Oxford Associates

©1998, William N. Goetzmann

Bruce Davis works for Oxford Associates, an investment consulting firm in Boston, Massachusetts. His client is Xerxes pension fund, a three billion dollar defined benefit pension plan for a firm that makes office business machines. Oxford has been hired to evaluate the performance of their equity managers, many of whom have been with Xerxes for most of the 1990's.

Over the past several years, the managers have each contracted with Xerxes to be compensated based upon whether they beat a style benchmark. Fixed fees are paid as a percentage of assets each year and an incentive fees is paid on the return achieved above the benchmark. One manager, Euclid Advisors, has a substantially larger incentive fee, however Euclid has a high water mark provision, which means that they must make up any loss incurred in previous years before they receive incentive payments.

Davis is charged with determining whether the benchmarks used are appropriate ones. In addition he has been asked to develop a quantitative basis for evaluating the manager's performance.

One particularly challenging problem he faces is that some managers are market timers, and Xerxes would like to determine whether they have timing skill.

Data: Returns for all funds and rolling weights for one fund are available in the tables.

Return data spreadsheet (same as Table III below)

Analyzer Folder

Attribution Folder

Rolling Sharpe coefficient weights for Brinson Partners (same as Table II below)

Using the data available in the case, and the Ibbotson Attribution software, answer the following questions.

1. Which managers have positive risk-adjusted returns? Answer this using the Sharpe ratio, the Jensen measure and a multi-factor model of your choice. Do any funds have significantly positive alphas?

2. Do the managers' earn their fees? Can you answer this question with certainty? Assume that the returns you obtain from the attribution software are net of fees.

3. Are the benchmarks appropriate to the managers?

4. Do any of the tactical managers have timing ability?

5. Consider Brinson Partners Global Securities Fund.

Has its performance been due to timing or superior security selection?

Brinson's implicit time-varying portfolio weights, based on 3-year rolling Sharpe coefficients are provided below. Are the EAFE weights correlated to leading or lagging returns to the EAFE index? Why would you care?

6. What is the source of Euclid's performance? Can you figure out what they invest in?

7. Would you keep all the current managers? Would you allocate funds equally across them? Do any of them provide any special diversification benefits.

Table I.   Managers, Dates they started, Benchmarks and Fees




Manager					Start Date	Benchmark	Fixed &Incen Fees

Brinson Partners  Inc.   Global Securiti		 3/82	EAFE Index              »% +2%
Equinox Capital Mgmt     U.S. Equity    		 3/86	S&P 500                 »% +2%
Euclid Advisors  Inc.    Euclid Advisors		 3/90	T-Bills	                »% +25%
Fidelity Mgmt Trust Co.  Select Small Ca		 6/89	Small Cal Value		»% +2%
First Quadrant Corporation Global TAA     		 3/89	50%/50% EAFE/TB		»% +2%
Harris Investment Mgmt   Systematic Equi		 3/86	S&P 500			»% +2%
J.P. Morgan Invt Mgmt    Tactical Asset 		 12/85	50%/50% S&P/TB		»% +2%
Lord  Abbett & Co.       Value Equity   		 3/77	Large-cap Value		»% +2%
PanAgora Asset Management Tactical Asset 		 12/85	50%/50% S&P/TB		»% +2%
Renaissance Investment Mgmt Three Way Asset	 	 3/78	50%/50% S&P/TB		»% +2%

Table II. Brinson Partners Global Securities

Rolling Estimated Sharpe Coefficients

Estimated using 12 leading quarters of returns

Table III: Returns Of Funds