Quality Money Management : Process Engineering and Best Practices for Systematic Trading and Investment

positions. Implied volatility of poor credit companies would spike, stock price of poor

credit companies would gap down, and you would look like a genius.

However, using a portfolio attribution system you would clearly identify that the

excess profit of above the benchmark for the volatility was provided by a credit bet. In the

bet explained above the trader won; however, the reverse could also be true. Therefore, to

make sure that a trading system built consistently outperforms its benchmark you need to

perform attribution analysis on the portfolio and the benchmark.

For attribution analysis, a risk manager must understand where the risk and return is

versus the benchmark. Metrics will include first standard statistical outputs, plus addi-

tional deliverables such as, for example, a linked chart showing the growth of $10,000 for

fund versus the benchmark, and calculations for:

● Alpha and beta.

● Excess returns/Sharpe ratio of excess returns.

● Average credit ratings.

● Average market cap.

● International exposure.

● Number of weeks top ranked stocks out-/underperform lowest ranked.

● Brinson–Fachler attribution metrics.

All of these calculations are useful for comparing how your trading/investment system is

performing relative to the benchmark. For charting purposes, the mean value of these metrics

will be that experienced during backtesting relative to the benchmark. However, these types

of calculations do not answer a fundamental question: are you making money by over- or

underweighting factors or are you making money by picking better trades, or by combin-

ing the two together. In an ideal world, a system should make money from both the over- or

underweighting of factors and by picking better trades. We recommend that where possible

all systems are analyzed using the Brinson–Fachler method to clearly understand returns.

27.2.1. The Brinson–Fachler Method

The best known approach to performance attribution is the Brinson–Fachler method, which

uses weighted sums, compounding, and value added for performance analysis. Performance

analysts use weighted sums to combine returns for a group of assets. Consider the following

table:^2

27.2. LOOP2: BENCHMARK ATTRIBUTION CALCULATIONS

TABLE 27-1

Portfolio sector returns Benchmark sector returns

Portfolio sector

weights

Q4

Portfolio

Q2

Active asset

allocation fund

Benchmark sector

weights

Q3

Active stock

selection fund

Q1

Benchmark

∑wriPjP

∑wriPjB

∑wriBjB

∑wriBjp

247