Recent times have seen a number of unprecedented events in the stock market. One event that has been somewhat overlooked has been the extraordinary rise in daily volatility in the markets. We have all gotten used to it now, but the numbers are startling.
From 1971 through 1997, there were 40 days in which the Nasdaq composite index rose or fell more than 3%. In 1998 alone there were 16. In 1999, 20. And this year through Oct. 5, there have been 50 days of 3% or more moves, more than in the first 27 years of the index combined.
Cross-sectional volatility also has been amazing. From 1985, when our daily data begins, through 1998, the Russell 1000 value and growth indexes had daily moves relative to each other of 2 percentage points or more seven times. In 1999, they had five such moves; and so far in 2000 through Oct. 5, 32 times - twice as many as in all our prior history.
And the moves do not appear to be in response to unprecedented events. There have been interest rate increases before. Inflationary expectations have risen and fallen far more than anything seen now. But never have we seen such a manic-depressive market, moving from despair to elation and back within a week, and sometimes even a day. So what has happened? What makes today so different from the relatively tame history of stock market returns?
One factor seems to stand out - valuation. The stock market is priced more highly, relative to earnings or assets, than ever before. That could mean the stock market is overvalued. But let us assume that the stock market is fairly valued and the fact that earnings and dividend yields are low today simply means growth will be higher in the future. Valuation and volatility are still inextricably tied, for a very simple reason: duration. Duration is a concept more frequently associated with bonds than stocks. Basically it is a measure of how long it takes, on average, to get each dollar of income from a security. It is not often mentioned in terms of equities, since the cash flows from stocks are not guaranteed. But they can be estimated. Even if the estimates are imprecise, they do tell an important story. Because as the price of a stock goes up, so does its duration. Historically, the Standard & Poor's 500 has sold at a price-earnings ratio of 15 and a duration of around 30 years. Today, at a p/e of 29, its duration is 61. For Nasdaq, which also has had a historical average p/e of around 15 and duration of 30, the current duration is more than190 because of a current p/e in excess of 100.
Why should we care about these numbers? After all, they are only estimates, and pretty imprecise ones at that. But fixed-income managers care about the duration of their portfolios because it tells them the sensitivity of their portfolios to a change in interest rates. The longer the duration, the greater the sensitivity. Equity duration tells us the same thing with respect to a change in either the discount rate or the expected growth rate. And frankly, there is little reason for equities to change in value apart from a change in one of those two factors. Given the current duration estimates, we should expect the S&P 500 to be more than twice as volatile today as it has been historically, and the Nasdaq six times as volatile.
So far this year, S&P 500 daily volatility is running at 1.9 times historical levels and Nasdaq at four times historical levels. Not spot-on the estimates, but pretty close. And given the extraordinary intraday volatility we have seen on the Nasdaq this year - notably a morning fall of 13% one day in April followed by an 11% recovery in the afternoon - perhaps even the daily data underestimate how volatile the market truly has been.
Something clearly has changed in the markets. Some investors would argue that this daily volatility is irrelevant because we are all in the stock market for the long term. Because there is no evidence (yet) for a rise in long-term volatility, there is no cause for alarm.
However, such protestations have a problem. One can either believe the stock market is a random walk or that prices are mean-reverting. If the market is a random walk, volatility should rise in proportion with the holding period, and the rise in short-term volatility should mean a similar (and frightening) rise in long-term volatility. Only if the stock market were mean-reverting would this not hold true. However, in a mean-reverting market, higher than average returns are likely to be followed by lower than average returns. So either the market is a random walk and is riskier than it used to be, or it is mean-reverting and due for a big fall.
Whichever is the case, this is not your father's Nasdaq. Expecting it to behave as if it were may be wishful thinking.
Ben Inker is director-asset allocation at Grantham, Mayo, Van Otterloo & Co. LLC, Boston.