Relative TSR peer group down to 14 names after two takeovers: rebuild or switch to an index?

Disagree a little with the above. We moved to the index five years ago and have had no questions on it since. The noise argument cuts both ways: with 200 constituents a single company cannot move the result. What matters is whether your share price actually correlates with the index. Ours does, about 0.7 over five years, so it was defensible.

Two things that worked for us. First, write the replacement rules into the plan before you need them: an acquired peer is frozen at the offer price and then tracks the median of the rest, a delisted one drops out from the start. Second, top the group back up to 20 or so with overseas listed peers and measure everything in a common currency.

We ended with 21 names, nine of them outside the home market, and the outperformance approach described above rather than ranking. The committee signed it off in one meeting.

Thanks all. We are going with a rebuilt group of 20 including six continental peers, written rules for takeovers, and outperformance of the median rather than rank. The index stays in the back pocket as a cross-check that the chair can see each year.