No abstract

This study investigates the consequences of dynamics in the term structure of Dutch interest rates for the accurateness of value-at-risk models. Therefore, value-at-risk measures are calculated using both historical simulation, variance±covariance and Monte Carlo simulation methods. For a ten days holding period, the best results were obtained for a combined variance±covariance Monte Carlo method using a term structure model with a normal distribution and GARCH speci®cation. Term structure models with a t-distribution or with cointegration performed much worse. Ó 2000 Elsevier Science B.V. All rights reserved.JEL classi®cation: E43; G28

No abstract

In an ageing society, defined benefit (DB) pension plans are increasingly difficult to manage by means of contribution policy only, as the contribution base is likely to shrink relative to total pension provisions. This development, together with an increased emphasis on market valuation in regulatory and accounting rules, has led to a switch of DB plans to defined contribution plans throughout the world. In the Netherlands, a different solution has been sought. The typical pension contract nowadays comprises an average earnings DB pension in which only nominal benefits are guaranteed, but with the intention to provide wage or price indexation. In the new supervisory regime, the guaranteed pension rights, based on market valuation, are subject to risk-based solvency requirements. Provisioning is not required for conditional pension rights, although contributions have to be consistent with the indexation ambition. In this paper, we analyse to what extent indexation is indeed likely, given various indexation and contribution policies. Simulations show that voluntary provisioning for indexation is to be recommended. Fully guaranteed indexation is virtually unaffordable under the new supervisory regime, because the real discount rate is generally both very low and highly volatile.

On the asymptotic distribution of impulse response functions with long run restrictions In this paper, the asymptotic distribution of the parameters of the moving average representation of structural VAR models with long run restrictions is derived. Moreover, it is shown that the structural model can easily be estimated in a two step procedure, where the reduced form model parameters serve as input for the structural parameters. The proposed model structure is very general, including the common trends model, in the sense that all identification schemes that can be expressed as linear zero restrictions on the contemporaneous or long-run impact of the structural shocks are allowed for. The asymptotic distribution of the moving average parameters can be used to determine asymptotic confidence intervals of impulse response functions or to perform tests of hypotheses concerning forecast error variance decompositions. The relevance of the correction for long run restrictions to the asymptotic distribution is shown in an example on German interest rates.