Advisor Perspectives welcomes guest contributions. The views presented here do not necessarily represent those of Advisor Perspectives.
Behavioral portfolio management (BPM) is based on two categories of financial market participants: emotional crowds and behavioral-data investors (BDIs). Emotional crowds are investors who base decisions on anecdotal evidence and emotional reactions to unfolding events. Human evolution hardwires us for short-term loss aversion and social validation, which are the underlying drivers of emotional crowds. On the other hand, BDIs thoroughly and extensively analyze behaviorally-driven price distortions and build portfolios based on these distortions.
Four weeks ago, I introduced the concept of behavioral portfolio management (BPM) as a way to build superior portfolios. Three weeks ago, I discussed the first basic principle underlying BPM: Emotional crowds dominate market pricing and volatility. The next week, I presented the second basic principle: Behavioral-data investors can earn excess returns. Last week, I introduced the third and final basic principle: Investment risk is the chance of underperformance.
In this final installment of my five-part series, I show how advisors and investors can implement a BPM-based strategy.
Implementing BPM
There are three key steps to implementing BPM: redirectingyour emotions, harnessing market emotions and mitigating the damage of client emotions on their portfolios. The first and third steps must be accomplished in order to successfully implement the second step. Many investment firms provide excellent materials to aid advisors in helping clients avoid emotional errors and improve the investment decision process.
But beyond an inventory of common emotional mistakes and antidotes, not much is available regarding how to harness market emotions. This is an important omission. Emotion-harnessing portfolios are key to earning superior returns. This article illustrates how to create them.
BPM-based asset allocation and portfolio construction
The standard approach to portfolio construction is to maximize return for a given level of volatility. This is often referred to as a risk-return analysis. In last week’s article, I argued that the typical measure of risk – volatility – is really a measure of emotion. So risk-return analyses are really emotion-return analyses. To avoid placing emotionally charged volatility at the center of asset allocation, we need to sideline it to the greatest extent possible.
BPM-based asset allocation uses a personal endowment approach to portfolio construction, a topic addressed in a recent article. Endowments are faced with the dual charge of providing an annual income stream to a university or other institution as well as growing the portfolio over a long-term horizon. To a large extent, endowment managers are insulated from the short-term performance pressures facing many other investment managers. For this reason, they are able to construct the best portfolios for meeting the dual charge of regular income and long-term growth.
Endowment fund behavior provides the basis for BPM-based asset allocation.
The first step is to divide the client portfolio into three buckets: short-term income and liquidity, capital growth, and alternatives. The short-term bucket is invested in low- or no-volatility securities that are sufficient to meet the client’s short-term needs with virtual certainty. This removes volatility from conversations regarding this bucket.
The capital growth bucket is built to maximize long-term wealth. Since the investment horizon is long for this bucket, the focus should be on expected and excess returns. Endowment funds do exactly this by overweighting the asset classes with the highest expected returns. Endowments heavily weight equities, with very little invested in bonds.
A significant challenge facing advisors is that clients have difficulty thinking long-term, as they are hardwired for short-term loss aversion. Instead of a 30-year horizon, for example, they see a series of 30 one-year time frames or a series of 120 one-quarter time frames. In each period, they apply short-term loss-aversion criteria. Some clients have difficultly staying the course with high-return, volatile investments such as stocks. Short-term loss aversion can undermine capital growth portfolio performance, as clients can make decisions based on current market volatility.
The obvious answer is to discuss investment performance infrequently, maybe once every 30 years. But regular meetings are an important part of client service, so the challenge is to talk to clients without triggering the emotions associated with unavoidable market gyrations. Two possible remedies are making investment performance a small part of the regular client meeting and emphasizing the long-term nature of the capital-growth portfolio. Another is to phase in and out of investments, so that a single price or total value does not become an anchor upon which the client focuses.
The alternative bucket contains those investments that do not fit into the other two, such as houses, favorite stocks, illiquid investments, jewelry and artwork. These are managed based on the unique features of the assets and as directed by the client.
The major benefits of breaking the portfolio into three buckets are sidelining volatility as a client issue and being able to construct each bucket to meet specific client needs. Volatility, correlations and other commonly used statistical measures, such as downside risk, play a diminished role in BPM-based asset allocation and portfolio construction. Instead, expected and excess returns are most important.
BPM-based fund selection: Strategy, consistency and conviction
Once asset allocation decisions have been made, the next step is to select the funds in which to invest.
The most common criterion for selecting equity funds is past performance. Funds that have performed well in the past feed on the emotional belief that they will perform well in the future. In fact, the most popular fund-rating system, Morningstar’s star system, is based on 3-, 5-, and 10-year past performance. There is a big problem, however: past performance is not predictive of future performance. This has been confirmed by numerous statistical studies. The fact that everyone in the industry continues to use past performance, in the face of overwhelming evidence against its usefulness, is a testament to its powerful emotional appeal. Counterproductive emotional habits are nearly impossible to break.
Rather than using past performance, BPM focuses on key manager behaviors: strategy, consistency and conviction. Strategy is the way a fund goes about earning superior returns through analysis, buying and selling. The strategy must be pursued consistently through time. The fund will move about the investment universe (based on its asset class mandate) in order to identify the most attractive securities in response to ever-changing economic and market conditions. Finally, the fund should take high-conviction positions in its best investment ideas.
These fund behaviors can be objectively measured and used to identify best-performing funds going forward. My firm, AthenaInvest, has done this for about 3,000 U.S. and international active equity mutual funds domiciled in the U.S. Average fund returns, since 1997, are reported in Figure 1, based on our a priori diamond rating (DR). The two highest-rated fund groups, those with the highest level of consistency and conviction, each outperformed the benchmark, while the two lowest-rated fund groups each underperformed and the middle-rated funds generated benchmark-equaling returns. The top diamond ratings are comprised of the most active funds, while the bottom is made up of closet indexers. On average, there is a gain of 1% in annual performance per diamond rating as we move from closet indexers to truly active managers. (See my article for more details.)
As Figure 1 demonstrates, active equity manager behavior is predictive of performance, while past performance is not.
Based on subsequent monthly returns for beginning of the month U.S. and international strategy identified, Diamond Rated (DR) active equity mutual funds April 1997-March 2012. DR is based on strategy, consistency and conviction, with DR5 being the highest on both scales and DR1 being the lowest. Fund returns are net of automatically deducted fees. The Benchmark is the MSCI All Country World Index. Data sources: AthenaInvest and Thomson Reuters Financial.
BPM-based stock selection: Best ideas of the best managers
In a previous article, I presented evidence that the top picks of active equity mutual fund managers earned superior returns. I argued that these were the result of fund managers (i.e., BDIs) taking high-conviction positions in stocks that were mispriced due to emotion-driven price distortions.
The direct way to tap into these behaviorally driven returns is to develop an investment strategy and manage a portfolio based on it, as active equity managers do. The evidence regarding individual investing success is mixed, though it is compelling for a fund’s best idea stocks.
Stock-selection skill can be measured by identifying the best ideas of the best managers. The best funds are those that are most strategy-consistent while at the same time taking high-conviction positions – the DR4 and DR5 funds described above. The stocks most held by those top funds are designated the best ideas of the best managers. (See my article for more details.)
The best idea results are reported in Figure 2. The best idea stocks, based only on data available at the beginning of each month, generated an ex post annual return that was 7.7% higher than the Russell 3000 index return (16.9% versus 9.2% from April 2003 to March 2013). The best-idea stock portfolio (made up of approximately 400 best ideas out of a DR universe of 5,000 stocks) represented the full range of market capitalizations, justifying the Russell 3000 as the benchmark.
The 7.7% best-idea return advantage exceeds the 2% return advantage of the best funds (i.e. DR5 funds), indicating that even the best funds hold a large number of non-best-idea stocks. Part of the difference is attributable to the average fund fees of 1.3%. But even accounting for these fees, best-idea stocks clearly outperformed the rest of the stocks held by the fund (a result confirmed by Cohen, et al.). This is further evidence that fund managers are superior stock pickers compared to the average investor and that BDIs are able to take positions in stocks characterized by emotionally driven price distortions.
Includes month beginning DR5 U.S. stocks for April 2003-March 2013, resulting in an average of roughly 400 U.S. stocks being held out of the DR universe of approximately 5,000 U.S. stocks. Subsequent monthly returns are simple averages across the stocks held. DR5 stocks are the best idea stocks of the best managers. Data sources: AthenaInvest, Thomson Reuters Financial, and Lipper
It may seem puzzling that active equity managers are superior stock pickers on the one hand, while on the other hand, they hold large numbers of non-best-idea stocks. The combination of incentives and investor behavior explain this inconsistency. Funds are strongly encouraged to grow, as they are paid a fee based on AUM. When they are small, it is easier for funds to hold concentrated portfolios of best-idea stocks, but as they grow, it becomes harder to stick with best-idea stocks. Many funds transition from BDI strategies to catering to investor emotions. Berk and Green argue that this is rational profit-seeking behavior on the part of funds.
BPM-based market selection: Which strategies are investors rewarding?
It is well known that returns from being in the right market at the right time dramatically exceed the returns from even the most successful stock-selection strategy. Along with investor’s short-term loss aversion, this explains why tactical market funds are so popular these days. Many of these are based on short-term price momentum and mean reversion. These patterns tend to be transitory in nature and thus are challenging to implement successfully. Another problem is that they appeal to investor’s short-term loss aversion, so it may be hard to determine if they are really generating superior returns or simply represent emotional catering.
When investors make cognitive errors that impact the market as a whole, the resulting price distortions are often measureable and persistent. A key is to identify objective measures of these distortions rather than relying on survey data, which is notoriously unreliable. One must understand what investors are doing, rather than what they are they saying. One of the first such measures was Baker and Wurgler’s sentiment index. The index is based on six objective measures of investor sentiment, such as the closed-end fund discount. The index is predictive of when small-capitalization stocks will outperform large-capitalization stocks and vice versa. Baker and Wurgler find that the more pessimistic investors are, the better it is for small stocks and the market as a whole. Investor optimism is a stock market return killer.
My firm has created two other measures of investor sentiment. Using the returns for each of the 10 U.S. and international equity strategies, we created a predictor of future U.S. and international market returns, dubbed market barometers. Both barometers are based on recent relative strategy return ranks versus long-term return ranks. Based on these comparisons, the U.S. and international markets are each separately rated strong, normal or weak (see my article for more details).
By combining the sentiment index with the U.S. and international market barometers, it is possible to implement a global tactical model that trades among U.S. large-cap, U.S. small-cap and international stocks, as well as cash. We have implemented the best markets methodology using a 100% investment in long or double-long S&P 500, Russell 2000, EAFE exchange-traded funds or Treasury bills for cash investments.
Trades into a 100% single-long or double-long exchange-traded fund for the S&P 500, Russell 2000, or MSCI EAFE or Treasury bills based on beginning-of-the-month U.S. and international strategy market barometers and modified sentiment index. Returns since September 2010 are GIPS-complaint actuals, with prior returns back-tested using the same month-beginning methodology as for the actual results. Data sources: AthenaInvest, Thomson Reuters Financial, and Lipper
The 10-year best-market results are reported in Figure 3. The best-market portfolio yields a 17.4% return advantage over the MSCI AC World Index return (26.8% versus 9.4%). This advantage is driven by being in the right market at the right time (of particular interest, it was invested in cash during most of the 2007-2009 downturn) as well as the timely use of leverage when behavioral measures signaled a strong market. As expected, the best-market return advantage is more than twice that of the best-idea stock advantage (17.4% versus 7.7%).
The resulting portfolio is not traded very actively, by tactical standards, with a 100% trade every nine months on average. This reflects the measurable and persistent market-wide investor behavior currents being captured by these measures.
Conclusion
BPM is based on the notion that if the advisor can redirect his or her emotions and mitigate the impact of client emotions, it is possible to build superior portfolios by harnessing market emotions. This article describes how this can be done and presents evidence of the superiority of focusing on investor behavior when constructing and managing portfolios.
The overall results are presented in Figure 4 for April 2003 through March 2013. They demonstrate the advantage of focusing on behavioral factors when constructing long-term portfolios, selecting the best funds, picking the best stocks and investing in the best markets. The return advantage grows from 6.9% by staying in the stock market versus investing in Treasury bills, increases another 0.9% by investing in the best (i.e. truly active) equity mutual funds, another 7.5% by investing in the fund’s best-idea stocks and another 9.9% by investing in the best markets.
Each of these return enhancements is based on currently available data that allow us to measure persistent behavioral factors. This data can in turn be used to build superior portfolios. The reward for harnessing these factors is worth the effort of redirecting your emotions while mitigating the impact of client emotions on their portfolios. This is the ultimate promise of behavioral portfolio management.
See footnotes in previous figures for more information on how each return is calculated. April 2003-March 2013. Data sources: AthenaInvest, Thomson Reuters Financial and Lipper
C. Thomas Howard is Professor Emeritus, Reiman School of Finance, Daniels College of Business, University of Denver and CEO and Director of Research, AthenaInvest, Inc. Contact information: [email protected]; (877) 430-5675 x100. A longer version of Behavioral Portfolio Management can be obtained at the Social Sciences Research Network website.
Read more articles by C. Thomas Howard, PhD
