Keeping the targeted active return at 2 percent, the marginal contribution to active risk for this portfolio is:
This clearly indicates that, because bonds are underweight relative to the benchmark, then marginal weight increments toward this asset class will lower active risk. Stocks, on the other hand contribute relatively more to active risk, both because they are riskier and because they are overweight relative to their benchmark. Moreover, we can attribute almost twice as much of the active risk (σa= 3.10) to stocks over bonds by estimating attribution risk:
Recall that the sum of these must equal the tracking error. Therefore, the active position in stocks comprises about two-thirds of the tracking error. What about the effects of active allocations that deviate from the benchmark? Allocative risk is estimated to be roughly 14.36 percent:
due primarily to stock overweighting to which we attribute 12.93/14.26, or 91 percent, to the position in stocks:
Moving to the program weights (recall, the view portfolio was 0.28, 0.72 stocks and bonds), we estimate portfolio risk to be
This is not active risk. It is the risk associated with the view portfolio. Decomposing it into the program level attributions give us:
Again, most of the risk is tied to stocks, but in this particular portfolio:
4.98/8.28 = 60 percent is accounted for by the 0.72 weight on bonds.