A reader responds to Joe Tomlinson’s article, Predicting Asset Class Returns: Recommendations for Financial Planners, which appeared last week:

Dear Editor,

This article on estimating long-term asset class returns emphasizes the importance of realistic assumptions. There are four critical areas, however, that the article touches on lightly or not at all.

First is the development of volatility assumptions. While Tomlinson referenced the difference between arithmetic and geometric returns due to volatility, it is important to highlight volatility and correlation assumptions in themselves as drivers of model conclusions concerning the success or failure of diversification. Where these come from and how you project them are therefore important.

Second, I generally subdivide equities into categories, including domestic and international, small and large. In this global investment environment, at a minimum, you also need to make assumptions on the US dollar exchange rate going forward and its volatility to assess the relative attractiveness of domestic and international equity.

Third, all of these attempts to predict returns assume steady states in some important variables that generally aren't even explicitly considered. A notable one is P/E ratios. You can't even try to predict equity returns without some assumption about whether P/E ratio expansion, contraction or stability is likely. We know from experience that P/E changes can be a critical medium-term determinant of equity returns (and therefore of volatility too). It's not just about income growth.

Finally, and possibly most important is the issue of consistency. Equity risks and returns, and those of fixed income and foreign exchange rates all exist in an interlinked economic and investment system that is also affected by government policy, commodity prices and other variables. It is counterproductive to develop assumptions by asset category without reference to how they interact with each other.

As an economist, it often looks to me like these efforts to develop asset-class assumptions are really efforts to informally develop working models of the economy without having to acknowledge that such a model is really what's needed to tie together all of the pieces.

Eric Stubbs

Global tactical portfolio manager and economist

A reader responds to Dan Richards’ articles:


I wanted to let you know how much I’ve enjoyed your website ClientInsights.ca and your articles from Advisor Perspectives. Your video presentations and writings are insightful, relaxed and genuinely client-focused. I’m pleased to have found someone in our industry with your breadth of knowledge and the inclination to share what they know.

Best wishes for your continued success,

John Shermer

Director of Business Development

Premier Financial Group

Eureka, CA 95501