In part I of this series, I recommended putting your 401(k) savings into an index fund (or more likely, a target date fund that gradually shifts assets from an index fund into bonds). In part II, I ran the numbers for a couple who did just that.
As it turns out, however, an index fund isn't the optimal investing strategy. Then why do I recommend it? Because it's simple, close enough to optimal, and doesn't require any further effort on the part of the investor after it's set up, making panic moves less likely.
On the other hand, what if you're not prone to panicking, can follow instructions without giving in to the fear of lost profits, and can commit to spending an hour each year updating your funds? Follow me over the fold...
When we ran our numbers previously, we assumed a whole market index fund - after all, that's how you make sure you're getting the same results as the market. But the market isn't one monolithic entity; we can divide stocks into different classes, which may behave in different ways. During a market rally, for example, the stock prices for both large-cap and small-cap stocks may be increasing, but small-cap stocks could be doing twice as well as large-cap stocks (or vice versa).
Aside: if these terms mean absolutely nothing to you, don't despair - you can think of them as just referring to the stocks of "large" and "small" companies. Large cap refers to companies that have a capitalization - that is, a total value of outstanding stock - of more than $10 billion. For example, a company with 100 million shares outstanding at a price of $2 per share would have a market capitalization of $200 million, and would thus be considered a small cap stock. Small caps (those worth under $2 billion) are riskier than large caps, but have greater growth potential.
Rather than buying a whole-market fund that owns everything, another option is to buy multiple index funds that concentrate on different segments of the market - one for large-cap, one for small-cap, and so on. In this way, we can still buy the whole market, but in discrete chunks.
Why would we want to do this? Well, everybody knows the first rule of investing: Buy Low, Sell High. Many people, however, end up doing the opposite: they get caught up in the excitement when the market is surging (Buy High) and then panic and dump their stocks when the market drops (Sell Low). As a result, they end up losing money when they could have done quite well by simply not touching their accounts aside from putting in more money each month - thus, the "buy an index fund and don't touch it" strategy.
But what if I could give you a rule that would not only free you from making any decisions, it would force you to buy low and sell high?
The "secret" formula: asset allocation and rebalancing
Let's take a step back and just consider two stocks, A and B. Both A and B tend to grow at a rate of 10% per year on average (this number is misleading, as we'll see in a second) but in any given year, will either gain 30% or lose 10% - they have high volatility.
Suppose we buy $100 each of A and B to hold for two years. In the first year, A gains 30% and B loses 10%, so we now have $130 worth of A and $90 worth of B. In the second year A loses 10% and B gains 30%, so we now have $117 worth of A and $117 worth of B. Notice that this is why we don't just take the average of the gains and losses to determine the real growth in the stocks; +30 and -10 sum to +20 over two years, so a simply average would lead us to believe that we should be earning 10% per year as opposed to the 8.5% annual growth we actually saw. This is why, in part II, we used the compound annual growth rate rather than the average growth rate.
Now let's try a slightly different strategy. We still buy $100 each of A and B, and end up with $130 in A and $90 in B after the first year. We now rebalance our portfolio; having determined when we started that we want to have the same amount of A and B, we sell $20 in A and purchase $20 in B, giving us $110 in both stocks. In year two, A loses 10% to drop to $99 and B gains 30% to improve to $143. Our annual rebalancing then moves some of the money in B back to A, giving us $121 in each stock - a 10.5% annual return!
Why did this work? A and B are non-correlated - that is, a good year for A doesn't imply a good year for B, or vice versa. While we expect both A and B to grow by the same amount in the long run (as we saw in the first example, when they both finished at $117), at any given time one of them may be at a relative low and the other at a relative high. Because we insist on rebalancing to our original proportions, we are forced to buy high and sell low.
Of course, in practice it's hard to avoid correlations; stock prices will tend to rise during a bull market and fall during a bear market. But different types of stocks will tend to react in different ways. If we hold multiple types of index funds, which track different types of stocks (big companies, small companies, foreign companies, etc) and occasionally rebalance back to our initial allocation, this forces us to sell whichever sector has been doing better than average and buy whichever sector has been doing worse than average. In this way, we benefit when one particular sector either rises or falls out of sync with the others.
How often do we rebalance? One extreme is to never rebalance - this is our original strategy of just buying and holding a whole-market fund. The other extreme is to rebalance daily or weekly, which would overwhelm any gains with fees. A one year period seems like a good compromise (and is easy to mark on your calendar). Another option is to rebalance when your funds deviate from the desired settings by a certain percentage or dollar amount; however, this does require more hands-on management as you have to keep an eye on your funds!
Note: I'm assuming here that everything is taking place within the confines of a 401(k), so we can ignore any tax considerations that might otherwise come into play when we turn over our portfolio (such as short and long term capital gains).
So now that we know the basics, how do we apply them to an actual portfolio? Until now, we've been simplifying a bit, using "total market" as shorthand for everything you might want to invest in, but that's not quite accurate. In practice, you'll probably want exposure to the US stock market (in the form of a total stock market index fund), to the international stock market (an international stock index fund), and the bond market (a bond market fund). Performance of index funds in a given category should be similar, as they should be buying the same stocks, so we're largely concerned with fees. For the US total stock market fund, for example, Vanguard's VTSAX charges 0.05% and Fidelity's FSTVX charges 0.07% gross / 0.05% net; over the last five years, VTSAX has had an annual return of 15.70% compared to 15.67% for FSTVX. In other words - which one you pick largely depends on which company is hosting your 401(k) and thus allows you to purchase their funds at no fee (otherwise, go for Vanguard). Note that while these funds require a $10k minimum investment, that minimum doesn't apply for 401(k)s.
Choosing your funds
So how many funds do you need? This is a matter of personal preference; arguments have been made for anywhere from two funds (one total stock market fund and one total bond market fund) up through more complex portfolios that include separate funds for large blend, large value, small blend, small value, international (possibly broken down into Europe, Pacific, and Emerging Markets), REITs (real estate), and short-and intermediate-term bonds. Obviously, the granular your fund choices, the better able you are to capture differences in behavior that only hit particular types of stocks. At the same time, the more funds you have in your portfolio, the more complex it is to choose the correct allocation.
To build your portfolio, you can start by deciding how much to hold in stocks and how much in bonds. The conventional wisdom is that if you won't need money for a long time, you should be more heavily weighted towards stocks, with a greater allocation of bonds as you approach retirement. Somewhat counter-intuitively, a portfolio containing both stocks and bonds will generally have both lower risk and higher return than a portfolio containing 100% of either! One guideline that many people use is to hold their age in bonds; that is, at 35 years old, 35% of your portfolio would be bonds, and this would gradually increase as you get older. There are two key things to keep in mind when making this allocation. The first is that you want to have a significant percentage of your portfolio in low-risk assets (which don't necessarily have to be bonds) so that in the event of a stock market crash, you can move in money to capture the upside of the recovery. The second is that if you're closing in on retirement, you want to have enough money in low-risk assets (that can be cashed out without taking a loss - if you were buying individual bonds, this would probably be a bond ladder) that in the event of a crash, you wouldn't need to touch the stock portion of your portfolio until after the market recovers. As a rule of thumb, I would say to put no less than 20% of your portfolio into lower-risk investments.
After choosing the percentage of bonds, you can decide how complex you want to make your stocks. The easiest option is to simply put all of your money for stocks into a total market fund and be done with it. I personally hold several different types of funds, including a REIT. I'm not going to advise you on what percentages to put into each type of fund, but see the reading suggestions below if you'd like to read what other people have to say. Here are a few sample allocations that you can make with Vanguard funds; of course, there's no rule that says you have to use international funds.
3-FUND
50% VTSMX - Total Stock Market
20% VGTSX - Total International
30% VBMFX - Total Bond
4-FUND
40% VTSMX - Total Stock Market
10% VISVX - Small Cap Value
20% VGTSX - Total International
30% VBMFX - Total Bond
5-FUND
30% VTSMX - Total Stock Market
10% VISVX - Small Cap Value
10% VGSIX - REIT
20% VGTSX - Total International
30% VBMFX - Total Bond
I took these examples from this discussion, which also included the results from 1995 through 2006:
The Results
Total
Return %
01/01/95 YTD
-thru- -thru-
Portfolio YE 1999 YE 2002 12/14/06 12/14/06 StDev
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
3-FUND 131.03% 81.67% 210.32% 14.59% 12.20
4-FUND 120.95% 85.84% 223.14% 14.96% 11.22
5-FUND 105.21% 88.52% 241.79% 16.91% 9.93
Notice that the results are fairly similar; there were actually portfolios listed (holding between 3 and 10 funds), with the 10-fund portfolio returning both the highest overall results (254% over the 95-06 period) and the lowest standard deviation (9.72). Again, the main point is that you set your allocation and stick to it (aside from possibly increasing the percentage assigned to bonds each year) so that the magic of rebalancing can work.
Aside: over this time period, the market returned an average of 13.44% per year, or 11.79% annualized return, with a standard deviation of 18.59%. Each of the portfolios above had both a higher return (19%/year and up) and a lower standard deviation.
And if you're pretty sure you can do even better than the market by picking your own stocks or sector funds? That's not a problem, as long as you keep it to a small percentage of your portfolio (or at least realize that the more money you put into a small area, the higher your risk). I personally enjoy finance, so I hold some stock in my 401(k)...but I keep it under 10% of the total.
Want to read more? Here are a few pages you might find interesting:
Boglehead page explaining rebalancing
Vanguard PDF on best practices for rebalancing
Boglehead page on lazy portfolios, meant to be easy to rebalance