January 18, 2011
Three Reasons to Ignore Market Performance
Capital Market Recap
Equity markets finished the year with a flourish. In the fourth quarter the S&P 500 was up 10.8%, small stocks (Russell 2000) were even stronger with 16.3% returns, while the international markets (EAFE up 6.7%) were held back by sovereign debt worries and uneven economic growth.
For the year, the story was similar. All markets had strong positive results, led by small company stocks, real estate, and emerging markets, with returns in the 20-30% range. Only developed international markets were at or below long-term expected returns.
Notably, intermediate term bonds lost money in the quarter as interest rates rose, but bonds did well over the year, with returns of 6.6%. A drop in interest rates increased the value of bonds during the first three quarters of the year, adding significantly to interest payments (yield). Falling rates have left the intermediate bond index with a yield of 3% or less, meaning rates will need to fall further for bond returns to exceed that level in the short term.
Three Reasons to Ignore Performance When Making Asset Allocation Decisions
Portfolio risk is determined by asset allocation, primarily the basic split between fixed income (cash and bonds) versus equities (small, large, and international stocks). Risk is inextricably linked to return, but performance data can blur our rational analysis of risk. Here are three reasons we should try to ignore performance data when determining our asset allocation.
1) Recency bias – Studies have shown that we tend to give disproportionate prominence to recent observations. In the early 1990’s we were busy telling people they should be willing to take stock market risk and get over the fact that bonds had outperformed stocks for much of the 70’s and 80’s. By 2000 everyone wanted to own technology stocks and we were busy telling people they should include bonds in every portfolio. In 2009 we were back to talking about maintaining a commitment to stocks.
Mark Twain said that history does not repeat itself, but it does rhyme. We need to be prepared for the rhyme and not base decisions on our recent experience that “feels” like a long-term trend.
2) The long-term can be very long – Bonds have outperformed stocks over the past ten years. This is contrary to long-term expectations. It is unusual but not unprecedented. In other periods of similarly dismal stock performance, the same thing happened. Bonds (5-year US bonds) outperformed stocks (S&P 500) for ten-year periods ending 1938-41, again in 1974-83, and now 2008 to the present.
I can’t remember what I had for breakfast, say nothing of how my portfolio was behaving ten years ago.
For most of us, ten years is a very long time, and yet market performance can be very far afield of our expectations over any ten-year period. On average since 1926, stocks have provided an annual return of 11.9% ($100 grew to $308) and bonds have returned 5.9% ($100 grew to $177). However, the ten-year period ending May 1959 was the best for stocks, when $100 grew to $697 (compared to $116 for bonds). These results are in contrast to the period ending February 2009 when $100 invested in stocks ten years earlier was worth only $71 while $100 in bonds grew to $182. The chart at left shows how much these returns can vary for a particular ten-year period.
3) It is difficult to make money predicting how the market will react to current conditions – One year ago it seemed inevitable that interest rates would rise and destroy bond values. It may still happen at some point in the future but the opposite occurred in 2010 – a costly error for anyone acting on that prediction last year. Similar examples are easy to find.
Conclusion – Ignore performance to make better decisions. Think of the asset allocation decision in terms of managing risk, and do not let your return expectations (long-term future expectations) be influenced by recency bias, a misperception of what is long term, or the tempting notion that someone can predict how markets will react to our current situation.