Market Benchmark





Kerry Back

Beta-adjusted market benchmark

  • We estimate the market (CAPM) beta by regressing

\[r - r_f = \alpha + \beta (r_m-r_f) + \varepsilon\]

  • We can rearrange this as

\[ r - [\beta r_m + (1-\beta)r_f] = \alpha + \varepsilon\]

  • The return \(\beta r_m + (1-\beta)r_f\) is a benchmark return.

Alpha and the information ratio

  • The return \(r - [\beta r_m + (1-\beta)r_f]\) is called the active return.
  • Its mean is \(\alpha\). Hence, “seeking alpha.”
  • Its risk is the risk of the regression residual \(\varepsilon\).
  • The ratio \(\alpha / \text{stdev}(\varepsilon)\) is the Sharpe ratio of the active return.
  • It is called the information ratio.

Compounding active and market returns

  • To see visually how much the active return is adding to performance, compound
    • the benchmark return \(\beta r_m + (1-\beta)r_f\)
    • the active return \(\alpha + \varepsilon\)
    • and the total return \(r\).

Example