Fixed-income attribution: Assessing readiness before the game starts

The purpose of a fixed-income attribution solution is to break down the total return of a fixed-income portfolio into returns of risks common across investments, and to compare and contrast fixed-income attribution results against stated portfolio investment strategies and styles.

For large institutional investors, fixed-income attribution has become an increasingly integral part of the investment management process. Growing acceptance and advances in the body of knowledge on FIA have prompted the emergence of several commercial solutions, as well as reports of successful implementations.

However, a number of institutions have rushed to implement FIA projects, only to be disappointed by project delays, higher than expected costs, or both. While fixed-income attribution practitioners invariably point to data quality as the scapegoat for project setbacks, the true causes run deeper and broader. Regardless of firms' differences in investment mandate/strategy, business model, and commercial and regulatory circumstances, those that have stumbled in FIA projects can usually trace mishaps to the early stages during which the business requirements are defined and analyzed, and when implementation strategies are set.

In large institutional investment firms, fixed-income investment management is a complex, multilevel decision-making process. For firms employing structured portfolio management methodology, there are at least two levels of decisions: total portfolio-level decisions, which involve allocating capital and/or risk across strategies, and managing aggregate, portfolio-level risks; and decisions made by individual managers specializing in different fixed-income investment segments or strategies.

In the final analysis, a well-designed, implemented and truly representative fixed-income attribution system should explain a portfolio's excess return in a way that recognizes the discrete responsibilities and strategies allocated among various decision-makers in the firm. Therefore, in presenting FIA results for a fixed-income investment strategy, the returns of aggregate risk factors outside of the manager's return and risk accountabilities should be removed from the total return of that strategy.

For example, in a fixed-income fund employing duration overlay strategy, the return of a corporate credit manager needs to be measured on a duration-adjusted basis — the removed excess duration returns should be allocated to the fund's overlay strategy for measuring its efficiency. A well-designed business-requirements collection and analysis process, therefore, would require the firm to first delineate risk factor returns in accordance with the fund's actual investment management framework — in short, “define the rules before the game starts.”

Many firms might find measuring returns and attributions for investment services, such as cash management or risk overlay, challenging, as the returns and attributions of an investment program must be split across several managers in accordance with internal return accountabilities.

For example, to measure the effectiveness of a duration overlay strategy, which manages the duration risk of the total portfolio, fixed-income attribution would require not only require examining portfolio-aggregated duration exposure and duration policy changes (targets, rebalancing dates), but also the duration returns from all strategies. Some firms might not have this kind of data in place prior to the launch of the FIA initiative.

Another source of complexity stems from the fact that for many firms, investment book-of-record systems, which provide critical information to fixed-income attribution solutions, were established when most risk exposures were captured in cash-market investments, and leverages could only be accomplished from borrowing and lending in cash markets. These legacy systems, while adequate in providing input for total return calculations at portfolio and strategy/manager levels, might fail to deliver reliable and necessary information at the instrument level. Fixed-income attribution teams unprepared for this fundamental difference between total return and factor-based attribution tend to underestimate the effort in their project planning and scheduling.

Widely employed derivative strategies by many firms make the deficiencies of legacy investment book-of-record systems more acute. From the FIA requirement point of view, information provided by an investment book-of-record system must reflect accurate risk exposure, embedded leverage and funding cost so that a derivative investment's performance can be measured consistently with its cash-market underlying or equivalence. For example, in a TBA and cash mortgage-backed securities strategy, the return of the TBA must be compensated by the return of cash or short-term investments, which are usually managed by cash/treasury manager(s). However, prior to fixed-income attribution initiatives, many firms probably did not view this as a requirement, and most legacy investment book-of-record systems will find this awkward, at best, to accomplish.

If operating on inadequately collected business requirements, a fixed-income attribution project implementation strategy then boils down to a judgment call of the project team in determining what metrics to include or emphasize — resulting in suboptimal implementation strategies.

For example, should the firm try to incorporate a list of fixed-income risk factors required by both current and planned investment strategies, or just focus on key risk factors that affect the majority of the firm's strategies? Newcomers to the fixed-income attribution game might feel the urge to be “complete,” only to realize later on that some ancillary risk factors have an insignificant impact on explaining investments' returns. For small organizations, devoting resources to chase returns from second-order fixed-income risk factors, such as MBS spread and interest rate volatility, is itself an investment with diminished returns. Limited resources are better deployed toward delivering reliable and consistent fixed-income attribution results on key risk factors, such as key rate durations, spread durations and effective yields, that explain a large share of investment returns.

A successful fixed-income attribution program provides valuable feedback not only to a firm's investment management professionals and asset owners, but also to the firm's senior management. Naturally, fixed-income attribution results find their way into firms' staff review and compensation process. Therefore, collecting FIA business requirements can be politically charged and personally sensitive. FIA business analysts responsible for requirement gathering and analysis need to be armed with adequate subject knowledge and interviewing skills, be aware of a firm's organizational culture and act in a consistent, disciplined and professional manner. A well-designed questionnaire approved by investment management leadership would greatly prepare a firm to launch this critical endeavor.

A high-quality fixed-income attribution business requirement document is a cornerstone for a successful implementation. The process of collecting and analyzing the requirements benefits not only investment management but also other investment processes such as investment book-of-record keeping and data management. In addition, the document prepares the firm in presenting and interpreting results from the fixed-income attribution solution for both internal and external audiences, ultimately to reap the maximum returns on the investment of an FIA system.

Bai Gu is the founder and president of Investment Measurement Services Inc., Duluth, Ga.