Iskander Karibzhanov is a Senior Economist in the International Economic Analysis Department at the Bank of Canada. His primary research fields are international trade, macroeconomics and computational economics. He received his PhD in Economics from The University of Minnesota.
How might one simulate a million realistic income paths and compute their statistical moments in under a second? Using CUDA-based methods to estimate the Canadian earnings process, I find that the distribution of labour income growth is sharply peaked with heavy tails—similar to that in the United States.
We use graphic processing unit computing to compare Bayesian and sample theory semiparametric binary response models. Our findings show that optimal bandwidth does not outperform regular bandwidth in binary semiparametric models.