For a country with a very small net borrowing requirement, Mexico has been a huge – and hugely important – issuer of global bonds. In the past year alone it has issued some $8 billion of bonds, making it easily the largest emerging-market issuer. Since it didn’t need most of the cash, it generally put it to use in liability management exercises; in doing so, Mexico has become the first country to all but eradicate its Brady bonds. Mexico also got its third investment-grade credit rating in September. But the biggest achievement came at the end of February, when the sovereign stunned the market with a 12-year bond with collective action clauses (CACs). The lead managers were JPMorgan and Goldman Sachs.
The bond was a true watershed moment in emerging-market debt. Until that day, the G7 and the US Treasury had been agitating for years for a country – any country – to bite the bullet and insert CACs in its bonds. No-one would do it: because the purpose of CACs is to make debt easier to restructure in a crisis situation, both bankers and issuers were convinced that being the first country to use them would signal an increased willingness to default, and would cost the issuer a lot of money.
Of all the world’s emerging markets, it was Mexico that stepped up to the plate. It was not because of any particular susceptibility to the US Treasury’s exhortations, but simply that Mexico was convinced of its own leadership position. That position gave the country the self-confidence to go ahead with the unprecedented deal, and also helped to convince its debt team that there were advantages as well as disadvantages in going first: Mexico set the terms of the deal itself, and didn’t need to use anybody else’s clauses. (South Africa, when it followed suit in May, simply borrowed Mexico’s carefully thought out clauses and didn’t change a thing.)
Agustin Carstens, Mexico’s under-secretary of finance and public credit, is in charge of the borrowing programme. He has nothing but good things to say about CACs, calling them “an important step forward” that “allowed us to strengthen the legal provisions in our documents”.
Even before February, however, Mexico was demonstrating great agility in the capital markets. It completed its entire financing requirement for 2003 on January 9, with a $2 billion 10-year global bond from JPMorgan, Goldman and UBS Warburg. Mexico always issues a new benchmark 10-year global bond at the beginning of the year, but this one was doubled in size from the originally announced $1 billion on overwhelming demand: the order book was more than $5 billion. The bond priced at 246 basis points over treasuries, inside the price talk of 250bp.
And back in September, a 20-year $1.75 billion bond from JPMorgan and CSFB was just as successful, albeit in much tougher markets. This deal, too, was upsized, from $1.5 billion, and the new 2022s came in at 353bp over treasuries, some 5bp inside the benchmark 2031 bond, and fully 35bp below the high-coupon 2026s. And all of this happened on a day when the most-watched instrument in emerging markets, the Brazilian C bond, plunged 2.5 points.
Mexico used $1.3 billion of the proceeds to buy back Bradys. As Carstens points out: “For the last two and a half years we have concentrated on liability management transactions.” They were roughly a quarter of the total Bradys then outstanding; but Mexico had much bigger ambitions.
On April 8, it announced yet another landmark deal: it would buy back all its remaining dollar Brady bonds, and become the first country to exit the Brady plan. Through amortizations, buy-backs and swaps, Mexico has now essentially extinguished all of the more than $30 billion in Bradys that it originally issued in 1989-90.
Bradys were always something of a mark of former default for Mexico, and the spread between them and global bonds never went away. Brady bonds – restructured defaulted bank loans – have complex structures and are often collateralized by zero-coupon treasury bonds. Investors simply prefer pure country risk, without any taint of default, and with a plain-vanilla bullet structure. So countries such as Mexico and Brazil have run active liability management programmes, swapping their Bradys for new global bonds, and realizing large net present-value savings.
There were always some investors that the countries couldn’t persuade to swap out of their holdings, however, and so in the end Mexico decided to call its Par bonds for cash. All that’s left now is a few small Italian lira- and Deutschmark-denominated Bradys that never trade; those, too, will almost certainly be called before long, bought back with the proceeds of a new euro-denominated global bond.
A strategic loan Mexico needed $3.84 billion to pay for its outstanding Bradys. Half of that it could get back more or less immediately by cashing in the collateral, but it wanted the option to hold on to those treasury bonds if it wanted to in order to sell them at a time of its own choosing. So the country borrowed $2 billion from JPMorgan and Barclays: a one-year loan at an undisclosed (but surely wafer-thin) rate. It was pure country risk for the banks: they have no recourse to the treasury bond collateral. But in return for the favour they got the mandate for two new global bonds issued to pay for the other half of the deal.
The 5.5-year bonds were hugely popular: between a strong local bid and enormous demand from US crossover investors, the leads quickly raised a $4.2 billion book, and sold $1.5 billion of bonds at 190bp over treasuries without any difficulty. Mexico also issued $1 billion in new 30-year bonds at 270bp over, which were less successful but still attracted $2 billion in orders.
In general, the past year has been fantastic for Mexico, which has lost no time in exploiting a historically unprecedented combination of low US treasury yields and low Mexican sovereign spreads. “We have been taking advantage of windows of opportunity in the market,” says Carstens, who says he’s excited by the “very low levels of interest rates and the strength of the balance of payments in Mexico”.
Mexico is now going to have to start building a whole new yield curve, this one with CACs. (The fact that all of Mexico’s future bond issuance will have the clauses means that the country can’t reopen any of its old bonds.) The present curve is very well defined, and very smooth, and it will be interesting to see whether there’s any noticeable difference in price between the CAC bonds and the old bonds, once there are enough of each to be able to make an empirically strong comparison. Either way, says Carstens, “the bottom line is that we will be substituting our maturities with new ones, with CACs”.
It won’t be alone. In Mexico’s wake, every sovereign that has come to the dollar market has followed suit, including similar clauses. So far, that’s only Brazil and Uruguay, although South Africa used Mexico’s CACs in a euro-denominated issue. But there’s no doubt that in the wake of that one Mexican bond, pretty much all future sovereign bond documentation has been fundamentally changed.
