AsianInvesterAsianInvester

Why Taiwan’s foreign investment cap should be lifted

Issuance in Formosa bonds has crashed following regulatory changes, but insurers still desperately need higher-yielding products to meet their obligations to policy holders.
Why Taiwan’s foreign investment cap should be lifted

Taiwan’s Formosa bond market is experiencing a sharp reversal of fortunes as recent regulatory and interest rate moves make it a far less attractive proposition for the international issuers that have been attracted to the sector in droves over the past few years.

This spells bad news for the country’s life insurance industry, which has been relying on Formosa bonds (domestically listed, foreign-currency-denominated instruments) to meet its portfolio yield targets ever since they were exempted from life insurers’ 45% cap on offshore investments in 2014.

Here FinanceAsia, a sister publication to AsianInvestor, looks at the ramifications of a recent Financial Supervisory Commission (FSC) ruling banning the market’s staple short-dated callable bonds.

Does the regulator need to consider lifting insurers’ offshore investment cap instead? Or will it now speed up approvals for new types of Formosa bonds, which might not only provide a better asset/liability match for insurers’ portfolios, but also give issuers enough of a pricing advantage to come to the Formosa market in the first place?

What is clear is just how immediate the impact of the May 24 ruling has been on issuance levels. Dealogic figures reveal a 68.3% drop-off during the second quarter to just $6.4 billion from $20.3 billion in the first three months of the year (see table below).

Activity was also almost half the level it reached during the second quarter of 2016 when borrowers raised $11.97 billion from Formosa bonds.

In particular, more sophisticated international corporate borrowers, which started arriving en masse towards the end of the third quarter of 2016, have almost completely disappeared from the scene.

The five biggest issues in the first quarter were all executed by US corporate borrowers, led by Verizon, which raised $1.475 billion, followed by AT&T with $1.43 billion, Pfizer $1.06 billion, Comcast $1.005 billion and Apple $1 billion.

In contrast, the five biggest deals during the second quarter mainly came from banks. The exception was a $750 million issue by Comision Federal de Electricidad in late June: the market’s first Mexican issuer.

Why insurers need Formosa bonds

At its heart the Formosa bond market is a regulatory artifice to aid Taiwan’s outsized life insurance industry: the result of local savers preferring to park their money in insurance policies rather than bank deposits.

As Moody’s vice president Frank Yuen pointed out, the combined investment portfolio of Taiwanese life insurance companies stood at $743.43 billion at the end of April and is growing by about $50 billion per year, dwarfing the country’s $289 billion domestic bond market (as of end May) and crimping the industry’s investment options. 

That supply/demand imbalance has long been compounded by a rates mismatch.

At the turn of the century many local life insurers offered products with guaranteed payout rates around the 6% level. Shortly after, government interest rates started dropping and today 10-year Taiwan Treasury bonds yield around the 1.075% mark.

South Korean life insurers suffer a similar problem with legacy products, but they have a much bigger domestic bond market to invest in. The latter is not only larger relative to their investment portfolios, but also deeper in terms of the credits on sale.

In comparison Taiwanese corporate structures are less debt-centric; companies are either cash-rich or in the technology sector, with the kind of volatile earnings streams that make bond issuance more difficult.

Allowing high-yielding foreign borrowers to issue foreign-currency bonds in Taiwan seemed to create a win-win situation for everyone. Life insurers were able to boost their investment returns and reduce negative investment spreads by augmenting the foreign-currency bond component of their investment portfolios.

According to PineBridge Investments research, Taiwanese life insurers held 34% of their investment portfolios in foreign currency-denominated bonds in 2010. Fubon Life’s recently published first-quarter results show that this ratio has almost doubled to 62% over the past seven years. Cathay Life is not that far behind on 60.4%.

Fubon also reported a recurrent yield of 3.01% during the first quarter. This still remains below the group’s 3.76% cost of liabilities, generating a negative investment spread. Cathay has an even more pronounced differential, having reported a 3% recurrent yield against a 4.21% cost of liabilities at the end of the first quarter. 

Moody’s Yuen says Taiwanese insurers have been able to bring their average cost of liabilities down by about five basis points to 10bp per annum over the past five years and estimates an ongoing drop at the same rate.

However, there is no doubt that the investment spread would look healthier still if there was no offshore investment cap at all. Being a captive investor base gives Taiwan’s life insurers less pricing power with issuers.

“What this effectively means is that faced with a low-interest-rate environment at home, Taiwan’s life insurers have traded access to higher-yielding US dollar paper by ceding some basis points, or effectively giving away a call option," said Rick Chan, a Los Angeles-based portfolio manager at Pimco. 

This pricing dynamic clearly benefited issuers in a second way, since they were also able to reduce their cost of funding by swapping proceeds from fixed to floating rate. Bankers suggest they saved upwards of 15bp per transaction.

In turn, banks made a lot of money (a couple of billion of dollars, according to some market sources) trading the volatility element of the call options.

Chan says issuers typically monetised the bond’s Bermuda-style call option by selling it to the issuing bank. In turn, the bank’s swaptions desk managed this long position by going short in the European swaptions market, profiting from any differential between the two.

One unintended consequence of all this activity ($48.28 billion Formosa issuance in 2016) is that volatility at the long end of the US rates market has stayed very low.

All good things…

The FSC’s decision to bring a halt to proceedings stemmed from the asset-liability mismatch this issuance activity created for life insurers.

For what were notionally long-dated issues with 30-year maturities were, in reality, extremely short-dated bonds because of the call options, which fell due after the first or second year.

In a declining interest rate environment it made sense for borrowers to call the bonds as quickly as possible so they could keep re-financing their debt at ever-lower yields.

Banks also benefited by playing the volatility element of the embedded call options.

“As time passes and rates move lower, the volatility of callable bonds decreases," Pimco's Chan explained. "In an extreme case where rates were to rally to 0%, the likelihood of a Formosa [bond] getting called is close to 100%, which means the volatility exposure is close to zero."

"Once a bond's called and a new one issued, there's an increase in supply of volatility to the market, which results in a lumpier supply profile," he continued. 

Therefore each time a bond was called and a new one issued, the banking community had a new supply of volatilty to play.

But in May the FSC stepped in by outlawing deals with call options of less than five years in the primary market and three years in the secondary market. Getting ahead of the ruling was one reason why issuance spiked during the first quarter.

However, Annemarie Ganatra, global head of medium-term notes at Standard Chartered in Hong Kong, believes issuance would have dropped off even if there had been no regulatory intervention during second quarter.

“The yield curve has been steepening so it made less sense for issuers to call the bonds away,” she noted.

Formosa bonds quarterly volume

Pricing date by quarter Deal value ($m)  Number
2015 Q1 12,334 54
2015 Q2 7,955 40
2015 Q3 9,289 41
2015 Q4 4,387 15
2016 Q1 13,546 51
2016 Q2 11,968 44
2016 Q3 15,765 43
2016 Q4 7,005 25
2017 Q1 20,318 47
2017 Q2 6,448 25
2017 Q3 to date 1,310 4

Source: Dealogic

¬ Haymarket Media Limited. All rights reserved.