Saturna Capital has an interesting take on the calculation of the Graham / Shiller PE10, otherwise known as the Cyclically Adjusted Price Earnings ratio (CAPE). Saturna argues that The Market May Be Cheaper Than It Looks because the Consumer Price Index (CPI) provided by the Bureau of Labor Statistics (BLS) understates the true rate of inflation, a key input to the CAPE calculation:
Potentially Understated Inflation
Given that inflation estimates play an influential role in the calculation of the P/E10, it is important to investigate the assumptions behind the calculation of inflation. Traditionally, inflation is measured using the Consumer Price Index (CPI) provided by the Bureau of Labor Statistics (BLS). The CPI estimates inflation by measuring fluctuations in the average price of a basket of consumer goods and services that is deemed to be typical of the average urban consumer. However, due to a variety of reasons, largely political, the methodology used to calculate CPI has undergone many changes in the past 10 to 20 years. One of the most controversial changes was to alter the composition of the basket to reflect changes in consumer behavior over time.
In doing so, the BLS hoped to remove biases that cause the CPI to overstate the true inflation rate. Former chairman of the Federal Reserve Alan Greenspan advocated this alternative methodology, arguing that if the price of steak went up, consumers would choose to eat more hamburger meat instead.² He therefore concluded that unless hamburger meat replaced steak in the basket, inflation would be overstated because consumers were not actually spending more money. Skeptics view these changes as government manipulation, the purpose of which is to understate the true inflation rate, as well as the wage and other rate increases indexed to it (think Social Security).
Saturna uses an alternative measure of inflation: the Shadow Government Statistics’ (SGS) Alternate CPI:
Over time this recalibration of the CPI has produced lower inflation estimates than the “old school” method. In fact, the discrepancy has become rather large… Unlike Mr. Greenspan, however, we prefer steak to hamburger meat. Accordingly, we tend to believe the truth lies somewhere in between the BLS’s CPI and the Shadow Government Statistics’ (SGS) Alternate CPI.
The implications for CAPE using Shadow Government CPI are as follows:
The wide gap between the government-sanctioned CPI and the Shadow Government CPI presents a competing set of assumptions about how to measure the effect of rising prices on the average consumer and the market as a whole. The relevance to investment analysts is that higher inflation figures can have a dramatic impact on the current P/E10 ratio. For example, if inflation is assumed to be 5% annually, $1 in nominal earnings from 10 years ago would be worth approximately $1.63 in today’s dollars. At 10% annually, $1 in nominal earnings from 10 years ago would be worth about $2.59 today. Using a higher inflation estimate therefore increases average real earnings over the 10-year period, and thus lowers the P/E10 ratio. If we assume the SGS figures are correct, then the current P/E10 based on the average closing price during the month of June is about 14x (see chart below). This ratio is much lower than the current P/E10 of near 20x using traditional CPI figures.
Hat tip Ben Bortner.
Here is link to a McKinsey article on how the accounting works and its effect on stock value https://www.mckinseyquarterly.com/Corporate_Finance/Capital_Management/How_inflation_can_destroy_shareholder_value_2520
LikeLike
[…] 2, 2010 by greenbackd My post on the Shiller PE10 ratio calculated using Shadowstats’ alternate to the BLS CPI generated some great discussion about the various flaws in the market-level PE and PE10 – […]
LikeLike
Hey Greenbackd!
Hope you are having a profitable and enjoyable year so far…. Not too many cigar butts laying on the street just yet, but they may be here in the next few months if your CAPE analysis is wrong — so if there is a silver lining, it is that you will be back to evaluating fantastic investment opportunities and making money for your readers…
I would look at the PE 10 inflation adjustment in terms of Wages and household income. Sure, corp. earnings are higher in absolute terms vs. the CPI reported, yet your argument that the CAPE is really 14 doesnt work in my humble opinion because the average worker is earning less or about the same as he was 10 years ago… Therefore the demand has not gone up as it would if this stealth inflation was improving our lives and making stocks cheaper…
Conversely, I would look at things like gold and the Rogers Commodities index and think that as our purchasing power has stayed the same, our raw materials input costs have become much more expensive and are rising after a substantial dip that has improved margins drastically.
So, earnings may be improving, but sales are not improving which would seem to suggest that further raw materials price increases will contract profit margins, and that the margin growth in the past year and half can be partially attributed to the fall in raw materials prices and the price of oil… The more money the system prints, the less oil there is per dollar, which theoretically should compress margins for just about every business besides the oil companies…
So for me, I would think the CAPE is very relevant and that stocks are reflecting higher margins due to lowered input costs or COGS… If we have to print more money to “save” housing and banking, this would it seem be to the detriment of most businesses that will continue to see their purchasing power erode vs. Oil and Food costs…
I really enjoyed reading your article as always… btw had a nice comeback since my failed proxy war… So things have improved over here. Lets get back to finding bargains! Hopefully we see some stocks on sale again — 2008 was like shooting fish in a barrel for your analytical strategy… All the best, nick.
LikeLike
BTW I am not a doomsdayer at all, just astonished by that chart you have displayed! Seems we were living in the seventies and didn’t even know it… How about some cheapies….
ACMR again
Tues
TA
CENT
PCC
voxx-covered calls are better here?
knd?
LEE
HQS
Anyways, thats what I’m staring at these days…
LikeLike
@ Parker, thanks for posting the link to the LSE paper — interesting and credible.
LikeLike
Gemfinder:
I like your analysis with one modification: Any company which can raise their prices as fast as inflation should do fine, and can be a good deal at PE10. As long as it’s expenses don’t rise any faster than inflation.
LikeLike
Dear Parker Bohn
I just want to bring to your attention is that there is two way to calculate the PE 10.
1-simple arithmetic average, which is falwed because of what you mentioned in your post.
2-ROE method which is the preferred one.
You take the average of ROE over 10 year and multiply it by latest BV.This would give you the normalized EPS that is adjusted for the entire factor you mentioned and without flaws of the simple arithmetic method.
Regards,
LikeLike
Thanks, Mohammed.
The ROE method makes sense. Are you aware of any sources using this, deriving long term PE averages from it, and so on?
For this to be used like the standard Graham/Shiller PE10, we would have to have long historical data series on market ROE. I’m not sure if this is available?
If anyone knows about this, please post. Thanks.
LikeLike
Another evident flaw in the PE10 ratio, other than fuzzy inflation numbers, is the failure of our earnings numbers to account for differences in capital reinvestment rates.
In decades past, companies paid out more of their earnings in dividends. This resulted in slower earnings growth, and thus a lower PE10.
In modern times, companies as a whole pay less dividends, and reinvest more capital. This results in a higher rate of earnings growth, which causes earnings from 10 years ago to appear small, thus resulting in a higher (less attractive looking) PE10.
Another way of saying this is, if company A invests only enough to maintain earnings at their present level, and pays out the rest as dividends, then its 1 year PE and PE10 will be the same number. If company B reinvests everything, and grows profits, it may offer the same prospects, 1 year PE, and total return to investors, as company A, but it will have a higher (less attractive looking) PE10, since earnings 10 years ago were much lower.
This is a shortcoming of the PE10 method, and adjustments should be made.
LikeLike
Interesting post and reply.
@ Gemfinder:
It seems like it would be impossible to simultaneously have a healthy economy and hyper-inflation, so, yes, stocks under hyper-inflation should be cheap.
However, assuming a more normal rate of inflation, shouldn’t corporate profits (and stock prices) rise along with inflation?
Or in other words, if you expected inflation to be 4% annually rather than 1% for the next 10 years, would it materially change how you valued the market?
LikeLike
@ Parker, during the 1970s inflation, share prices did not keep pace with inflation; some of the lowest PE10 ratios of the past century occurred in the late 1970s. This is because inflation requires yields to rise.
So when inflation is 8%, investments must yield at least 8% to deliver a real return. Under those conditions, someone judging investments only via the crucible of P/E would not pay more than 12.5 times earnings, minus risk premium. To pay more than that is to bet on either earnings growth or lower inflation — neither of which is contemplated by the PE crucible.
@ Mohammed, that is interesting, and true, ROE gets around the problem of average earnings, but only if you adjust the value of assets, particularly PP&E, for inflation. I’m borrowing this idea from a Berkshire Hathaway annual report from the 1970s: ROE results can mislead under high inflation, because the replacement cost of PP&E is much higher than shown on the books.
LikeLike
Gemfinder, thanks for the reply.
It is true that stocks tend to do poorly during times of high or rising inflation. My opinion has always been that this is an irrational market reaction, ie stocks were underpriced during much of the 70’s.
A theoretical explanation of this is offered by the Modigliani-Cohn hypothesis, which posits that investors use incorrect nominal discount rates during times of inflation, when inflation-adjusted discount rates should be used. Thus stocks are typically too cheap when inflation is high.
Another way of saying this is, it is still reasonable to have real GDP growth of, say, 2%, even if inflation is 8%. In nominal terms, this is GDP growth of 10%, which is so far above the average GDP growth that investors fail to anticipate it, and therefore misprice equities.
A recent wonkish review of this theory:
Click to access Money%20illusion%20in%20the%20stock%20market.pdf
LikeLike
Clever data mashup, and mathematically true, but should draw exactly the opposite conclusion.
Why? Because with high inflation, 14x earnings is not cheap.
To appreciate this, imagine you are investing in the stock market of imaginary, hyperinflationary Zimbabwistan.
Say a Zimbabwistani dollar of earnings from 10 years ago is worth Z$10 million today. Using CAPE, you determine that every stock is trading at less than 1x PE10. Bargains galore?
No. If hyperinflation persists, then a Zimbabwistani dollar of earnings 10 years from now is worth just Z$0.0000001 today. How much should you pay for that? A lot less than 1x PE10.
Yes, the above does not factor in growth. But neither does PE10, so the comparison is apples-to-apples. Coming back to the post, it should be clear that if inflation really is 10% (half the dollar’s value is lost every 7 years), then 14x earnings is not cheap.
This brings up an interesting historical point about the “bargain” stocks of the late 1970s. The reason stocks were such a good deal is that they were pricing in a sustained inflation that, thanks to Volcker, did not in fact occur (or is widely believed not to have occurred). But if inflation had continued at 15% through the 1980s, then the low PE of the S&P500 in 1978 would not have turned out to be such a bargain.
A rational investor, unable to predict the future, would have begun buying in 1981, when Reagan kept Volcker on. This made it abundantly clear the Fed was committed to low inflation, but shares had not yet priced that in. That was the rational entry point.
Today’s situation is almost the opposite. The Fed has signaled a rock-solid commitment to expansionary policy. Markets are not pricing that in. So today’s PE10 is likely much too high.
LikeLike