Steve TeSelle, CFA, CFP ™
February 2002
Asset allocation. In addition to the alliterative allure of the phrase, the concept is a useful investment approach for the average investor. The basic idea is to figure out how much of your hard-earned money to put into different classes of assets. Why? Because your investment portfolio is likely to perform better if you have a reasoned approach to asset allocation than if you, willy-nilly, pick your investments from the range of dishes available on the investment buffet table.
What are the different asset classes? Well, you can get pretty nutty here. I would definitely start with stocks and bonds. I think you could safely say that real estate is a separate asset class. From there, who knows? Stocks can be split into domestic and foreign stocks. From there, foreign stocks can be diced into developed and emerging markets; and domestic stocks can be sliced and diced into large, mid, and small capitalization (market value), and perhaps into growth and value, though I think this last distinction is over-rated. Bonds can be sectioned into various maturities and can also be segregated into US Treasuries, corporates, mortgage-backed, and municipals. Then there are hedge funds and various limited partnerships, and commodities and on and on. It's enough to make your head spin.
If you're not careful, you can find yourself being talked into all of these categories and you end up with an unwieldy and incomprehensible portfolio of assets. I prefer to start from the simplest approach and force myself to justify adding any asset classes from there. For me, the starting point is the justification for having different asset classes to begin with: a better risk-return payoff than you would have otherwise.
Let's use an example. Stocks are expected to return about 10% annually, based on historical averages. But stocks don't just increase at 10% every year with the reliability of a Swiss train. No, you get a little bit of a roller coaster ride -- +33% one year, -20% another year, +7%, and so on. Volatility, we call it. In the 1980s and 1990s, if you had allocated 20% of your portfolio to bonds and rebalanced every year, you would have given up 1% and 2%, respectively, in annual returns versus a 100%-stock portfolio; but you would have benefited from smaller fluctuations in your portfolio. The standard deviation was 2% lower during both periods for the 80/20 portfolio (standard deviations become unreliable for small samples, so I don't want to put too much emphasis on the amounts). However, in the 1970s, an all-stock portfolio would have returned 5.9%, while an 80/20 portfolio would have returned 6.3% -- with lower volatility! [Source for return data: Ibbotson.]
If you have an asset that is perfectly negatively correlated with stocks (when stocks go up, it goes down, and vice versa), then you benefit by adding this asset to your portfolio. You can get a comparable (and sometimes higher) return with less volatility. Whenever you get that option, take it. It's as close to a free lunch as most of us will ever get. Don't confuse direction with size. Correlation has to do with direction only, not size.
Stocks and bonds aren't negatively correlated, but they have a relatively low correlation (.11 from 1926-1999). Why? Because some factors cause stocks and bonds to move in tandem, while others cause them to move in opposite directions. For example, greater-than-expected inflation tends to be bad news for both assets, whereas faster-than-expected economic growth tends be good for stocks and bad for bonds.
The tricky thing about correlation is that just because two assets have had a certain correlation in the past does not mean that those assets will maintain that same correlation in the future. For example, returns on stocks and bonds showed a higher correlation during the 1980s and 1990s than they have in the past (.45 in the 1990s). This could be due to the fact that falling interest rates throughout this period produced positive returns for both stocks and bonds. So any time you're thinking of adding an asset class, think about why you're doing it. If your reason is based on a low correlation among assets, is that low correlation likely to persist? What are the factors behind the low historic correlation?
Here's another example. Junk bonds are often considered a separate asset class.
They have a low correlation with bonds, but they have a high correlation with
stocks. Why? A key factor for junk bonds is the health of the economy, same
as for stocks. So if I were tempted to add junk bonds to my portfolio due to
high expected returns, I'd want to substitute them for a portion of my stocks.
If instead, I substitute junk bonds for regular bonds, then I've just increased
the risk of my portfolio. 
[top]
So now that you have the idea behind asset allocation, you can start to build a portfolio that meets your return and risk requirements. I'll use just stocks and bonds; you can take it from there. For stocks, I mean the S&P500 index; for bonds, I mean intermediate-term bonds (a duration of 4 to 7 years).
Benjamin Graham, one of my investing heroes, wrote many years ago that you could vary your exposure to stocks between 25% and 75%, depending on your view of expected returns. So, if stocks look quite expensive, you might reduce your exposure to 25%, with the remainder in bonds; and you would increase your stock exposure to 75% if stocks looked relatively cheap. If you have the framework to make those kinds of prognostications, this is a nice, simple approach that could work well over time. In my experience, however, the investment world seems a lot more topsy-turvy when you're in the middle of making your decision. 'Tis a difficult task to decrease your exposure to stocks when everyone else is saying that only an idiot would take money out of the market. We tend to want to close the barn door after the horse has bolted. Only after stocks have performed poorly do we want to reduce our allocation.
An alternative is to set an asset allocation of stocks between 50% and 80%, depending upon how well you can stomach volatility, and put the remainder in bonds. Then each quarter you check your portfolio. If the assets are more than 5% off your target allocation, you rebalance. This approach relieves you of the need to change your allocation based on expected returns, and gives you an identifiable rebalance mechanism. Also, this approach forces you to sell the asset that has increased in price - you are forced to buy low and sell high.
I took a little bit of everything…at first it looked exciting, now it looks yucky.
People's situations vary. You may want to add other assets because you're familiar
with them. You may want to reduce the number of times you rebalance because
of taxes. But whatever you do, I encourage you to keep it simple.[top]
© Dorato Capital Management, LLC. All rights reserved.