Waiting For The Fed

While the whole of Austrian economics is too much to be covered in this article, there are some quick basics that will make the concepts easier to understand. The Austrian model starts with a market that is thoroughly free from government intervention, save for the protection of life, liberty, and property. Any form of intervention in the free market, states the Austrian approach, is an artificial obstruction (or accelerant) that will eventually lead to a misallocation of resources.

A simple example of this is in farm subsidy programs. When governments decide that agricultural crop prices have fallen too low for farmers to harvest profitably, there has always been a temptation for the government to subsidize crop production. As critics of the farm bill that was working its way through the US Congress in 2002 noted in a Washington Post editorial: "The basic flaw of the bill is that the lion's share of its money goes to subsidize farm output, a policy that stimulates supply, drives down prices, and hurts the farmers it is meant to help." Stephen Moore of Club for Growth, an organization that lobbies for limited government and low taxes, blamed this "price-support system" for rewarding "the largest producers, because the more bushels you produce, the bigger your check from Uncle Sam."

Unfortunately, like with most government handouts, it is difficult to wean recipients from the state's largesse. And so, market interventions tend to be deployed too soon and repealed too late. (Something similar, you might recall, was said of the Fed funds rate policy from August 1999 to December 2001. More on this later.) Those who are prepared for the changes in policy are often in a position to take great advantage. Unfortunately, those who are not prepared — or cannot prepare themselves — are often hit hard.

WHY FOLLOW THE FED?

There is no greater form of government intervention in the economy than the intervention of the Federal Reserve Board. While officially responsible for merely setting reserve requirements and the discount rate, the Fed's control of monetary policy is such that its influence over the price of money (that is, interest rates) is unmatched by any other government agency. It is true that ultimately the market system sets interest rates. But the ability of the Federal Reserve Board to provide powerful incentives for rates to move in one direction or the other should never be underestimated.

As such, stock market timing models based around both this notion of Austrian economics (government intervention in the marketplace as artificial) and the role of the Federal Reserve Board in directing interest rates have been among the more popular ways for economists, investors, analysts, and traders to keep track of when stocks might be more or less likely to outperform. Because stocks (and the economy in general) tend to do well when interest rates are either low or stable, having a sense of how interest rates are moving can be invaluable for those participating in the financial markets. As Martin Zweig writes in his classic, Winning On Wall Street:

In the stock market, as with horse racing, money makes the mare go. Monetary conditions exert an enormous influence on stock prices. Indeed, the monetary climate — primarily the trend in interest rates and Federal Reserve policy — is the dominant factor in determining the stock market's major direction.

The role of the Federal Reserve Board and of the current trends in interest rates both take on added importance for investors in 2002. After the Fed's historic campaign to lower interest rates — a campaign that began in January 2001 and accelerated after the terrorist attacks in New York City and Washington, DC, in September 2001 — economists and investors find themselves nervously waiting for the first signal that the Federal Reserve Board will begin bringing the fed funds rate back up from its exceptional, sub-2% lows. "Nervously," because the track record for stocks once the Fed begins "tightening," or raising interest rates, is especially poor. In its March 25th issue, Barron's cited 20 years of data compiled by Morgan Stanley that suggested that none of the main industry sectors — from technology and basic materials to utilities and financials — do well in any of the first three, six, or 12 months after the Fed begins to draw interest rates back up.

MARKET TIMING AND TIGHT MONEY

So what happens to those monetary policy­based stock market timing models when the Fed begins to tighten? Do they get investors and traders out of the market altogether? Or do they attempt to pick through the periods of underperformance for the few opportunities when stocks, inevitably if temporarily, will rally? Here, we'll look at how a few monetary policy­based stock market timing models have fared in the two and a half years since the Federal Reserve Board began lowering rates, and try to anticipate how some of these timing models might perform over the balance of 2002.

One of the first monetary timing models we will look at comes from Boucher's book The Hedge Fund Edge. This particular model, according to Boucher, has featured an 18.8% annual rate of return since March 1947. Simply stated, the model calls for buying the Standard & Poor's 500 when the year-to-year rate of change in the yield of the three-month Treasury bill is less than or equal to 6%. Again, note that the model is referring not to the nominal yield, but to the rate of change year-to-year in the yield. The theory behind this model is as stated earlier: Stocks tend to outperform when short-term interest rates are not rising (or, as we will see, falling) swiftly.

Boucher's survey stops at the end of 1997, with a long position opened in October 1995 that remained open as of December 1997 and was up over 66% for that period. This, incidentally, is the second-largest percentage gain in the history of the model, the largest being in excess of 94% from buying the S&P 500 in November 1984 and selling the S&P 500 in August 1987 (the model is long only). Extending the study into 1998 and beyond gives us the results displayed in Figure 1.

Figure 1: Boucher's monetary timing model.

One of the most remarkable aspects of this extended study is the performance of the model from October 1995 to September 1998, an 84% return. But also remarkable is how it kept investors and traders out of the market for the entirety of 2000, a year in which the S&P 500 was down 10%. The model suggested buying the S&P 500 in January 2001, but this trade was closed out for a small loss after less than a month (the S&P 500 was down 13% in 2001).

Finally, it is interesting how this model treated the old canard "Don't fight the Fed." This truism has for years suggested that when the Fed starts lowering rates, it is time to stop asking questions and start buying stocks. Conversely, when the Fed begins to raise rates (referred to in good times as "taking away the punch bowl"), it is usually time to head for the hills. As Figure 1 shows, the model gave a buy signal at the same time that the Fed began lowering the fed funds rate in January 2001, but got back out of the market a month later with a 2% loss. Meanwhile those investors who held on, who put their faith in the Federal Reserve Board's ability to engineer a liquidity-based rally in stocks, were punished by an S&P 500 that lost 13% for the year.

PRIME RATE TIMING

The second monetary policy­based stock market timing model we will examine and update is one provided by Martin Zweig. This model is based on changes in the prime rate that interest rate banks charge their most creditworthy customers. Using the prime rate in a stock market timing model has a variety of advantages, according to Zweig:

The prime rate usually falls only after a drop in Federal funds rates or in the yields on certificates of deposit or commercial paper. But that's exactly what an investor wants to keep his eye on, because changes in interest rates generally lead changes in the stock market. An interest rate that moves a little behind other interest rates can often mark just that point when stocks finally begin to respond to the changes in rates.

Zweig examined the performance of this model over the 30-year period from 1954 to 1984 and discovered an annualized return of 17.2% (both long and short). The rules for this timing model contain four conditions, two for buys and two for sells. With regard to buys, the model calls for buying the S&P 500 on any initial drop in the prime rate if the prime rate is less than 8%. The other buy condition comes when the prime rate is already 8% or higher. In this case, the buy signal comes either after the second of two cuts or after a full 1% drop in the prime rate.

As for sell signals, the model calls for selling the S&P 500 on any initial rise of the prime rate if the rate is already 8% or more. The second type of sell signal comes when the prime rate is already less than 8%. When this is the case, the sell signal comes after either the second of two rate hikes or on a full 1% increase.

Even with the final losing trade (which is still open, actually), Zweig's Prime Rate Indicator would have turned a $10,000 investment in October 1984 into $19,300 by May 2002, a gain of 93% after 16 years (Figure 2). It is telling that the short end of the trades between 1984­2002 tended to be significant losers. Zweig's own survey from 1954 to 1984 revealed that the model's long trades tended to be correct 83% of the time, compared to 59% for the short trades. So there perhaps should be little surprise that the long trades in the compressed time frame presented in Figure 2 were significantly more successful than the short trades.

Figure 2: Zweig's model.

MAKING MONETARY MODELS WORK

It is impossible to determine where the S&P 500 will be at the end of the summer of 2002, or in October, or at year's end. However, if we believe in the robustness of the monetary policy stock market timing models presented, then they can be used to determine when the models will either issue new buy signals for the S&P 500, or turn currently losing signals into winning ones.

This task may be harder for the second model, which relies on the movement of the prime rate. In many ways, the prime rate model is a classic "Don't Fight the Fed" liquidity model that posits strong gains in stocks when interest rates are low or are in the process of being lowered. The prime rate model gave a buy signal one month after the Federal Reserve Board began lowering the fed funds rate. Unfortunately, 18 months and 21 losing percentage points later, the model has yet to produce the sort of equity gains that usually accompany a "loose" monetary policy. A lower prime rate might support the buy signal from this model. But as long as the S&P 500 fails to respond strongly to the liquidity provided by the Fed, the prime rate model will be helpless to provide a winning signal.

Boucher's model is more interesting in this regard. One of the most significant differences between the two models is that where the prime rate model sees the lowering of the rate as a qualified good and the raising of the rate as a qualified bad, Boucher's three-month Treasury bill model is concerned with the rate of change — up or down. From the text that accompanies Boucher's description of the model ("when T-bill rates aren't rising rapidly, stocks rise nicely"), it is clear that Boucher tends toward the "Don't Fight the Fed" philosophy — even as his model successfully does just that by staying out of the market while the Fed is pumping money into the economy.

What would cause Boucher's model to issue a buy signal after getting out of the market altogether in February 2001? Believe it or not, rising short-term interest rates. This is because Boucher's model would look favorably on a T-bill rate in 2002 that was less than 6% different from the T-bill rate at the same point in 2001. For example, the three-month T-bill yield in summer 2001 hovered between 3.7% and 3.4%. A buy signal would be initiated if, during the summer of 2002, three-month T-bills yielded something in the neighborhood of 3.9% on the high end, with 3.2% representing a minimal (less than 6%) year-to-year change. Of course, to get to a 3.2% yield on the three-month T-bill from current levels (about 1.83%) would require an increase of about 130 to 140 basis points. Such an increase in a period of months is not without precedent: In January 1994 the yield on the three-month T-bill was 3.04%, and by May of that year the yield was up 123 basis points to 4.27% (on its way to a peak rate of 5.94% in February 1995).

In addition, it has been widely rumored that the Federal Reserve Board is determined to "take back" the rate cuts in the wake of the September 11th terrorist attacks in the US, which would leave a fed funds rate of about 3.5%. Paul McCulley, Pimco Funds' Fed watcher, has suggested that he expects an average 25 basis point tightening each quarter — once the Fed gets going — until 3.5% is reached, most likely sometime late in 2003. In order for such a sustained monetary tightening campaign to be feasible, of course, the 5% Gross Domestic Product growth from Q1 2002 would have to be the harbinger of something truly positive in the stock markets. And there are few market timing models — monetary or not — that would quarrel with that.

David Penn may be reached at DPenn@Traders.com.

SUGGESTED READING

Doherty, Jacqueline [2002]. "On Your Mark. . .", Barron's: March 25.

Greider, William [1987]. Secrets Of The Temple, Simon and Schuster.

McCulley, Paul [2002]. "Eating Crow With A Dr. Pepper Chaser," Fed Focus: April, Pimco Bonds Commentary.

Moore, Stephen [2002]. "Farming The Taxpayers," National Review Online: April 23.

"Stop The Farm Bill" [2002]. Editorial. Washington Post: May 2.

Zweig, Marty [1986]. Winning On Wall Street, Warner Books.

Current and past articles from Working Money, The Investors' Magazine, can be found at Working-Money.com.