Terry Tian: Hi. My name is Terry Tian. I'm an alternative investment analyst with Morningstar. Today, we have Vikas Kapoor, portfolio manager of Ramius Dynamic Replication.
Vikas, thank you very much for joining me today.
Vikas Kapoor: Thank you, for having me Terry.
Tian: We have seen a number of hedge fund replication mutual funds launched in recent years, including your fund which was launched in 2010.
Tian: Could you just explain a little bit about the idea of hedge fund replication?
Kapoor: Sure. The core idea behind hedge fund replication is that in a diversified portfolio of hedge funds, the majority of risk and return is driven by the set of systemic long and short betas, let's call them alternate betas. And a lot less of the return is driven by idiosyncratic security selection called alpha.
The idea is very similar to what has happened in the long-only world, where you own IBM shares and you [look at how] replicable IBM is with the S&P 500. And they are very different. However, you can create a portfolio of 40 to 50 U.S. large-cap stocks, and so long as you can measure the beta of that portfolio to S&P 500 that explains a lot of the risk and return. It's the same logic, the same theory [with hedge fund replication].
The difference is that the alternate betas are different because hedge funds are different. They go long and short, and they try to capture a different set of risk and return. What has happened in the last few years is that a lot more of those instruments are available in a liquid tradable way, and, hence, you see a lot of explanation of a portfolio of hedge funds coming from liquid factors because people can measure them and trade them. That's really the core principle in that if you can own in a liquid way the majority of the risk and return of a diversified portfolio, you don't have to pay 2 and 20 or take the illiquidity risk and the nontransparency risk of actually investing in hedge funds.
Tian: But there are several limitations of the strategy. For example, with the selection bias, a replicator can only be as good as the underlying pool of hedge funds he tries to replicate.
Tian: How does your fund deal with this issue?
Kapoor: Terry, you're touching on I think the core principle of what we will call as version 1 of the advancements in replication that have come around. If you look at the original set of products and original set of research, that was much more driven by looking at indexes like HFRI, which is one typical example that gets thrown around, and then using some long-only tools like backward-looking regressions to try to capture some of the factor risk.
To your point about selection bias, HFRI indexes and hedge fund indexes in general have a number of biases. But beyond that, if you believe hedge fund indexes are good to replicate, you have to believe that all hedge funds have equal skill. We don't believe it, and we believe there is a very small subset of hedge funds that actually possess skill in really deciding the set of good and smart long and short betas to deliver.
So rather than replicating the nonefficient hedge funds, one example that you and I have talked about before is if HFRI has 2,000 hedge funds and you believe 50% of them are good and 50% of them are bad, why do you want to replicate the 1,000 bad ones? What we do is try to focus on the good ones that have gone through our very detailed due diligence and research process, but then also impose our top-down asset-allocation thinking and try to construct an efficient portfolio because replication is only a means to the end.
What you are replicating should be worth replicating and then you can gain access to it in a liquid way given all the liquid tradable instruments that are out there.
Tian: Another drawback of the strategy is the way most of these hedge fund replications are designed, which is backward-looking, and they lack the predictive power. Could you talk a little bit about how you approach this challenge?
Kapoor: That's a very good point, and if you look at lot of the practitioners in the space, they use this approach as generally referred to as regression-based approaches. What regressions are trying to do is, they are looking back in history and saying given a set of factors, and given a set of things you're looking to explain, how can I explain a combination of factors to explain what happened to what I'm trying to explain.
The issue with that is it assumes that those factors remain constant; that is not what hedge funds do. For example, if you get a beta to the S&P 500 of 25%, that assumes that this group of hedge funds took the risk to the S&P 500 of 25% three years ago, using a three-year regression, and they have not changed. Hedge funds are not in a style box; they are dynamic beasts in some way because they have flexibility. And you need to have a right set of quantitative tools that allow for long-short [betas], a change of factors, and a change of leverage to maximize predictive power.
So we use an approach called FLS, flexible lease square, which is an approach that is a dynamic, self-learning algorithm that tries to maximize predictive power not explanatory power. But also its very important when you use sophisticated approaches to have a qualitative input, a qualitative common-sense starting point. Because our target portfolio is research-driven and we understand the type of hedge fund strategies we are investing in, that allows us to provide the right qualitative input.
Tian: So what should investors expect of these hedge fund replications in terms of risk and return profiles and diversification benefits?
Kapoor: That’s a key question that gets asked around a lot. We think hedge funds provide a different source of risk and return. In terms of the upside and a downside capture, hedge funds, because they are dynamic and they change their exposures and they take different set of exposures, provide you with some protection on the downside while allowing you to capture an upside coming from different sources of risk and return.
When we look at our portfolio and look at it over a long period of time, the returns are generally in a 10% kind of range with a 6% type of a volatility. But most importantly, what we find useful in the context of a client portfolio, which is generally a 60-40 or a 50-50 portfolio, is there are two major sources of risk and return, either coming from equities or fixed income. We think alternatives as an additional source of risk and return, where it is a different source of risk and return built in a thoughtful research-driven way, provide the diversification so that the total portfolio is more efficient, both in terms of risk as well as return.
Tian: In terms of diversification, how much do you suggest investors should allocate to a hedge fund replication in their portfolio?
Kapoor: I think the answer to that question by definition should be specific to an investor's need. However, there is a lower bound. We think allocating less than 10%--especially because it's a generally a [6% volatility] type of a strategy relative to equities with midteen volatilities as well as fixed income--does not really make a significant marginal difference.
We think one should think of investing in a core and in a satellite way, where the core source of risk and return coming from a diversified portfolio can be owned via a fund similar to ours, and then, one can build around that a number of very specific satellites if one has a high level of forward-looking conviction. But the core different sources of risk and return coming from hedge funds can be sourced without actually investing in hedge funds. And a thoughtful, next-generation set of products, that allow you to not just capture that in a blind-correlation mining way but in a thoughtful way, we think is a better approach.
Tian: Alright. Thank you very much for sharing your insight with us today.
Kapoor: Thank you very much, Terry.
Tian: Thank you.
Kapoor: Appreciate it.