Combining Strategic and Tactical Asset Allocation
In a recent article, I discussed Tactical Asset Allocation [TAA] and its relationship to Strategic Asset Allocation [SAA]. To take advantage of both TAA and SAA, it is necessary to build a highly diversified portfolio that has a projected future return that is greater than the returns that the portfolio has generated in recent years (i.e. the assets in the portfolio have generated less-than-expected returns in recent years). The recent under-performance means that the portfolio is due for a reversion to the mean and provides the potential for a performance boost beyond long-term expected returns (the TAA component). The core Strategic Asset Allocation ensures that the portfolio derives the maximum available diversification benefit and has a total volatility that is consistent with the needs of the investor(s).
In this article, I will discuss a “real life” example that combines the Strategic and Tactical aspects of asset allocation. A little over two years ago, I analyzed a portfolio that exemplified both of these characteristics. In this article, published on September 20, 2006, I analyzed a portfolio made up of Berkshire Hathaway’s (BRK.A) top twenty holdings of public companies (in the same relative proportions as in Berkshire’s portfolio). My goal was to examine whether there were any notable statistical features of this portfolio. What I found was that the Berkshire portfolio was both very well diversified [SAA] and the holdings appeared to be due for a Reversion To the Mean [RTM] on the upside [TAA].
To many, the idea that the Berkshire Hathaway portfolio is “well diversified” will seem paradoxical. After all, Berkshire is famously concentrated. When I analyzed the portfolio in the original article, it held substantial positions in Coke (KO), Procter and Gamble (PG), Wells Fargo (WFC), Budweiser (BUD), American Express (AXP), Conoco (COP) and others. From the perspective of our quantitative portfolio planning model (Quantext Portfolio Planner, QPP), Berkshire appeared well diversified, however. The quantitative measure of diversification comes down to generating a portfolio with the maximum return relative to its risk—i.e. the portfolio is expected to generate a high risk-adjusted return. Berkshire’s portfolio appeared to accomplish this, despite having very concentrated positions.
I used three years of trailing historical data as input to Quantext Portfolio Planner [QPP]. QPP generates expected returns for the entire portfolio (the default setting). At the time of the analysis, I found that the Berkshire portfolio had a trailing three-year average annual return of 11.8%. QPP projected that the Berkshire portfolio had an expected return of 13.2%. For the Berkshire portfolio to average a return of 13.2%, the coming years would need to generate returns higher than average to offset the lower returns of the recent three years. As such, this portfolio was positioned to exploit the benefits of TAA. By contrast to the Berkshire portfolio, the S&P 500 appeared over-valued at the time of the analysis. The S&P 500 had a trailing three-year average annual return of 9.56% (before adding in dividends) and QPP’s default setting is for an expected total annual return from the S&P 500 of 8.3%.
Aside from being well positioned to exploit TAA opportunities, QPP suggested that the Berkshire portfolio was very effectively exploiting diversification benefits between the stocks in the portfolio (i.e. it was successfully capturing the Strategic Asset Allocation component). QPP generates a statistic called the Diversification Metric [DM] and the Berkshire portfolio exhibited a very high value of DM:
A portfolio with a high value of DM is doing a better job of strategic asset allocation to exploit offsetting risks between investments. The Berkshire Hathaway Top 20 portfolio yields DM=67% when we use the trailing five years to initialize the model and DM=66% when we use the trailing three years. These are very high values for the diversification metric and show that this portfolio is an effective mix and relative weighting of stocks to exploit non-market correlations between components to manage total portfolio risk levels.
- Source: Original analysis from 2006 (See footnote 2)
In the two years since writing this article, I have found that a range of resources suggest a practical estimate of the maximum diversification benefits available in a well diversified portfolio. The Berkshire portfolio was not quite optimal in this regard, but it also did not include any asset classes like REITs, commodities, or bonds. By adding a mix of these to the Berkshire equity holdings, it would have come very close to estimates of optimal diversification.
How Has It Worked Out?
In the two years from September 20, 2006 through September 30, 2008, Berkshire Hathaway (BRK.B) has generated a cumulative return of 39% vs. -8.4% for the S&P 500. Back in the original analysis, QPP projected a standard deviation in annual return of 15.75% for the Berkshire portfolio (using three years of trailing historical data as input)—with the default assumption that the S&P 500 was have an annualized standard deviation in return of 15.1%. In other words, QPP projected that the Berkshire portfolio would be more risky than the S&P 500. This has been the case. In the past two years, Berkshire has been more volatile than the S&P 500.
The fact that a quantitative analysis of Berkshire’s equity holdings about two years ago suggested that Berkshire was well positioned both tactically and strategically has been of considerable interest to a range of readers. While it may not be rigorously significant, I took comfort that QPP “liked” Berkshire’s portfolio. In this case, QPP passed the test. The fact that Berkshire has substantially out-performed in the years since that article was published adds (for me) additional confidence in QPP’s results.
I am strongly in favor that more testing and validation is better and that single points of validation—like this Berkshire case—must not be treated as overly significant on their own. The interested reader will find plenty of historical “back tests” of QPP on our website that go over long periods of history. A model which generated wonderful results in testing must still bear up to real life scrutiny, and this example with Berkshire is a useful data point along the way.
QPP projected that Berkshire’s portfolio combined both Strategic Asset Allocation and Tactical Asset Allocation effectively—and these portfolio characteristics can be described by quantitative metrics. This is not to say that QPP is a ‘robotic Buffett’ but rather to suggest that investors will benefit by taking a more objective approach to both sides of the asset allocation challenge. The fact that the two years since the original article have Berkshire substantially outperform the S&P 500 (as QPP projected) may serve to motive some interest in this type of approach.