Archive for August, 2010

Top-Level Outlook for Stocks: Weak returns for a while yet

Saturday, August 14th, 2010

On DeLong’s Blog, I got some replies to my comment (see previous post below, where I expanded on the comment).  Robert Waldmann responded: “Brad didn’t assert that stocks are reasonably valued — he asserted that they are undervalued from the perspective of a long term investor. His claim is that stocks have always been undervalued from the perspective of a 35 year old saving for retirement.”

But even with a 30-50 year time horizon, one should not put money into an investment that is quite likely to stagnate for the first 10-20 years. The opportunity cost is too high. With regard to the stock market, one can write out a trio of trivial identities which, when combined, have some distressing import:

Stock Prices = (P/E) * Earnings.

Earnings = (Earnings/GDP) * GDP.

GDP = Population * (Employment/Population) * (Productivity/Employee).

From which:

Stock Prices = (P/E)*(Earnings/GDP)*(Population)*(Employment/Population)*(Productivity/Employee)

The value of this is that each of the components on the right-hand side is fairly well measured, allowing some insight into the broad trend for Stock Prices on the left hand side.  Now, before we dive in to stock price trends, note that total return is of course related to both dividend yields and capital gains (growth in Stock Prices).  But currently, dividend yields are historically very low, so any 30-something (or any other age) investor seeking a comfortable retirement needs capital gains.  So we must consider the factors affecting prices:

1) P/Es have stayed on the high side since the dot-com bubble, and have not yet reverted to anywhere near their historical low points, despite the two market crashes in the last decade.

2) Earnings/GDP is still on the high side and has refused to revert to historical norms, despite a sharp nudge in the right direction in late 2008, thanks to government giveaway policies. (Some might even say we’re robbing the taxpayer to pay boardroom leeches…)  But revert it will; go higher it cannot, if history is any guide.

3) Population growth is slowing and the rising anti-immigration sentiment will not help.

4) Employment/Population is on a downtrend and faces secular headwinds from aging demographics (and not just in the U.S., either).  There is some room for near-term improvement from uptake of the unemployed (and the overseas military who are not necessarily contributing to domestic economic production), but the long-term scenario is not positive.

5) Productivity/Employee has been squeezed hard already during the recession and, absent another huge technological revolution which is not yet evident to investors, is unlikely to grow at anything more than the usual 2-3% for a while to come.  (There is some room to grow hours/employee back to recent highs, though, and productivity here is a combination of hours worked and output per hour.)

So where exactly are we to find the factors that will even yield positive returns, much less 6% returns, within the next 10-20 years?

Footnote:  The above statements are all backed with hard data, which I will endeavor to (re-)post in future notes here.

Critiquing the P-E ratio (earnings yield) valuation approach

Friday, August 13th, 2010

Brad DeLong, for whom I normally have more respect, recently wrote a takedown of a poorly-argued Atlantic Monthly piece by Megan McArdle. But DeLong himself does not live up to the standards of rigor that we need to adopt if we seek to invest for sustainable gain.  DeLong argued that since P/E ratios are currently in the 5 or 6% range (depending on the timeframe details), and bond yields are lower, stocks cannot be said (as McArdle claims) to be poor long-term investments. However, as I illustrate below, it’s not clear that todays P/E ratios are meaningful.  Nor is it true that just because bonds are currently horrifically overpriced, stocks are not also overpriced.  We could be facing (Japan-style) a very long period of subpar returns in both investment classes.

Here is the comment I posted to DeLong’s article:

C’mon, sir.  It’s trivial to find poorly-written financial articles.  Even your own qualifies!

For starters, it’s a non-sequitur to expect, in a world of creative destruction and transformational technologies, that your Atlantic Monthly should bear any resemblance to your great grandmother’s.  Nor Harper’s, nor even Newsweek.

More importantly, you start with the mathematical truism that “The return on stocks is the dividend yield plus the capital gain.” (but neglect both inflation and taxation) and then proceed to assume that “If the P/E ratio is stable, the capital gain is equal to the growth of earnings.”

The reason I call that an assumption is because, once more, we live in a world of creative deception and transformational technobabble. “If the P/E ratio is stable, the capital gain is equal to the growth of earnings” is only true in a world where “earnings” has a constant meaning.
In point of fact, it appears that ‘earnings’ is a very loosely defined term whose meaning has been subtly shifted during the credit bubble of the past decade(s).  Today, “earnings” as used in the “P/E of 16” or “P/E of 20” metrics, manages to exclude a whole host of one-time costs, and poorly accounts for incentive pay schemes.  Even if we shift to GAAP earnings (where P/E ratios have recently been both negative on a quarterly basis, or in the hundreds on an annualized basis — both wildly beyond historical norms and not at all “stable”), we still find fantastic obfuscation and all manner of mathematical mangling by managements.  My third favorite is corporations reporting as “earnings” the capitalized, computer-simulated change in value of mortgage securities holdings, despite the fact that no cash is actually flowing nor can reasonably be expected to flow.  Second best is when a dysfunctional company recognizes the devaluation of its own debt (sometimes due to imminent insolvency!) and then claims the reduction in debt valuation as “earnings” too!  But best of all is when corporations simply lie outright, and stuff things they don’t want to talk about into physically-fictional “off-balance sheet” legal creations… both on the earnings/assets side (to hide management’s side-dealing, personal siphons into what should be secure corporate profit streams) and on the liabilities/losses side…

Back in the glory days of Atlantic Monthly, owners — shareholders — were in most cases much more closely involved in managing the businesses they owned.  I think the current system has led to a lot of bezzle — destruction of shareholder wealth not yet revealed to shareholders — and we will find that recently claimed credit-bubble “earnings” are far from sustainable.

Sir, it used to be that “capital” had physical meaning and represented, usually in a tangible way, the surplus of actual production over actual consumption.  That era ended with the demise of the gold standard and the slow, steady, Chinese-water-torture devaluation of the dollar since.  Capital gains were things you could take home with you and believe in.  And so were earnings, because dividend yields were much higher.

Today, who knows?  But I sure wouldn’t consider P/E ratios as currently reported to be historically normal, nor stocks to be reasonably valued, given what we understand about legally fraudulent and/or morally bankrupt behavior by corporate managements in so many economic sectors.


Let me quantify my point a bit further:  GDP growth can be factored in terms of population growth, employment/population ratio, and productivity of employees.  The first two are bounded demographically and the latter is limited technologically.  Meanwhile, corporate profits are currently near historic highs (as a fraction of GDP) and unlikely to grow faster than GDP going forward.  And P/E ratios are just coming off historic highs even including what appear to be more-than-usually fraudulent earnings.

It is quite likely that we will have (a) low population growth (due to birth rate and sustainability issues), (b) a secular flattening or even outright decline in employment/population ratio (due to an aging population), (c) limited productivity growth (it is always limited), (d) no growth in corporate earnings as a share of GDP, and (e) some reduction in reported “earnings” as accounting practices are cleaned up (to match actual sustainable earnings, as fraud-driven losses are inevitably recognized).

Under those assumptions there is no reason to expect stock market returns to grow much at all, for a long time to come, until one or another of those factors shows a secular change.  And (d) and (e) both point to a significant downside risk to stock prices.  Plus it is also quite possible that given the negative outlook, P/E ratios will mean-revert to the 10-12 range from the current 16-20 range as investors demand a higher risk premium.

It is therefore entirely plausible that money placed in the stock market now may decline in value by 50% over a fairly long time horizon (10-20 years) before the factors above improve.

And this does not include either inflationary losses in purchasing power, or tax-code changes.  One must keep in mind that the Federal Budget cannot remain this imbalanced for long, and those seeking to boost tax revenues will be forced to go where the money is…