Back to the brink

By Danielle Robinson

Citibank vice-chairman Bill Rhodes is proud of his Latin American debt restructuring achievements. He is said to keep letters expressing gratitude and respect for his efforts, framed and hanging on his Manhattan office wall. Brazil is the blight amid the haul of trophies.

It's been a tough year for Rhodes and his bank negotiators. Trying to conclude the Argentine Brady plan has been hard enough without political upheaval forcing their attention on Brazil every other day. Since July, trading on Brazilian MYDFA (multi­-year deposit facility agreement) paper, the country's benchmark bank debt, has swung wildly, finally crashing through 30% in mid-August as political problems deepened. US banks took a direct hit when fears that the Brady plan would not go ahead contributed to a 1% dip in US bank stock prices in a single day's trading last month.

At first the stories of president Fernando Collor de Mello's misdeeds amused Brazilians. Businessmen would giggle as they recounted how Collor's younger brother spilled the beans on the president's penchant for life in the fast lane. The smiles quickly faded when allegations of widespread corruption, influence-peddling and news of Collor's attempts to buy political survival surfaced. The banks, owed $44 billion in medium- and long-­term public-sector debt plus more than $7 billion in interest arrears, never saw the funny side. They have refused to allow one man's political ego to destroy two years of exhausting talks on the most complicated Brady plan to date.

As soon as it became clear that Collor's misdemeanours were not confined to a few wild parties, bankers and Brazilian debt negotiators moved quickly to wrap up the deal and signed the in-principle agreement on July 9. "We started working on this two years ago but we did not really get anywhere until [June]," says a banker involved in the Brady talks. "It's no coincidence that Brazil decided to sign this at the same time as allegations of political corruption surfaced."

"You certainly want at least to get what's owed to you today," adds the chief debt negotiator at a US money-centre bank, referring to the fact that the signing of the in-principle agreement triggered an agreement to securitize about $7 billion worth of 1989-90 unpaid interest, including interest on unpaid interest.

Bank negotiators say it was the Brazilian side that made the decisive move that enabled the signing to go ahead. "When [Brazilian economics minister] Marcilio [Marques Moreira] came to the States in June [treasury secretary Nicholas] Brady and [treasury under-secretary David] Mulford had a conference with all of the Latin American finance ministers," explains the banker. "Marcilio came just before that to New York, and he went around and talked to everybody. Then he went down to Washington and talked to his fellow ministers as well as Mulford and Brady. Basically he checked with all the sources here and they all said, 'Hey, you aren't getting a better deal, and you had better grab it now because even this might disappear'."

Both sides have been anxious to get the deal done, as evidenced by an overly­ enthusiastic joint press statement on July 9, extolling how the in-principle agreement marked the end of the debt crisis. Moreira described the deal as "closing a long chapter overshadowed by foreign debt problems and opening a new chapter full of opportunities".

Rhodes, meanwhile, seems to have nurtured dreams of going to this year's IMF meeting with two more Brady deals to his credit – Argentina and Brazil. A joint press release had Rhodes stating that the in-principle agreement signalled "the final close-out of the debt crisis among the major economies of Latin America".

Unlike any other Brady scheme, Brazil's involves a phasing-in of collateral for defeased options. The phase-in procedure – based on Brazil's arguments that it cannot come up with more than $3.2 billion to buy zero-coupon treasuries as collateral – is still a major bugbear for some banks.

"We spent about six months talking about it," complains one senior US banker involved in the negotiations. "With more than $20 billion in international reserves, you would think they would be willing to pay more for enhancements, but they feel this is a group effort. To me it is a case of pure political willingness. They are sitting on this money and are not willing to pay it all up."

As the crisis deepened, the bank negotiators moved quickly to keep the economic team's focus on the Brady deal and the need to go to the senate for approval this month. "As I look at it Brazil is going along two tracks," says one senior US banker involved in the negotiations. "The first is still working relatively OK – meaning the documentation. The more fundamental track is the one we are worried about and which no one has the answer to. We need to know if Brazil can get the tax and fiscal reforms necessary to get back on track with the IMF, because if there is no IMF, there really is no Brady deal. In the meantime we continue to pursue the first track. The idea is that if things sort out soon we don't lose a beat. If they don't, then it becomes dangerous."

Throughout the crisis, the banks appeared to play an important role in ensuring that Moreira got the pep talks necessary to keep him in the ring fighting for Brazil and the Brady deal. This is where Rhodes is such an asset. He is renowned for his "personal touch": his ability to make contacts and act as friend and confidant to political and business leaders. Moreira is the latest to come into that fold.

When rumours spread in early August that Moreira and other members of the economic team might resign, Rhodes is understood to have had a long conversation with the Brazilian and got as strong an assurance as possible that the minister would stay – at least to see the Brady term sheet through the senate. Rumours of resignation subsided after an emergency cabinet meeting in August, when ministers decided to stay on while indicating the decision should not be construed as support for Collor.

Senate approval will bind Brazilian politicians to the Brady plan. As a US banker explains: "We think that once the Brazil deal is negotiated and there is a political commitment, there is a very strong commitment to honour them, because... no country wants to be the first one to blow up a Brady deal."

Such political commitment has always been a major point in the Brady negotiations. Pedro Malan, Brazil's chief bank debt negotiator, notes that "the obligation of servicing the debt will be with the Treasury, not with the central bank", thereby avoiding the risk of the country printing more money to pay debt. "We want to have a lean, clean central bank, so the Bradys will be issued by the federal republic of Brazil, which has the implication that the servicing of this debt will have to be budgeted and... fiscal targets being met are implicit in the Brady agreement."

The urgency behind the banks' push for the Brazilian Brady owes more to timing than to the banks' desire to show immediate benefits on their balance-sheets. A major difference between this and other earlier exchanges of Latin American debt is that US banks have written down their Brazilian medium-term debt heavily, to the point where the maximum carrying value is about 40% of the principal amount of claims. Some analysts believe the Brazilian exchange potentially could release between $500 million and $1 billion of loan loss reserves in the US. But conservative accounting strategies adopted by the banks mean they are not expected to record any bottom line improvements in the near term.

"The reason why they are pushing ahead with this, despite all the problems, is because they don't have a whole lot of choice," says an LDC debt analyst at Salomon Brothers in New York. "They have to do something and get whatever they can."

Brown Brothers Harriman bank analyst Raphael Soipher adds: "It's also a case of the banks wanting to strike while the iron is, well maybe not while it's hot, but before it's cooled off too much. As long as there is a willingness in Brazil to do a Brady, you have to get it done while the stars are in the right house."

A senior source admits that there is a danger of losing momentum if the Brazilians are not locked into a Brady now – and getting the Argentine deal done is crucial to the success of the Brazilian process, he argues. "As we look at it, getting Argentina is a very important step in also getting a Brady for Brazil. If they see Argentina moving ahead, it's an example to them."

It seems the plan was to make Brazil squirm by announcing the Argentine deal at the IMF meeting – leaving Brazil as the only major with economy in Latin America out of step with the international banking community. But delays in the Argentine selection process have made it almost impossible to reach that deadline, according to a US banker.

"What is going on in Latin America is a reversal of the perception of what was going on six months ago," he says. "Most of the Latin American countries have hit their highs. Earlier in the year the most interesting place was Latin America, now you have seen stockmarkets there lose as much as 50% (in the case of Argentina). Investors are becoming very leery and as a result these deals are now very important."

Once the Brazilian term sheet is approved by the senate, the focus will then be on the Brady bond enhancements issue. Some commercial banks are so anxious for a Brady that there has been serious consideration of doing a deal even if the IMF agreement – and therefore guarantees for $1.6 billion in enhancements from the fund, the Inter-American Development Bank (IADB) and World Bank – falls apart.

"The enhancements and the IMF are today considered one and the same issue, but that is not necessarily the case," says a senior bank negotiator. "Under the present arrangement if you don't have the IMF, you don't have the enhancements – that may be right, but it also may be wrong. Brazil is sitting on $20 billion-plus in foreign reserves. They could use these reserves for collateral. I am not saying they would, but they could if they really wanted to get this deal done. But of course there could be some banks that could say, ‘We don't want to do it unless we know they are in compliance with an economic reform programme as endorsed by the IMF'."

Pedro Malan is fighting to keep the issue of the country's foreign reserves off the negotiating table. Under the current plan, the exchange of debt for bonds will coincide with Brazil producing at least $3.2 billion in collateral – about $1.6 billion from the IMF, IADB, World Bank and new money bonds, and a further $1.6 billion from Brazil's international reserves.

“We have a formula here where we match dollar-for-dollar contributions of other sources with our international reserves,” says Malan, arguing that this shows Brazil has made “substantial” contributions from its reserves.

“I think they have a lot of money,” complains one banker involved in the Brady negotiations. "$20 billion goes a long way and it was not that they didn't have enough money for collateral, it was their reluctance to use the reserves."

Malan refuses to discuss whether Brazil will use the reserves should the IMF deal fall through. "Let's cross that bridge when we come to it," he says. "Basically we feel confident money from the IMF, World Bank, IADB and new money will come to this deal – it is in our interest that it does and it is also in the interest of the banks as well."

The prospect of doing a Brady without an economic reform plan ensuring that an IMF agreement is in place may spell disaster in the LDC debt market. "I would not like to see them do that," says a senior LDC debt analyst. "Brady plans are very good for the market. They're characterised as the last restructuring. I think if they put a whole lot of Brady bonds without an IMF economic reform plan in place, there is a chance they could ruin the market for the other Bradys." The final restructuring characteristic of Bradys has improved the bonds' attractiveness among a wider investor base, he argues. "I'm not saying that if the Brazilian Brady falls apart then Mexican Brady bonds will go down as well. What I'm saying is that these bonds have just started to get a wider distribution and we want to see that stay and expand. It's a lot better than the banks holding on to the stuff and passing it around to one another in the secondary markets."

The amount of enhancements needed will depend on what options the banks choose, and how many include the new­ money option in their selection. The new ­money bond has the highest net present value with an interest rate payment set at Libor plus ⅞. It also has a short tenor of 15 years with a seven-year grace period. For every $1. 00 of debt tendered under, this option $0.1818 must be provided in new money. The underlying eligible debt will be exchanged for a debt conversion bond with an 18-year tenor, a ten-year grace and an interest rate of Libor plus ⅞. Only banks that are bullish about Brazil's future, or which have long-term commitments to the country, are likely to go for this option. Such creditors are scarce.

"It's a more attractive instrument at net present value, and the conversion is at par," says Malan. "We have had indications there will be some banks choosing this as an option." On the creditors' side, a top negotiator admits that some banks have refused to increase their exposure to Brazil and so far only Citibank has publicly stated it is seriously considering the new-money option, although other bankers say it is doing so reluctantly.

For accounting and regulatory reasons, there will be a strong inclination towards defeased instruments, particularly the discount and par bonds. The collateralized discount bond involves a 35% cut in eligible debt tendered for this option, but does not provide interest relief for Brazil. The interest rate is Libor plus 13/16 and the tenor 30 years with bullet repayment. The principal collateral comprises 30-year zero­coupon US treasury obligations and a 12-month rolling interest guarantee.

The par bond is a 30-year bullet maturity bond paying annually 4%, 4.25%, 5%, 5.25%, 5.5% and 5.75% in the early years and 6% thereafter. Collateral covering principal and 12 months' interest is the same as the discount bond.

Current low interest rates favour the par option, but if at the selection date the availability of enhancements is still an issue, analysts believe banks will have a better chance of obtaining more collateral if they choose the discount. The phase-in system involves Brazil providing $3.2 billion up front on the exchange date, and then phasing in the balance over two years. Debt holders who select the par or discount bonds can choose between two options for allocating the available collateral at the exchange date.

Under option A, holders of discount bonds are assured of receiving the entire amount of principal collateralized up-front, with any remaining funds applied to its interest collateral. If there is not enough for interest collateral, the discount bond holders will receive such enhancements in semi-annual instalments over two years. Holders of par bonds under option A will receive only as many par bonds with full principal collateral up-front as the available funds allow, with the remainder in "phase-in bonds". These phase-in bonds will be exchanged for par bonds as additional principal collateral is delivered during the two-year phase-in period. Interest collateral for option A par bonds will also be delivered during the phase-in period.

Option B provides banks with a "cleaner hand", in that both the par and discount options bonds with full principal and interest collateral will be issued to the limit of the available funds and the remainder of the debt will be exchanged for phase-in bonds. Just how many fully-­collateralized discount and par bonds a bank can have at exchange date will depend on the amount of enhancements available, and how many choose option A, which gives collateral preference to the discount. Again, as in option A, collateral not paid at the outset will be phased in over two years and phase-in bonds will be exchanged for par and discount bonds as principal collateral is delivered.

The phase-in bonds under both options will accrue interest at the rates of the par or discount bonds. If Brazil misses any phased-in collateral instalment, the phase-­in bond will be converted to a bond paying Libor plus 14/16, with a final maturity of ten years and two and a half years grace.

The two phase-in options evolved because some banks want to be fully defeased at the outset in terms of both interest and principal, while others want to be covered as much as possible on principal risk.

Says one US bank negotiator: "The par and the discount bonds are for the risk­averse players and are particularly attractive to US banks who want to eventually improve their balance-sheet and free up reserves for other problem areas."

On a net present-value basis, the most attractive is the new-money, followed by perhaps the restructuring option – the so­called front-loaded interest reduction with capitalization bond (or C bond), the front-loaded interest reduction bond (the Flirb), and then the par and discount bonds.

What mix a bank chooses "depends how much Brazilian risk you want to take", says one bank negotiator. "Obviously the more the risk the higher the net present value."

At this stage, it seems that the restructuring option will be taken up mostly by German banks, some of which voiced a particular desire for the inclusion of such an option. Says one banker: "Some of the German banks come from the philosophical point that Brazil doesn't need or deserve a Brady – that is, that it doesn't deserve forgiveness. The Germans think Brazil just needs more time, not forgiveness. As a result they don't want to take any hits on Brazil and that's the point of the restructuring option. It gives Brazil more time to pay, and it gives them reduced rates in the near term, but they don't forgive any of the interest due because the outstanding is eventually capitalized."

Under the restructuring option, interest rates between years one and two are at 4%, three and four at 4.5%, five and six at 5% and seven to 20 at Libor plus 13/16. The difference between the six-month Libor rate plus 13/16 and the interest rates in years one through to six will be capitalized. The tenor is 20 years with a ten-year grace and increasing principal instalments of 1% of principal in instalments one to four, 4% in instalments five to eight, 5% in instalments nine to 12, 6% in instalment 13 and 6.75% in instalments 14 to 21.

Some US banks consider the C bond as an attractive option. "The C bond has some advantages," says one bank negotiator in New York. "You don't forgive anything, you get a large cash interest payment capitalized, it has a shorter grace, it is a bearer instrument, and once the debt load comes down this undefeased instrument has more upside potential than the par."

The C bond offers Brazil temporary interest relief in the near term with interest rates ranging from 4% in years one and two, 4.5% in three and four, 5% in five and six, and 8% in years seven to 20. The difference between 8% and the interest rates in years one through six will be capitalized; the tenor is 20 years and the grace ten years.

At this stage the Flirb looks like the orphan. "It's everyone's third or fourth choice," says one banker. "It's in the middle and it has some collateral – 12 months rolling interest guarantees, but you are giving some forgiveness for that on the interest rates. It does have some elements to it – a shorter term and grace – but it doesn't have the greater defeasance advantages of the par or discount or the large upside potential of the undefeased instruments."

The Flirb has annual fixed interest rates of 4% in years one and two, 4.5% in three and four, 5% in five and six, and floating rate at Libor plus 13/16 in years seven to 15. It has a tenor of 15 years with a nine-year grace and interest collateral of 12-month rolling interest guarantees for six years, after which unused collateral returns to Brazil.

A $7 billion debt market is expected to be created in the next few months when Brazilian interest due and unpaid (IDU) bonds are scheduled to begin trading. The bonds securitize money that Brazil owes to the banks in 1989-90 interest arrears, documentation for which was on September 10.

The issue of the bonds was triggered by the July 9 signing of the in-principle Brazilian Brady plan agreement and provides some compensation for bank creditors while they wait to see if a debt deal is possible for Brazil. Added to these bonds is a 1991-92 past-due interest agreement, also announced on July 9, in which Brazil has agreed to increase its current cash payments from 30% of interest due in this period to 50% if the country's senate approves the term sheet.

The two schemes will provide "significant cash for the banks", say analysts. It is calculated, for instance, that it could make a $40 million to $50 million difference to Chase this year. According to public documents, Citibank will be the biggest beneficiary, with about $400 million of interest bonds.

The IDU bond agreement goes back to April 18, 1991, when Brazil and the bank advisory committee reached an accord covering the interest due and unpaid on its foreign debt from July 1989 to December 1990. The agreement required Brazil to pay back 25% of that in cash – about $2 billion in instalments.

Traders and Brazil's creditors hope for active trading and a liquid market. Some US banks say they do not expect to sell­ off large quantities of the bonds to record profits in 1992. While some non-­US banks may seek immediate bottom-­line results, American banks have indicated that they are more likely to hold a fair amount in trading portfolios, thereby helping ensure a strong market.

Many traders recommend investors buy the IDU bonds, on a "cash" or a "when-and-if issued" basis because they have a relatively short average life of about 6.3 years compared to other sovereign restructured-debt bonds in the LDC market. With $7 billion of paper available, liquidity should be ensured.