Arvid Raknerud

Using the Helmert-Transformation to Reduce Dimensionality in a Mixed Model: Application to a Wage Equation with Worker and Firm Heterogeneity

A model for matched data with two types of unobserved heterogeneity is considered – one related to the observation unit, the other to units to which the observation units are matched. One or both of the unobserved components are assumed to be random. This mixed model allows identification of the effect of time-invariant variables on the observation units. Applying the Helmert transformation to reduce dimensionality simplifies the computational problem substantially. The framework has many potential applications; we apply it to wage modeling. Using Norwegian manufacturing data shows that the assumption with respect to the two types of heterogeneity affects the estimate of the return to education considerably.

How Do Banks’ Funding Costs Affect Interest Margins?

We use a dynamic factor model and a detailed panel data set with quarterly accounts data on all Norwegian banks to study the effects of banks’ funding costs on their retail rates. Banks’ funds are categorized into two groups: customer deposits and long-term wholesale funding (market funding from private and institutional investors including other banks). The cost of market funding is represented in the model by the three-month Norwegian Inter Bank Offered Rate (NIBOR) and the spread of unsecured senior bonds issued by Norwegian banks. Our estimates show clear evidence of incomplete pass-through: a unit increase in NIBOR leads to an approximately 0.8 increase in bank rates. On the other hand, the difference between banks’ loan and deposit rates is independent of NIBOR. Our findings are consistent with the view that banks face a downward-sloping demand curve for loans and an upward-sloping supply curve for customer deposits.

Indirect inference methods for stochastic volatility models based on non-Gaussian Ornstein-Uhlenbeck processes

An indirect inference method is implemented for a class of stochastic volatility models for financial data based on non-Gaussian Ornstein-Uhlenbeck (OU) processes. First, a quasi-likelihood estimator is derived from an approximative Gaussian state space representation of the OU model. Next, data are simulated from the OU model for given parameter values. The indirect inference estimator is then obtained by minimizing, in a weighted mean squared error sense, the score vector of the quasi-likelihood function for the simulated data, when this score vector is evaluated at the quasi-likelihood estimator obtained from the real data. The method is applied to Euro/Norwegian krone (NOK) and US Dollar/NOK daily exchange rate data. A simulation study reveals that the quasi-likelihood estimator may have a large bias even in large samples, but that the indirect inference estimator substantially reduces this bias. The accompanying R-package, which interfaces C++ code, is documented and can be downloaded.

The Relation between Banks' Funding Costs, Retail Rates and Loan Volumes: An Analysis of Norwegian Bank Micro Data

We use a dynamic factor model and a detailed panel data set for six Norwegian bank groups to analyze i) how funding costs affect retail loan rates and ii) how retail rate differences between banks affect market shares. The data set consist of quarterly data for 2002Q1-2011Q3 and include information on loan volumes and retail (interest) rates for loans to firms and households. The cost of market funding is represented in our analysis by the three-month money market rate and a proxy for market risk { the credit spread on unsecured senior bonds issued by Norwegian banks. Our estimates clearly suggest incomplete pass-through: a 10 basis points increase in the market rate leads to an approximately 8 basis points increase in retail loan rates. We also find that credit demand from households is more elastic with regard to the loan rate than demand from businesses.