Fixed-income portfolio management has become a complex business in which managers and clients trade off competing goals and objectives. The trade-offs involve benchmarks, manager constraints, risk tolerances, and alpha production. All of these factors require quantitative modeling. In fact, navigating the modern bond market’s fundamental issues is difficult for someone who is not a quantitative manager.

Quantitative Management of Bond Portfolios  provides clear, empirically based solutions to many of the practical challenges of running a bond portfolio. It is not an introduction to the subject, but most investors with a grasp of basic fixed-income math will find the book readable and useful. The authors are not academics but practical experts whose professional duties require them to make their research comprehensible to buy-side clients.

Much of the book’s content has been presented previously to investors through Lehman Brothers research. Here, however, the major topics are brought together and organized according to their common themes. Some of the ordering of the sections could be better (e.g., putting benchmark replication first), but this is a minor complaint.

The book is divided into two major parts: (1) empirical studies of portfolio strategies and (2) benchmark design and portfolio management tools. The first part, which is the stronger of the two, is further divided into a section on broad benchmark issues and a section on managing specific portfolio performance factors, such as credit risk and mortgage prepayments. The portfolio management discussion focuses on risk budgeting and on performance attribution, which warrants a whole book in its own right.

In the first section, the authors show how indices can be replicated with a minimum of tracking error through the use of derivatives. This replication framework is applied to many different benchmarks, which allows the reader to develop a feel for how tracking error changes in response to the level of unsystematic risk. The authors also progress beyond the simple dynamics of index replication to the more complex issues associated with benchmark customization.

Quantitative Management of Bond Portfolios  highlights credit risk at the portfolio level, which is truly an underresearched area relative to the effort many fixed-income managers devote to credit analysis of individual companies. The authors argue persuasively that credit decisions and mortgage security selection have a large bearing on fixed-income performance, even though duration may represent the single biggest portfolio risk. The chapters on credit portfolios address the degree of diversification that is necessary for effective management, the risk of extreme moves, and the impact of financial distress. With mortgage portfolios, the authors focus on the difficulties of measuring duration risk. All of these discussions take the reader beyond the basic task of analyzing a security to address the intricacies of building and managing a fixed-income portfolio. This extra level—analysis of portfolio structuring—is what makes this book stand head and shoulders above all rivals.

In the second part of the book, the authors present their multifactor risk-budgeting model, which takes into account the potential for exploiting manager skill. They describe and quantify the major factors driving fixed-income returns and examine the key trade-off between return and risk. This section is especially useful because risk budgeting has not been as fully developed or as widely embraced in fixed income as in equity. In admirably accessible language, the authors provide a comprehensive guide to global fixed-income management, including how to model risk and conduct performance attribution.

These tools highlight the concepts underlying the fundamental law of active management, which is essential to enhancing return relative to a benchmark. The fundamental law states that the information ratio for a portfolio (that is, the manager’s excess return divided by the amount of risk taken relative to a benchmark) is a combination of skill times breadth, or number of active bets taken. The authors take this framework, so well developed for equity markets by Grinold and Kahn, and rigorously apply it to bonds. ¹  The result is not simply presentation of ideas but powerful analysis of what actually works.

In this approach, the major types of fixed-income portfolio risks—duration, sector differences, and security selection—are measured to quantify potential gains. Within this useful empirical framework, key investment management questions are addressed. For example, is it more effective to run a portfolio top down, with duration and sector bets, or bottom up, through security selection? The authors make a convincing case that security selection is the more rewarding method because it offers a wider range of bets than trading on intersector relative value or the direction of interest rates. Choosing the strategy that generates the most chances of success, given a certain level of skill, makes the most sense. Even successful top-down managers can be made better by adding some skill in security selection, as opposed to enhancing existing skills. The authors also show convincingly that there are limits to how much return can be achieved through a given strategy.

The authors also confront the challenges of benchmark replication and return enhancement relative to a benchmark. Because fixed-income indices contain so many securities, measuring the value added by portfolio components is much harder for bond than for stock management. Measuring tracking error and determining the number of securities needed to match an index are critical tasks. Helped by their strong command of benchmark structure and risk budgets, the authors quantify the cost of portfolio constraints quite effectively. For example, they produce measures of the impacts of leverage, of varying the credit exposures, and of constraining the allocation to mortgages.

The authors generate these results through the systematic use of “imperfect foresight” modeling and indexing dynamics. Empirical imperfect foresight modeling provides valuable insight into how returns can be generated, especially for complex portfolios, at varying levels of skill. This novel approach calculates returns over a stated period on the basis of an assumption that an investor can predict future returns perfectly and then compares those returns with random returns. From these extreme scenarios, an analyst can calculate returns for various skill levels as a percentage of returns that could have been earned with perfect foresight.

Despite the authors’ shared background at Lehman Brothers, the quality of the chapters is somewhat uneven. The book could have been shortened without sacrificing any of the essential points. For example, the article on managing central bank reserves is probably not a topic for a broad audience. Additionally, the portfolio tools section could have been folded into the first section of the book without much loss to the reader.

Footnotes and references are limited, and the book manifestly emphasizes the Lehman Brothers approach to the markets. There is an advantage, however, in having access to Lehman’s index database. Moreover, because it reflects the authors’ years of experience in answering practical questions from clients, the advice is well considered, with a basis in the market’s standard for benchmarks and the support of solid empirical evidence.

At 978 pages, this book is weighty. It justifies the reader’s effort to tackle it, however, by realizing the authors’ goal of creating an all-encompassing tool kit for managing fixed-income portfolios. No other book on the subject is comparable to this one in depth and breadth.

—M.S.R.