In my recent post, I tracked margin debt on the NYSE back to 1970 and found a consistent pattern across bull and bear markets. During cyclical declines, we tend to see year-over-year declines in margin debt. During bull market peaks, we tend to see substantial annual gains in margin debt. It is this tendency to borrow money to buy stocks when markets are already high and refrain from borrowing when markets are weak that makes margin debt a worthy contrary indicator.
From the vantage point of market psychology, margin debt is one of the purest measures of fear and greed. When investors are greedy, they will go beyond their cash balances to buy stocks. When they are fearful, they will refrain from leveraged positions. The big question, however, is whether future market returns are impacted by the fear and greed of margined investors.
Going back to 1970 (1818 weekly periods) in the Dow Jones Industrial Average, we find that, when annual changes in margin debt exceed 30% (N = 445), the next 52 weeks in the Dow average a gain of only 1.39% (209 up, 236 down). That is much weaker than the average 52-week gain of 9.05% (1285 up, 533 down) for the entire sample. In short, periods of greed lead to subnormal market performance.
How about fear? When the annual rate of change in margin debt has been -25% or less (N = 128), the next 52 weeks in the Dow average a gain of only 1.14% (57 up, 71 down). Again, that is much weaker than the average yearly gain for the Dow since 1970.
Indeed, when we have neither fear nor greed--when the annual rate of change in margin debt is less than +10% but greater than -10% (N = 427), the next year in the Dow brings an average gain of 13.91% (357 up, 7o down)--considerably stronger than average. Given that we are neither seeing extreme fear or greed at present in margin debt, we'd have to say that there are no immediate bearish indications.
In short, extremes of investor fear or greed have led to subnormal returns in stock prices. It is when sentiment has been moderate that returns have been best.
Let's however, consider the trajectory of margin debt at recent market peaks. The annual rate of margin debt peaked in March/April, 2000, near the peak in the NASDAQ but prior to the Dow peaks later that year and in May of 2001. The annual increase in margin debt peaked in May/June, 1998, a month or two ahead of the sharp decline. The peak in August, 1989 preceded the price peak of 1990 by nearly 11 months. The margin debt rate peak of December, 1986 preceded the price peak of 1987 by about 9 months. The annual peak in November, 1983 preceded the price peak of 1984 by about 2 months. The late April peak in 1981 was only about 2 months ahead of the price peak.
What is clear from this little excursion is that annual rates of change in margin debt are not precise market timing tools, but there is a tendency for these rates of change to top out ahead of the Dow stocks. In other words, investors reduce the growth in their margin accounts as markets are topping. This is relevant because we saw our maximum annual rate of change in margin debt all the way back in April, 2004. The most recently reported rate is well below that recorded in May of 2006. The slowing of growth in margin debt following a multi-year rise is a yellow--not a red--light for stocks going forward.