How should you benchmark your manager?

Carl Hitchman: It comes back to reflecting on the purpose to which pension funds are actually investing. On the basis that they are looking to generate either a gilts-plus return consistent with their funding plan or a Libor-plus return to support the derivatives programme, it has to be recognised that these are not investible benchmarks. To me, keeping the return target in mind and then monitoring against that will draw out for the trustees the risks they are actually running.

You could derive some sort of strategic benchmark against which to monitor the manager. However, this is likely to drive different behaviours to an absolute mandate. Further, while this may have the appearance of lower risk in terms of a lower tracking error relative to that benchmark versus cash or gilts, what does that matter if the real purpose is to generate Libor-plus?

Therefore, I think where the purpose is to deliver absolute returns or returns relative to liabilities, you really should focus these metrics as the benchmark rather than trying to come up with some alternative benchmark. It means you will really develop an understanding of the asset classes invested in and the risks relative to what you are trying to achieve.

William Nicoll: Tracking error depends on what you are using as a benchmark and varies through the cycle. We see this every time. There are points in the cycle where we will be told that we are not taking enough risk. And yet the funds may still offer good returns compared with the benchmark because the tracking error is not a good measure of idiosyncratic risk. We find that we produce great Sharpe ratios and other statistics. However, we believe there is a limited value to them. But again, it depends what you are using the benchmark for.

David Felder: I think interest rates will stay low for a long time. Pension funds will continue to come out of equities, at varying paces, and go into things which look like fixed income, whether it is leases, infrastructure or directly into the sort of products we have been talking about.

Hitchman: Fixed income is getting more complex.

Felder: Yes.

Nicoll: One of the key challenges for the industry, other than trying to read the markets and make returns, is around education of all concerned. It feels as though things will continue to get more complicated in terms of the market dynamics. I think that we, as an industry, have to get the education around that right to help our clients.

Felder: Right. That means further focus on governance. At Law Debenture we are not advocates of any particular governance model. We work with all frameworks. That includes effective delegation of all of this to a fund manager right through to having enhanced investment committees with chief investment officers and outside academics and other experts on the committees.

Ian Smith: It is what is right for the client.

Felder: Yes. And there are four or five governance models for delivery. However, one thing schemes do need to do is in sync with the development of their derisking path and the movement into these assets, is to make sure the governance process is in place to service that.

Nicoll: Pension funds do have the advantage that the insurance companies have gone down this route already. The experience, certainly with Prudential [M&G’s parent company], is it started off just investing in UK corporate bonds, then it did European corporate bonds, then once it had bought those markets it went into the private markets. Therefore, there is a route. It is not unknown what needs to be discussed. However, I do agree there is quite a lot of extra education that needs to happen.

Felder: Smart beta certainly has a role to play in fixed income, as it does in equities; your fixed income indices are dominated by the bigger issues.

John St Hill: I think it is an emerging area. Some of the things that have been done are very interesting. One of those things is around GDP-weighted global bond indices. The stuff that has been done on corporates is interesting, although there is still work to be done in order for those sorts of techniques to become accepted across the board.

Felder: The biggest wake-up call on the issue of indices came through in the financial crisis with the bank exposure in iBoxx and similar indices. Banks had been big issuers for senior and subordinated debts so they were a large proportion of indices. Therefore, derisking funds that thought they had derisked by moving out of equities into UK corporates found they were heavily exposed to subordinated bank debt and so on when it formed in line with falling equities.

St Hill: I would like to come back your point about the problems of index construction. Not every single bond in a fixed income index is traded every day. The prices of the majority of bonds in fixed income indices are inferred. That is to say, a model is used to estimate the price of the target bond by observing the prices of other bonds with similar maturity, credit quality, sector and coupons. Therefore, you do not really get a ‘real’ price return from indices. It is not necessarily biased, it is just not necessarily tradeable on any particular day or week.