Deciding the Trade-Off Between Volatility and Return

In yesterday’s Question of the Day post, I stated that I didn’t like lifecycle funds because most have a 15% allocation to bonds, no matter what your age. A reader named Duane left the following comment regarding that statement:

I seem to recall reading that a 90/10 balance between stocks and bonds has on average the same rate of return with less volatility. By that measure, a lifecycle fund with a minimum of 15% in bonds may not be a bad thing.

That statement got me to thinking…

So I fired up Excel and went to work comparing a 100% large cap stock portfolio with a 90/10 portfolio using the numbers found in Ibbotson’s Stocks, Bonds, Bills, & Inflation. I looked at yearly returns, and 5, 10, and 20 rolling period returns. I did not look at inflation.

Over the entire 81 year period (1926 – 2006), a 100% large cap stock portfolio outperformed the 90/10 portfolio 61% of the time. From 1926 – 2006, the compound annual growth rate for 100% stock portfolio was 10.42% compared to 10.10% for the 90/10 portfolio. It seems pretty close. However, that .32% difference can really add up over the years.

Because all I had to go by were yearly numbers, it was hard to judge volatility. So it is hard to judge the tradeoff between the increased return of the 100% stock portfolio and its increased volatility. My thought: if you’re investing for the long-term, who cares about volatility? Naturally, as you age, you should move your money into some bonds for stability purposes. However, it’s important to keep in mind that retirement itself can be 20-30 (or more) years, which makes you a long-term investor.

10 and 20-Year Rolling Period Comparisons

Over the 72 10-year rolling periods, the 100% stock portfolio outperformed the 90/10 portfolio over 72% of the time. The chart below was put together using the growth of a $10,000 investment over each 10-year rolling period.

100% Stocks vs. 90/10 Portfolio
Click on the chart to see a larger version

And over the 62 20-year rolling periods, the 100% stock portfolio outperformed the 90/10 portfolio 57 times (nearly 92% of the time). And, all 5 times that the 100% stock porfolio was outperformed by the 90/10 portfolio occured in the first 5 20-year rolling periods. Again, the chart below illustrates the growth of $10,000 over each 20-year rolling period.

100% Stocks vs. 90/10 Portfolio - 20-Year Rolling Periods
Click on the chart to see a larger version

I’m certain that the 90/10 portfolio would offer some decreased volatility. However, one has to decide how important that decreased volatility is to them.

17 thoughts on “Deciding the Trade-Off Between Volatility and Return”

  1. I know volatility becomes more of an issue when cash flows are introduced. Few people actually invest by socking away a lump sum for 10 or 20 years, they usually are either regularly adding money or taking it out. When they do that, the ups and downs along the way have the potential for a much greater impact on ending values.

    Since you already have the spreadsheet set up, how about taking a look to see what the difference might be if you invested $1,000 each year for 10 or 20 years into a 100% stock versus the 90/10 portfolio?

  2. When you did this calculation, did you rebalance back to 90/10 every year or two? Rebalancing was essential to Bernstein’s conclusion (i.e. the source of the original comment).

  3. Don,

    Yes, the results assume annual rebalancing. I plugged in the annual returns for the 90/10 portfolio found in the Ibbotson Book and then used the geometric average to arrive at the ending value for each rolling period.

  4. Jeremy,

    I’m normally pretty giving with my spreadsheets and stuff. However, since this one was derived using the stats out of SBBI, I wouldn’t feel right about giving it out.

  5. Saying a 90/10 portfolio has the same rate of return with less volatility is not quite right. Mathematically if stocks return 10% over the long term and bonds return 5% (hypothetical numbers), it would be impossible for a 90/10 portfolio to have the same return as a 100% stock portfolio. But a 90/10 portfolio has a much better risk/return ratio than an all-stock portfolio. This means that while your return is slightly better with all stock, the risk is much higher and not worth the extra return. The 90/10 idea is about having the most “efficient” portfolio in terms of risk/reward; it won’t give you the same return for lower risk.

  6. Actually after I made the above post I had a nagging feeling that I had also read somewhere that you can actually increase your return for the same risk by diversifying. So I did a little internet searching and found what I was thinking about. There is a “90/10 rule” that lets you increase your return for the same or even LESS risk, but it is at the other end of the spectrum. You can actually add stocks to an all-bond portfolio and get this effect (i.e. a portfolio with 10% stocks can actually have a higher return with less risk than an all-bond portfolio). It is not true at the other end though. You can’t reduce your risk without lowering your return by adding bonds to a stock portfolio (though you can make the portfolio more “efficient”). Here is a good explanation that shows graphs of the concept:

  7. I don’t know what I’m talking about but I have had a negative opinion of the lifecycle funds because I’d prefer to control the timing that I move from fund to fund. I have worried that these would end up timing the moves during a down cycle etc.

    Maybe they move a tiny bit at a time? so timing isn’t as much of an issue?

  8. Return is return, and risk is risk -> they are NOT related except that one would prefer to get paid more by taking more risk. The free market brings about a consistent “price” for risk. You can find a portfolio that contains the least amount or risk possible for any given return. If you choose to collect that same return but take on more risk than the minimum, by not diversifying for example, then you are dumb 🙂

    As probably everyone here knows there are different grades of bond risk too. Taking a risky bond would likely yield a higher return. Then there are essentially the risk free bonds of the government that are dirt low returns. Who knows what kinds of bonds make up the 90/10? Who knows what type of stocks either . . .

  9. Actually, according to your graph, the current decade so far is inverse to all of the others. In fact, adding some bonds to a stock portfolio both increases return and lowers risk.

  10. Don – looking only at the current decade skews the results because of the bubble that formed at the end of the last decade and burst at the beginning of this one. When the market is down then the more bonds you have will almost always be better. 100% bonds would have done better in the current decade than 90/10. Does this mean 100% bonds always has a better return for lower risk? No. There is no particular reason to look at the 6 year results that happen to make up the “current decade” as opposed to any other 6 year period. Those time periods are too short. The important part of the graph is the black part in the middle that shows the cumulative results over a really extended period.

  11. Ahh, no problem, JLP. I didn’t even think about the fact the data was from a paid resource. Thanks for the post though, I have always wondered what some hard data would prove in regards to this topic.

  12. As the person who instigated the discussion, thank you to everyone (and JLP in particular) for such a thorough investigation of my off-hand remark.

Comments are closed.