Some hours later than expected, the Singapore government completed a monetization of stakes in ST Engineering and Capitaland on Friday. On full conversion of the S$350 million ($203 million) and S$451 million ($261 million) deals, Singapore Technologies will have divested a 4.3% stake in STE, bringing its ownership down to 50.7% and an 8.2% stake in Capitaland, bringing its ownership down to 51.8%.
The Singapore government's intention to offload the stakes has been well flagged for some time and on the surface, completion of the two exchangeables appeared to be a job well done. Both transactions were trading around issue price by the end of Asia's trading day, neither stock had cratered under short selling pressure and the issuer had got itself exceptionally aggressive terms.
Yet, criticism of the terms and execution of the deals has been deafening from just about all quarters and at the beginning of Asia's trading day it had all seemed so very different. At this point, neither deal had been officially priced or allocated, both stocks had been suspended and rampant speculation the leads were long about 80% of the bonds, pushed grey market trading levels straight down to 97%.
Few would disagree there was far too much confusion and confusion is never good for equity markets, nor the reputation of those involved.
An initial post mortem of the deal throws up two points. Firstly, the Singapore government set out with very good intentions and was determined not to fall victim to the bait and switch tactics, which often win banks deals. What it did not appear to have thought through quite so thoroughly were the finer logistical points of how the deals would actually be executed and this led to delays after launch.
Secondly and more importantly, was it wise to select a mandate purely on price in the first place? What if the deal had performed badly in the secondary market because the terms were far too aggressive? Was it worth risking such a reputational hit with investors? Most would say not and particularly since the government had been considering monetizing stakes in five stocks and is, therefore, likely to come back to the market at some point
A mandate process which pitched one bank against another made aggressive terms a given. Forcing banks to buy deals outright and put their capital at risk would have made every house think long and hard. But such is the intensely competitive nature of Asian investment banking, that a number would inevitably still pitch terms that left no room for any slippage and would need a strong prevailing wind to successfully clear the market.
A mandate too late
In the two months running up to the deal, SingTech sought proposals from up to 15 banks, before whittling the number down to a final five and then three banks, which went head to head on the day of launch. These three were Citigroup and Credit Suisse First Boston, which were subsequently awarded joint books and JPMorgan. Prior to this, Goldman Sachs and Merrill Lynch are said to have been in the frame as well.
SingTech's main concern is said to have been an all-in funding cost of about 1.5% per annum. It did not care what fees banks took from the deal, as long as the company met its targets. In the weeks running up to the award of the mandate, it also because clear officials wanted a long dated structure and had very precise conversion prices in mind. The main variable was the coupon.
What happened once the mandate was awarded is subject to argument. Some say the leads hoped to bait and switch over the weekend, while the preliminary documentation was being sorted out. Others say it was always clear the leads owned the deal once terms were shook upon.
Were the latter the case, this left Citi and CSFB in a very difficult position. Had the mandate been awarded at lunchtime, they would have had the afternoon to start sorting out the documentation. But by the time SingTech put them together as joint bookrunners, it was early evening Asian time. Local investors had gone home and the European day was half over.
If they had waited until Asia's morning Friday to launch the deal, they faced building a book while the most important investor base (Europe) was asleep. Were they to wait until London's open Friday, they faced a whole day of market risk and the fear their competitors would try and sabotage the market for them.
So it was decided to turn the deal around like a placement and a fixed price term sheet was sent out a few hours before London closed on Thursday. Books were built on a first come, first served basis and accounts were told to be precise with their orders, as they would get filled.
STE is said to have been covered within three to four hours and Capitaland about half an hour later. When the two leads merged their order book, there was said to be an S$380 million book for STE and an S$480 million book for Capitaland.
SingTech had said it would make stock borrow available and accounts subsequently went on to receive about 35% of overall demand. However, it is at this point that legal issues appeared to gain the upper hand. In particular, outside observers say a stock borrow agreement and underwriters agreement were not signed until lunchtime on Friday. There were also issues with various reps and warranties.
In retrospect, the combination of a rushed timeframe, a first time borrower and one with a reputation for meticulousness was probably not ideal. As a result, the leads asked for a suspension of the stocks and the deal was not free to trade until the stock exchange was in a position to make a public announcement early in the afternoon.
Pricing of the S$350 million offering for STE came with a seven-year final maturity and a five-year put. With a par in par out structure, the deal had a coupon of 1.56% and a conversion premium of 45.8% to the stock's S$1.92 close.
There is also a three-year call option with a 120% hurdle. At the call option, the issuer has the option to force conversion into either shares or cash. At maturity, it is simply cash.
Underlying assumptions for this deal are a bond floor of 92.6%, theoretical value of 99.5% and implied volatility of 33%. This is based on a credit spread of 55bp over Sibor, which is equivalent to about 65bp over Libor. The stock borrow assumption is 0.5%, cash dividend assumption of 6.3% and volatility assumption of 24%.
The S$451 million Capitaland deal has the same seven-year maturity, with a five-year put option and par in par out structure. Here the coupon was set at 1.08% and the conversion premium at 46.66% to the stock's spot close of S$1.50. The call option falls in year four.
Underlying assumptions comprise a bond floor of 90.6%, theoretical value of 100% and implied volatility of 33.5%. This is based on the same credit spread assumption, 0.5% stock borrow cost, 3.33% dividend yield and volatility assumption of 32%.
Both order books are said to have been fairly concentrated, with two investors together taking about half of each deal. The leads reject speculation that bonds were sold below par and state that both deals were fully distributed at issue price.
With STE, a total of 35 investors are said to have participated and with Capitaland about 45. Geographically, the book had a 60%/40% split between Europe and Asia.
Outside observers highlight three particularly aggressive aspects to the terms. All agree that a credit spread of 55bp is too tight and while the market diverged between 55bp and 125bp over Sibor, the consensus settled around the 85bp level.
In the case of Capitaland, this would have lowered the bond floor to below 90% and to about 91.9% for STE. It would have also reduced theoretical value even further below par (99.5% in the case of STE and 98.98% in the case of Capitaland).
Secondly, volatility was very aggressive. With STE, implied volatility was well through historic levels.
Thirdly, a number of bankers criticized the "hazy" dividend protection language and said the 6.3% dividend yield assumption for STE would have been very off-putting given that it has traditionally paid a high special dividend and runs a 1.6% ordinary dividend. It is currently yielding 9.64%.
"Accounts would have hated it," said one observer. "Not only will they not be compensated on their long positions, but they'll also have to pay out the dividend as part of their stock borrow agreements. The dividend risk for arbitrage driven accounts is huge."
Others counter that while the bond floors are low, eight points over five years is not an unreasonable amount to pay. A similar argument is used to justify the high premium. "Compound growth of 9% per annum over five years is pretty realistic and SingTech certainly believes the stock will get there," says one participant.
But what did investors think?
One of the world's largest CB investors based in London, says his firm purchased both deals.
"Yes the deals should have been one to two points cheaper," he comments. "The credit spread was too aggressive and the conversion premium could have been a couple of percentage points lower. But we didn't think the theoretical value of the deals was that important because we like the market tone.
"We have a very favourable view of the underlying fundamentals of each stock and these deals are a good way to get access," he adds. "We particularly like Capitaland because we think it will benefit from a re-valuation in property prices similar to what's been happening in Hong Kong. We've asset-swapped half our position and put the rest in our long only portfolio."
Secondary market trading
While there is little consensus on just how far-fetched terms were, everyone agrees the deal ultimately came good because of the strong performance of the SGX on Friday. Property stocks were up 6% to 7% on the day and far from being shorted down, Capitaland actually closed the afternoon up at S$1.53.
Volume in the stock was huge, with 18 million shares traded over the course of the afternoon compared to a six-month full day average of five million. STE, by contrast, ended the day marginally down at S$1.89 and also saw large volume. A total of 7.9 million shares traded, compared to a 4.3 million six-month average.
Year-to-date Capitaland is up 39.64% and STE 13.94%. Most houses have a buy recommendation on both.
Some believe the completion of two large, Singapore-dollar denominated exchangeables in such a rapid timeframe is nothing short of remarkable. Combined, the two stand second to Mega Financial's convertible as the largest from Asia ex-Japan this year.
"This is a deal, which could have so easily gone either way, but was saved by the market and let's hope the issuer learns something from that," says one specialist. "Certainly none of the banks expected to make any money from the transaction, but their gamble did seem to pay off."