“It’s not that easy to shift a debt as large as Belgium’s,” says economist Kristin Vandenbergen at Bank Bruxelles Lambert (BBL), as she flicks through the pages of Belgium’s public debt statistics looking for the latest figures. Under the terms of the 1992 Maastricht Treaty, prospective entrants to European economic and monetary union (Emu) must meet, or make “satisfactory progress” towards meeting, a ratio of debt to GDP of 60%. They must also convince the European Commission, and ultimately the council of ministers and heads of state, that debt levels are sustainable.
Belgium’s boule de neige – as the country’s debt is known in Brussels – “is beginning to melt”, says Vandenbergen. But at 127% of GDP the ratio is the highest in Europe. When Wilfried Martens (the prime minister before Jean-Luc Dehaene), foreign affairs minister Mark Eyskens, and finance minister Philippe Maystadt signed the Maastricht Treaty in 1992, Belgium’s debt managers must have felt that they had an impossible task ahead.
By agreement with its European counterparts in 1993, the government is intent on reducing total debt by 10 GDP percentage points by the end of 1997. Hiliana Coessens, general auditor of the Belgian treasury and head of the treasury’s debt department, says confidently: “We will meet this reduction by the end of the year.”
Bankers in Brussels believe that in doing so the government will satisfy the council of ministers and heads of state that eventually decide which countries will take part in Emu. Says BBL’s Vandenbergen: “Belgium should qualify because of the way the debt has been restructured. Most of it is in Belgian francs and held in Belgium’s banking system. And also because the debt was so high when the criteria were set.”
In terms of debt-to-GDP reduction (between 1993 and the forecast for the end of 1997), Belgium ranks third after Ireland and Denmark. At the end of 1995 the European Commission issued a statement that Denmark and Ireland, both over the 60% debt-to-GDP ratio, would meet the criteria. Although this was challenged unsuccessfully by the Germans, it was taken as a precedent by the Belgians. At that time, Ireland’s debt stood at 86.6% of GDP, down from 95.2% of GDP in 1990.
“It’s toughest in the beginning,” says Baudin Richard, a senior adviser at the treasury. As the debt comes down, a corresponding reduction in interest rates will create a virtuous circle by reducing constraints on the budget. Assuming the country’s primary surplus is maintained at or near 6% of GDP, Maystadt’s treasury advisers believe that they will hit the 60% ratio by 2020. On current forecasts the primary surplus will reach 5.9% to 6% by the end of the year.
Says BBL’s Vandenbergen: “Dehaene should be confident about Emu. He has been granted special powers to meet the criteria.” If there is the slightest chance that Belgium could miss its Emu targets, Dehaene can make adjustments without having to put another budget through parliament. Under guidelines laid out in the government’s budget presented in October 1996, actions that could be implemented to reduce debt levels include privatization and gold sales, running a substantial primary surplus, and the netting of local authority accounts.
One problem is that Belgium’s public sector does not fit easily into Maastricht’s accounting criteria. Public money is held in several semi-autonomous organizations and regions which support the government. However, these funds are not accounted for in the country’s central accounts – much of this money is invested in government paper.
But, since the Maastricht criteria states that debt-to-GDP ratios are calculated using gross debt minus liquid assets, these funds are not included in the country’s debt statistics. This means that a sizeable chunk of the Belgian public sector is not accounted for under the Maastricht guidelines.
By consolidating the accounts for one day the government reduced its total gross debt, although, as the treasury points out, the net debt remained constant. By entering into a Bfr121 billion ($3.4 billion) swap with the semi-public organs (SPO) – which effectively compensated them for the loss of interest over the duration – the SPOs handed over their excess liquidity at the end of last year by depositing it in government securities. Says BBL’s Peter Vanden Haute: “It’s public money, but yes, it was an accounting trick of sorts. If the state didn’t have this particular system, public money would be consolidated anyway. Other countries without centralized accounts may do similar things to get around the rules.”
Equal footing
Says a treasury spokesman: “Countries with centralized accounts report the net position anyway. No-one has criticized this. Belgium has its provinces and different semi-autonomous bodies. If we had done nothing to centralize their accounts, Belgium would have been penalized. We asked for their cooperation to deposit excess liquidity in state securities at the end of the year. We then asked parliament whether we could do this and Eurostat has accepted the temporary centralization of accounts. It’s not a clever accounting trick, since we are simply trying to put ourselves on an equal footing with other countries.”
Although commentators say the Belgian government has exploited loopholes in Maastricht’s terminology, it is also rumoured that some of these measures were suggested by the European Commission.
Other reforms have had longer-lasting effects, not just on the debt figures, but also on Belgium’s financial markets in general. For example, before Belgium’s financial reforms in 1990 (known, like London’s, as The Big Bang), treasury certificates paid a higher interest rate than interbank deposits. “This immediately disappeared,” says Vanden Haute. “Bigger trading volumes followed Belgium’s Big Bang and Belgian debt quickly became more liquid. Interest payments came down. The treasury also appealed to the retail market through a series of new instruments.”
At the same time the Belgian treasury changed its policy towards the management of its debt portfolio by introducing new technology and ideas to reduce interest payments. One strategy it pursued in the late 1980s and early 1990s involved swap transactions, underwritten by the Belgian treasury. These were designed to reduce interest payments on Belgium’s debt. The swaps took the form of currency and arbitrage swaps, which had the effect of borrowing in hard currencies while simultaneously lending in weaker ones. According to the treasury, “the foreign exchange rate risk was, at the time, deemed smaller than the potential interest rate saving following the expectation”. The positions were taken in expectation of currency convergence, particularly after Maastricht was signed. Effectively the treasury accepted exchange rate exposure in return for a “locked-in” profit on the interest rate differential between the currencies.
But according to one Belgian banker, the trades turned sour when the whole European monetary system experienced problems in the summer of 1992. “There were huge potential losses on these so-called currency contracts. The transactions had been presented as a free lunch and the treasury simply didn’t have the systems or staff to price these contracts at that time,” says the banker.
The positions, staggered as knock-out options, mature between 1998 and 2002. Although the contracts paid a regular income on the interest differential between the currencies in the first year of their duration, possible losses will only be realized from next year onwards, when the government has to repurchase hard currency as part of the trades.
Many in Belgium attributed the trades to the former head of the treasury’s debt department, Jean Basecq, claiming he alone took a bet on European convergence. Others who knew Basecq say this is impossible.
Once the convergence trades turned sour, the treasury turned to the derivatives market again, this time to hedge its potential losses. The treasury sold foreign exchange puts on the US and Canadian dollar to finance put options on the soft currencies to which it was exposed. The treasury says that many of the puts, which all have knock-out clauses, have been knocked out with the rise in both the US and Canadian dollar.
Losses on these swap positions were revealed by the president of the Flemish Liberal party, Herman De Croo. When the government presented the 1997 budget to parliament, De Croo took the stand saying he did not accept the budget figures because the treasury was hiding capital losses. According to one source: “The Court de Compte [Court of Accounts] told Maystadt to publicize the potential losses in early 1996, but Maystadt didn’t say a word of this to parliament. He knew that the population would not accept the tax proposals if they knew the treasury was losing money.”
Cutting its losses
One Brussels-based banker, questioned by the Belgian press as to whether the government had been speculating, says his answer was “yes”. He says: “De Croo wanted to play the same trick again two weeks ago when the government presented its mini budget. He is trying to come up with a new scandal. The Court de Compte refused him information this time, but there is more. It is a parliamentary right for an MP to get information.”
According to Coessens, who succeeded as head of the treasury department’s debt portfolio when Basecq died last year, the treasury has been unwinding its position using the treasury’s new portfolio model. De Croo says about Bf15 billion has been recuperated from the earnings in the first phase of the contracts.
Vanden Haute argues that the losses, now greatly reduced, have to be taken in the context of the gains made over the period. “There is no doubt that debt levels have come down since the start of the reforms. Any losses have to be taken in the context of the gains that have been made,” he says.
Coessens says the impact of these second trades is minimal – just 0.2% of the country’s debt. But even these estimates are unlikely to correspond to the final net profit or loss on these trades. Although market players discuss the positions on these trades, no-one would comment on how much money the Belgian treasury stands to lose.
Certainly the treasury marks the positions to market each day using its new portfolio system. Says Louis de Montpellier, general adviser at the Belgian treasury: “We’ve isolated this portion of the portfolio to see whether it adds to the overall risk – and it doesn’t because of the size of the portfolio. The press and opposition focused on one particular trade. No client on earth would do this on a profitable account.”
The debate now revolves around public access to information that the treasury deems to be market-sensitive. “This was discovered piece by piece,” says De Croo. “Then at the given moment, I revealed the potential loss. My question is, is it normal for a government to speculate? Any normal bank has to communicate its positions to the national bank, but there is no law for the treasury to report to the national bank.”
Model strategy
Confidentiality is now the bone of contention. The treasury has suggested three levels of access to information. The first level would be fully public. The second level would be controlled by a sub-committee with access to ministerial guidelines for Belgium’s benchmark portfolio, including risk tolerance levels. This information would be available to parliament but would not be fully public. The third level of information – full access to treasury files including positions and counterparties – would be given to the Court de Compte. Parliamentarians would gain access to this information only through appeal.
Says the treasury spokesman: “We have been trying to formalize our strategy since Maystadt launched the reforms. They began in 1989, and we are now at a stage of defining and quantifying the risks in our portfolio.” Broadly speaking, the treasury has tried to reduce its exposure to short-term market movements. “One way to lower our debt costs is to decrease the interest revenues,” says another senior treasury manager. “We’ve built a model portfolio to measure exposure to foreign-exchange risks and interest rate movements. The model also takes into account duration and value.”
Following competition for the contract in August last year, the treasury began to work with UBS and Morgan Stanley to develop a benchmark portfolio to guide the country’s debt management. “Traditional asset liability models don’t really apply well to sovereigns. We took what we could from CAP [capital asset pricing] models and tried to integrate this with the constraints of monetary and budget policy,” says the treasury manager.
In a way it’s a marriage of financial and political constraints on the country’s debt management. The treasury manager points out that “market risk and debt management are two different things. For example, fixed-rate debt can be easily matched to predetermined budgetary cashflows, but from a market point of view this represents an unacceptable risk”.
In a study prepared during the second half of last year, treasury officials proposed a series of guidelines to Maystadt. Says the treasury manager: “We had several reviews, including constructive criticism from Maystadt.” The treasury and the central bank have calculated annual savings of Bf70 billion in managing the debt since the reforms started.
In the last few years the Belgian treasury has clawed back its foreign-currency exposure. Says the treasury spokesman: “Foreign currency is often a risk for sovereign debt managers.” But, as Coessens says: “The swap window in Belgian francs is pretty limited.”
Says Rupert Brindley, director of quantitative finance at UBS: “No government on earth, even if they wanted to, could convert all its debt. Nor would they want to buy back all their debt from the markets.”
UBS says it won the mandate to develop the benchmark portfolio system on the idea of robustness. It was an idea which Rob Baston, who once taught applied mathematics at Oxford University and is now managing director of quantitative finance at UBS, expresses as the dictum that: “Conclusions must be stable. Assumptions about the future economic environment are always varied.”
What the Belgian treasury particularly wants to avoid is assumptions about short-term market moves. Says Baston: “No-one wants to make too many choices where you are sticking your finger in the air to see which way the markets are going. But sovereigns often know the structure of their own economy better than anyone else. They are also as well placed to judge the long-term structures of the market. Sovereigns talk to each other all the time and many have a deep understanding of the markets.”
Before banks pitched for the contract to help design Belgium’s benchmark portfolio, treasury officials prepared a six-page document outlining basic guidelines. According to Baston the priority was to “minimize the cost of borrowing subject to budgetary and monetary constraints, and acceptable levels of risk”. Debt portfolio models look at the maturity spectrum, the ratio of fixed-rate to floating-rate debt and currency composition.
Nigel Fox, executive director of fixed income at Morgan Stanley, says: “We’ve been evaluating whether a lower-cost portfolio exists, given the constraints imposed, in relation to the existing portfolio, and attempting to define a road map to move between the two.” According to Brindley of UBS, the markets think Belgium will enter the next stage of Emu in the first wave. He says: “Just look at Belgian franc spreads over the Deutschmark, and compare them with other countries. The benchmark is predicated on Belgium entering the third stage.”
In the past, liability managers used imprecise tools to judge market risks. “Quadratic optimizers were pretty unstable since you never knew if your assumptions were correct. We are trying to quantify the stability of our conclusions against the assumptions which are made in looking at the markets,” says Brindley. Now the rocket scientists have created three-dimensional simulations which represent any number of underlying risk and cost constraints. As the markets, or assumptions, change the model rotates and changes shape in real time. Certain areas of the shape represent the mathematically most efficient portfolios, depending on the guidelines imposed on the model. Says Baston: “There are an infinite number of ways to reach any given point on this model, and each point could represent an ideal portfolio. It gives the client a possibility to fly their possible portfolios and assumptions in real time.”
Belgium is one of many countries looking at different ways in which to value its debt portfolio. The way an actuary or accountant values a country’s debt is different from the way a debt manager looks at his portfolio on a weekly basis. Says Baston: “The market value of countries’ debt is of secondary importance. But the change in market value, together with interest cost, is extremely important. How this changes throughout the year is as important as absolute level changes.”
Sovereigns are beginning to care more about changes in their day-to-day market value of their debt. “The cost of their debt is not just the cost of the instrument. It has to include the changes in market value of the portfolio as a whole,” says Brindley.
With political pressures to meet the Maastricht criteria, Belgium’s treasury has its work cut out. But, as BBL’s Vandenbergen points out: “I don’t think Belgium is unique in this respect.”
