Views from the bridge

Risk looks at the different approaches taken to risk management by thegovernments of Sweden and Denmark.

Two of the Nordic region’s largest economies, Sweden and Denmark, may now be linked by the Øresund Bridge between Malmö and Copenhagen, but their outlook on many financial matters remains far from close.

Although both countries are members of the European Union, have held referendums on joining the euro (in Sweden’s case twice) and have rejected the option, the risk management methods of Sweden’s national debt office (the Riksgäldskontoret) and Danmarks Nationalbank in Denmark are quite different. The policies of Sweden’s national debt office (NDO) – specifically its large use of foreign currency swaps for funding purposes – appear racy compared with the conservative approach of Denmark’s Nationalbank.

As one veteran banker in the region – admittedly a Swede – notes: “The Swedish national debt office is fairly adventurous. It still has an active mandate. There is nothing adventurous about the Danes. They have strict rules and follow them to the letter.”

That is not to say that both are not admired by bankers throughout the region. Compliments about the “professionalism”, “dedication” and “stability” of both institutions are almost interchangeable.

In Sweden, the NDO sets out how it intends to fund and manage the national debt burden. The government tends to issue fairly specific strategic guidelines after propositions from the NDO. For instance, the NDO was instructed in 2004 to amortise the foreign currency debt by Skr25 billion (€2.7 billion), achieve a duration of 2.7 years and increase the share of government debt borrowed through inflation-linked bonds.

Charlotte Lundberg, head of debt management at the NDO in Stockholm, says the main objective of her strategy is to raise cost-effective funding while taking into account the risk that involves. “We look at our funding and debt as a portfolio, which we try to optimise,” she says. “We can choose to deviate from the strategic guidelines if there are market trends that we could benefit from or, for example, if there is a strong case for strengthening the Swedish krona.”

The NDO is one of the largest end-users of derivatives in Sweden and uses interest rate futures, interest rate options, interest rate swaps, swap options, forward foreign exchange contracts and currency options. A benchmark is applied to the foreign currency debt of Skr300 billion (€33 billion), which reflects Sweden’s long-term requirements – for example it is currently 65% in euros – and minimises risk and volatility. Two-thirds of the currency debt is then actively managed using derivatives, which are positions based on that passive portfolio.

The size of the NDO’s portfolio limits what it can achieve in the domestic market. “It would be very difficult for us to lengthen or shorten duration according to our view on Swedish interest rates as we are such a major player,” she says. There is a self-imposed cap of Skr30 billion (€3.3 billion) a year for swap use to avoid distorting the market – most of which is used for currency funding. “It’s not an ideal situation but we, at certain times, can’t do everything we want because the Swedish market is limited in size.”

However, in the international market, swaps, futures and options are commonly used by the NDO to take interest and currency positions. One example of this policy was the decision to move a greater percentage of outstanding debt to dollars and out of euros in the expectation that the dollar would weaken and effectively decrease the debt during the year. Although the passive portfolio remained the same, the NDO sold dollars for euros using forwards. That position was closed in 2003 and proved highly successful from a cost reduction perspective, although the NDO declines to provide figures.

The NDO has a medium-term horizon for its active portfolio and takes positions with a horizon of up to 12 months. Its value-at-risk (VAR) mandate of Skr220 million (€24.3 million) is supplemented by traditional absolute position limits. For example, the position in a single currency may not be larger than the equivalent of 6% of the managed debt amount. Similarly, interest rate positions may not affect the total duration of the managed debt amount by more than 0.4 years in each individual currency and 0.6 years overall.

A team of just five people is largely responsible for what positions are taken but six portfolios are outsourced to external managers – though their risk mandate is a small portion of the NDO’s total. “They get a small proportion of our overall risk mandate of Skr220 million in value-at-risk,” says Lundberg. Each manager has a mandate of Skr6 million (€660,000) in VAR. The rationale for the external mandates is to provide benchmarking for internal performance, hopefully to achieve a better return, and to aid the transfer of knowledge to staff at the NDO.

In Denmark, Danmarks Nationalbank – which runs government debt issuance and the balance sheet of the bank – has recently taken the unusual step of publishing a book detailing its risk management strategy (see Risk February 2004, page 12). It has two main business lines: its transactions as a monetary authority (essentially buying and selling currency and making currency interventions); and add-on risk (the taking of interest rate risk).

“We have investigated whether we can take more interest rate or currency risk,” says Ib Hansen, head of risk management at Nationalbank. “With currency risk we concluded that we only want to take the risk associated with our role as the monetary authority of Denmark. It is difficult to find convincing arguments from a revenue perspective to assume, for example, more dollar risk than we really need to.”

Denmark is currently the only member of the fixed exchange rate policy ERM II, and has an agreement with the European Central Bank to limit currency fluctuation to just 2.25% above or below a central rate of Dkr746.038 per €100. Given this position, euros – as could be expected – make up much of its reserves, with holdings also in dollars, sterling, Swedish kronor and a small gold holding. “The gold holding for us contains two types of risk – dollar versus gold and dollar versus krone; so we put that risk into dollars as well,” explains Hansen.

All of these risks are put into a rolling forward book, so that the Nationalbank is left with euro exposure and a small dollar and gold exposure. The rationale for this is that the standard deviation of the Danish krone against the dollar is about 10% whereas against the euro it is almost nothing. Even a slight increase in Nationalbank’s dollar exposure would dramatically increase VAR, according to Hansen.

On interest rates, Nationalbank maintains exposure primarily in Danish kroner, dollars and Swedish kronor, and all exposure is considered in krone terms – what the impact on krone reserves will be given a 1% change in interest rates. In 2003 the exposure was Dkr2.4 billion (€322 million). The board of governors meets every quarter to set so-called neutral limits and information on exposures is published every year in retrospect. Hansen says that the main goal of interest rate risk management is to change the krone duration of currency reserves, although “in the long term we presume that taking interest rate risk gives additional yield so we maintain a long position”.

The neutral limits are used as the basis for fluctuation bands of 20% for the maximum deviation of krone duration, which are altered if interest rate expectations change significantly. These are then used to set tactical margins, which reflect changes in interest rates expected in the near future by the Nationalbank.

For instance, if interest rates are expected to fall, the tactical margins are set at the upper end of the neutral fluctuation band – the krone duration of the portfolio is increased to achieve a higher capital gain in the event of falling interest rates. Active positions, based on views on short-term interest rate changes, are only taken within the tactical margins. And as Hansen points out: “In order to assess our ability to absorb losses and the impact on solvency due to interest rate changes we use VAR, stress tests and Monte Carlo simulations.”