A Case For Convertible Bond Arbitrage In 2014

By Paul Sansome, Ferox Capital LLP

Part One of this article discusses the main reasons to own convertible securities in the current environment. Part Two looks at why convertible bond arbitrage is an attractive investment strategy now.

Part One

We see four main reasons to own convertibles:
1. Strong new issuance
2. Dispersion in convertible valuations
3. Rising dispersion in equity returns
4. Macro conditions that may favour the asset class

Strong new issuance

One of the key reasons we were enthusiastic about the asset class last year was that we could see issuance accelerating from a low base in 2012 and issuance is usually a precursor to strong performance from the asset class.
Convertible issuance was remarkably strong in 2013 and we believe it will be vigorous again in 2014. 2013 was, with $97bn of issuance, the best year since 2010. Furthermore, the trajectory is seemingly relentlessly up – the November total was the biggest month of issuance that we have seen since 2008 globally and a top decile month looking back over 20 years. It was the fifth best month of issuance ever for European issuers. Issuance levels bottomed in July 2012, as the 12-month moving average monthly issuance hit $3.4bn. As at December, that monthly issuance number sits at $8.0bn – a 133% increase. It has risen in every month (bar one) since that point – when, incidentally, we wrote a note to investors entitled “Convertible New Issuance – The death of the market is greatly exaggerated”.

There is a myriad of reasons why a healthy issuance calendar is a positive for convertible issuers:

• New issue premium Many convertibles are issued at discount to fair value in order to clear the market. As a result, they tend to appreciate immediately above the issue price. This is the benefit most non-investors focus on when thinking about the importance of the convertible issuance calendar – and it is possibly the least significant. The premium is “free alpha” and, as such, we do attempt to maximise our share of this. However, such an obvious trade is popular – and not only with convertible investors. Obviously, the premium is the largest in the “hottest” deals. These deals attract equity funds and straight bond funds looking for their share of this premium. By the very definition of a “hot” deal, no one gets enough of this “free alpha” to make a significant difference to their return (though, it must be said that we feel we punch above our weight in terms of allocations).
• Refreshing the convertible universe We exist in a “term” asset class, where a significant part of our universe disappears every year through bond maturities. It therefore needs to be constantly refreshed. We would obviously like the universe to remain constant or grow. The level of issuance we experienced in 2013 was larger than the expected maturities in 2014. The market capitalisation of the universe rose in the year.
• Companies issuing when they see growth potential Convertibles are usually issued when the company is looking for capital. This tends to happen, for publicly listed companies, for three major reasons: to raise money to fund expansion, to fund an acquisition or to recapitalise the balance sheet. Each of these are moments that can be transformational for the company’s share price. In short, often, convertibles are issued at interesting moments for investors. This is not necessarily true for all capital raisings. With high-yield, it is becoming increasingly common for the company to issue bonds and dividend the proceeds to the private equity investor.
• Creation of a more complex universe We have seen recent new issuance move down the credit spectrum. This is a positive for our style of credit research intensive investing in the asset class. Not only will fewer participants be able to make sense of the universe but also the underlying equities are higher beta – always interesting when one has an asymmetric position (limited loss potential and infinite upside). That is not to say that all of these issues are interesting today – far from it. Poor credits are difficult to analyse when they have many years to maturity. However, when they are, in the future, much shorter dated, our experience of looking at security of repayment should throw up numerous opportunities that few will have the expertise to exploit.
• Dislocation in convertible portfolios Even if the company or pricing of the convertible is not interesting (to us, at least), new issues provide other opportunities. Most notably they create turnover in other convertible portfolios. Long-only investors run unleveraged and so, if they want to buy a new issue, they will need to sell something else in their portfolio. This is particularly true of those who index or track a benchmark closely when a new issue is likely to be included in the index. Very often, what they chose to sell is something that can be interesting to us.
• Cheapening of valuations The extension of this dislocation is that, as supply rises, inevitably the price drops. The large amount of issuance in November cheapened considerably – we would estimate that on average, convertible bonds in general cheapened about 3 bond points. Many cheapened by more. Clearly, cheaper convertibles is good long-term (though it will of course hurt mark-to-market NAVs).

Perhaps one of the biggest surprises of the current spate of issuance is that it is happening at the pace it is.

Issuance should be correlated with interest rates and credit spreads. When the absolute cost of financing is high, then the attractions of reducing that cost via a convertible deal also would naturally be high. When interest rates and spreads are low, as they are now, then the attractions of convertible bonds should be lower too. In short, companies should just issue straight debt.

There are three reasons, in our view, why issuance is strengthening:

1. Lack of alternatives When we wrote 18 months ago about the reasons that new issuance would accelerate, it came down to the attractions of convertibles, the shortage of bank financing and the lack of depth to the high-yield market (particularly for European and Asian companies). All of these remain true. However, the need for capital is accelerating relative to 2012. Growth is happening and companies want to expand to take advantage of that opportunity.

2. Opportunistic financing The market might just have hit the best moment for issuers to lock in their most attractive ever terms. Rates are at all-time lows and, with Bernanke’s speeches about tapering, CFOs can be forgiven for believing rates are unlikely to go much lower than they are today. If one wants the lowest ever rate, now is the time to capitalise on the current pricing environment. Furthermore, with long-dated implied volatilities relatively high, convertibles are most likely to deliver the best possible pricing, if low interest costs are one’s goal.

3. Diversify investor base With the convertible market now dominated by long-only funds, companies can now access a new, distinct and differentiated investor base. With the Solvency II regulations pushing insurers clearly towards the asset class, this is only likely to strengthen. It should also have a snowball effect: if the asset class becomes larger, it will be more important for all investors to allocate and the depth of the investor base will increase further. Consequently, more issuers will want to tap this source of capital.

We would conclude that, if issuance can accelerate when rates and spreads are low, they are likely to strengthen further if rates rise.

Thus, convertible new issues have historically been a precursor to excellent returns. The reasons for this are clear and logical. The conditions for issuance look good in the short- and medium-term.

Dispersion in convertible valuations

We touched briefly on convertible valuations cheapening when the market is faced with a wave of new issues. It is perhaps worth pointing out that, prior to November, convertible valuations had moved to being (in our view) on average at or slightly above fair value – which is the upper-end of their normal range.

Most of the time, convertibles are the cheapest source of long-dated equity optionality available. There are several reasons for this:

• The issuer is a company, not an option seller and they have other priorities;
• The company is not “short the tail” – they do not have to hedge the downside risk of being short the option, as a seller might;
• The volume and scale of optionality is far larger than most option markets, requiring a discount to clear the market; and
• The complexity of the instrument makes accurate valuation difficult and so investors err on the side of caution in their assumptions.

Thus, trading, on average, at fair value is an unusually rich valuation for the market as a whole.

However, the key words in this statement are “on average”. Averages in financial markets are spectacularly useless. If the dispersion in valuations is large (and one is not forced to buy the expensive bonds), the average becomes not only meaningless but also misleading.

On studying the universe, one can quickly identify very cheap bonds that offset very expensive ones. Looking at some UBS data, their valuations are throwing up bonds that are 17 bond points expensive and 23 bond points cheap – a 40pt range.

Even the range tells one very little, as it could be by just two significant outliers. However, one standard deviation is 5 bond points, which indicates that the market is particularly dispersed – almost 10 points around the mean. It is noticeable that the narrower (and more followed UBS Global Focus index is both more expensive and less dispersed).

Clearly, this dispersion is valuable, if one is an active manager. However, it is also important to be running an unconstrained mandate in the space. The reason for this is most likely that this inefficiency is driven by the lack of hedge capital in the market.

Without the cold rationality of hedge funds and their ability to arbitrage out this inefficiency, anomalies can persist almost indefinitely. This anomaly in the high degree of dispersion in valuations is, in our view, caused by the dominance of long-only capital. Long-only capital, tends to be judged relative to a benchmark. The natural instinct is therefore to minimise the tracking error to that benchmark. In short, the most expensive bonds tend to be the bonds that have large weightings in most long-only benchmarks.

For a benchmarked manager, this is apparently of little concern – they are not at risk owning expensive bonds, if they are neutrally weighted. However, it is of huge importance for an investor – in fact, more important than owning expensive equities for a very simple reason: equities can remain expensive indefinitely convertibles cannot. At maturity, the option embedded in the convertible must be at fair value – the equity option will have either expired worthless (so the bond redeems for cash) or will be at intrinsic value (i.e. its value will be determined by the conversion value of the underlying equity).

Imagine two scenarios. In the first, one overpays for an option. In the second, one underpays. In each case, we will consider two outcomes: one where the share price falls and the option expires worthless; the second where it expires in-the-money.

The difference is dramatic. In this example, the profit is four times the loss potential for the cheap option and yet just a quarter in the case of the expensive option.

In conclusion, there remains sufficient value in the convertible market for long-only and hedged investors to find alpha. However, that value seems to us to be clustered in off-benchmark situations. It has never been so important to have an active and benchmark-agnostic manager in the asset class.

Rising dispersion in equity returns

The correlation of equity returns (or rather lack of correlation) is important to convertible returns. The whole premise of convertible investing is that one is buying an asymmetric trade – one can make large, unlimited gains if the equities rise but, if the stock falls, generally the losses are small and contained by the bond element of the security.

The importance of this convexity is that, in a market with a high dispersion of equity returns, it should make it easier to make profits in a convertible portfolio.

To explain this, imagine a simple scenario: the equity index is unchanged but half its constituents are up 20% and the other half are down 20%. If one owns equities (which have a symmetric return profile) in order to make a profit, one needs to be able to select more of the ones that rise than fall. However, in a convex portfolio, the winners will still make significant positive returns but the losses will have only muted negative returns – in short, one does not have to be anywhere near perfect in stock selection to make money.

Correlation has been extraordinarily high since the global financial crisis. This is probably understandable. Markets were driven by crisis announcements by politicians and central bankers. Fundamentals meant very little. Equally, companies, terrified by the chaos unfolding around them, were unlikely to make important long-term strategic decisions (M&A activity fell 56% from the pre-crisis peak to the 2009 low).

However, today things are different. As central bankers try to disentangle themselves for the direction of equity markets, political announcements are becoming less important (compare Bernanke’s most recent comment on tapering with those that went before). In addition, companies are now starting to make the type of transformational business decisions that drive share price movements (M&A activity has risen 32% from that 2009 low). We believe that fundamentals have taken over and, as the chart above shows, correlation is falling (dispersion is rising). It is only currently, in our view, at “normal” levels. It certainly does not look stretched in either direction.

This has important implications for convertible investing. It should make it easier for convertible managers to make returns generally. However, the best returns should be found by focusing on the highest convexity situations – those with the highest “gamma”. Gamma is the property that makes convertibles gain or lose equity exposure – the higher the gamma, the faster it becomes equity sensitive when share prices rise and the faster it becomes bond-like (and therefore the faster it protects your capital) in a falling market. It is therefore the key measure of convexity in a convertible.

Whilst we try to stay away from talking about “greeks” in these pieces, this is important for understanding the types of bond that benefit the most from dispersion.

In conclusion, the manner in which equity markets are delivering their returns seem to be returning to a more dispersed style of distribution than we have seen for a number of years. This should make it easier to find good risk/return in convertibles relative to equities. However, it remains important to understand the convexity of the individual trades one takes on – expensive and longer dated bonds will be less asymmetric. Again, in our view, one needs to select a manager who is unconstrained and willing to maximise the convexity to find the best risk /reward.

Macro conditions that may favour the asset class

We don’t claim any particular insight into macro factors. However, the scenarios that other commentators are proposing would seem to have attractions for convertible bonds.

There are two main elements that are often highlighted:

1. Interest rates are unlikely to fall further; and
2. Bullish predictions for equity markets.

Both of these have attractions for convertibles.

Therefore, with a unique combination of little/no interest rate risk, less default risk and equity upside potential, convertible bonds are one of the few places in the bond market where one should be reasonably confident of some potential capital gains.

The equity market would appear to be one of the markets with the highest chances of delivering positive returns in 2014. However, it is nevertheless becoming an increasingly risky place to source profits. Convertibles deliver one potential solution: participation in equity upside with limited risks.

Pages: 1 2