London (HedgeNordic) – 2018 saw a synchronized downturn nearly all liquid asset classes and markets, with 90% of liquid asset classes losing money – the highest proportion since 1901, according to Deutsche Bank data. There were very few places to hide. The majority of broad hedge fund strategy indices lost money, including some non-directional strategies, such as many equity market neutral funds, partly because factor returns were unstable. There were exceptions however: one market neutral Nordic Hedge Index member, Formue Nord A/S, made 13.59% in 2018.
The search for uncorrelated strategies can be categorized into four areas: illiquid assets; different markets; different factors and different models. The articles following this introcution will introduse some of these strategies with concrete examples.
Illiquid Assets
Starting with illiquids, strategies such as direct lending and asset- based lending generally held up well in 2018. This is partly because they are valued using models rather than market prices. Many direct lending funds will not recognize losses until and unless a borrower has been in default for a number of months – even if credit spreads have widened out. That said, asset-based lending strategies lending at a decent discount to collateral values can be in a “heads you win, tails you win” situation: even if the borrower defaults, they can still make a profit by foreclosing on and selling the collateral.
Still, these strategies may often involve a fund lock up of at least three to five years, which is too long for many investors. Where can diversification be found in the liquid space?
Different Markets
The obvious area to look is different markets, including insurance linked securities, catastrophe bonds, life settlements, electricity, or cryptocurrencies. These markets should not be correlated to financial markets – but of course there is no guarantee that they will actually profit during a down year for equities and bonds.
CAT bonds have had a rough two years, partly due to more natural catastrophes occurring, and some say this is partly due to climate change. Payouts for insured losses were $79 billion in 2018, following $150 billion in 2017. The Eurekahedge ILS Advisers index lost 3.92% in 2018, after losing 5.6% in 2017, but there is a wide spread of returns and some managers made money in both years. The Swiss Re CAT bonds index actually made 2.52% in 2018. But as the CAT bond market reached record size of $34.8 billion in 2018, the head of Swiss Re publicly suggested that investors would be disappointed by returns. It is possible that too much capital has flowed into the space, depressing returns.
Life insurance related investments have fared better. The AAP Investable Life Settlement Index made over 10% in 2018, and has been rising steadily since the index was created in December 2012. Life settlements involve buying life insurance policies, paying premiums, and collecting payouts upon mortalities. The distribution of mortalities should not be correlated to financial markets, and discount rates used to value the policies have also been fairly stable in recent years.
As rainfall is a key driver of pricing for Norway’s hydro- electric industry, there should not be a financial market correlation. It was an unfortunate coincidence that a Norwegian electricity trader, Einar As, lost USD 114 million in September 2018, as the spread between Norwegian and German electricity prices blew out. A breakdown in historical correlations can of course cause losses in alternative as well as traditional financial markets. But as always, manager selection matters and some electricity traders did well last year: Nordic Hedge Index member, Shepherd Energy Portfolio, who have been trading electricity for over 16 years, were up 13.78%.
As bitcoin prices crashed in 2018, it is no surprise to see the Eurekahedge Cryptocurrency index down 71.68% last year. But even in this asset class, there are some relative value traders, essentially making markets in cryptocurrencies without taking a directional view, who profited – mainly from bid/offer spreads, and sometimes from arbitraging pricing discrepancies between different market venues.
CTAs in general had a challenging 2018, but some of those trading less well followed commodities, and trading Chinese commodity markets, which are starting to open up to overseas investors, did much better. For several years, those CTAs trading OTC (Over the Counter) non-exchange traded markets (such as Brummer seed Florin Court) have outperformed, though many of them did not profit in 2018.
Relative value commodity traders, doing trade types such as calendar spreads, often did better than directional traders in 2018, which leads us on to the topic of factors.
Different Factors
Hedge funds trading conventional commodity markets could be wagering on the shape of the curve, or looking to pick up positive carry from roll yields, rather than taking a directional view. Curve gradient and carry are clearly different from the momentum factor that lies behind many traditional, trend following CTA approaches.
Within the CTA universe, short term traders – who follow a variety of approaches including mean reversion – outperformed, with the SG Short Term Traders index flat in 2018, while Nordic Hedge Index member, Estlander and Partners Presto, made 7.99%.
Most asset managers have some kind of factor exposures, and Alternative Risk Premia managers are explicitly seeking to capture factor returns. On a risk-adjusted basis, most ARP managers performed poorly in 2018, but the spread of returns was wide. The devil is in the detail of how managers define factors, and apply them.
Merger arbitrage partly profits from deal break risk, and this was a bright spot in 2018, with the strategy index posting low single digit returns as spreads widened but most deals still completed.
Models and Data
Last but not least, some hedge fund managers may trade only a handful of the largest and most liquid markets, but do so with very different models, that may not easily fit into any standard factor labels. One example here is Sweden’s IPM – Informed Portfolio Management, which has a return profile that does not show correlation to anything we have identified. The manager’s flagship systematic macro strategy made 1.86% while its systematic currency strategy was up 3.35% last year.
Some managers are using alternative data, ‘big data’ and advanced techniques including machine learning, statistical learning, and alternative intelligence, to build new types of models, with different inputs and different mechanics. The skeptics argue that there is a lot of ‘hype’ around these new approaches, and may point out that the Eurekahedge AI Index lost 5.18% in 2018, very close to the average hedge fund. Nonetheless, that index has shown a Sharpe ratio of 1.63 since 2011, and 2018 was its only losing year. If institutional investors are taking a five or ten year view on hedge fund allocations, it may be better to look at multi-year periods rather than obsess about a one year period.
Calendar years are also somewhat arbitrary, so it may be more useful to look at a range of rolling 12-month periods. Shifting from calendar year 2018, to February 2018 to January 2019, changes the sign from negative to positive for many managers.
Multi-Managers
Fund of hedge funds or multi-manager vehicles could be a useful way for some investors to access more unusual and uncorrelated managers, including some of those highlighted above and other approaches.
For example, Sweden’s Merrant Alpha Select, which follows a unique and very selective approach to picking uncorrelated managers, made 2.2% in USD in 2018. In the USA, SkyBridge Capital’s fund of funds, which manages c.$9 billion, made between 3.6% and 4.5% depending on vehicles, outperforming the broad hedge fund industry by 9-10%. SkyBridge has exposure to a number of specialist credit managers. Global fund of hedge fund assets of around USD 600 billion still represent c.20% of hedge fund industry assets of roughly USD 3 trillion, and the fund of fund industry should not be written off.
Picture: (c) fuzzbones-shutterstock.com
