Bob Veres is a highly respected columnist for Financial Planning magazine. He’s been in the forefront of advocating good practices in financial planning. He had an interesting article about the dangers of tactical management last month–and the longer I chew over that article, the more problems I see with forecasting, explicit and implicit.
First, let me set the scene for you. Mr. Veres indicated that he had spent a month doing data analysis of a survey of over 1000 financial planners. One of the most interesting takeaways:
Perhaps the most striking thing I learned is that, post-2008, activities once labeled “market timing” are now solidly in the mainstream. No, planners are not moving into or out of the market based on reading the entrails of animal sacrifices. But they seem to be taking a much more active approach to protecting clients from downside risk. The survey asked whether advisors planned to raise or lower their allocations to each of 35 different investment classes or vehicles in the next three months. A remarkable 83% are anticipating at least one tactical adjustment – and the great majority expects to make several.
Mr. Veres raises a legitimate question about how accurate planner’s forecasts are, and what the possible consequences of poor forecasts could be. As an example, he cites inflation forecasts:
One of the most provocative set of responses came when advisors were asked to forecast the inflation rate and the real (after-inflation) return of both equities and 10-year Treasuries over the next 10 years. On inflation, the majority of responses clustered between 3% and 5%, and the remaining responses had a center of gravity on the 6% to 7% part of the chart. If there is wisdom in the crowd, the crowd of advisors seems to believe we will experience above-average inflation for at least the remainder of this decade.
But we also had responses as low as -4% a year (projecting severe Japanese-style deflation), several as high as 10% and one advisor who anticipates annual inflation of 13% over the coming decade. One or the other group on the fringes is going to be spectacularly wrong (or both will), and I suspect that damaging consequences will show up in the portfolios they recommend.
He is absolutely correct in his belief that a strategic allocation based on a wildly incorrect forecast will be damaging to a client. And, he indicated that expectations for real returns were even more widely dispersed. It’s where he goes next that made me think. He writes:
Say what you will about the buy-and-hold ethos that lasted until September 2008. It may not have been an ideal strategy during the worst of the bear markets, but it did keep the members of the herd from straying too far from the center – and, more important, it kept the profession from getting the kinds of black eyes I think advisors are going to encounter in the future.
I think there is a critical error in this line of thinking. Buy-and-hold strategic allocations are typically based on historic returns. Those historic returns, of course, are the same for everybody and they do have the effect of keeping everyone in the middle of the herd.
This is the critical error: basing your allocation on historic returns is also a forecast–it’s simply an implicit forecast that historic returns will continue along the same path! If that doesn’t happen, you will end up just as horribly wrong as the advisors on the forecasting fringes! Yes, you will fail conventionally along with everyone else in the middle of the herd, but you will fail nonetheless. (How do you think most clients feel right now, with their equity allocations based for the last decade on an 8-10% historic return, a return that has not materialized? And there’s always Japan.)
Frankly, if you’re going to do strategic asset allocation at all, research shows that naive equal-weighting performs as well as anything else. The reason advisors are gravitating toward tactical management in the first place is because traditional strategic asset allocation has so much trouble with tail events, and 2008-2009 was a big tail event. Clients have memories, and advisors are simply responding to client demand for a more active form of risk management. Now, like Mr. Veres, I’m not very confident in the forecasting abilities of financial planners. I’m not very confident in the forecasting abilities of anyone, for that matter, including me.
All forecasting is flawed, whether it is explicit or implicit. To me, there are only two realistic choices for asset allocation. Either 1) equal-weight a large number of asset classes and rebalance periodically or 2) commit to a systematic method of tactical asset allocation. There are funds that use valuation triggers and funds that use relative strength to rotate among asset classes–and I suspect either will perform acceptably over time if it is systematic and disciplined.
Posted by Mike Moody 
