The convex payoff of convertible bonds is well-suited to these uncertain times. But Martin Steward asks how easy - or desirable - it is to maintain optimal convexity
As quantitative easing (QE) depresses bond yields and spills liquidity into risk
assets across the globe, investors need to consider possibilities for combining yield and growth characteristics to achieve convexity - the tendency to respond positively in either risk-on or risk-off markets or, as is the case today, both at the same time.
The purest form of convexity comes from convertible bonds issued by growth companies that find it difficult to access straight bond markets. By offering investors exposure to their infinite supply of stock volatility they keep a lid on their bond yields. While the yields are often comparatively low and the options comparatively expensive (the ‘conversion premium’), the contribution offers equity exposure with downside limited to the ‘bond floor’, the value of the remaining cash flows to the fixed income instrument.
Right now this enables you to ride waves of QE liquidity into equity risk while insuring against bearish anxiety about liquidity traps or double-dips. “It’s a bit of a sweet spot for convertibles,” says Mike Reed, a senior portfolio manager at BlueBay Asset Management. Convertibles specialist Peter Warren at CQS agrees: “The natural convexity of convertibles should give them a place in people’s thinking, particularly as they look at an uncertain world.”
But there is not a constant quantum of convexity in convertible bonds: the level changes with the value of the underlying equities. If the option is a long way out-of-the-money (OTM) the convertible behaves like a straight bond, and if it is a long way in-the-money (ITM) it behaves like equity. Delta expresses this relationship: a delta around 0.5 reflects maximum convexity, or equal exposure to bond and equity characteristics.
In 2008, plummeting stock markets left deltas as low as 0.20. Some yields, lower than government bond yields for nine out of 10 convertibles before the crisis, leapt into double digits: they were being priced as straight bonds - ‘busted convertibles’. But because they are short-dated, once the market had sorted out which names were real default risks, the rest quickly began to recover as investors started pricing-in recoveries at par.
“As spreads came in and bond floors strengthened equities also recovered, and the market started to think about optionality in again,” says Aberdeen Asset Management’s head of convertibles Pierre-Henri De Monts De Savasse.
Today yields have pulled back to about 2-3%, delta to around 0.40, and we appear to be moving right into the sweet spot for convexity. But if one of the premises for convertibles is right - QE will pump up stocks - what is to stop delta from creeping up and eroding our convexity?
This is what happened in 2006-07, and when the crashes came convertibles behaved like equities. “You have to manage that type of risk actively,” says Martin Coughlan, a senior portfolio specialist with Calamos Investments. “Sometimes you may have to do that by constructing synthetics to bring your equity sensitivity back down. You can’t buy a converts index and expect convexity: it doesn’t always get delivered, which might explain the high attrition-rate among convertible bond managers.”
To some extent the market has a natural ebb and flow around delta 0.4-0.6 because new issues tend to be priced around 0.5. Moreover, at the trough of the business cycle when bonds are generally OTM, companies are most likely to issue new convertibles because they are unable to raise equity; and when bonds move ITM, issuers begin to call or force conversion to avoid paying the remaining coupons. Delta collapsed in Q1 2009, but the taps opened up again by April as cyclical companies sought to restructure debt, and at over $100bn (€72.5bn), 2009 turned out to be a bumper year. “Investors have the opportunity to construct very high-convexity portfolios because we saw such a lot of new issues last year,” says Fabienne Girard-Tokay a convertibles fund manager with Mandarine Gestion.
This dynamic provides the logic for De Monts De Savasse’s strategy, which targets market delta rather than convexity: “The naturally changing demographics of the universe means that your portfolio will find itself most exposed at the best possible time, and the universe will offer the optimal delta at each point in the cycle.” During 2009, if you held busted old names like Air France, Peugeot or WPP, you could ride their spread-tightening performance without becoming skewed to higher-delta new issues. But if, like F&C Asset Management or Schroders and many others, you target convexity, you had to act.
“We needed to re-position because deltas in the portfolio had fallen so low,” says Alan van der Kamp from F&C’s convertibles team in Amsterdam. “By summer 2009 we regained our balanced status - 0.40 - with the combination of market movements and repositioning.”
That delivered a 36% return for 2009. But investors who picked up new issues also saw equity sensitivity rising higher than those who sat them out. Deltas in many 2009 issues are now in the 0.80-1.00 range. F&C’s portfolio delta has jumped from 0.4 to 0.50 in just a couple of months, significantly higher than the market’s. Schroders’ is similar.
At this point of convexity there is not yet a problem but if the market continues to rally it will become one. Convexity players will need to rebalance using whatever they can get from primary markets or any temporary dips in secondaries. “It’s not on the agenda now,” as Swisscanto portfolio manager Peter Meier says, “but that might mean liquidating the high-delta names and where possible we looking for new issues at 0.5.”
What does that mean in an environment of falling yields, falling equity volatility and rising equity markets? Well, the good news is that while you probably won’t get as much yield per unit of delta as you did in 2009, yields do get reset to levels well above those of the same credits in the secondary market. Vedanta’s July 2009 issue came with a 5.5% coupon: eight months later the bond’s delta hit 0.72 and its yield had plummeted to 1.5%, but it issued a new bond at delta 0.5 with a 4% coupon. Less encouraging is that, ceteris paribus, options will be more expensive: the initial conversion premium on Vedanta’s 2009 bond was 35%, but 37.5% on its 2010 issue.
“The issue is not maintaining convexity, it’s the price you pay to maintain convexity,” says Steve Roth, manager of the GLG Convertibles fund at Man Group. “At the beginning of this year there was a lot of cheapness in the investment grade universe, but because of the lack of issuance most of that cheapness has disappeared. With AAA names like Microsoft you’re actually 3-4 vol-points rich to [straight] options.”
Tom Wills, manager of Morgan Stanley Investment Management’s Global Convertibles Strategy, agrees. “There are several blue chip companies which, when modelled, are overpriced,” he says. “The delta may seem attractive, but you are overpaying for that option because the market is now dominated by outright funds rather than arbitrageurs. However there are opportunities on the other side of this - attractive names which are up to 10% cheap partly because they do not feature in benchmarks.”
Indeed, complaints about the paucity of new issuance (and rising conversion premiums) are loudest from those constrained to European or US investment grade. Roth notes that $8bn of investment grade issue is due to mature in the US next year. “But if you have the flexibility to invest in non-investment grade - where the universe is migrating - you can still find interesting bonds.”
Wills makes the case for a global mandate. “Europe has 25% of the 1,500 or so liquid issues,” he says. “It is difficult to maintain a sensible delta profile while also picking your favourite equities and credits from those 300 or so issuers.”
Everyone with a mandate to do so has been all over the Asian markets, which since August have seen a boom in new issuance, including a record-breaker from China Unicom. Carrell at Schroders observes that Asia delivers optionality twice as cheap as the US. He also points to big industrials like Tata Steel or Mahindra & Mahindra that still offer 4-5% yields in the Indian secondary market. “Our standalone credit team sees a lot of value in that area. But they’ve never had a default: you have to know your credit analysis to get any kind of yield in convertibles now.”
This extra yield is crucial, because buying higher-priced options to maintain convexity exposes you to a spike in implied volatility, which has to be compensated by improved bond floors.
Convexity is a precious commodity, but it is not ready-made. If this is your objective in convertibles, you should first confirm that this fits with your manager’s style; and then be comfortable that he has the credit and equity skills so crucial to maintaining convexity - because it is at just those times when it becomes so valuable, like today, that managing it safely becomes most demanding.