Do you think that the crash is over, as certain former bears do? This question arises as we have breached the first downside target, of Dow 7000, based on my proprietary investment value model, that was first published in thestreet.com October 24, 2007. It was less a forecast than an evaluation. The Dow has now vindicated this model by reaching "fair value," as one would expect from a simple definition. Does that represent a base for a new bull market? Or is it just one more stop to the nether regions?
To understand my model, note that a stock can be analyzed as a combination of a bond plus a call option. My proprietary investment value metric for a stock is book value plus ten times dividends. That is a Ben Graham like construct that treats stocks almost like bonds, and gives no effect to growth over and above the pro rata return from the reinvestment of retained earnings. On the other hand, many investors prize stocks, particularly tech stocks, for their "optionality," the hypothetical ability to generate "positive surprises" over and above what economic theory would support. At bottom, the belief in the new economy was a belief in "optionality," that random positive events that occur from time to time, and did so with particular frequency in the 1990s, will become a recurring fixture of the economic landscape.
But such a process can also work in reverse, as it has recently. We are now experiencing what my colleague Robert Marcin calls the Great Unwind. A turbocharged economy is most likely to become "unstuck" when the conditions that initially favored it no longer exist. When this happens, an economy can grow as much below trend as it was formerly above trend, a fact that is likely to be reflected in the financial markets. History is not very encouraging on this score. In past downturns, such as those of 1932 and 1974, the Dow troughed at one half of my investment value metric, reflecting then-prevailing investor beliefs for negative optionality; that the economy will be worse than normal economic forces would dictate. With investment value at 7000 (actually a rounded version of 6600) on the Dow, half of that would be 3300. And during the 1930s, this metric actually fell, meaning that the "ultimate" low could be half of a number lower than 6600.
So having completed a first downleg, the market is now working on a second one. And this would be fully reflective of economic forces. For instance, financial earnings used to represent some 40% earnings (if you count the financing arms of some old line "industrial" companies such as General Electric and General Motors). Thus, they made up $32 of what used to be normalized S& P earnings of $80. But most of those financial earnings have disappeared. That, by itself, would take the S&P earnings into the $50s.. But how many of those non-financial earnings (of $48) were tied to the finance bubbles such as the homebuilding and the "housing ATM?" At least 10%, or around $5, and that is being conservative. Thus, normalized S&P earnings are likely to be no more $50 a share, if that.
The problem comes at payback time. For instance, much of the borrowing was tied to the housing market, on the bogus theory that houses could be made twice as valuable (as a multiple of rent) as they were for all of American history if prices could be kept on steady incline. The problem was that valuations collapsed when house prices fell, or even failed to rise, bringing down the market with it. To make up the shortfall, the U.S. economy now has to consume less than it produces, for a time. But the formerly virtuous circle became a vicious circle when falling prices (and consumption) led to falling production in a self-reinforcing process of the kind best described by George Soros in the Alchemy of Finance. This is a process called underabsorption, which in its strongest form, is called disintermediation. When a major part of the economy becomes "unstuck, the rest of it doesn't merely go into retrograde. It has to fall apart also to keep pace.
But I can live with $50 trough earnings, say many. And at historical multiple of 14-16 times trough earnings, the S&P should stop its downside in the 700-800 range. But the point is, they're not trough earnings, they are the "new normal." And in the current "slow" (zero or worse) growth environment, a trough P/E of 6-8 times earnings is more likely. Put another way, we are about to get the worst of all worlds; below trend earnings, below trend growth from a depressed base, and below trend P/E, after having gotten the best of all worlds, astronomical P/Es on above-trend and rapidly growing earnings, about a decade ago. Warren Buffett now agrees, saying that we will get "almost the worst of all possible worlds..."
The bears-turned-bulls have taken the latter stance because the market now reflects at least a severe recession. One such commentator likened the recent market to 1938-1939, and feels that the latter represents a bottom. But the 1930s bottom was 1932, not 1939, which is to say that the market probably has further to fall. Having correctly dodged the "overvaluation" bullet earlier, the new bulls pin their hopes on the prospect that the current market represents everything bad short of the 1930s Depression. Unlike us, they aren't willing to grasp the nettle that the current crisis will likely be as bad as anything including the Great Depression.
A Stock Market Rebound Closely Linked with Economic Data Surprises
by William Hester, CFA - April, 2009
There are several ways to interpret the economic data in March, most of which came in above what economists were expecting. Some analysts concluded that the worst is over for the economy, and a rebound is ahead. Others suggested that the economy is still contracting, but at a slower rate for now. In any case, economists have overestimated the economy's rate of contraction lately. The rebound in the stock market has been at least partially fueled by economic data that consistently came in better than expected last month. Some part of this rally is likely relying on the continuation of these "positive" surprises.
To track the trends in economic performance, we keep an ongoing tally of how data is announced relative to expectations – a method of analysis originally inspired by Bridgewater Advisors . Economic data that surpasses expectations gets added to a 3-month running total. Data that comes in weaker than expected gets subtracted. A rising line means that economic data is generally coming in above expectations, while a falling line means that the data has disappointed. A descending line could be the result of an economy that is not expanding as quickly as economists predict or – like in 2008 – it could be the result of an economy that is contracting at a faster rate than expected. In the first graph, and the others below, I've isolated only the data that measures the growth in the economy, leaving out measures that track the rate of inflation and sentiment. The first chart below shows the surprise line for growth-related economic data since last August, just prior to the passing of the Emergency Economic Stabilization Act, from which the first version of the TARP was born.
There's nothing quite like pointing out how bad a shape the economy is in to get people acting like the economy is in bad shape. During the early part of last summer the economy was actually holding up better then what was generally expected. But during the final quarter of last year, the economic surprise line (in blue) collapsed. Data persistently came in below expectations, which created the steepest drop in the line tracking economic performance versus expectations in the available data.
The red line in the graph above tracks the S&P 500 Index and it shows that stocks have recently closely tracked the trend in data surprises. The market fell along with the deteriorating surprise line last year, rallied slightly prior to improved news in December, and then rolled over again as the news weakened versus expectations in late January. In March the market rebounded along with a more pronounced persistence in favorable economic news versus expectations.
The data released in March was better (or less negative) than expected on a number of fronts. The slowdown in spending eased, there was temporary relief in the new and existing homes sales data, and sentiment measures mostly halted their steep decent of recent months. But while much of the data was surprising relative to expectations, it's difficult to point to any piece of data that was surprisingly strong (outside of some of the volatile data series like, for example, durable goods). New homes sold at an annual rate of 337 thousand versus 300 thousand (and a peak of 1.4 million). GDP was revised to -6.3 percent versus an estimate of -6.6 percent.
Much of the excitement in the stock market – at least that is related to the current performance of the economy - seems to be centered on an economy that is performing less badly than expected. The risks here seem to be that if the trends in data surprises change, so could investor's attitudes toward stocks that are currently overbought on a number of measures.
There are a couple of reasons why the trend in the rate of data surprises could change. The first is that trends in economic surprises are very prone to reversals. The chart below shows a longer-term picture of the changes in the trends in economic data surprises. The one thing that stands out looking at the graph is that the trends in surprises often reverse abruptly. When the estimates of economists fall behind in an expanding economy - underestimating its strength - expectations are adjusted upward. These estimates eventually become too optimistic. The same can be said of an economy that is contracting more quickly than expected. And the data shows that the more pronounced their forecast errors – the more abruptly economists begin to overestimate the economy's recent trend.
Another reason why the economic news may begin to disappoint at some point is that recoveries rarely proceed smoothly. The trends in month-to-month and quarter-to-quarter data tend to lurch forward and backward as the economy regains its footing (and at times, like in 1982, the economy can fall right back into recession).
One recent example of this was in 2002, which is shown in the graph below. The trends in economic data versus expectations were persistently better than expected from late 2001 as the economy emerged from recession that year through late spring of 2002. The S&P 500 surged by more than 20% from its 2001 low as the economy began to regain its footing and offer up positive data surprises. But by the summer of 2002 the rebound proved not robust enough when compared with economist's expectations, and the surprise line rolled over. With stocks not yet at valuation levels that were attractive to investors, the S&P plunged along with the data surprise line.
It's important to note that this was during a period where the economy was, in hindsight, no longer in recession, and where there were many measures that showed the economy was growing again. But the market was still tripped up at least partly because expectations had moved ahead of the economic recovery. The bear market remained unfinished, and stocks fell to new lows. This may turn out to be an important risk over the next couple of months. The economic data is certain to be uneven, which in turn may cause investors to begin to question whether an economic recovery is really at hand. Risks will likely be higher at points where the market is overbought.
Investors tend to punish economic disappointments much more strongly during bear markets than during bull markets. The graph below, which shows the S&P 500 and the surprise line from 1998 to 2002, highlights this tendency.
Although it's just a portion of one cycle – the late stages of an expansion and a mild recession, it's worth noting how stocks performed in response to economic data surprises. In the last part of the 1990's bull market, a rising economic data surprise line mostly fueled rallies. Data worse than expected weighed on performance – often causing shallow declines like in 1998 and late 1999. Conversely, during the 2000-2002 bear market, disappointing economic data coincided with steep declines in equity prices, while positive surprises usually eased the market's deterioration.
These trends were also evident during the market's advance from 2003 through 2007, but were somewhat less dependable. During that period, the trends in the surprise data were shorter and more variable than the market's slow, persistent advance. Since last summer, the correlation between the two has tightened considerably. In fact, the correlation between the S&P 500 and the data surprise line has climbed above .80, implying that investors are keeping a close eye on how data comes in relative to expectations.
If the high correlation between stock prices and data surprises holds, the recent rally in stocks might be tested. Even if the economy has bottomed, it's very likely that the eventual recovery will prove to be uneven, causing the flow of positive surprises to be uneven. During these periods, the risks to stocks will be greatest when the market is overbought and investors have priced in high expectations of positive data surprises continuing.
No comments:
Post a Comment