Interest Rate Risk in the Icelandic Index-linked Bond Market. Effects of the Pension System’s Actuarial Valuation

• Main Author: Húni Jóhannesson 1982-
• Other Authors: Háskóli Íslands
• Format: Master Thesis
• Language: English
• Published: 2016
• Subjects: Fjármál fyrirtækja; Lífeyrissjóðir; Verðbréfamarkaðir; Fjárfestingar; Vextir; Viðskiptafræði; Iceland
• Online Access: http://hdl.handle.net/1946/26164

This thesis attempts to evaluate how the rules regarding the actuarial valuation of the pension funds in Iceland affect interest rate risk on the index-linked bond market. It was discovered that not only does actuarial valuation affect the management of interest rate risk in the index-linked bond market, but it also causes abnormal price formation. A few factors are discussed regarding how and why the pension funds might influence the index-linked market. These factors include the size of the pension system, especially on the bond market where its share of index-linked HFF bonds is steadily increasing, and the possibility of how the limiting effect of capital controls on investment options might play a role. It is probable that these factors contribute to the influence of actuarial valuation on the index-linked market. The valuation seems to create an unusual incentive for the pension funds, with the 3.5% discount rate rule at its center as an investment benchmark for the pension funds, with the repercussions of failing to meet an ever-looming actuarial valuation standard. A hypothesis was formulated, stating how actuarial gains, or losses, should be the same on the index-linked market. This was then further developed into a theoretical model, which established the importance of the 3.5% discount rate rule by implying a constant forward rate of 3.5%. By placing the theoretical model in the context of interest rate risk and duration immunization strategy, it was discovered that there is no such risk, as any interest rate shock would affect each bond equally. The theoretical model was tested in four different ways using market data, over several time intervals, with and without the HFF 14 bond. This was done against two null-hypotheses: that the market moves according to parallel interest rate movements, and that the market moves according to the proposed model. It was discovered that the shape of the yield curve conforms to a large extent to the model’s prediction, and a test of variance of the price changes ...