Plan Architecture and Portfolio Engineering

The articles listed here focus on Jacobs Levy Equity Management's philosophy of portfolio management, including the scope of the security selection/portfolio engineering problem, the goal of portfolio management, and the place of an individual portfolio within the investor's overall investment scheme.

As the “Security Selection” introduction noted, our process considers a wide range of return predictors designed to capture economic and behavioral effects, as well as company-specific information and events. But the power of these predictors can differ across different types of stock. The selection process must thus include breadth in terms of coverage of stocks, as well as return predictors. This does not mean that one should ignore the very real differences in price behavior that distinguish particular market subsets, or that one cannot choose to focus on a particular subset, such as value, growth, or small-capitalization stocks. It simply means that the model used for analyzing individual stocks should incorporate all information available from a broad universe of stocks.

“Engineering Portfolios: A Unified Approach,” which appeared as the lead article in a Special Technology Issue of the Journal of Investing (1995), discusses the many benefits of taking a broad, unified approach to the investment problem. Such an approach offers a coherent framework for analysis, one in which each stock in the universe has one and only one alpha, and in which each can be related to every other stock in the universe. A unified approach can also take advantage of more information than a narrower view of the market can provide. The effects of interest rates on value stocks, for example, may have repercussions for growth stock prices, which a focus on growth stocks alone would not indicate. Of practical importance is the fact that a broad, unified approach allows the investment manager to “engineer” portfolios designed to outperform various client-specified mandates.

A broad, unified approach, combined with the power of a security selection system based on an appropriate multivariate analysis of a large number of return predictors, allows for numerous insights into profit opportunities and improves the goodness of those insights; this in turn can lead to superior portfolio performance. The process of translating the insights into the performance is the process of portfolio engineering.

A portfolio optimization process that is customized to include exactly the same dimensions found relevant by the stock selection process helps to ensure that all the opportunities detected by the modeling process are exploited, while all the risks detected are accounted for and controlled. The aim of portfolio engineering should be to provide the maximum possible expected return for the desired level of risk.

“Residual Risk: How Much Is Too Much?” the lead article in the Spring 1996 issue of the Journal of Portfolio Management, considers the portfolio engineering problem within the broader context of the investor's risk policy. In particular, it demonstrates that the investor must factor into the portfolio selection decision the level of manager skill--the manager's ability to deliver incremental return for each unit of incremental risk taken. Taking too little risk may end up costing as much as taking too much!

· “Alpha Transport With Derivatives,” by Bruce I. Jacobs and Kenneth N. Levy, The Journal of Portfolio Management, May 1999; and abstracted in The CFA Digest, Fall 1999.(1) article
Investors can use derivatives to transport the excess returns available from the selection of securities within a given asset class or subclass to virtually any other asset class. For example, an investor can pursue the return possibilities in small-cap stocks, while using futures or a swap to neutralize exposure to the small-cap asset subclass and establish exposure to the large-cap segment. The investor can thus benefit from both the security selection opportunities in small-cap stocks and the asset class performance of large-cap stocks. Using derivatives in conjunction with market-neutral long-short portfolios can offer further performance enhancement.

· “Investment Management: An Architecture for the Equity Market,” by Bruce I. Jacobs and Kenneth N. Levy, Chapter 1 in Frank J. Fabozzi, Ed. Active Equity Portfolio Management. New Hope, PA: Frank J. Fabozzi Associates, 1998. Also in Fabozzi, Ed. Handbook of Portfolio Management. New Hope, PA: Frank J. Fabozzi Associates, 1998.
A blueprint of the U.S. equity market reveals three basic building blocks--a comprehensive core representing all U.S. equity issues; static style subsets, comprising large-cap growth stocks, large-cap value stocks, and small-cap stocks; and a dynamic entity reflecting differing relative performance in different market environments. Investment approaches, too, can be categorized into three groups--passive, traditional active, and engineered active. Engineered active management has the potential to provide the best match between client risk/return goals and investment returns, because it can offer consistent performance relative to the equity market core or its various subsets.

· “Residual Risk: How Much is Too Much?” by Bruce I. Jacobs and Kenneth N. Levy, The Journal of Portfolio Management, Spring 1996; and abstracted in The CFA Digest, Winter 1997. article
The optimal level of residual risk for a portfolio is the level that allows the portfolio to provide the highest expected return the manager can generate within the limits of the investor's risk tolerance parameters. As it is not always easy to determine investor risk tolerance or manager ability to add value, portfolios are often pigeonholed” according to residual risk levels alone. “Enhanced passive” or “index-plus” portfolios, for example, are expected to offer excess returns of up to 1% at residual risk levels not to exceed 2%. But such artificial constraints as a 2% bound on residual risk can lead to selection of suboptimal portfolios. In particular, they can lead investors to assume too little risk, hence allow too little expected return, for their actual risk tolerances, or to accept less skillful managers when more highly skilled managers are available. They may also encourage suboptimal manager behavior.

· “How to Build a Better Equity Portfolio,” by Bruce I. Jacobs and Kenneth N. Levy, Pension Management, June 1996.
Investors in U.S. equity can choose among a variety of selection universes, from the broad core including all stocks to various style subsets. They can also choose from a variety of investment approaches, from passive to traditional active to engineered active. Investors may be able to make more informed decisions if they understand the “architecture” of investing that links selection universes and investment approaches to their potential risks and returns.

· “Engineering Portfolios: A Unified Approach,” by Bruce I. Jacobs and Kenneth N. Levy, The Journal of Investing, Winter 1995; and abstracted in The CFA Digest, Summer 1996.(2) article
Many traditional equity managers focus on particular subsets of the investment universe--value or growth stocks, for example--and structure their portfolios from preselected groups. By contrast, a “unified” approach starts with a blank slate, having no built-in biases regarding any particular type of stock, and searches the widest possible stock universe and the largest number of investment variables. At the same time, it recognizes differences in stock price behavior across different types of stocks and over time, as well as possible nonlinearities in stock price response to gradations in exposure to a given variable. A unified approach to stock valuation is poised to take advantage of more information and to discover a greater number of potentially profitable investment opportunities. These opportunities are maximized by a portfolio optimization process that is customized along the same dimensions as the valuation process. This ensures a portfolio whose risks and return opportunities are balanced in accordance with the insights garnered from the unified valuation approach. Given its range and depth of coverage, a unified approach provides a firm with substantial flexibility to engineer portfolios to meet a variety of client risk/return requirements.

· “The Law of One Alpha,” by Bruce I. Jacobs and Kenneth N. Levy, The Journal of Portfolio Management, Summer 1995. article
Firms that use one valuation model for their core portfolio and different models for subsets of that core may end up with multiple estimates of alpha. But as every asset has only one price, doesn't it follow that the asset should have only one mispricing? It is argued here that it hardly makes sense for a single firm to begin the investment selection process with an approach that allows for the possibility of multiple mispricings for a given stock over a given horizon.

· “What's A Pension Officer to Do?” by Bruce I. Jacobs and Kenneth N. Levy, Financial Analysts Journal, March/April 1990.
Pension officers deliberating between the multitude of products available to them should apply four key criteria--the plausibility of the approach; whether the underlying concepts have been published and survived public scrutiny; the proprietary value of the insights and implementation; and the quality of the people involved.

· “Broader Indexes Widen Horizons,” by Kenneth N. Levy and Bruce I. Jacobs, Pensions & Investments, August 20, 1984.
The S&P 500 is not truly representative of the broader U.S. equity market. It is biased toward large-cap stocks, for example, and exhibits less earnings variability, growth and market variability than the broader universe. This has implications for passive investors in search of a proxy for the U.S. equity market return.

Other Research Categories:

Security Selection

Long-Short Investing

Portfolio Optimization Including Short Positions

Market Simulation

_________________________________________

(1)The Journal of Portfolio Management Special 25th Anniversary Issue.
(2)The Journal of Investing Special Technology Issue, lead article.