Giorgi Nikolaishvili

I am currently on leave as a Visiting Scholar at UNC Chapel Hill (Fall 2026).

I am a macro-econometrician interested in the estimation, inference, and decomposition of macroeconomic impulse responses. Some of my recent work develops confidence intervals for impulse responses while controlling false discovery and false coverage rates. Other work studies nonlinear dynamic effects using semiparametric local projections and develops methods for decomposing dynamic effects into channel-specific contributions in reduced-form settings. I have applied these methods to study the role of bank lending in monetary policy transmission and the total dynamic effects of monetary policy and oil shocks, among other empirical applications. In a complementary line of work, I develop computational methods that make solving large-scale heterogeneous-agent models more accessible.

At Wake Forest, I designed and teach a course on Prediction and Machine Learning in Econometrics. In this course, I teach students high-dimensional statistical learning theory and how to apply modern methods to real data using contemporary computational tools and reproducible workflows.

I am a proponent of open-source and reproducible research and regularly contribute replications to the Institute for Replication. I was one of many coauthors of I4R’s first meta-study, now published in Nature.