Over the past year (in case you hadn't noticed), the prices of technology stocks have fallen more than the prices of nontechnology stocks. That means the market capitalization of tech stocks -- as a percentage of the total market cap in a cap-weighted index like the S&P 500 or the Russell 1000 -- has fallen, too.
Jim Cramer predicted in a commentary Thursday that this loss of relative market cap means that
"Tech's Slope Downward Will Just Get Slipperier." That's because, in Cramer's analysis,
[I]t is deadly for tech and will cause us to swing from radical overvaluation to the possibility of severe undervaluation. Billions upon billions of dollars are indexed to the S&P. It has a tremendous lagging impact, like an anchor on the companies that have been getting smaller in the S&P. Firms that index money will be allocating fewer and fewer dollars to the shrinking names, once the S&P rebalances its stocks based on market capitalization. It is a self-fulfilling prophecy: The smaller get smaller.
As someone who spent years running portfolio management and trading at the world's largest index fund manager,
Barclays Global Investors, I'd like to offer a few perspectives on Jim's prediction. I'm going to take a few shots at some of Jim's assumptions, but then I'll give him a little intellectual ammunition to use, too.
How It Works
Jim is correct that index funds both hold and trade fewer dollars in shrinking names. If
Cisco's (CSCO Quote) price is cut in half, its market cap is automatically and instantaneously cut in half -- simply because market cap equals stock price times shares outstanding. That means that Cisco's weighting in the S&P 500 is automatically and instantaneously cut in half, too, because the index is cap-weighted.
But no administrative change by
Standard & Poor's is required, and no trades by indexers are required. Neither has to do anything to change Cisco's weighting on a particular date. It happens all by itself. In fact, it happens any time and every time that Cisco's price changes, even by the smallest amount. And of course the same thing applies to any stock, and it applies proportionately and in the appropriate direction whether the stock goes up or down.
So, yes, the tech weighting in the S&P has fallen, and yes, tech's value in index funds has fallen, too. But that doesn't mean index funds have had to sell tech or will have to sell tech in the future to maintain their index-tracking weight in it. By holding the same number of shares of tech stocks as they did before their price drop, their portfolio weighting in tech has already self-adjusted.
But Jim's point is that, in the future, any time an index fund does a buy program (and these programs often happen a billion dollars or more at a crack), it will be buying less tech now than it was a year ago. Jim's argument is that this takes away an important element of demand from tech -- and that it's a vicious cycle. The less demand, the lower tech's weighting will get. And the lower the weighting, the less the demand. And so on -- as Jim calls it, a self-fulfilling prophecy.
But here's a strange paradox that cuts Jim's prediction: When the prices of tech stocks drop, all else equal, an index buy program of a given size will actually have to buy
more shares of tech stocks, even though it is spending
fewer dollars on them.
Let's work through an example to see what I mean. Let's say that a year ago Cisco was at $60, and that it had a 3% weight in the S&P 500. A billion-dollar index buy program would be 3% Cisco, or $30 million. That would be 500,000 shares at $60 per share.
Now let's say that a year has gone by and the price of Cisco has dropped to half of what it was -- it's now only $30 -- but everything else has stayed the same. Now Cisco would have a 1.52% weight in the S&P 500 instead of 3%. A billion dollar index buy program would be 1.52% Cisco, or $15.2 million. That would still be 507,614 shares at the new $30 price -- more shares than a year ago when the price was twice as high.
Don't believe me? Get out your calculator and work it out for yourself.
Now despite this seeming paradox, Jim still has a point. In
dollar terms, today's buy program is spending just a little more than half the money on tech than it did a year ago. And if you believe that raw-dollar flow into a stock is what builds its price, then you probably agree with Jim's prediction.
But there's really no rigorous evidence for this belief (nor -- to be fair -- against it). It's equally plausible to believe that what builds a stock's price is the number of shares that are bought as a fraction of its shares outstanding. And as my example shows, that way of looking at the world would lead you to believe the opposite: that indexers will now have about the same effect on tech stocks that they've always had, and maybe even a little more.
Watching the Russell Rebalance
An upcoming index event does unambiguously tend to support Jim's general observations. That is the June 30 rebalancing of the
Russell 1000 and Russell 2000 indices. Unlike the S&P indices, which get new members every once in a while ad hoc, and whose weightings are rebalanced quarterly (and then only to adjust the weightings of stocks whose number of shares outstanding have changed), the Russell indices are radically reconstituted once a year, adding new stocks and throwing out old ones by the dozen.
Merrill Lynch, whose Equity Derivatives Strategy department puts out some of the best index research on Wall Street, has published a forecast of what the Russell reconstitution will look like this year -- and it's not pretty for tech.
Merrill's forecast predicts that 13 new companies will be added to the Russell 1000 (defined strictly as the top 1,000 stocks by market cap), and that 136 will graduate to the 1000 from the Russell 2000 (defined as the next 2,000 stocks by market cap).
Of the 13 new companies joining the 1000, only five are techs. Of the 136 graduating, only nine are techs -- that means that only nine technology stocks have grown over the past year into the top 1,000 market-cap club.
At the same time, 99 stocks are falling from the 1000 to the 2000, and 62 of them are techs. And two stocks are being deleted from the 1000 altogether (they didn't just fall from the 1000 to the 2000, they fell off the face of the earth) -- and they're both techs.
This means that the tech weighting in the 1000 will fall from today's 22% to 21%. But even 22% is surely a big drop from the levels of a year ago. Tech weighting in the 2000 will actually increase from 12% to 18%, because this small-cap index will serve as a refugee camp for all those fallen angels from the 1000.
Merrill reports that the 13 stocks joining the 1000 makes the smallest number of newbies in at least four years. That's because of the sudden drop-off in spectacular tech IPOs, in which a brand-spanking new tech company springs into existence and immediately vaults into the ranks of America's largest-cap companies.
Nowhere near as much money is indexed to the Russell indices as to the S&P 500. But come June 30, when the Russell reconstitution takes place, there will definitely be some tech selling. When the date gets closer, I'll file a report naming the names of the victims.
RSS
Don Luskin
Get Cramer's portfolio stock picks at Action Alerts PLUS —