Taiwan doesn’t seem like a country that should weigh heavily on the US interest rates market. Yet the callable bonds that the small island nation’s life insurers buy in bulk are helping power the US swaptions market, holding down US rates volatility – and generating an estimated $200 million a year in profit for some banks' rates desks.
“This is probably the single most profitable part of the rates options market-making business that exists, and globally this is arguably all of the money that comes in on the US dollar swaptions desk. I would estimate it accounts for a few billion dollars of profit across the Street per year,” says a portfolio manager at a hedge fund in New York who formerly worked at one of the dealers.
The Taiwanese insurers have regulatory incentives to buy US dollar-denominated bonds known as Formosas, which contain embedded, Bermudan-style US dollar swaptions exercisable only on certain dates. Banks’ rates desks offer hedges to issuers, then turn around and sell European versions of these swaptions to hedge their volatility exposures, known as vega, to the US market, profiting handsomely from the basis between the products.
But this cash cow could now be under threat. New regulation from the Taiwanese Financial Supervisory Commission (FSC) introduces a mandatory five-year non-call period on the bonds, which could impact appetite from issuers and, in turn, hit the supply of US dollar swaptions. According to Dealogic, bank debt issuance has made up 83% of Formosa bond issuance year-to-date, and some think that is where the regulation could hit hardest.
“The profit for the banks comes from buying the swaptions, hedging them out and using that vega position to either sell the vega back into the market or to create a positive gamma profile,” says a senior rates trader at a European bank in London. “It is worth around $200 million for a lot of American banks, but the supply of US swaptions to them could be set to drop off now.”
The new rules are already having an effect. Issuers rushed to get Formosas out before the regulation came into force in May, with callable issuance hitting record highs of $22.5 billion during the first quarter, up 42% on the same period last year, according to JP Morgan. Some analysts, however, predict a 36% year-on-year fall in the second quarter. At the same time, the rising strength of the New Taiwan dollar against the US dollar makes Formosas less attractive as an asset.
A decline in Formosa issuance could mean volatility goes up because these corporates are a big supplier of volatility to the marketChief investment officer at a US-based hedge fund
Others are more sanguine. “In terms of the notional amount issued, I don’t think the regulation will have a big effect,” says Albert Lin, head of Americas structured rates trading at Barclays in New York.
If the swaptions flow from these products is upset, it could have an impact on US rates volatility. Traders point to the Formosa bond market as one of the reasons long-dated US rates volatility is so low. The Chicago Board Options Exchange's 10-year US Treasury note volatility index is near all-time lows, at 4.49% as of June 1. While at-the-money forward implied volatility for five-year options on 25-year swap rates – a proxy for callable bond activity – has as yet shown no signs of moving higher, some estimate the FSC's regulation could eventually hit volatility supply and see long-dated US rates volatility rise.
“A decline in Formosa issuance could mean volatility goes up because these corporates are a big supplier of volatility to the market. Rates volatility has been at its lows, which is crazy considering the Federal Reserve has been on this hiking path all on its own. It doesn’t make sense to be that way, and one of the reasons for that is all this volatility selling from corporates in Taiwan,” says the chief investment officer at a second US-based hedge fund.
The lure of Formosas
Since 2014, Taiwanese life insurers have been investing heavily in Formosa bonds because, despite being foreign currency-denominated, they do not count towards a 45% regulatory cap in investments in foreign assets as they are issued onshore and trade at the Taipei Exchange. Formosas are also callable by the issuer, so they attract a comparably higher yield – at least 4%, typically – than regular bonds with no optionality.
“These insurance firms have large stocks of legacy policies with relatively high guarantee rates compared to local bonds, so US dollar assets are always in the mix when considering possible alternatives. And if you add call features to a bond, its yield goes up more. As long as they have this asset-liability gap, then you will have demand for US dollar callable bonds, especially if their assets continue to grow,” says Joshua Younger, an interest rates strategist at JP Morgan in New York.
As US rates have declined, Formosas have also been popular for issuers. Corporations have been able to call their long-dated bonds after just one year and reissue them again at a lower yield. For example, in July 2016, the 30-year US Treasury rate dropped from 2.98% to lows of 2.11%, sparking calls of 20–25% of bank-issued Formosa bonds, according to a senior banker in Taiwan.
“In 2014, Taiwan changed its regulations so that if a company issued an international US dollar bond, and listed it on the Taipei Exchange, it qualified as a local portfolio bond. That opened the door to international firms, at first mostly bank treasury departments looking for flexible long-term financing, but then treasurers at major US and European corporates also entered the market,” says Rodrigo Gonzalez, head of debt capital markets for the Americas at Standard Chartered in New York.
All the dealer volatility desks wind up long Bermudan swaptions, which typically have a lock-out period, so they’re not callable for a certain timeRichard Mazzella, Hutchin Hill Capital
The Formosa bonds are attractive to issuers because, after swapping the fixed rate risk and embedded callability to floating, they achieve more attractive all-in funding costs compared with issuing vanilla floating rate debt.
To do this, the issuer executes a package trade, entering into a receive-fixed interest rate swap and simultaneously selling a Bermudan-style receiver swaption, which is only callable on a given date per year. The swaption matches the optionality embedded in the Formosa and the premium reduces the effective all-in floating rate payable by the issuer.
“If a corporate issued a 30-year bond and swapped it to Libor using an interest rate swap, it would probably pay somewhere around a 4.5% fixed rate and receive a floating rate of around 200 basis points over Libor. If the corporate decided to enter into a swap and monetise the call options through the swaptions market, it would be able to have a 30-year, non-call five-year that swapped to Libor plus 130bp, saving 70bp. Typically, the savings are between 60bp and 80bp,” says Gonzalez.
If rates fall, the issuer calls the Formosa, and the sold receiver swaption would be exercised by the counterparty cancelling out its initial pay-fixed swap to the issuer.
The banks have been quicker than other institutions to take advantage of the lower floating rate exposure available through the swaptions market, but Gonzalez points to increased interest coming from other sophisticated US firms, such as Pfizer, AT&T and Verizon. According to JP Morgan, corporates doubled their Formosa activity in the first quarter compared to the 2016 average. However, the majority of corporates leave their fixed interest rate exposure unhedged.
Dealers’ rates desks act as the swap counterparty to the bank treasury desks or corporates issuing the Formosas. When they enter these package trades they become structurally long rates volatility from the Bermudan swaption exposure.
“All the dealer volatility desks wind up long Bermudan swaptions, which typically have a lock-out period, so they’re not callable for a certain time,” says Richard Mazzella, partner and managing director at hedge fund Hutchin Hill Capital in New York.
Typically, banks do not want to be structurally long Bermudan volatility as it attracts higher market risk capital requirements. In the future, the illiquid position would be treated more harshly by the Fundamental Review of the Trading Book, and would attract a residual risk add-on.
To reduce their vega exposure, banks can’t simply turn around and sell these positions due to their illiquidity, so instead they sell calls on European-style swaptions to a combination of hedge funds and asset managers. The banks then need to manage the basis risk between the long Bermudan and short European swaption positions. Market participants say this is where the big money can be made.
“Let’s say a European option trades at 60 volatility points per year. The Bermudan itself would typically trade at a further 15 volatility points per year below that, which is a gigantic number. That number has been bigger in the past, ranging from between 17 and 25 points. You don’t immediately monetise it, but instead make gains over the life of the trade,” says the New York-based portfolio manager.
The wide basis has attracted some hedge funds chasing a relative value play, but managing the position is tricky and the pay-off is far away in the future (see box: Hedge funds eye Bermudan-European swaptions trade).
Managing the position is made more complicated due to the variable US interest rate. If rates go down, the Formosa bond issuer is likely to call its bonds so it can reissue again at a lower rate, leaving the bank holding the European swaptions positions, which can only be called at maturity. According to one senior trader at a third hedge fund, this means banks are not able to price solely using a Black-Scholes formula, but instead have to use a simulation-based pricing method.
“The pricing model will spit out a set of hedges that are European swaptions, and depending on how sophisticated your model is, it will have a set of at-the-money options and hedges for the vega. The bank will have some skew exposure too, so it will suggest some out-of-the-money options to sell to hedge that as well,” says the trader.
Alternatively, if US interest rates go up, the expected duration and volatility of the exposure increases, and the position needs to be dynamically re-hedged.
“Say at the time traded the expected maturity is five years on a 10-year non-call one-year bond, and, as rates go up, the weighted-average life of the trade might extend to nine years. Essentially, you’re suddenly longer volatility and therefore you need to sell more options. As rates go up, volatility goes down because all the dealers have this big exposure,” says the senior trader at the third US-based hedge fund.
A swaptions trader at a European bank in New York agrees. “The hedging activities change depending on whether rates go up or down, which we call the vanna impact. For example, when rates go lower, the Bermudan swaption is more likely to be called. Then the dealers are more likely to sell shorter-expiry vega versus buying longer-expiry vega,” says the trader.
“The bank does not necessary make direct profit from the hedging. For us it’s more about reducing the illiquid risks and the risk-weighted assets, which helps to increase the return on capital,” he adds.
This hedging activity by the banks is said to be part of the reason long-dated US rates volatility remains subdued. “Formosa bonds are absolutely a factor in keeping US rates volatility low,” says Barclays’ Lin.
“There aren’t that many natural users of long expiry vega and there is this huge amount of supply coming from callable issuance, so it definitely has a big impact on the shape of the vol surface. You can extract a correlation between the movements of long expiry vol versus issuance.”
Bracing for regulatory impact
This might all change following the FSC’s introduction on May 23 of a requirement for life insurers to only buy Formosa bonds that cannot be called for at least five years. The New York-based portfolio manager says this could affect the frequency banks and corporates issue the bonds.
“With rates having moved lower, it’s been economically beneficial to call it at the very first call. That’s been the case for past few years. If you change the dynamic of that so it’s only after five years, the likelihood of the bond being called frequently and reissued is going to change dramatically,” says the portfolio manager.
However, a structured products specialist at an Asian bank in Hong Kong says large international bank issuers may actually welcome the FSC’s change as it allows them to lock in funding for at least five years. But the banker says corporates that like the flexibility of being able to call and reissue may fall away as a result.
“For people who genuinely need long-dated funding, it won’t be a problem. But those who tap this market on a more tactical basis might lose interest. I think that is the reason why there was a glut of corporate issuance in the first quarter. People were expecting this regulation to come in the second quarter, so a lot of corporates jumped in and issued early on in the year,” he says.
Emmanuel Lefort, head of global markets for Asia-Pacific at Natixis in Hong Kong, says there could be an impact but it is likely to be small.
“Issuer demand may be impacted as the change will result in less flexibility for issuers, and a minimum non-call of five years may not meet their funding plan. But the impact on investor demand will likely be minimal as the benefits of Formosa bonds remain the same,” says Lefort.
Formosa issuance volumes should remain stable because Taiwanese life insurers are very conservative in terms of the credits they can considerLorna Greene, National Australia Bank
On top of all this, the New Taiwan dollar’s nearly 7% rise against the US dollar is making the bonds more expensive to hold as an asset (see box: New Taiwan dollar rise hits Formosas).
Despite everything, however, market participants believe Formosas are here to stay. The combination of the higher yield still offered in Formosa bonds and the fact that the issuer tends to be highly rated makes the instruments uniquely appealing to life insurers.
“Formosa issuance volumes should remain stable because Taiwanese life insurers are very conservative in terms of the credits they can consider. There are not many other similarly yielding options available, so the Formosa market remains the most attractive option,” says Lorna Greene, director of debt syndicate and origination for Asia at National Australia Bank in Hong Kong.
In addition, Taiwanese life insurers have naturally moved to instruments with longer non-call periods, also known as lock-out periods, over time. In 2014, 60% of Formosa bonds bought mainly by Taiwanese life insurers had a non-call period of one-year or less. By 2017 that figure had declined to less than 10%, according to JP Morgan. Lock-out periods of between two and five years now make up over 40% of the notional balance, with five-year non-calls representing 25%.
“Over time there has been a lot of reinvestment risk and a lot of capital returned and reinvested at lower yields, so [the insurers] lost out on returns in that scenario. As that has occurred, without the intervention of the regulator, the Taiwanese lifers have extended their typical lockouts quite a bit. So what was initially very biased towards a one-year lock-out flow has since been much more easily split between one, three, five years and longer,” says JP Morgan’s Younger.
New Taiwan dollar rise hits Formosas
While all eyes have been fixed on the changes being made by the Taiwanese Financial Supervisory Commission (FSC) and the record issuance of Formosa bonds in the first quarter, the strength of the New Taiwan dollar against the US dollar complicates things further.
Since the turn of the year, the US dollar has lost 6.8% of its value against the New Taiwan dollar, dropping from 32.318 on January 3 to 30.121 on June 1. This could potentially make it less attractive for Taiwanese life insurance companies to invest in US dollar assets, such as Formosa bonds, according to Joshua Younger, interest rates strategist at JP Morgan in New York.
“All else being equal, these currency moves could make it less appealing to invest in US dollars. At the same time, hedging costs have increased on a forward basis as these companies have accumulated US dollar assets, more than doubling over the past year or so,” he says.
Rodrigo Gonzalez, head of debt capital markets for the Americas at Standard Chartered in New York, says that while this currency situation may not last, it could raise hedging costs for Taiwanese investors.
“Our economists believe the government wants the currency to be more competitive, so this will likely be aimed at reverting this tendency. For now though, it makes Formosas less attractive compared to domestic investments, and makes the hedge a little more costly,” says Rodrigo Gonzalez, head of debt capital markets for the Americas at Standard Chartered in New York.
Others feel that due to the attractiveness of Formosa bonds, the strengthening of the Taiwanese currency against the US dollar will not have a significant impact on demand.
“Given that they have so many assets that they need to hedge, there is therefore a lot of liability that they need to find product to invest in. So looking at the bigger picture, the strengthening of the Taiwanese dollar is not that significant,” says an investment banker based in Hong Kong.
Hedge funds eye Bermudan-European swaptions trade
With big profits to be had managing basis between Bermudan and European US dollar swaptions, some banks have noticed a handful of hedge funds becoming involved in buying cheap Bermudan US dollar swaptions and selling more expensive European swaptions as a relative value play.
“They realised the cheapness of the Bermudan risk versus the European risk, and they are engaged in trying to absorb some of the inflow from Asia,” says a senior rates trader at a European bank in New York.
“It’s hard to manage the Bermudan risk, but I think a lot of the fast-money accounts are just looking at the relative value. Some of them see it as a tail hedge play, so they are buying some forward optionality for cheap. A lot of them are not even dynamically managing it by not hedging all the vega out from the Bermudan risk, but they see the cheapness in the volatility and the trades get positive carry if volatility stays at these levels. So it’s a view on their part on future volatility levels,” says the trader.
Two of the US-based hedge funds that spoke to Risk.net said they were active in this market.
“It’s a component of what we do,” says a portfolio manager at a hedge fund in New York. “Long-dated volatility in rates tends to be low because of all the structural supply from Formosa issuance and other factors, and that creates an opportunity for funds to buy some of this long-dated volatility that doesn’t exist in most other asset classes. Equities and foreign exchange long-dated volatility tends to be quite high, so it’s nice to find an asset class where you can find it. However, it does take a long time for it to pay off.”
Richard Mazzella, partner and managing director at hedge fund Hutchin Hill Capital in New York, says that, like banks, it is also difficult for funds to manage the Bermudan options risk. If they’re holding a Bermudan-style option it qualifies as a so-called level three asset on a hedge fund’s books, meaning it is not particularly liquid or easy to price, which can be unpopular with investors.
“Everyone in the Bermudan swaptions market is structurally long the volatility, so if you’re a buyer then that’s great. If you’re a seller, then it’s not like there is a ton of people looking to take the other side,” he says.