There was a place to hide in the perfect storm...
Posted: 15 Mar 2009 10:19
I started this exercise to prove a foregone conclusion using Norm's Crystal Ball: you need a significant weighting of bonds in your portfolio as a dampening effect against equity market swings and as protection against shit happening, specifically the perfect storm. For the equity weighting, I chose the old RRSP 30% foreign content limit. For the bond weighting, I chose short term bonds for a couple of reasons. We've been in a 25 year bull market for bonds and long rates have gotten about as low as they can get. Similarly, inflation risks - as opposed to inflation - have been rising steadily for 10 years although they aren't reflected in long bond prices. I've included a couple of twists such as a 10% gold weighting and RRBs instead of short term bonds.
And then I switched the asset mix from 50/50 bonds to equity to 25/75 bonds to equity. Average returns and total returns don't change much but variability almost doubles and 2008 becomes a disastrous year rather than just a bad year. Disastrous years have long term and even permanent effects. IIRC, correctly GeorgeH has alluded to the effect the 70s bear had on him.
The results were as I expected. They may be even better than the results indicate. If Norm's model doesn't rebalance annually, then the returns are understated and the variability may be overstated. They are nominal not real which does have an impact. It's worth the exercise for people to run real numbers.
We can't say we didn't know. Norm's tool and ones like it (although not as good) have been around for a long time. His cautionary note says it all: This tool is much like a scalpel. In competent hands it is useful but the amateur may wind up cutting themselves. The amateur should seek professional advice. I suspect that most people would use this tool to find the optimal asset mix that gives the highest return over the longest period of time and ignore the variability of the returns. I suspect also that people would accept the numbers without trying to understand what was happening. The numbers say that long bond annual returns of 10% from 1970 to date are great. Adjust for inflation and 10% becomes 5.5%. Looking inside the numbers shows a more detailed picture of inflation. From 1982 to the present returns are 12% but only 6% from 1970 to 1981. Adjust for inflation and the results become 8.9% and -1.8%. Using numbers without trying to understand them is a recipe for disaster.
Let me put a behaviourist's spin on things. As with all animals, we herd. When market returns are good, we focus on returns and think they will continue and even get better. So we chase the returns. We buy dotcons; we buy income trusts; we buy oil trusts; we buy dividend stocks at inflated prices; we buy houses at inflated prices. Unlike other herd animals, we tend to ignore risk until it becomes a reality. Sheep run when they smell a bear; we don't. Neither do we learn very well. Sheep learned a long time ago what the smell of a bear means. We read about tulips, the South Seas and other bubbly times, and then smile and think that we've learned because we're so much more knowledgeable and sophisticated now.
Bottom line: A diversified portfolio that focuses on risk rather than return would have produced a loss last year of 10% (11.4% real). The impact of tha kind of number can easily be offset to zero with a bit of cost cutting on the consumption side.