The sector that arrived late to the party

Listen to this article

00:00
00:00

The concept of index investing may be old hat in most asset classes but, in the world of hedge funds, it is still work in progress.

Hedge fund indices have, of course, been around for years but there is no simple underlying measure such as a FTSE 100 to recreate.

Traditional investable hedge fund indices simply provide the post-fee performance of a basket of available hedge funds. But these indices suffer from negative selection bias – few of the better managers are interested in index money – as well as erosion from the industry’s high fee structure. In short, they invariably provide sub-industry returns.

Non-investable indices, which merely aggregate and average out performance data, are more accurate but, of course, offer no way of participating in the sector.

For the past few years, a select handful of academics on both sides of the Atlantic has been pondering an alternative notion – that of synthetic hedge fund replication models, which would serve as an investable index either for the broader hedge fund universe or a specific hedge fund strategy.

In the past year, this concept has finally seeped out of the realms of academia. Banks such as Goldman Sachs, JPMorgan, Merrill Lynch, Deutsche Bank and Bear Stearns have unveiled commercial products that aim for market-like returns at a fraction of the price, in the manner of a passive index-tracking equity fund.

The late arrival of the hedge fund sector to the indexing party is understandable. The key problem is that hedge funds are not really an asset class. Although high net worth and institutional investors increasingly allocate assets to hedge funds, what they are really getting is a series of exposures to underlying asset classes, from equities to volatility. This led many to conclude that hedge fund returns were essentially pure alpha, ie dependent on the skill of the manager.

But the academic work suggested returns were often more to do with movements of underlying markets than manager skill. For some trading strategies at least, this suggested that similar returns could be generated mechanically.

The commercial strategies unveiled so far broadly fall into factor-based and pay-off distribution approaches. Most investment banks have opted for factor-based models, which rely on proprietary algorithms that determine the net exposure of the industry to a basket of liquid factors such as the S&P 500 and US dollar.

“If broad-based passive exposure is what investors are looking for, perhaps they can get that more cheaply by replicating the broad underlying investments of hedge funds. If you can get that right with a reasonable degree of accuracy, you can then replicate the aggregate impact of their trading strategies without management fees,” said Edgar Senior, executive director in the fund derivatives structuring team at Goldman Sachs, which launched an Absolute Return Tracker Index in December.

Others such as Harry Kat, professor of risk management at London’s Cass Business School, have opted for the pay-off distribution approach. This aims to use a dynamic basket of futures to replicate the distribution of hedge fund returns in terms of correlation with assets and level of volatility. However, returns in any given month may differ significantly from those of the industry.

Having analysed 2,000 hedge funds, Mr Kat believes he can replicate all but
17 per cent, where the manager has “so much skill and so many tricks in his bag that he can add so much value that he can more than offset the fees”.

However, not everyone is convinced that these replication strategies will be so successful.

Jack Schwager, an author and fund manager at Fortune, part of Close Brothers, likens them to bad systematic hedge funds with lower fees. “They are a hedge fund in disguise,” he says. “I think they will wither and die.”

Edhec, the French business school, also concluded in a recent report that none of the replication strategies unveiled thus far was capable of accurately recreating hedge fund returns.

Backtesting of factor-based models produced results that were “bad to acceptable” in replicating hedge fund returns. Edhec was more upbeat about the pay-off distribution approaches but cautioned that, as monthly returns can vary significantly from those of the underlying hedge fund industry, it would take years for an investor to know whether the strategy had succeeded.

Copyright The Financial Times Limited 2017. All rights reserved. You may share using our article tools. Please don't copy articles from FT.com and redistribute by email or post to the web.