2&20 adds up to hardly anything

Remember 2&20? Probably only too well. As most of the wealthy, private clients of asset management funds know, this was the base rate many were charged for the privilege of handing over their money.

There were variations on the 2&20 theme, but the general principle was to hit clients up for two per cent of the net asset value of the investment, plus 20 per cent of the returns above a pre-determined benchmark such as Libor. The "performance fee" was fixed on both realised and unrealised or paper gains, and most of it went in bonuses.

The core argument for charging 2&20 or versions thereof was that these funds, generally hedge funds, were staffed by such brilliant people – the quants with their whizz-bang investment models – that they achieved dazzling, alpha returns regardless of what the markets were doing. "Two and 20 is a bargain for alpha", argued a hedge fund manager at the height of the boom, likening it to paying hundreds of dollars for dinner at a top restaurant. Indeed some funds charged more than 20 per cent, as high as 35-50 per cent.

The interesting thing is that nearly all these funds were claiming alpha-like skills and charging 2&20. On top of that, the hedge fund fees tended to drag up those of pension funds and other sectors of the asset management industry. However, while a very few funds achieved spectacular returns for their clients – at least for a while, most of them did not.

Their cover was blown by Bostonian Mark Kritzman in early 2007 with a study called "Portfolio Efficiency With Performance Fees." As president and chief executive of Windham Capital Management, as well as a lecturer in financial engineering at MIT's Sloan School of Management, the sceptical Kritzman analysed the effect of 2&20 on 10 hypothetical but typical portfolios. He assumed a Libor of 5.4 per cent, which was the rate prevailing around that time, and that 10 imaginary hedge funds would earn seven percentage points above that.

That gives 12.4 per cent. Now we come to the fees. Kritzman calculated that 2&20 would knock 3.8 per cent off the returns. Subtract 3.8 from 12.4 and you get 8.6 per cent, hardly an alpha return. His conclusion was that the "asymmetry penalty" of the 20 per cent slice of the gains essentially ensured less than stellar returns. As Warren Buffet would say, it generally comes down to simple addition and subraction.
 
Robust calculations by others endorse Kritzman's findings – at least 90 per cent of the money entrusted to such funds posted dismal results. Most investors would have done as well with a common-and-garden index-tracker. As it happens, nearly all the quant models, the basis for these stratospheric fees, were busted from the start. Computer rather than market-driven, they relied by definition on prior data to predict future events, rather like weather-forecasters saying it will rain in five years time after consulting ten year-old meteorological patterns.

David Swensen, the doyen of endowment investors, observed recently: "How can you tell the difference between a quant model that works and one that doesn't? If you don't understand, how can you justify putting your money there?" Swensen, who heads the Yale fund, has returned 16.3 per cent a year after fees, putting him in the top one per cent of institutional investors. And that's without the benefit of quant models.

The good news is that fees have fallen back to earth in the widespread collapse of hedge funds. Instead of applying a blanket rate, many private investment offices tailor charges to the client's needs, pegging them to assets but with extra fees added for investment complexity, time spent on managing the assets and other factors.

The main thing is that if the fund makes a profit on a client's assets, it's the client who collects it. As Yale's super-investor David Swensen points out, this amounts to a long overdue rediscovery of the first principle of managing other people's money. Namely, the primacy of fiduciary responsibility over annual bonuses.

Mastering these social, organisational, and human issues is not easy.

As one observer said with regard to collective human behaviour: "these people issues, these so called soft issues, aren't 'soft' issues at all. Maybe we should call them the 'even harder' issues."