The following are in response to Rob Arnott’s commentary, The Glidepath Illusion, which was published on September 25:
Dear Editor,
Mr. Arnott may be interested in this discussion on his article.
Regards,
Paul Crafter
Rob Arnott replies:
Thanks for bringing this to my attention. It’s fun to see that my simple backtest stirred so much controversy. While I didn’t much enjoy some of the ad hominem comments (it’s funny how some people just naturally assume nefarious intent in others … kind of sad to live with that mindset), I did thoroughly enjoy seeing such an active dialogue, in just a couple of short days.
Part of the dialogue points out that the failing of glidepath is that the dollar-weighted average allocation is well under 50%. That’s the point! Having less risk once you have more money reduces your end-point wealth, because your dollar-weighted average exposure to risk is lowered with a conventional glidepath strategy.
The comments about my failure to discuss risk tolerance are on-target. A larger study would have to include this discussion, as well as a more careful return attribution. This was a deliberately simple backtest of a widely-held bit of conventional wisdom, which flunks that simple backtest.
One person did a careful replication of the study, using Shiller data, adjusted to reflect a zero equity risk premium, was interesting. He/she assumes that you can bail when you reach your target level of wealth and lock in that level … a dangerous assumption, given the current example of negative real interest rates.
Also, it’s interesting to note the possible effect of herd behavior. As the boomers mature and – collectively – decide to de-risk, the appetite for stocks goes down and for bonds goes up. This happens at a time when bonds offer a negative real yield. How, exactly, is that going to help the boomers retire on schedule, with the retirement lifestyle that they were wanting?
Thanks again for forwarding this to me. All the best.
Dear Editor,
What Rob doesn’t mention (in the article at least), is how the annuity is calculated. He also doesn’t mention what I think is the most important reason for strategies like a glidepath, and which any actuary would be able to confirm. The annuity that can be purchased (and I’ll explain why this is important) from a life office, will be based on nominal yields or real yields depending on whether the annuity is level (or fixed increases) or inflation-linked. That creates a link between assets and liabilities at the time of retirement. As always, your assumptions are critical in any modelling work, and sometimes, your results flow directly from your assumptions. This creates the problem of drawing incorrect conclusions by not understanding the assumptions. If the annuity has been correctly calculated at retirement, then I’m happy that the conclusions are valid, but I can’t establish this from the article.
Too often, I see criticisms of strategies from people who don’t understand the basics around the thinking that went into creating these strategies. I therefore think that it is important to understand the subject intimately, before criticising it. The arguments presented for glidepath investing demonstrate a lack of understanding of the strengths of the strategy. Other assumptions, like how much investors will take in cash at retirement for settling of debts or other short term uses may also inform how these strategies were designed i.e. these strategies could be created to address actual investor behaviour. The final point relates to how strategies are crafted within pension funds to try to address the needs of the most people without the requirement for individual investment advice. It is always beneficial to have the opportunity to create individual investment strategies / solutions based on individual circumstances, but typically this is not viable within pension funds, and typically comes at a fairly high cost.
