Firm Productivity and the Foreign‐Market Entry Decision

We use Japanese firm‐level data to examine how a firm’s productivity affects its foreign‐market entry strategy. The firm faces a choice between exporting and foreign direct investment (FDI). In the case of FDI, the firm has two options: greenfield investment or acquisition of an existing plant (M&A). If it selects greenfield investment, it has two ownership choices: whole ownership or a joint venture with a local company. Controlling for industry‐ and country‐specific characteristics, we find that the more productive a firm is, the more likely it is to choose FDI rather than exporting and greenfield investment rather than M&A.     

Not Quite the End of Tightening

Today's release of China's May data suggests that economic growth has held steady (pushing back worries about a hard landing), while inflation accelerated a little from April (meaning debate about another interest hike will continue). All in all, the data has a mild positive bias-inflation was as expected, despite lots of scary stories of escalating pork prices in recent weeks, while the demand side looked solid.     

Pathwise CVA regressions with oversimulated defaults

We consider the computation by simulation and neural net regression of conditional expectations, or more general elicitable statistics, of functionals of processes $(X,Y)$. Here an exogenous component Y (Markov by itself) is time-consuming to simulate, while the endogenous component X (jointly Markov with Y) is quick to simulate given Y, but is responsible for most of the variance of the simulated payoff. To address the related variance issue, we introduce a conditionally independent, hierarchical simulation scheme, where several paths of X are simulated for each simulated path of Y. We analyze the statistical convergence of the regression learning scheme based on such block-dependent data. We derive heuristics on the number of paths of Y and, for each of them, of X, that should be simulated. The resulting algorithm is implemented on a graphics processing unit (GPU) combining Python/CUDA and learning with PyTorch. A CVA case study with a nested Monte Carlo benchmark shows that the hierarchical simulation technique is key to the success of the learning approach.