Within the Institute, and even outside it, we (I) have been pushing the idea of money-weighted returns. I would characterise the response as broadly supportive, despite acknowledgement that more effort is involved in calculating them. I wonder whether it is now time to see if we can turn talk into action, especially as this is likely to be another case of ‘the devil lying in the detail’. This could be a situation ideally suited to ‘rapid and distributed prototyping’ or, in English, the best approach might be for a few members to experiment on their own and report back for common learning.
At the 2015 Cambridge roundtable we presented a slide showing returns for a hypothetical hedge fund which demonstrated it is possible to post a positive time-weighted return while losing significant value (dollars) for investors. We could now formalise and extend that idea.
To seed the thinking, I would float the following idea: we could propose a new metric to be included in key information documents for funds – “cumulative dollars earned for investors”. At the risk of appearing somewhat aggressive, we could also suggest “cumulative fees earned for manager” (CFE). The relative scaling of the two numbers could be interesting… The former idea is already calculated by LCH Investments for hedge funds (link), but we haven’t seen anything for the latter.
As noted, the devil is likely to lie in the detail. How should cumulative dollars earned (CDE) be calculated? Presumably beta- and leverage-adjusted – but what if part of the value the manager adds comes from deliberate management of beta through the cycle? Should CDE be quoted as a monetary amount, favouring larger, longer established (and successful) funds? Or quoted as a percentage of assets – in which case, how to calculate the appropriate asset value over time?
It should be relatively easy to calculate an estimate of the CDE (and CFE) for public funds using monthly AuMs and monthly returns. Segregated accounts would be a different matter altogether, and may only be calculable by the asset managers, or asset owners directly. I have a hunch that these calculations could provide additional information that we are currently not seeing – but it is only a hunch, and I could easily be wrong. I would appreciate hearing views on whether throwing a bit of effort at this is felt to be worth it.