Summary: 5% of tech stocks have historically accounted for the entire sector's returns. Technology mutual fund managers, who are forced to limit position sizes, are therefore destined to be poor technology investors.
Technology is a unique area of investing for one simple reason: tech companies seem to have the ability to grow faster than, and to reach market shares greater than, almost any other type of company. New technologies that provide compelling functionality and return on investment spread rapidly. As a tech company's product becomes the de facto standard (interoperability is often critical in technology), it can dominate its market. Economies of scale - partly due to the networking effects of standardization - are particularly intense in technology. So the companies that win (think Intel and Microsoft) win big. And it follows that the tech companies that don't win either fail completely or are at best lackluster.
This observation is born out by profit data from tech stock investing. The seminal work on this is The Technology IPO Yearbook, compiled annually by Morgan Stanley analyst Mary Meeker. In it, she shows that of the 1,705 tech companies that went public over the last 22 years, 5% of them accounted for 100% of the wealth creation (as of end 2001).
This presents interesting problems for investors.
First, you can make significant profits in technology stocks, but stock picking is hard. If 5% of the stocks made 100% of the profits over the last 22 years, then broadly speaking you would have lost money with the other 95% of them. Second, to make those outsized profits from the handful of winners, you need the resolve to hold on to them for long periods of time, even when the market is falling or the company's business doesn't look good. How many of us have the guts to do that? How many people are not tempted to take profits on a stock that is up over 100% when doubts emerge about the company's competitive position or short-term profitability?
Third, if you stay the course and hold on to the stocks that will turn out to be winners, they will dominate your portfolio. If you'd invested in tech stocks over the last 22 years, holding on to your winners, you'd have ended up with an account massively dominated by Microsoft, Dell, Oracle and Cisco. Goodbye diversification.
Well, let's assume that you, as an individual investor, decide to go for tech investing, and you're prepared to stomach the enormous portfolio concentration that true success requires. Here's the next problem: When do you sell? You see, even though your portfolio would be filled with stocks of dominant tech companies, they would still be susceptible to unforeseen and disruptive competition. And if you don't sell them in time, you'll lose a lot of your compounded profits. Remember US Robotics, which dominated the modem market? Its business is now commoditized. Or what about Sun Microsystems? Only a couple of years ago, Sun was the undisputed leader in the market for computer servers. As the company said, it was "the dot in dot-com". But now the combination of Linux with Intel microprocessor-based hardware has thrown Sun's market position and future into uncertainty, and reduced its enterprise value above net cash to little more than, well, a dot.
My gut feeling is that it is precisely these difficulties that repel Warren Buffett from this sector. Buffett is the king of long-term investing, and understands the power of after-tax compounding better than almost anyone else. And Buffett never buys technology stocks or technology companies.
The moral? Individual investors can make money *trading* tech stocks. But they're playing against a lot of smart hedge funds, will likely realize short-term capital gains, and will incur significant trading costs. In contrast, if as a retail investor you're looking to invest in technology for the long haul, you'll need nerves of steel, acceptance of a highly concentrated portfolio, and a real appetite for company risk and volatility. This is consistent, by the way, with the last issue of this newsletter. There, I argued that you can beat the market if you focus on the areas of greatest information inefficiency (in this case: small cap technology stocks). But for the core of your portfolio, where you'd prefer less risk, less volatility and greater diversification, you're better off buying and managing a basket of exchange-traded index funds in an ultra-low cost online brokerage account.
What about mutual fund managers?
The technology portfolio concentration problem is particularly acute for mutual fund managers. Hedge fund managers are happy to trade short-term movements in tech stocks. Their volatility is exactly what makes them attractive. (I'm referring to the tech stocks, not the personalities of the hedge fund managers, though perhaps the statement is true of both.) But mutual fund managers generally want to pick stocks for the long run. And this poses a critical problem for technology mutual fund managers. The data, as we've seen, demonstrate that relatively few stocks account for most of the long-term profits in technology. An individual could buy a basket of tech stocks, hold on to the stocks of the companies that turn out to be winners, and end up with a highly concentrated (but successful) portfolio.
But a mutual fund manager could never do that. The trivial reason is that allowing a portfolio to become highly concentrated would likely breach the mutual fund's diversification guidelines (for good reason). But a more serious reason is that the technology mutual fund manager faces conflicting demands over time.
Here's what I mean. Imagine that a technology mutual fund purchases a basket of technology stocks and holds the "winners" as they progress to market dominance. The fund's portfolio becomes increasingly concentrated. Now, early investors in the fund would be happy with this: the fund manager's unwillingness to sell the "winners" would lead to a highly tax efficient mutual fund. The portfolio manager would sell only the tech stocks that did badly; so there would be no capital gains tax distributions to the fund's shareholders in the early years, and the fund's gains would compound over time.
But now look at the fund - say in year five - through the eyes of a new investor. Would the investor want to buy shares in a mutual fund with a highly concentrated portfolio with significant unrealized capital gains? With mutual funds, if an investor owns shares and the fund sells appreciated stocks, the investor is allocated a proportionate share of the capital gain for tax purposes even if the investor purchased the fund shares after the stocks had already appreciated. In other words, you can buy shares in a mutual fund and then get hit with a capital gains distribution even though the mutual fund shares have not appreciated since you purchased them. So when a new investor considers this technology fund for a taxable account (as opposed to a tax-free retirement account) in year five, it is highly unattractive. The investor is being asked to purchase mutual fund shares that are in fact worth less than the asking price, once unrealized taxes are taken into account.
Unrealized taxes are a problem for any mutual fund, but the problem is particularly severe for technology funds. Technology investors (as opposed to traders), we've seen, should hold winners for the long run. Successful technology funds should therefore have ever increasing unrealized capital gains. And that means that the interests of early investors in the mutual fund fundamentally conflict with the interests of later investors in the mutual fund, who don't want to buy shares of a fund with large unrealized capital gains. There is no way for the fund manager to reconcile these conflicting demands.
In practice, tech mutual fund managers seem to opt for the interests of later, rather than earlier, shareholders. They maintain diversified portfolios and thereby avoid egregious unrealized capital gains. Perhaps the fund managers do this because of simple diversification rules. Or maybe they are rated on their ability to attract new money into their funds. Mutual fund inflows have grown each year, so perhaps fund managers generate larger fees by attracting new money than by maximizing the return on money already in the fund.
In any event, by respecting the interests of new (and potentially new) shareholders, technology mutual fund managers always sell the stocks of successful technology companies too early. And that makes them intrinsically poor technology investors.