Hedge fund replication
Hedge fund replication is the dumbest idea since "portable alpha" or "130/30". You can buy a Rolex "clone" anywhere here in Beijing but there's a big difference between them and the PERFORMANCE of a genuine Rolex. Same with clones trying to mimic hedge funds. There are always cheap inferior knock-offs and there is no threat to REAL hedge funds. Shame on any investor that breaches fiduciary duty by allocating to such garbage.
Perhaps AVERAGE hedge funds can be cloned but who wants to invest in "average" managers? Why waste money in "aggregate" hedge fund beta? Amazing some rookie investors are falling for such nonsense. Reminds me of the human clone hoax. It's alternative alpha you want not alternative beta. The typical hedge fund is too dependent on underlying market factors and does NOT add diversification. The ONLY reason to invest in a hedge fund is to access rare idiosyncratic skill. Less than 5% of hedge funds exhibit skill.
Hedge fund cloning is a SALES trick that no competent investor spends time on. The clones have fallen into the familiar traps of spurious correlation, inadequate sample size, non-predictive data mining and random pattern recognition so prevalent in much of the quasi-academic financial marketing literature. It is cheap and easy to find factors that "forecast" the past but it's difficult and expensive to identify drivers of future performance. Hedge fund clones are marketing POSTdiction as PREdiction.
Alternative beta is similar to traditional beta. Avoid those risks and focus on alpha. There is a mistaken perception that ANY positive return from a hedge fund is alpha. And also an assumption that skill and alpha are synonymous. But by the strict definition of alpha, the excess returns above the RELEVANT index, that is not always the case. Cash has generated significant alpha over technology stock indices for 7 years or Japanese equities for 17 years for example, but cash is not skilled.
Alpha and absolute returns are also NOT the same thing; you can have high positive alpha and still lose money. There is also some recent confusion that hedge funds somehow have an "exclusive" on alpha; but alpha is a traditional performance metric originating firmly from long only territory. "Cheap" hedge fund replicators might offer alternative beta but have no chance of producing ANY alpha.
Unfortunately some hedge fund clones have only backtested 20 years which is too short and temporally biased. That particular time period has seen exceptional global equity and bond returns highly unlikely to be repeated in the next 20 years. A clone might think it would have made 12% or so historically but that was in a generally very conducive market conditions. So the hedge fund replication models themselves may be mistaken that such "replication" will continue to perform.
While benchmarks exist to measure traditional managers, what the hedge fund industry needs is a way to differentiate and evaluate strategy alpha from strategy beta. With hedge funds, outperformance can only be claimed by comparing a fund's performance to a naive replication of that strategy and to the underlying returns of the assets in which the fund operates. You can cheaply clone the market beta and risk factor dependence beta of a strategy but not the alpha.
A long biased emerging market fund needs to be evaluated in the light of the Peru, Vietnam, India and China stock indices this year. An emerging markets "hedge fund" that loses money when emerging markets equities drop is NOT a hedge fund. A concentrated "10 best ideas" stock fund should be compared to the best-performing 10 stock basket NOT the broader index. A leveraged credit fund should be compared to the appropriate leveraged credit index. A manager trading oil needs to be benchmarked to energy price gains. A short biased fund to the inverse of the stock index. A highly leveraged fund requires comparison to the LEVERAGED performance of any dependent factors not simply the raw index return.
The underlying climate for a particular strategy also needs to be factored in. A Japan long/short fund up 10% in 2006 was a genuine alpha producer while a Japan fund up 30% or more in 2005 was likely just leveraged beta. In 2000 almost every CB arbitrage fund had a great year as the conditions were ideal, but few outperformed a naive CB arb replication. CB arb alpha can only be identified if we first have a way of measuring CB arb beta. The best managers are those who still make money in poor conditions for their strategy. A skilled CB arb fund that made money in 2005 is worth studying unlike the many lucky managers from 2000.
This is why efforts at replicating generic hedge fund returns are a welcome development. They will provide a baseline "alternative beta" for each hedge fund strategy. It will allow the truly skilled hedge funds to demonstrate their REAL alpha generation capabilities within their category. There has been crackpot comment that hedge fund clones pose a "threat" to the current industry. On the contrary and the same was said about investable hedge fund indices. As we have seen with real cloning, all you get is a cheaper but INFERIOR version of the original. But as a benchmark, they have value for the better managers AND especially investors by identifying who was skilled and who was lucky. To actually invest in a hedge fund clone would be silly.
Replicating the return for a hedge fund sub-sector for a public domain plain vanilla strategy is pretty trivial. Naive implementation of most well-known equity and credit long/short, common arbitrages, mean reversion and trend following strategies won't present much problem. If hedge fund replication can be done, great as such work would provide alternative beta benchmarks, but investors will be disappointed if they think the returns from hedge fund clones will mirror the performance of the best hedge funds.
The main benefit will be a possible way to differentiate true skill and real alpha from luck and strategy/asset beta. Clones will also have a detrimental impact on the many managers with a high risk factor dependence. No longer able to claim genius trading abilities or ANY performance fees when they make 20% and the hedge fund clone or the asset in which they operate makes 30%. But replicating a brilliant stock picker, commodity trader or a black box system developer is simply not going to happen.
No matter what occurs in the markets, well-managed "expensive" hedge funds operating proprietary strategies with skilled traders and robust risk management WILL perform, even under pessimistic economic scenarios. That is why it is worth paying the 2 and 20. As with all businesses, cheap generic things are cheap for a reason. A hedge fund is a quality purchase and if investors want cheap they can get cheap but at the heavy cost of poor performance. However those dependent on asset beta might have some explaining to do as hedge fund clones flood into the market.
Perhaps AVERAGE hedge funds can be cloned but who wants to invest in "average" managers? Why waste money in "aggregate" hedge fund beta? Amazing some rookie investors are falling for such nonsense. Reminds me of the human clone hoax. It's alternative alpha you want not alternative beta. The typical hedge fund is too dependent on underlying market factors and does NOT add diversification. The ONLY reason to invest in a hedge fund is to access rare idiosyncratic skill. Less than 5% of hedge funds exhibit skill.
Hedge fund cloning is a SALES trick that no competent investor spends time on. The clones have fallen into the familiar traps of spurious correlation, inadequate sample size, non-predictive data mining and random pattern recognition so prevalent in much of the quasi-academic financial marketing literature. It is cheap and easy to find factors that "forecast" the past but it's difficult and expensive to identify drivers of future performance. Hedge fund clones are marketing POSTdiction as PREdiction.
Alternative beta is similar to traditional beta. Avoid those risks and focus on alpha. There is a mistaken perception that ANY positive return from a hedge fund is alpha. And also an assumption that skill and alpha are synonymous. But by the strict definition of alpha, the excess returns above the RELEVANT index, that is not always the case. Cash has generated significant alpha over technology stock indices for 7 years or Japanese equities for 17 years for example, but cash is not skilled.
Alpha and absolute returns are also NOT the same thing; you can have high positive alpha and still lose money. There is also some recent confusion that hedge funds somehow have an "exclusive" on alpha; but alpha is a traditional performance metric originating firmly from long only territory. "Cheap" hedge fund replicators might offer alternative beta but have no chance of producing ANY alpha.
Unfortunately some hedge fund clones have only backtested 20 years which is too short and temporally biased. That particular time period has seen exceptional global equity and bond returns highly unlikely to be repeated in the next 20 years. A clone might think it would have made 12% or so historically but that was in a generally very conducive market conditions. So the hedge fund replication models themselves may be mistaken that such "replication" will continue to perform.
While benchmarks exist to measure traditional managers, what the hedge fund industry needs is a way to differentiate and evaluate strategy alpha from strategy beta. With hedge funds, outperformance can only be claimed by comparing a fund's performance to a naive replication of that strategy and to the underlying returns of the assets in which the fund operates. You can cheaply clone the market beta and risk factor dependence beta of a strategy but not the alpha.
A long biased emerging market fund needs to be evaluated in the light of the Peru, Vietnam, India and China stock indices this year. An emerging markets "hedge fund" that loses money when emerging markets equities drop is NOT a hedge fund. A concentrated "10 best ideas" stock fund should be compared to the best-performing 10 stock basket NOT the broader index. A leveraged credit fund should be compared to the appropriate leveraged credit index. A manager trading oil needs to be benchmarked to energy price gains. A short biased fund to the inverse of the stock index. A highly leveraged fund requires comparison to the LEVERAGED performance of any dependent factors not simply the raw index return.
The underlying climate for a particular strategy also needs to be factored in. A Japan long/short fund up 10% in 2006 was a genuine alpha producer while a Japan fund up 30% or more in 2005 was likely just leveraged beta. In 2000 almost every CB arbitrage fund had a great year as the conditions were ideal, but few outperformed a naive CB arb replication. CB arb alpha can only be identified if we first have a way of measuring CB arb beta. The best managers are those who still make money in poor conditions for their strategy. A skilled CB arb fund that made money in 2005 is worth studying unlike the many lucky managers from 2000.
This is why efforts at replicating generic hedge fund returns are a welcome development. They will provide a baseline "alternative beta" for each hedge fund strategy. It will allow the truly skilled hedge funds to demonstrate their REAL alpha generation capabilities within their category. There has been crackpot comment that hedge fund clones pose a "threat" to the current industry. On the contrary and the same was said about investable hedge fund indices. As we have seen with real cloning, all you get is a cheaper but INFERIOR version of the original. But as a benchmark, they have value for the better managers AND especially investors by identifying who was skilled and who was lucky. To actually invest in a hedge fund clone would be silly.
Replicating the return for a hedge fund sub-sector for a public domain plain vanilla strategy is pretty trivial. Naive implementation of most well-known equity and credit long/short, common arbitrages, mean reversion and trend following strategies won't present much problem. If hedge fund replication can be done, great as such work would provide alternative beta benchmarks, but investors will be disappointed if they think the returns from hedge fund clones will mirror the performance of the best hedge funds.
The main benefit will be a possible way to differentiate true skill and real alpha from luck and strategy/asset beta. Clones will also have a detrimental impact on the many managers with a high risk factor dependence. No longer able to claim genius trading abilities or ANY performance fees when they make 20% and the hedge fund clone or the asset in which they operate makes 30%. But replicating a brilliant stock picker, commodity trader or a black box system developer is simply not going to happen.
No matter what occurs in the markets, well-managed "expensive" hedge funds operating proprietary strategies with skilled traders and robust risk management WILL perform, even under pessimistic economic scenarios. That is why it is worth paying the 2 and 20. As with all businesses, cheap generic things are cheap for a reason. A hedge fund is a quality purchase and if investors want cheap they can get cheap but at the heavy cost of poor performance. However those dependent on asset beta might have some explaining to do as hedge fund clones flood into the market.
© posted by Hedge Fund on 7:29 AM
<< Home