Sign up for our weekly newsletter!
Professor Justin McCrary is the Faculty Director of D-Lab. He joined the Berkeley faculty in 2008. An economist by training, he teaches business law, statistics, and law and economics in the law school.
In this workshop, I use the Trade and Quote (TAQ) and Center for Research in Securities Prices (CRSP) data to examine U.S. financial markets. TAQ data contain intra day pricing information, and CRSP data provide daily pricing information. These files are usually analyzed in SAS. The plan of the workshop is to use them as a vehicle for teaching effective SAS programming. We will spend 30 minutes reviewing SAS data steps and procedures in the context of CRSP data, including some more advanced ideas such as the double-DOW loop, which can reduce computational time and often simplifies your code. We will then turn our attention to the vastly larger TAQ data, where we will discuss strategies for efficiently calculating the national best bid or offer (NBBO) from the quote file, interleaving the quote and trade files, how to restructure the data so that the unit of analysis is a security (rather than a quote or a trade), how to use SAS macros, strategies for parallelizing your analyses, and strategies for compressing files to stay under quotas. Time permitting, I will give a brief introduction to analyzing these data in Berkeley's new condo cluster http://research-it.berkeley.edu/services/high-performance-computing and compare compute times between it and the Wharton Research Data Services (WRDS) machines.