Just a few weeks back, I was chatting on the phone with Warren
Lammert, portfolio manager of the Janus Mercury fund, to get his take on Janus' performance. We got to talking about more academic stuff -- modern porfolio theory and stuff like that -- when he told me, "Studies have shown that you can achieve adequate diversification with 10 to 15 stocks."
Now, Lammert is one of the smartest fund managers I know, and a nice
guy to boot, but his words just didn't sit quite right with me. First of all, he's probably the 479th manager who's told me that chestnut about diversification over the years, and I get nervous whenever everyone in the investment business decides that something is true. Second, the phrase "studies have shown that" is right up there with "this time is different" as Wall Street's four most dangerous words. As far as I'm concerned, studies have shown that the only thing we can be sure of is that people who like to say "studies have shown that" probably haven't read the studies.
What's more, the studies Lammert was referring to (which, unlike most study-quoters, he has probably read) were published more than 20 years ago. So I decided to take another look and found that much
newer research shows something entirely different: It turns out that
10 to 15 stocks are nowhere near enough to diversify you.
The latest article, published in the Winter 2000 edition of the
Journal of Investing, is "The Truth about Diversification
by the Numbers," by Ronald J. Surz and Mitchell Price of
Roxbury Capital Management in
Santa Monica, California.
Surz and Price point out that the question of how many stocks you
need in order to diversify depends on how you're defining
diversification. Stocks have two basic kinds of risk: specific and
market. In plain English, the first kind means the stock you own
could go kerflooey while the rest of the market keeps going up; the
second kind means the entire stock market could go kerflooey and pull your stock down with it.
Now, there's not a whole lot you can do about the latter, but if
you're trying to reduce specific risk, Surz and Price show that you
need a whale of a lot more than 15 stocks. In the accompanying table, Surz and Price use three indicators to show how raising the number of stocks lowers specific risk. As you can see, to eliminate roughly 90 percent of your specific risk over the period Surz and Price studied (1986-1999), you'd have needed at least 60 stocks.
More stocks = less risk
|
|
1 stock
|
15 stocks
|
30 stocks
|
60 stocks
|
Total stock market
|
Standard deviation
|
0%
|
93%
|
97%
|
98%
|
100%
|
R-squared
|
0%
|
76%
|
86%
|
88%
|
100%
|
Tracking error
|
0%
|
82%
|
86%
|
88%
|
100%
|
The percentage reduction in three key risk measures.
Source: Surz and Price article, Exhibit 3
|
Surz and Price aren't alone. In their March, 2000 article
"Have
Individual Stocks Become More Volatile?", finance scholars John
Campbell, Martin Lettau, Burton Malkiel and Yexiao Xu find that in
recent years, individual stocks have grown much more volatile than
the market as a whole. (And since their paper only looked at stocks
through the end of 1997, Campbell and crew weren't even counting such disasters as priceline.com and NetZero!) Because individual stocks can easily move by at least 25 percent in a day, these researchers reckon you now need at least 50 stocks to eliminate 90 percent of your specific risk.
Considering the way most people actually invest, this is scary stuff. According to the New York Stock Exchange's latest survey of
shareownership, 15 percent of all investors owned only one stock, and the average investor held just 3.2 different stocks. All I can say is, if this sounds like you, then you'd better hurry up and buy yourself an index fund. That will give you a permanent stake in hundreds of different stocks and eliminate your specific risk. Maybe three's company in a Volkswagen Beetle, but it's plenty of nothing in the stock market.
|