Investment Preference Theory
Overview
Los Angeles Capital Management employs an innovative investment process for security selection, portfolio construction, and trading that adapts to changing market conditions. Rather than making static assumptions about the characteristics of superior investments, the firm recognizes that investor preferences for specific risk characteristics evolve with changing economic and market conditions.
The firm's investment process addresses the fundamental questions, “How do investors view key investment risks in the current market environment? How are investor decisions influencing the price of risk?” Twenty five years of quantitative research suggests that shifts in valuations, business, and market conditions create new equilibrium prices for a variety of risk factors. Do investors care more about the risk of inflation or deflation? Forward looking earnings estimates or reported results? The safety of strong balance sheets or the opportunities of an improving economy on weaker balance sheets? These are just a few of the many factors we consider.
Whereas most pricing models (e.g., CAPM or APT) provide a framework for estimating risk and return based on a fixed, long-term definition of risk, Investor Preference Theory provides a flexible framework for estimating returns and risks as market conditions change.
Why Adopt a New Approach?
Most managers focus on the identification of characteristics which have generated excess returns historically. By the end of the 1970's, the dominance of small stocks led investors to conclude smaller companies were better able to adapt to changing market conditions and provided a small stock premium. In the 1980's, value's dominance over growth led to the broadly held view that investors overpay for growth stocks. A decade later, investors shifted their views as global prosperity, a benign inflationary environment and falling trade barriers created an environment where multi-national growth companies held a competitive advantage. By the end of the decade with the capitalization of internet companies reaching $2 trillion, investors declared that the information revolution had just begun, that the destiny of brick and mortar companies was bleak, and that the valuations of the post venture sector of the market were in fact warranted. Today, investors talk about the impact of higher oil prices as capital chases alternative energy sources just as it did thirty years earlier.
History suggests that investors are overly influenced by the past in forming their investment views regarding tomorrow's investment opportunities and market inefficiencies. What's the risk with such thinking? Embracing the consensus view as to why past factor returns will likely repeat, regardless of how rational the argument, frequently contributes to under-performance in subsequent periods.
The problem extends beyond the obvious one of chasing yesterday's winners. In a more subtle fashion, changes in market structure can impact an investor's risk management process, forcing higher exposures to risks which have become over-priced. This in fact happened to many portfolio managers during the 1990's, who, despite their misgivings over the valuations of internet companies, succumbed to the pressures of risk management, buying internet companies at the peak to decrease their previously unprofitable bet against the sector.
In 2003, investment managers unprofitably increased their exposures to higher earnings quality companies just as the lowest quality companies rebounded on improved economic and business conditions. Today investors continue to increase allocations to real assets despite the historic run up in commodity prices over the past five years.
What are Investor Preferences?
Investor preferences include any characteristic which investors believe represent either a favorable or an unfavorable risk. Over the market cycle, investors assign values or prices to a variety of risk characteristics which include market factors (e.g., beta, market capitalization, price momentum), income statement and balance sheet measures (e.g., E/P, CF/P, B/P, ROIC, operating margins), analyst expectations (e.g., estimate revision, earnings surprise, Torpedo (stocks with the highest earnings expectations)), and sector risks (e.g., Technology, Internet, Health Care, Biotechnology, Energy, Basic Materials, Telecom, Services, Utilities, and Retail.)
To better understand how these preferences change through time, one may look at historic factor returns. By analyzing these returns it becomes quite clear that investor preferences are not random, but follow patterns as investor views evolve over the market cycle. Price movements are in fact quite similar to the more familiar pricing behavior of systematic risks in the fixed income markets, such as credit and prepayment spreads or duration risk.
Who Shapes Investor Preferences?
Investor preferences are shaped by Wall Street analysts, academics, institutional separate account managers, retail mutual fund managers, and private equity investors. While stock prices may display random walk characteristics, investor preferences evolve through time, creating discernable pricing patterns.
Research shows that investor preferences are generally applied across the market with one noticeable exception. While capital flows more freely across the growth and value spectrum, plan sponsors generally control the allocation between large and small cap stocks. In fact, given the segmentation in the management of large vs. small capitalization portfolios, it is not surprising that preferences would differ.
How Can Investor Preferences Be Measured?
For twenty-five years, quantitative managers have measured factor returns through a cross sectional analysis of monthly stock returns. By disaggregating common factor and sector effects, as much as two-thirds of a stock's extra-market return may be explained for any period. While traditional methods are useful for performance attribution and risk measurement, L.A. Capital has identified a number of enhancements to extract greater information content from the analysis, providing more stable forecasts.
The key enhancements include: estimating returns independently for large and small cap stocks, removing non-linear effects, minimizing the impact of data outliers (univariate and multivariate) on forecasts, adjusting forecasts for model and measurement errors, mitigating the effects of exogenous events, and dynamically changing sector classifications as warranted by changes within the industrial composition of the market.
Once factor returns have been estimated for the previous week, the velocity, acceleration, and statistical significance of each series is examined over the recent past to determine the new equilibrium price of each risk. After making adjustments for both measurement and model errors, this information is translated into discrete alphas for each security in our universe. This alpha is derived by multiplying each security's exposure to forty-five risk characteristics, by the price of each risk.
How Accurate is the Forecasting Process?
The following charts look at projected vs. actual returns for three factors: Profit Margin, Debt-to-Equity, and return on invested capital (ROIC). The solid line depicts the normalized ex-post factor return, the grey area the standard error around the return series, and the dotted line represents the forecast which is derived from a factor's velocity, acceleration, and statistical significance.
The following chart shows that companies with higher profit margins have been consistent outperformers over the past decade. While traditional theory states that higher margin industries and businesses will attract capital and that any advantage should be priced into the stock removing the opportunity for excess returns, this clearly has not been the case over the past decade.
The next chart looks at the pricing of leverage. What is most interesting about the graph is its high correlation with Fed policy. When the yield curve is flat due to Fed tightening, leverage receives negative pricing. When the yield curve is steep as the result of Fed easing, leverage is priced favorably. While intuitive, rarely do public equity managers discuss balance sheet structure as a key determinant of a stock's expected return. For private buyout managers, this is their focus.
The last chart shows the pricing for return on invested capital which is a classic cyclical factor. While the longer-term trend has favored companies which invest their capital more efficiently, clearly there are periods when less profitable companies are deemed to have superior prospects, leading investors to de-emphasize the metric in such an environment. With the spike in market volatility from turbulence in the fixed income markets and an increasing inflation threat, investors have again shown a strong preference for stocks with high ROIC.
By not making judgments on the long-term return or fair market price of any particular factor, not only is the risk of being wrong on any individual view mitigated, the investment process does not require the future to resemble the past for success. No investor can consistently anticipate tomorrow's market conditions. Embracing an adaptive process, however, enables an investor to consider new developments not previously experienced.
What is Unique about L.A. Capital's Process?
Understanding how investor views are changing regarding risk and return opportunities is the cornerstone of the firm’s investment philosophy and drives the research, stock selection, portfolio construction and trading process. Twenty-five years of experience managing institutional portfolios has provided unique insights which have been built into the process to generate consistent results.
- Focus on the Factors Important to Investors. The research team monitors three information sources to better understand the types of information investors believe to be important: (1) Wall Street and vendor research, (2) academic research, and (3) private equity research. Wall Street research, both quantitative and traditional fundamental research, remains the best source for understanding what institutional and retail investors find valuable. Academic research helps keep the firm abreast of new valuation techniques, while following the private equity markets helps the firm better understand new corporate strategies for increasing value to shareholders.
- Diversify Bets Across a Large Number of Investor Preferences. Because risk is multi-dimensional, focusing on a broad number of factors decreases model risk which includes:
- The risk of being over-confident on a particular view
- The risk of unexpected events creating negative results
- The risk of being early or late on a particular view
- Blend Qualitative Thinking with Quantitative Analysis. Traditional research helps define the problem while quantitative methods help extract better information from the analyses. The investment team purposely avoids over fitting problems to the past and does not simply estimate future mean factor returns based on long-term history. L.A. Capital is also careful to avoid forecasting time horizons for factor returns based on historical cycles, and does not tailor forecasting tools for individual factors based on historic explanatory power. Most importantly, the modeling process does not weight factors based on a "best fit" analysis to gain the highest historic information coefficient. In fact, factors may be added to the model before they are statistically significant if the research team makes a determination that they represent an important investor preference in the current environment. The focus of L.A. Capital’s quantitative analysis is to remove noise and the effects of exogenous events in the hopes of gaining a better understanding of investor preferences.
- Adopt a Dynamic Approach to Risk Management and Trading. Managing returns or risk in the rear view mirror frequently leads to undesirable results. Understanding and recognizing that no active management process works in every possible environment will lead to better and more stable results. As such, the portfolio construction and risk management process seeks to understand both changing return opportunities and levels of risk. By maintaining generally low levels of active risk relative to the client’s benchmark, style risk is maintained, enabling sponsors to better manage unwanted style risk across their multiple manager teams. For trading, the implementation team embraces neither an aggressive nor a patient strategy but employs a proprietary “wave optimization” algorithm that is more aggressive when prices and liquidity are favorable, but is patient when they are unfavorable. This dynamic approach to trading has led to a 20% reduction in implementation shortfall across all portfolios.
What is L.A. Capital's Comparative Advantage?
The firm believes it has two comparative advantages: a proprietary investment philosophy based upon our insights, and superior investment technology. Both advantages are the result of the three decades of experience and the tenure of the investment team together.
The investment philosophy is contrarian to conventional philosophies which describe static market inefficiencies. We believe that the nature of inefficiencies is always changing and that prices of key risks are not based on the past but based on current market conditions. Estimate revision, sales to price, earnings quality, Telecom, Internet, and Energy are all examples of risks which have been significantly re-priced over recent time periods.
The firm's investment technology has been developed under the leadership of its Director of Research, Dave Borger, CFA. Mr. Borger has focused on the development of quantitative tools for security selection, portfolio management and trading since 1977 when he started his career in the Research Department at the National Bank of Detroit. The National Bank of Detroit is also where Mr. Borger began his collaboration with Tom Stevens. The research effort is focused in three areas:
- Improving the attribution and forecasting models to more accurately identify the forward looking price of each investment risk;
- Enhancing the portfolio construction process to both eliminate unwanted risks and maintain appropriate exposures to risks which have a high probability of offering reward; and,
- Developing trading algorithms to trade risk exposures at the lowest possible costs.
Summary
Investor Preference Theory acknowledges the dynamic nature of the equity markets. By not embracing the past, the portfolio management process is able to adapt to changing market conditions. Since its implementation a decade ago, the process has successfully priced key investment risk over dramatically different market environments.
Generating consistent performance requires that managers consistently identify a changing set of investment opportunities. The one thing investors can say about the decade to come is that it will present challenges and opportunities beyond what investors can imagine today.
