MILLER’S MUSINGS: February 2009

February 6, 2009

Here I go again, dear readers. Sorry, this is much longer than I intended.

A Quick Summary:

1. We are in the midst of the most damaging, global, economic and financial crisis since the Great Depression. A significant reversal is unlikely until later in 2009.

2. The major precondition for a stabilized economy is that asset prices (residential and commercial real estate, stocks, credit card debt, mortgages, asset backed securities of all kinds, high yield bonds, commodities) must stop declining.

3. The U.S. stock market is not yet fully reflecting an extended period of slow growth and a major decline in corporate profitability. It can be argued that the credit market is a better and safer way to invest in an eventual economic recovery.

4. The financial services industry will be much more regulated and significantly different.

5. The future economy will be different in ways we don’t completely understand at this juncture, but it’s likely that the high level of confidence that led to lower savings and higher leverage over the past twenty years will be absent for a considerable time.

6. Investors will be scarred by their recent losses, so that protecting against the risk of extreme financial/economic outcomes will become a more important part of investors’ future investment objectives.


There is little room for argument about statement #1. The global economy is contracting rapidly and sharply, and will continue to do so for some months ahead.
The U.S. banking system is basically insolvent, even before the full impact of deterioration in credit card debt, commercial real estate debt, growing troubles among the leveraged buyouts of the past few years, and the usual cyclical problems with regular loan portfolios.

The fourth quarter’s GDP offered scant comfort. The 3.8% decline masked the involuntary inventory accumulation that took place, which will have to be undone in the current quarter. Employment statistics are even more dismal than the official numbers, which were buoyed by use of the birth/death model of smaller businesses and a seasonal adjustment factor that was misleading. If there is any good news it is centered in lower oil prices, which effectively are a large tax cut for consumers.

The effects of the expansive monetary policy, unprecedented federal incursion into financial markets, and a hastily and poorly designed stimulus package from Washington are largely unknown, but are likely to give the economy some traction by yearend.

Regarding asset price stabilization as a precondition for a better economy:
The wealth wipe-out that has occurred over the past year-and-a-half has been simply immense; in real terms it’s equal to at least 1.5 times GDP (see footnote #1). And it is continuing. This has not only made the people with shrinking 401k’s feel poorer but it has also hugely raised the ratio of debt to assets throughout the economy, leaving us in a vulnerably overleveraged position. As the economic players attempt to reduce leverage, they not only reduce spending but also further damage the prices of the assets they must sell. It becomes a nasty feedback loop. Only by stabilizing the asset prices can we begin a more orderly servicing and/or reduction of debt (see footnote #2).

I am not sure what the policy response to this problem should be. Perhaps a direct purchase of assets can be implemented. The “bad bank” idea has some merit as a way to stop the attempts to liquefy within the financial system. As things now stand, banks and others who want to raise liquidity and lower leverage sell what they can sell, not what they may wish to sell. This brings down prices all along the quality scale. If a bad bank can take on the lower quality assets at fair prices, it might go a long way toward stemming the feedback loop. But it is also true that regardless of what is done directly, the problem will take time, perhaps a long time, to be resolved. But resolved it must be if we are to see the light at the end of the tunnel.

(I should note here that some people think this is a chicken-egg problem; that asset prices will not stabilize until the economy stops declining. But in the aftermath of immense wealth destruction of the past two years, I strongly believe that spending, both consumer and business spending, will await some confidence that asset price declines are over.)

The financial services industry will be more regulated and significantly different.
The business models of highly leveraged and excessively unregulated banking and investment banking that evolved over the last 20 years, and that facilitated the bubbles and excesses are dead, dead, dead!

The financial engineering that developed and exploited securitization and enabled what Paul McCulley of PIMCO calls the “shadow banking system”, the group of non-deposit, unregulated, unreserved businesses that became major lenders, is in disrepute and non-operative. Structured Investment Vehicles (SIV’s) by which banks were able to make loans and place them off balance sheet financed by commercial paper have disappeared. Hedge funds are undergoing major transformation ranging from complete disappearance to less leveraged models. And private equity firms have been forced into inactivity by the unavailability of credit and the tenuous position of many existing investments. My guess is that losses in private equity will be huge, particularly among investments that were made from 2005 through 2007, the height of the craziness.

Strengthened government regulation of all financial institutions is being demanded and will happen. And the use of credit default swaps as “side bets” on credit quality will be seeing major reforms and changes to govern potential excesses.

The federal government now owns huge swaths of major banks and insurance companies. It’s unclear just how this will play out in addition to specific regulative actions. However, the heavy hand of government will surely be felt in many areas, such as executive compensation and dividend policies.

Also, how about the rating agencies which gave AAA ratings to tranches of various asset backed securities? They must be spanked and reformed to remove the conflict of interest that exists when a borrower pays the fee for a rating. I hope also that they will be made more transparent, so both borrowers and lenders have a clear idea of what the parameters of each rating are.

The financial services industry will never be the same. It took a generation for Wall Street to come back after the depression. I know. My father was a broker who lost it all in the depression, and had several years of essentially zero income while I was a child. Depression phobia carried well into the postwar period. It may disappear sooner this time, but that’s far from a safe bet. Scars are very durable things, and many players have been very badly wounded (see footnote #3).

I don’t mean by this that the economy will never recover, or that prices in the stock and credit markets will be forever depressed. I do mean that the willingness to use financial leverage and the willingness of lenders to facilitate the use of leverage will be moderate and careful for some years to come. As a result, economic growth will be less credit reliant and probably less buoyant. It likely will create a less profitable environment for business in general and financial services firms in particular.

Let me turn now to the question of whether the U.S. stock market is adequately discounting slow economic growth and lower corporate profitability.
U.S. stocks must be judged in relation to a concept of normal earnings, or, if you prefer, basic earning power. At cyclical peaks, stocks should reflect skepticism about the sustainability of earnings, and at cyclical troughs they should reflect an eventual recovery to a more “normal” earnings level. To be sure, things don’t always work that neatly. It wasn’t many months ago that professional investors were forecasting over $90 earnings for the S&P 500 in 2009. By late fall the forecast had dropped to about $60. Now the forecasts are running under $50. At the October 2007 peak for the S&P, and using the eventual peak earnings forecast of $90, the price-earnings ratio that investors then thought they were paying was 17.2. At the current price, using a 2009 forecast of $45, the PE ratio they actually paid was 34.4.

To make matters more difficult, the sharp decline in both the price and the earnings of the S&P 500 is importantly due to the collapse of earnings and prices in the financial sector. Thus, the S&P 500 may be a misleading way to look at the whole stock market’s valuation level. So let’s look at some other methodologies.

Value Line calculates PE ratios by looking six months forward and six months backward, then finding the median PE among all stocks with earnings. The last issue of Value Line shows the median PE to be 11.3 and the median yield to be 3.3%. The highest level reached by these metrics was in July 2007, when the median PE was 19.7 and the median yield was 1.6%. Perhaps more relevant is where they were at the lows of 2002: 14.1 and 2.4% respectively. My guess is that the six month forward earnings forecast of Value Line is still too high and the resulting PE too low, but the point is that by this metric stocks are undeniably much, much cheaper than at the past market peaks and also much cheaper than at the valuation low of 2002.

So what is cheap and what is expensive? And where do interest rates enter the equation? Today’s record low risk-free rates should, other things being equal, raise the valuation of future earnings. But, is that necessarily so when the record low rates are centered and concentrated in the Treasury market and are the result not only of monetary policy but also plain fear of other assets? Clearly, investors are struggling to gain some idea of “normalcy”.

There are many methods to estimate normal earning power. One that has proved very valuable as a gauge of overvaluation and undervaluation is to average the previous ten years of real earnings and divide this average into an inflation-adjusted S&P 500 price. This is what Robert Shiller of Yale does. I like this methodology because it compensates for the tendency for profit margins to revert to a longer term mean. That may be especially true today, following a period of record high profit margins.

Below is the Shiller chart showing data since 1881. (Click on chart to enlarge.)

shiller_pe_9571_image0011

Source: Robert Shiller

You will see that the current Shiller PE of 15.3 is slightly below the long-term average and median levels of the past. Reasonable, but not what one would call very cheap. And notice that in past economic contractions it became considerably cheaper.

Another method is to look at past relationships of profitability and valuation. What do past relationships of net profit margins to the valuation of sales show us?

It should, of course, be true that the price investors pay for sales (P/S) is dependent on the earnings derived from those sales (E/S, or profit margin). If we divide the P/S by E/S, the result is P/E, the price-earnings ratio.

In order to minimize distortions from the disaster in financials, we looked at the S&P 400 Industrials, an index that eliminates financial services. The average net profit margin for industrials has been about 5.0% over the past 40 years. The problem is that recent margins have been at the high end of longer term experience. Consider these numbers:

S&P 400 Industrials Average Quarterly Net Profit Margin
1968-2008 4.98%
1980-2008 4.86
1995-2008 5.63
2003-2008 6.70

Of 25 quarters with margins >6%, 17 occurred between 3/31/2004 and 12/31/2008. Of 159 quarters examined, 7 had margins > 7%. All have been since 9/30/2005. Did the record margins of recent years derive importantly from an economy whose growth was largely unregulated, consumer driven, wealth driven, credit driven, and trade driven? I think so. Can we use recent margins as a guide to normalcy? I don’t think so.

Arbitrarily, (is there another way?), I averaged the margins of 1996-2008 and those of 1980-2008 to arrive at my assumption of 5.25% as “normal”.

Please click on each chart to see a larger view

image001

Quarterly Net Profit Margin, December 1968 - September 2008

image002

Monthly Price to Sales, January 1969 - January 2009

Source: StockVal

The current price-to-sales ratio is about .80. With a P/S of .80 and a normal margin of 5.25%, the P/E on normalized earnings is (.08/.0525) or 15.2. Interestingly, this is close to the Shiller P/E. Again we see a reasonable valuation but certainly not a cheap one. From the early 1980’s to 1990 the P/S ranged from .6 to .8, producing a P/E range of 10.7 to 14.3. There is no logical way to dismiss the possibility of that sort of cheapness appearing again. A P/S of .6, should it occur, produces a P/E of 11.4. Now we are getting interestingly cheap. But only after reaching a P/S down 25% from here!

Bottom line? Stocks are reasonably priced compared to past history but not cheap. And I still question whether valuations reflect either the current crisis or an extended period of low growth and lower profitability than the experience of the past few years.

Do credit markets offer a safer alternative to invest in anticipation of an economic recovery?
The credit markets, while having shown some improvement from the complete shutdown of
October and early November, are still disorderly and frightened. Junk bond yields are at 18%, and investment grade corporates are at or over 7%. Financing in the debt markets ranges from impossible to very expensive.

Interestingly, the credit market seems to be offering attractive possibilities for investors. It is inconceivable to me that the stock market can rise significantly unless confidence returns to the fixed-income market. Both markets are suffering from the same set of worries about what a bad economy will do to corporate cash flows. In the meantime, the bond holder gets paid rather handsomely. True, defaults are and will continue to rise for some time ahead, but significantly higher default rates are well baked into current prices. If one assumes a recovery rate of 30%, and uses 5.0% as the current default-free rate, one can argue that current prices are forecasting a default rate of 13% and a loss rate of 9-10%. Is that a conservative calculation? In this environment we can’t be sure. Are lower quality bonds and other fixed-income instruments greater bargains than stocks? I believe the two asset classes will move together when markets improve. If markets continue to decline, then bonds offer a better risk-return relationship than stocks.

What will the future global economy look like?
I think that it’s much too early to begin answering that question except in broad generalities.. There are some rather glib and unoriginal answers such as: It will probably be less U.S.-centric and more reliant on the developing world and China. Growth will probably be importantly concentrated on infrastructure building, especially communications and transportation, on energy production, including non-fossil fuel development, and a continuously expanding reach of electronics into all aspects of life.

But these same old answers, while perhaps accurate, are getting tired and don’t capture what the essence of the global economy might be. The past twenty years have been extraordinary in terms of trade expansion with minimal government hindrances; in terms of free markets spreading to major parts of the world that had not seen them before; and in terms of peaceful relationships between and among the larger players. These conditions fed confidence in economic growth and stability. That confidence was at the core of willingness to reduce savings rates and take on leverage, for all players from businesses to consumers. Will it still be in place after the economic crisis? I very seriously doubt it.

An additional influence on spending/savings that is fast approaching is the funding of entitlements or, which seems unlikely, the revision/reduction of them. Funding these obligations will entail a substantial “wedge” of personal and corporate taxes well in excess of what we have now. In an important sense it will be a reduction of the future estimated real net worth of asset owners.

Meanwhile, governments have entered the global economic scene in a big way. That alone will make the future very different. The U.S. continues to take extraordinary monetary and fiscal actions. Federal deficits will be off the charts. Unwinding the array of guarantees and commitments, and facing up to reforming entitlements must eventually take place lest we face a seriously inflationary situation. Will we have the political courage to do this as the economy recovers? An open question at best.

Also the global economic cooperation that marked the past decade will erode. Can increased protectionism be avoided? Here is a news item from Bloomberg that appeared as this is written:

“As world leaders in Davos called last week for international cooperation to tackle the financial and economic crisis, businessmen were complaining that the stress is only aggravating national divides.”

Finally, let’s take a look at how investors may view risk after the current crisis.

Relevant to this question and for those of you who have not read Nassim Teleb’s The Black Swan, or Benoit Mandelbrot’s (Mis)Behavior of Markets, I enthusiastically recommend them as good for both their intellectual and practical content.

Teleb, a disciple of Mandelbrot, says, “History does not crawl, it jumps.” The really important events are rare and unpredictable. Those who design policies and objectives based on the assumption that the world’s developments are distributed in a normal bell curve are doomed for disaster. They will miss the really big events and changes.

Mandelbrot, a former IBM researcher, mathematician, Yale professor, and inventor of fractal geometry has similar conclusions. The problem, according to Mandelbrot, is not only that using the bell curve leads to an underestimation of volatility and understates the odds of loss. It also leads to understating, perhaps substantially, the magnitude of loss.

Modern Portfolio Theory (MPT), the Capital Asset Pricing Model (CAPM), and the Black-Scholes option model have at their heart assumptions that prices move continuously (see footnote #4) , and the bell curve distribution of past returns is a good picture of probable returns that can be used to project future risks.

I don’t want to get too mathematical here, in fact, I doubt that I am capable of that, but the bottom line is that such tools as the efficient frontier analysis used widely for establishing asset allocation guidelines are, according to Mandelbrot, deeply flawed.

Asset allocations based on bell curve distributions of past distributions is flawed in yet another way. They are often used without reference to the starting point. It’s one thing to say that long-term returns from an asset have been, say, 10%, with a standard deviation of 10%, and another to then assume that the asset will behave the same way during the coming ten or twenty years, or more to the point, during the tenure of the current composition of an investment committee, without reference to the valuation of the asset class at the starting point.

The bad markets of the mid-1970’s, the 1987 market crash, the bursting of the dot-com bubble, and the events of the past two years lead me to bend a sympathetic ear to Messrs. Teleb and Mandelbrot. And even if investors never read or know what these guys say, they now know in their guts that investing is far riskier than they believed.

Henceforth, I think that portfolio design will be significantly adjusted in attempts to accommodate the possibility of Black Swans. Truncating or limiting the downside will be an increasingly popular objective. The financial consequences of a Black Swan are unacceptable to most investors.

Certainly more attention will be paid to absolute return than in the more recent past, when “relative performance madness” engulfed most institutions and individuals. But whether that means hedge-fund-like operations will become more common is not clear. More market timing? Perhaps. It certainly has been completely out of favor among institutions in recent years because they thought it was a fruitless exercise. How about more emphasis on current income? Ingenious minds are, I’m sure, already looking for something that seems to work in stabilizing returns (see footnote #5).

The one thing that seems certain to me is that in setting asset allocations, the dynamics of investment committees and the way individuals view risk will be very different from the past twenty years.

Footnotes:

1. GMO’s Jeremy Grantham discusses the arithmetic of wealth and debt at this link: http://www.gmo.com/websitecontent/JGLetter_4Q08.pdf

2. For an interesting discussion of the importance of asset price stabilization and see Pimco’s Bill Gross’s February comments click here

3. See a fascinating paper by Stanford and Berkeley professors, Nagel and Malmendier, Depression Babies: Do Macroeconomic Experiences Affect Risk Taking. Click here to view this paper. Here is part of their conclusion:
“Our results show that the stock returns and inflation experienced over the course of an individuals’ life have a significant effect on the willingness to take financial risks. Individuals that have experienced high stock market returns report lower aversion to financial risks, are more likely to participate in the stock market, and allocate a higher proportion of their liquid asset portfolio to risky assets. Individuals that have experienced high inflation invest a lower proportion of their non-stock liquid assets in bonds. Differences between cohorts in their life-course stock return and inflation experiences appear to strongly predict heterogeneity in willingness to bear financial risks at a given point in time and controlling for age, wealth, income, and other demographics. While individuals put somewhat more weight on recent stock market returns and inflation than on more distant realizations, the impact fades only slowly with time. According to our estimates, even experiences several decades ago still have some impact on current risk-taking of older households.”

4. That is, prices move in a way similar to Brownian Motion, the motion of molecules in a uniformly warm medium. A simple plot of the standard deviations of changes in the Dow-Jones Average in vs. standard deviations of changes in Brownian Motion, clearly reveal huge differences in terms of frequency of extreme moves . See pp 87-93, The (Mis)Behavior of Markets

5. For those interested in how investors tried to protect from the downside after the depression, read about so-called formula plans, originated at Yale and Vassar in 1938. The plans automatically reallocated assets by a formula that related to the level of a stock market average. These plans were hopelessly outdated by the post-war bull market. Here is a link to the Third Edition of Graham and Dodd’s, Security Analysis that discusses formula timing: click here.

cropped-img_2147.jpg