedly. I'm really not sure why you think otherwise.

But valuations matter. Higher valuations imply lower expected returns and lower valuations imply higher expected returns. I would have thought that was obvious but perhaps not.
 
Gordon Gekko said:
You are incorrect regarding CAPE; it is a valuation metric which looks at the position relative to long-term averages; i.e. it’s based on the idea of mean reversion.
Click to expand...
I had to review the "angels dancing on pins" debate. This is where it began. This is quite clearly a criticism of the CAPE methodology per se on the grounds that it is based on the "idea of mean reversion". That is blatantly incorrect. GG has subsequently generalised his criticism against any statement that a current ratio is historically high or low. That is not what he meant in the first place. Let's move on.
 

Duke, I’m not sure why both you and Sarenco have a fascination with telling people what they mean.

My point remains as follows:

It is silly to dismiss mean reversion and the historical performance of markets on the one hand and to then go on about CAPE. CAPE is just a number that has to be compared with something. If one ignores the concept that the future will mirror the past, all metrics and analysis becomes clouded.
 
Throwing out numbers like CAPE or P/E is completely meaningless unless you’re comparing them to something; and what do we compare them to? Long-term averages. “Valuations are high” “Valuations are low”. All nonsense unless one buys into the notion that over the long-term markets will do similar things. And why it’s inconsistent to talk about CAPE on the one hand and the uncertainty around future market performance on the other.
 
Gordon Gekko said:
(CAPE) is a valuation metric which looks at the position relative to long-term averages; i.e. it’s based on the idea of mean reversion.
Click to expand...
Gordon Gekko said:
Duke, I’m not sure why both you and Sarenco have a fascination with telling people what they mean.
Click to expand...
The first of the above statements requires no guesswork as to what the author meant. It is an incorrect statement, end of. It subsequently transpires that what you really meant was something quite different.
 
@Gordon Gekko

With respect, I really can't figure out what you mean - that's why I asked for clarification.

I don't think that CAPE is predictive of the future direction of the stock market if that's the implication. I brought it up in the context of a discussion on the relationship between bond and equity pricing.

Again, I think it's perfectly legitimate to refer to historic asset returns when framing your expectations about future returns. That doesn't mean I think the future will mirror the past!
 
Duke of Marmalade said:
The first of the above statements requires no guesswork as to what the author meant. It is an incorrect statement, end of. It subsequently transpires that what you really meant was something quite different.
Click to expand...

How is it incorrect? It’s a ratio. What are ratios? Means of comparison.
 
Gordon Gekko said:
How is it incorrect? It’s a ratio. What are ratios? Means of comparison.
Click to expand...

In another thread you will read that Ireland has a ratio of English speakers of 93%. By comparison the US has 79%. Who is arguing that such ratios do not make for good comparisons? But I presume that in this case you will accept that it has nothing to do with reversion to mean.

A. Mean reversion means the tendency to revert to "long term averages".
B. CAPE is a ratio calculated using "long term averages".

You stated that the use of B implies a belief in A because they both are linked to "long term averages". That is incorrect.

Look, even economics professors can have a "deposit selling" moment. 'Fess up and move on.
 
You must log in or register to reply here.