In this paper we would like to focus on the benefits of using hedge fund strategies in an institutional portfolio. We believe that treating hedge fund strategies as a distinct asset class is not optimal. We will show that hedge funds add more value when they are combined with long-only managers according to their underlying investments. Also, we will show that using hedge fund strategies outside of the main asset classes may offer a good opportunity to improve the quality and performance of an overall portfolio.
Originally, most institutional allocators invested only in two main asset classes: Equities and Fixed Income. The "60/40" asset allocation model became the cornerstone of the asset allocation process. Over time, institutions searched for ways to improve portfolio diversification and added a number of other asset classes and strategies, which are broadly defined as alternative investments and include private equity, real estate, commodities and hedge funds. Generally, institutions are allocating 15-20% of their portfolios to alternative investments and based on the review done by J.P. Morgan Investment Analytics & Consulting. The following is a good estimate of the most common asset allocation mix among institutional investors:
Effectively, each allocation is a basket of managers or securities that focus on a specific asset and have a specific benchmark, which makes it possible to monitor and evaluate their performance. Generally, most of the performance is expected to be delivered by traditional investments while alternative investments are needed mainly to enhance portfolio diversification due to their historically lower market correlation.
There are a number of weaknesses in this approach. First, there is an over-reliance on the performance generated by traditional investments. In the last 10 years, a bet on market beta has not paid off in equities, as the major indices have been mostly flat. Fixed income investments have done better, but current yields are low and the risk of higher rates in the not too distant future is significant.
Second, alternative investments (PE, RE and Commodities) exhibit lower correlation during normalized markets, but their correlation to the remaining balance of a portfolio tends to increase during sharp market declines. In the case of hedge funds, the analysis becomes even more complicated, as hedge funds are not homogeneous and do not behave the same across all strategies. Some strategies are heavily depended on beta of an underlying market while the others tend to focus mostly on Alpha generation. An asset mix also varies from one strategy to another. Therefore, even more so than for the other alternatives, historic correlations of hedge funds tend to be a poor predictor of the future relationship between hedge funds and underlying markets. As a result, the diversification benefits that are expected by the addition of hedge fund holdings may be greatly reduced when needed the most.
Third, while most institutions construct their portfolios with a longer-term view, short-term market volatility may result in a significant drawdown, as seen in 2008 and the early part of 2009. An available option to mitigate this volatility is the use of outright market hedges. However, as with other types of insurance, this one comes at a cost, which in the current low interest rate environment may be too high to bear. Another option is to use tactical portfolio overlays. While this option implies more active management, it is also a form of insurance and, as such, adds additional costs during normal market conditions.
We believe that a better approach would be to combine both traditional and alternative managers according to the asset class in which they invest, rather than to their investment style. It makes investment sense, as both long-only and long/short strategies are impacted by the same market factors. For example, a long/short equity manager would be allocated to the same portfolio basket as a long-only equity manager while credit and interest rates managers would be part of a fixed income allocation. This approach is better suited for building a long-term portfolio since, contrary to long-only managers, hedge funds have the ability to actively adjust their market exposures depending upon changing market conditions. Due to generally net long bias of hedge funds, an overall portfolio would still be able to capture most of the long-term market performance (beta), but with the benefit of reduced volatility.
Deconstructing allocations to hedge funds from one basket into individual strategies may allow for more efficiency in employing the various strategies on an as-needed basis. For example, an overall portfolio may benefit from an exposure to Global Macro and/or CTA strategies, which are able to provide positive performance during rising markets while exhibiting negative correlation to major markets during the times of market declines. This also opens up more opportunities to creating customizable portfolios.
We have tested five different portfolios:
- 60/40 - The original and still quite popular asset allocation mix.
- 60/40+ - We added hedge fund managers into the asset class allocations without changing the split between Equity and Fixed Income.
- Institutional - The extended version of 60/40 model, which includes allocations to traditional alternatives and hedge funds and is based on JP Morgan research.
- Institutional+ - We added hedge fund managers directly into the asset classes and took the hedge funds out of the alternative investments to avoid double counting.
- Enhanced - In our view, this portfolio is the most optimal from the stand point of utilizing advantages offered by hedge fund strategies.
All the portfolios were rebalanced monthly. We used Hedge Fund Research, Inc.'s ("HFRI") Indices as the proxy for hedge fund strategies' performance. Since we used the hedge fund indices, an issue of a survivorship bias needs to address before we can move on. Survivorship bias occurs when hedge fund managers stop reporting their numbers to hedge fund databases and from that point on, the Index captures only surviving managers' performance. Various researchers on this subject have come up with estimates of the survivorship bias across different indices, ranging from as low as 1% and to as high as 6% per annum. These estimates are not easy to incorporate into analysis as they are conditional upon an exact strategy, period and index that were used to calculate them. This explains why the range of estimates is so wide. Also, there are a number of successful hedge funds that have never reported or stopped reporting to databases after reaching their target asset sizes. Some researchers argue that adding these managers' returns to current indices will result in a higher overall performance by hedge fund strategies that measured by the indices. Overall, while we agree that survivorship bias exists, we believe that given all of the above it is still reasonable to use HFRI indices as proxies for hedge funds' performance.
Cumulative (3 yr)
Annual Compound Return
Cumulative (5 yr)
Annual Compound Return
Cumulative (10 yr)
Annual Compound Return
Annual Compound Return
- The results show that portfolios that use hedge funds as a part of their asset class mix performed better. Comparing 60/40 to 60/40+ and Institutional to Institutional+, we can see the difference in performance, both on absolute and risk-adjusted bases, which can be explained by the inclusion of the hedge funds. The difference becomes more prominent as we increase the length of the tested periods.
- Adding an individual allocation to Global Macro strategies significantly improves risk-adjusted returns. As the result, the Enhanced Portfolio consistently has the lowest drawdown and lowest beta among the presented portfolios over all tested periods and has the highest absolute performance over all tested periods with an exception of 2009-2011.
Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.