Convertible Bonds

Why does the least loved hedge strategy deliver the best returns?


Perhaps one of the biggest dichotomies in the hedge fund industry currently is the dramatic outperformance of convertible strategies in recent years compared to their continuing unpopularity with investors. Convertible strategies have been the best performing of the major hedge fund sub-indices since 2008 and have outperformed the broader index in every single year.

However, in surveys of investor attitudes to sub-strategies, they consistently rank last, as the least favoured asset class. The fall from grace of convertible hedge fund strategies has been extraordinary. When Ferox Capital was founded in 2000, most hedge fund investors would build their portfolios around a core allocation to convertible strategies of circa 20%. Today, if investors even allocate at all, it is often 1-2%.

Much of the blame must lie with the convertible managers themselves. Convertible funds scared – and scarred – investors with their returns in 2008 as a result of the pre-crisis leverage levels and the scale of the forced unwind. However, traumatic as that year was, does it have any relevance to understanding the market today? We genuinely believe that the convertible world bears little resemblance to the environment that existed before 2008. Everything has changed: investors, leverage, liquidity, valuations, strategies, opportunities.

The first and most important of those shifts is in the investor base. Pre-2008, hedge funds and proprietary desks dominated – most estimates place their participation at over 80% of the market. Those investors were squeezed from the market by leverage restrictions and margin calls. However, in the macroeconomic environment that followed, where credit contraction and struggling growth dominated investment decisions and a flight to quality and highly liquid investments became a priority, the shift by long-only investors into fixed income strategies caused a fundamental change in the investment structure of the convertible market. Today, most convertible dealing desks are doing the vast majority of their business with long-only funds, particularly on European and Asian markets.

It should be noted that there are strong reasons to buy convertibles on a purely long-only basis. The asset class has comprehensively outperformed equities over the last 20 years on an absolute basis and has done it with a fraction of the volatility. A 20-year analysis of the efficient frontier shows that there is no combination of bonds and equities that outperforms a combination of bonds and convertibles. If one puts long-only convertible returns into a portfolio optimisation tool, it will not let one own any equities. Convertibles have also outperformed bonds. If there was any silver lining to the dark clouds of 2008, it is that long-only investors were forced to look at the asset class again and liked what they saw. As a result, the investor base should be stable and deep from here.

One could speculate that the outperformance is partly correlated with the lack of demand from hedge investors: an absence of allocations to the asset class leads to a lack of technical capital, in turn creating an inefficiently priced asset class, thus generating a rich source of alpha. The subsequent domination of long-only funds, particularly those with an index-tracking mandate, has created further opportunities for the arbitrage community. Those funds that survived the crisis of 2008 have incorporated the lessons learned into current trading strategies and have been well placed to collect alpha.

Finally, one should not read too much into the low proportion of multi-strategy capital allocated to the asset class as being a signal for the opportunity set. The enormous growth of assets in multi-strategy funds since 2008 now makes the asset class simply too small for them to allocate significant capital or resources.

The asset class today is a very different place in terms of leverage, concentration and efficiency, but work still needs to be done to convince investors that the profile of risk has fundamentally changed from that prevailing in the market going into the Lehman’s bankruptcy. It is rare today to find a convertible fund that has more than 2x leverage. Pre-2008, it was difficult to find a fund that had less than 4x. Given that, in the post-Lehman crisis, prime brokers forced funds to 1.5x (and given that hedge funds no longer dominate the space), it is clear that the scale of any future forced deleveraging is unlikely to have the same impact. In short, the whole market was probably leveraged 2.5x (80% hedge participation at 4x leverage); today it is at 1.1x (25% participation at 2x).

The convertible market’s reputation for liquidity has been consistently – and extraordinarily unfairly – poor. As the convertible can be traded whole or broken into its constituent parts, the liquidity of the market is excellent relative to its overall size. Indeed, on a size-adjusted basis, it may well be one of the most liquid major markets. However, it is, at circa $500 billion globally, the smallest of the major markets. Its market capitalisation equates to about two-thirds of the size of the US high-yield market. To conclude, even in the dislocated markets of 2008, where convertibles were caught in a tornado of selling, the market could clear the forced selling (albeit at highly distressed prices).

However, the nature of liquidity has changed post-2008. It was a market where proprietary trading desks controlled the market and hedge funds ensured that it was efficient. Now, long-only funds dominate and the market makers take little or no position risks. This is both opportunity and risk. The opportunity is that one can often be the price setter, taking on risk at attractive levels. However, the counterpoint to that is that the bonds will price illogically in a crisis. One aspect that offsets this risk is that, in the post-2008 meltdown and recovery, many long-term investors made an enormous amount of money in buying cheap convertibles from forced sellers. It is an asset class that should therefore have a strong underpinning of fundamental capital at distressed levels.

Without sufficient technical capital to keep convertible valuations efficient, we have seen a number of trends. Firstly, the market is exhibiting a wide range of valuations from the very “cheap” to the very “expensive”. One does not have to buy the “average” bond. This range of valuations is about as large as it ever has been – the absence of technical capital means that many, many bonds are attractively priced. The “expensive” bonds tend to be caused by the long-only preference for certain types of bonds – index constituents, investment-grade, at-the-money. Furthermore, the regional markets exhibit a large range of valuations with, in general, the US today being more expensive and both Asia and Europe displaying the most obvious value. It is a strategy that needs to be played globally to deliver the best returns. Being non-benchmark has never been so important.

Secondly, the market tends to exhibit rising valuations in rising equity markets and falling valuations in bear markets. This is counter-intuitive as, generally, equity volatility is gapping higher when markets fall and is low when markets rise.

Finally, at maturity, all bonds will be at “fair value”. By buying them “cheap”, even without trying to arbitrage this “cheapness”, one has a substantial tailwind. One of the structural advantages of the convertible market is that it is one of the very few option markets in which one is not entering a “zero sum game” with a professional seller of options. One is not taking a market counterpart “short” of an option – the options have been created by companies who have a very different agenda to any option players. It is therefore possible to source options that one simply could not source (sometimes at any price) in other markets, such as five-year call options on companies that have seen their share price fall 65% in a year (Sony) or long-dated out-of-the-money put options on a speculative mining company that had seen its share price triple in a year (Hochschild).

Whilst we cannot speak for everyone in our space, it would seem that the strategies being pursued by market participants have changed. We have almost completely abandoned traditional volatility arbitrage and instead concentrate on risk/return and special situations trades. As valuation metrics have changed, historic volatility is now of little use in assessing the long-term pricing of a convertible. Furthermore, if anything, convertibles move in a semi-predictable anti-theoretical manner. The old “volatility extraction” model of arbitrage seems of little value in such an environment.

However, the range of valuations makes interesting trades available. Trades rich in gamma can be held for their convexity. It becomes a trade not about volatility but about a low-risk way of exploiting trends and sudden moves. Expensive bonds can be shorted. More widely, bonds held hedged with credit default swaps (CDSs) have dramatically reduced. CDSs proved to be almost useless in a liquidity meltdown by not only failing to provide adequate protection but also making the trade challenging to exit. They have been replaced by asset swaps and deeper credit research.

However, there would seem – to us, at least – to be techniques that clearly deliver absolute returns in such an environment. Most notably, combining low-priced put options with low-priced calls, both of which can be extracted from convertibles, has been a winning strategy over a very long period. It even worked in 2008.

In short, there are many ways to exploit the current market from long only, to short-biased. We have found ways of successfully exploiting moves such as the recent “upside crash” in Japan and yet have a tail-hedge protection product that delivered a 31% positive return in 2008. We are delivering low-volatility absolute returns with class-leading Sharpe ratios and high-vol returns that are as good as our best years. Now that convertibles no longer trade in a narrow band around fair value, it is the dispersion of valuations and the range of exposures one can extract that drive the returns.

The opportunity for hedged investors should benefit from many of the aspects outlined above. Inefficient option valuations, misunderstood credit and an active new issue calendar should all be reasonable precursors for strong returns. Furthermore, the continued absence of leverage in the sector and the lack of technically-driven capital (hedge funds and proprietary trading desks) should leave plenty of alpha to be collected in a relatively low-risk manner.

However, as noted above, convertible bonds have become a plaything of long-only investors. Without technical capital to force them to fair value, the inefficiency is likely to persist or even increase. We see the value not in terms of arbitrage but in terms of the value being realised by the move in the underlying equity. Owning high-gamma securities with little risk in markets that are moving sharply ought to be a recipe for generating positive returns.

We are therefore optimistic for hedged returns. Investor appetite for leverage remains low and the willingness of managers to use gearing is perhaps even lower. This does not necessarily mean that the returns will not return to the 2000-2004 levels, as one can take directional views on equity markets with well controlled risk. Indeed, our returns this year have been some of the best that we have achieved in our 13-year history.

Given the natural asymmetry available in convertibles and the ease with which one can adjust one’s exposure with effective equity hedges, this strategy may actually be one of the few areas where one can take directional views quite sensibly. The fact that the convertible market, on a long-only basis, has outperformed both equities and bonds since 1994 might well be a clue to the future attractions of the asset class to absolute return investors.

When one analyses the opportunity set, it would appear that the outperformance should almost be expected. The asset class is rare in that it offers convexity – bond-like returns on the downside and equity-like returns on the upside. Other mainstream hedge fund strategies, such as CTAs, global macro, equity long/short, simply do not start with this basic advantage. This asymmetry is further enhanced by the structural cheapness of a product sourced from corporate issuers rather than investment bank derivative desks. It is skewed even more by the actions of the long-only investor base clustering into benchmark deals. Finally it is enhanced by our ability to focus on the most extreme risk/reward trades that offer the highest degree of asymmetry. None of these factors seem to be at risk in the medium term.

The interesting point for convertible strategies here may be that this tide of negativity towards the asset class feels like it is turning. The lack of returns in broadly held hedge strategies such as global macro and CTAs is forcing investors to look at all asset classes again. The outperformance of convertibles has not gone totally unnoticed. Many of the indicators that investors (sometimes erroneously) consider (such as new issuance of convertibles) are looking positive. As more investors begin to embrace the asset class, others will be forced to reconsider their zero allocations, especially if the outperformance continues.

Paul Sansome joined Ferox Capital in 2001. Prior to that he had been at JP Morgan since October 1998 where he was responsible for equity and equity-linked origination within the investment bank. During that time he was actively involved in a number of major flotations, equity-linked transactions and complex equity issues. He has 26 years of experience in the convertible business having started in the convertible sales operation at Merrill Lynch in 1987. He graduated from Cambridge University in 1985.