Steve TeSelle, CFA, CFP ™
November 2003
Benchmark for Both Risk and Return
A benchmark is supposed to tell you how you're doing. If you're beating the benchmark, you're doing well; if you're not beating the benchmark, you're not doing so well. For example, if I told you that my portfolio increased 12% in a year, you might think that was pretty good. Until you found out that stocks in general increased 40%. Then you might think I'm not so good. Or if stocks in general decreased during the year, then you might think I'm a great guy. The point is, you don't really know how to measure my performance until you have something to measure it against.
[top]
Dorato uses the S&P500 Index as an equity benchmark because it is a widely followed and easily understood measure of how the stock market is doing. I could use other indexes, or a combination of other indexes, but the key is to have an easily identifiable measure of performance.
Also, a benchmark should be easy to invest in because then an investor can always choose to invest in the benchmark rather than an actively managed alternative. The S&P500 meets this test - a number of mutual fund companies offer index funds with annual costs as low as 0.18%.
The Dorato portfolios do not look like the S&P500. In my view, there wouldn't be much point in using Dorato if the portfolios mimicked the S&P. Rather, Dorato's approach is to create a diversified portfolio of 30 to 35 stocks - a portfolio that is diversified by the size of the company and by the type of business in which the company operates.
If I always invested in small companies or in Japanese companies, then using
the S&P500 as a benchmark doesn't make much sense. However, if I only occasionally
own small or foreign companies, depending on whether they offered good value,
then the S&P500 might still be a good benchmark. Common sense, which is
sometimes in short supply, must prevail.
[top]
Some investors get carried away with creating asset classes and associated benchmarks. They'll split the stock market into small, mid and large capitalization, and split it further into growth and value investing. Then they assign a benchmark to each of those classes. The problem with this approach is that it can get awfully complicated, and, sometimes, the forest is nowhere to be seen. You can find yourself measuring managers within each asset class without an overall view to the primary goal: earning an investment return that meets the client's needs.
Institutional investors (pension funds and the like) have fired managers for
style drift. The manager may have outperformed the index, but the returns were
not correlated enough with the benchmark to save the manager's job. This forces
managers to look a lot like whatever index they're measured by. So a mid-cap
manager should sell a stock if it gets too big and is removed from the mid-cap
index, because the stock's future performance could be different from that of
the index. For an individual, that seems like an arbitrary and nutty approach.
[top]
While most everyone focuses on return relative to a benchmark, you also want
to keep track of risk. People will use the word risk to describe different things.
I use risk here to describe volatility. To measure risk, you might use the beta
of the portfolio, or you might use the standard deviation of returns, or you
might use both. But you want to include risk in your evaluation of how you're
doing. For example, say a fund returns 15% versus a 10% return for the benchmark.
Good, except that maybe the fund has a beta of 1.5, versus a benchmark beta
of 1. This means that if the benchmark dropped 10%, you'd expect the fund to
fall 15%. So is the fund manager particularly good, or is she just getting a
higher return because she's taking on more risk. I don't think the information
I've given you is enough to say one way or the other whether the fund manager
is good, but it should get you thinking about the risk/return tradeoff.
[top]
This is a tricky question. Generally, the longer the time period over which a manager beats a benchmark, the more confidence you can have that the manager is good. But how long is long enough? Statistically, you probably need something like 30 years of data to prove that a manager is good, but by then everybody who cares is dead. My less-than-scientific approach is that a three or five-year record is a decent indicator, and a ten-year record is a very good indicator. Economic cycles (expansion and recession) and investing cycles (growth vs. value) tend to go in multi-year periods. Someone who does well in one environment, might not do so well in another. The longer the track record, the more cycles they've had to deal with.
[top]
If your money is in a taxable account, taxes will affect your return. The S&P500 is an index, so Standard and Poors does not provide an after-tax calculation. I use the Vanguard S&P500 Index Fund as my source for after tax returns. Vanguard is very good about providing this information on their web site.
You'd like to know your after-tax return in a taxable account because it could be quite different from the pre-tax return. Suppose a mutual fund gets a 15% return versus a benchmark return of 10%. That looks pretty good. But what if all of those gains are taxable as short-term gains? If you're in the highest tax bracket, and in a high-tax state, you could lose nearly half that return to taxes. Not so good anymore.
[top]
Managers tend to crow like a rooster when they beat their benchmark, and have a pocketful of excuses when they don't. The individual investor's job is to keep a level head, understand that a good benchmark is tough to beat and that performance over time is more important than what happens over short periods.
"I think I found my benchmark."
© Dorato Capital Management, LLC. All rights reserved.