Proposals for an exchangeable were originally due Wednesday after being sprung on bankers, who were only given 48 hours notice to put together a structure that many believed to be completely unworkable from the outset. After vociferous complaints, the government was said to have reluctantly agreed to extend the deadline on the final afternoon, only to inform the market that it had changed its mind 15 minutes later, before finally conceding defeat for a second time and allowing banks until today (Friday) to arrange their submissions.
At the heart of the issue lies a belief that the idea is a non-starter and that putting together any kind of proposal is complete a waste of time when JPMorgan, the originator of the Opera acronym (outperformance equity redeemables in any asset), must be a shoe in for the mandate anyway. The Korean government's efforts to bring greater transparency to the selection process appear to have rebounded, as it becomes clear that one bank has had a six month head start devising a structure that no-one else has had time to understand.
"JPMorgan's original idea was probably based on quite a sensible strategy," says one banker. "The idea that a government can maximize privatization proceeds through an exchangeable when the underlying shares are trading at depressed levels is a good one. When markets are bad, it also makes sense to offer investors a range of assets for conversion. The problem with this transaction lies with the assets the government is trying to sell - one listed bankrupt bank and one unlisted bankrupt bank."
Critics of the idea cite three major structural flaws. Firstly, investors have remained extremely wary of Korean banking stocks ever since their shirts got blown off by Hanvit Bank's $1 billion GDR of September 1999, which was sold at Won 8,250, a 21% discount to the stock's then close. In April this year, Hanvit along with Peace Bank, Kyongnam Bank, Kwangju Bank and 11 other struggling financial institutions were merged into Woori.
As one Korean specialist puts it, "Investors are aware that these banks are undercapitalized and will be seeking further equity funding. Cho Hung has already indicated that it hopes to launch a GDR next year. There is a high likelihood that exchangeable investors will witness their investment diluted to almost nothing."
Secondly, bankers point out that the government currently holds an 80% stake in Cho Hung and a 100% stake in Woori.
"Cho Hung Bank has a current market capitalization of about $1 billion and who knows where Woori should be valued," explains a second banker. "Raising $500 million represents roughly 2.5 times Cho Hung's freefloat and perhaps one times that of Woori. No dedicated outright convertible buyer in their right minds is going to accept this."
Thirdly, the government has also indicated that it does not want to sell stocks in either bank below their par value of Won 5,000 - the level at which it injected funds back in 1998. Bankers have, therefore, concluded that since Cho Hung is currently trading at Won 1,835 per share (Wednesday's close), any deal would need to have a huge conversion premium attached to it.
"This has to be just about the dumbest idea the Korean government has ever come up with," argues one Korean specialist, whose view appears to echo the many. So why is it doing it? Detractors ascribe the sudden revival of JPMorgan's Opera proposal purely to politics and the government's need to show the IMF that it is doing something to aid the privatization of the Korean banking system. The exchangeable is thus viewed as a window dressing exercise that will be sold as pure debt, for which there is likely to be a ready market.
"There will have to be an exceptionally high bond floor and virtually no equity option," says one convertible specialist. "I can't see investors paying more than 0.5 points for the equity option."
Korean spreads, on the other hand, have been among the strongest performers of the Asian credit universe during 2001 and few doubt that there would be strong demand for the credit of Korea Deposit Insurance Corporation (KDIC), in whose name the exchangeable would be issued. The Baa2/BBB-rated credit is currently trading at about 150bp to 170bp over in the asset swap market on a 2005 maturity, about 20bp wider than where it was pre September 11.
This is based on the outstanding $1.001 billion KDIC exchangeable into Kepco that was launched in September last year. This is currently bid at 111, with a bond floor of 107.33 and yield of 5.07%. With a 2005 maturity, the deal has a put in 2003 at 114.52 and a call in 2003 subject to the 135% trigger.
Had the government decided to pick any of the other assets in which it has stakes, many bankers argue that the potential for success would be a lot greater. These include stakes in stronger banks such as Kookmin and Housing & Commercial Bank of Korea, as well as corporates including Kepco and Korea Telecom.
However, those who favour the structure argue that if the government can get it right, a successful deal will have a beneficial impact over the whole privatization programme. As one explains, "The Korean government has come in for a lot of criticism from those who say that because it still owns large stakes in the Korean banks, corporate restructuring is being driven by policy rather than market economics. Hynix is a strong case in point.
"We believe," he adds, "that if the government can successfully divest itself of the banks, it will get rid of this perception and re-invigorate the privatization programme."
In terms of structuring a deal, the key is to delay setting a conversion price for Woori until it is listed in the first half of 2002 and for Cho Hung, when the stock becomes more liquid. Thus while the conversion premium will be set at the outset, the conversion price on which it is be based will come later after the banks have been further re-capitalised.
The deal will also need to have a three-year final maturity, because this is the timeframe set by the government for its disposal of the bank stakes. In essence, JPMorgan's Korean Opera structure is expected to closely mirror its previous Eu2.5 billion issue for Telecom Italia, with additional tweaking to take into account the difficulties of hedging Cho Hung and impossibility of hedging Woori.
The Telecom Italia structure hinged on giving both issuer and investor as much flexibility as possible. Launched in late January, the five-year deal had a one-off investor put at the end of year three and a call option after year three subject to a 120% trigger. Issued at par, the deal has a 1% coupon, exchange premium of 31.22% over TIM (Telecom Italia Mobile) and 102.83% over Seat Pagine Gialle. Yield-to-maturity was 4.25% and the bond floor came in at 91.
Much of the flexibility in the structure, however, is derived from the delivery options. The issuer has the choice of giving investors the shares, the cash value of the shares, or such shares as can be purchased with 105.3% of that cash amount.
As an analyst explains, "If the deal performs well and the exchange is triggered, the issuer could decide to either deliver shares in TIM, or if it wanted to hold onto the TIM shares, then the equivalent economic value of shares in Seat. If it decided it wanted to hold onto both stocks, then it would give the equivalent economic value in cash."
Investors could, therefore, want TIM shares, but end up with Seat. To compensate for the fact that they would then have to sell the unwanted shares into the market, the issuer would pay a 5.3% penalty charge in cash.
JPMorgan launched and closed the Telecom Italia deal within a day. Launch of a prospective Korean deal, should it win a mandate, could also be equally as quick. However, it is believed that the government wants to clear KT&G from the market first.