A Most Important Rule
John P. Hussman, Ph.D., Hussman Funds
A few weeks ago, I noted that the return/risk profiles that we identify for stocks, bonds and precious metals had shifted abruptly. Since then, a decline in bond prices has modestly improved expected returns in bonds, but not yet sufficiently to warrant an extension of our durations. Precious metals have become more overbought, and while we are sympathetic to the long-term thesis for gold, intermediate term risks are now elevated. Finally, we have observed a further deterioration in market conditions for stocks.
Since the early 1980's, I've examined and tested an enormous range of analytical techniques and investment rules, including various versions of "Don't fight the Fed," and "Don't fight the tape." If I had to choose between only these two, "Don't fight the tape" would win, hands down, as Fed-based investment approaches typically endure excruciating drawdowns - even those that succeed in slightly improving long-term returns. That said, there are numerous refinements that perform far better than these simple rules-of-thumb; especially those that reflect broader considerations such as valuation, sentiment, economic pressures, yield trends, market internals, and so forth. If I had to pull a single rule from these combinations, one particular admonition - "Don't take risk in overvalued, overbought, overbullish, rising-yield environments" - is one of the single most important in terms of avoiding major, sometimes catastrophic losses.
Our investment stance is not defensive here based on our concerns with the appropriateness or legality of various Fed actions, nor based on our concerns about the underlying state of balance sheets in the financial sector, nor based on the elevated vulnerability of economic outcomes to small shocks. We are defensive because stocks are presently overvalued, overbought, and overbullish, and this combination is joined by rising yield pressures despite Fed actions.
The lessons that we've learned over the past two years do not include disregarding overvalued, overbought, overbullish, rising-yield conditions. Nor are we convinced that it would have been appropriate, except in pure hindsight, to ignore the evidence of past credit crises in the U.S. and internationally. It is undoubtedly true that it would have been more rewarding, in hindsight, to treat the most recent economic cycle as a standard, run-of-the-mill recession. I remain convinced that this is not a reasonable assumption, and that it would have been improper to ignore other post-credit crisis data given that the past two years were "out of sample" from relative to post-war experience.
At the same time, I believe that we've learned a great deal that will be of long-term value. Most importantly, the experience of recent years has fueled research that recently convinced us to expand the range of Market Climates we identify, defining their robustness and measuring their "model risk" by testing them in multiple subsets of historical data. As I noted last week, the first outcome that shareholders are likely to observe will be an increased tendency to accept moderate, transitory exposures to market fluctuations. Large shifts in market exposure will most likely still require a large shift in valuations or a broad combination of improved economic fundamentals, sentiment conditions, yield pressures, and other factors.
While we do keep certain aspects of our approach proprietary, somewhat more detail may be helpful. Generally speaking, our investment exposures are proportional to the expected return that we can expect per unit of risk in any given Climate. Each Climate is built from a set of criteria, and is therefore sensitive to the particular set of criteria being used. The challenge in broadening the range of Market Climates is that we require those Climates to be "robust" across time, and not highly sensitive to the way they are defined. Accordingly, it's useful to measure not only the average expected return and variability of returns (risk) in a given Climate, but also the variability of those outcomes across multiple criteria and multiple subsets of data.
Consider the whole set of observable conditions on any given date in market history, which might include valuations, market action, interest rate behavior, economic evidence and so forth. There are billions of ways that you could potentially combine this data. If you would assign very different expected returns depending on the specific set of criteria you examine, you've got a great deal of model risk. In contrast, if a given set of market conditions results in uniformly positive or uniformly negative expected returns, without much sensitivity to the specific way you define the "Market Climate" or the subset of data you are examining, your return expectations are more robust, and it's reasonable to have a greater sensitivity to the expected returns on that basis.
Of course, at the heart of everything, we still depend on strong valuation models, reliable indicators, and the like, but we've extended our approach by broadening the range of Climates we identify, while explicitly estimating the associated "model risk" that we face at each point in time. Again, the main practical effect will be an increased tendency to accept moderate, transitory exposure to market fluctuations on a more frequent basis than we have in recent years.
I recognize that investors are eager to move on to the thesis of sustained economic recovery, with no need for any risk management at all. However, it appears unwise for investors to rest their financial security on faith in a recovery that relies on the government running a deficit of 8.5% of GDP, simply to keep the existing 6.3% gap between actual and potential GDP from widening further. It appears equally unwise to rely on Fed purchases of Treasury bonds to sustain ever greater exposure of investors to risk, when the creation of financial bubbles does nothing to increase the underlying cash flows deliverable by the securities that are increasing in price.
While we have succeeded over time in outperforming our respective benchmarks with smaller periodic losses than a passive investment approach, we have certainly been penalized for not taking "enough" risk in a world where formal U.S. policy has been singularly focused on bailing out private lenders at public expense while the Fed's "more, bigger" policies aim at encouraging stock market speculation in hopes of creating a wealth effect. The frantic "risk on" attitude of investors over the past several weeks has been difficult for us. Given that a large proportion of our stocks lean toward what we consider "high quality" - consistent revenue growth, stable profit margins, and an emphasis on sound balance sheets - the embrace of leveraged, cyclical companies by investors in recent weeks has made our holdings appear to have far smaller "betas" than is typically the case over longer horizons. This has made market advances somewhat uncomfortable, while market declines have been more profitable than they "should" be. I am convinced that this is short-run behavior, but we continue to carefully modify our hedges in response.
A Trifecta of Reckless Central Bankers
I continue to view Bernanke's apparent objective for QE2 - to create a "wealth effect" by encouraging speculation in risk assets - to be dangerously misguided. Historically, the elasticity of GDP to changes in the stock market is on the order of 0.03 to 0.05, and is transitory at that. In plain English, this means that even large changes in the value of the stock market do not translate well into changes in GDP. This is because consumers correctly consume on the basis of what they see as their "permanent income," and are well aware that changes in volatile assets tend to be transitory when they are not accompanied by growth in real output and incomes. Bernanke is not thinking as an economist in this regard. He is thinking like a witch doctor calling on animal spirits (ooh, eee, ooh-aah-aah, ting, tang, walla-walla bing-bang).
With regard to Sunday's "60 Minutes" piece, I sometimes enjoy seeing Barbara Walters do a celebrity puff piece, but I expected more from a show on investigative journalism. Proper questions might have included, "Chairman Bernanke, how do you justify the fact that all of Bear Stearns' bondholders stand to get 100% of their money, plus interest, while at the same time, the Fed still holds $30 billion dollars worth of Bear Stearns' questionable MBS junk in an off-balance sheet shell company called Maiden Lane, which you justify by appealing Section 13-3 of the Federal Reserve Act, despite that this section deals explicitly with "discounting" - which everywhere else within the meaning of the Act allows nothing but a short-term check-cashing function, in nearly every case for paper of less than 90 days in maturity?" Nothing about Fannie or Freddie, or the fact that Treasury yields have increased since QE2 was announced.
