Sammendrag
This thesis, though twofold, focuses on the financial accelerator, procyclical credit and the counter cyclicality of banks interest margins. All analysis and data transformations were undertaken in either Oxmetrics (PcGive) or Excel. The first part of the thesis looks at the financial accelerator mechanism in the Norwegian economy through a reconstruction of the work of Hammersland and Jacobsen (2008). In this section, I estimate a Structural Vector Error Correction Model (SVECM) of the simultaneous relationships between asset prices, production and credit, using the original methodology and sample period of Hammersland and Jacobsen s paper. The methodology used differs from other, more mainstream, methods in that it identifies the structural equations of the model through so-called structural dummies, influencing only one of the equations each. In following this method, I found it to be somewhat unstable and the battle to get to the final structure was hard-fought. In the end, the evidence from this reconstruction gives reason for second-guessing the results of Hammersland and Jacobsen. Nonetheless, and even with less significant results, it is hard to discard the possibility of the existence of a financial accelerator working in the Norwegian economy. Though these results are interesting in themselves, this part of the thesis should also serve as a remark on the lack of reproduction and validation of economic work. In the second part of the thesis, the focus is shifted over to one of the key mechanism in the financial accelerator framework of Bernanke, Gilchrist and Gertler (1999), namely banks interest margins. Following the theory in their paper, the interest margin of banks should show sign of countercyclical movements, enhancing the reinforcing mechanism of the financial accelerator. I start this section of by operationalizing my own measure of the interest margin. In this, I use the average lending rate of banks and develop, through a weighted average cost of capital approach, a measure for banks funding cost, including the costs of deposits, wholesale funding and capital. The interest margin is defined as the difference between these measures. I later find that the development of this margin as well as the development in the net interest margins, defined as net interest income over total interest earning assets, shows signs of non-stationary negative movements from the beginning of the 1990 s until today. This however, in light of historical events and general developments in the banking market, not surprising. Specifically, I point at higher quality in credit analyses, capital regulations, more competition and higher productivity, as possible explanations. Having thoroughly discussed and analyzed the developments in the interest margin, I turn focus to the cyclical behavior of the measure. First, I analyze at the correlation between the interest margin and the output gap, unemployment rate and the number of bankruptcies, which are all highly cyclical variables. This analysis shows that there are clear and significant sign of counter cyclicality in the interest margin. Secondly, to get a more holistic view and be able to look at some dynamics, I specify an econometric model through a Robust OLS method. In this analysis, I separate the funding cost and the average lending rate to get a more realistic model, where the funding cost follows the business cycle, while the lending rate follows the funding cost. The final Vector Error Correction Model shows that the banks funding cost (as defined in this thesis) can be considered procyclical. Moreover, the lending rate, which follows the movements in the funding cost, will typically change with less than the funding cost, linking us back to the countercyclical movements of the interest margin. At the end, different long-term cyclical effects of the margin are also considered. Assuming a negative connection between the margin and bank lending, the result gives support to the use of banking regulation as means for dampening the procyclical movements in credit and the financial accelerator effects. Specifically, the results can be connected to the newly introduced countercyclical buffer, and its goal of creating a more stable financial system.