Using our approach for developing mid-term expectations for US Stocks, Bonds, and Cash, we find a 21% asset growth advantage when we backtest a dynamic asset allocation approach vs. a strategic asset allocation approach over two full market cycles (backtest xls included below).
As illustrated below, we know long-term stock gains come via valuation (P/E) cycles.

We also know, mid-term (10 yr) valuations have an inverse relation to subsequent returns.

By utilizing mid-term valuations (Shiller’s 10 year cyclically adjusted P/E (CAPE)) to develop mid-term stock return expectations, we can utilize mid-term inputs to make dynamic shifts to more efficient allocations to therefore improve cumulative asset allocation gains.
To read the previous posts on developing mid-term expectations go to:
- Enhancing Our Traditional Approach with a Mid-Term Perspective
- Developing Mid-Term Stock Return Assumptions
Let’s look at an important dynamic asset allocation example…when stocks were at their highest valuations (CAPE) ever.
Efficiency = Return/Risk

The dynamic approach shifts an allocation with 26% Efficiency (Moderate) to an allocation with 85% Efficiency (Conservative) in August 2000. In subsequent periods, mid-term expectations are developed and assets shift again to a higher efficiency allocation. The number of overall shifts is based on the frequency of the mid-term expectations calculation and variance shift parameters.
Utilizing the mid-term expectations and efficiency, the below chart illustrates a comparison of a moderate allocation strategy (strategic asset allocation approach) vs. a dynamic moderate allocation strategy (dynamic asset allocation approach). Note: dynamic allocation shifts occur when efficiency varies by more than 5%.

Here is the backtest xls (which includes the allocation shifts) to download and view … Dynamic AA
Summary
By utilizing mid-term expectations, one can protect client assets and achieve greater cumulative asset growth over time vs. a traditional strategic asset allocation approach.


Comments
Log in or sign up to join the conversation.