High frequency data for heterogeneity in macro
What is your research about?
This research is divided in two streams.
In the first stream, we quantify precisely the wealth redistribution generated by the current inflation shock in the US - and find it to be an order of magnitude larger than other distributional impacts of inflation, such as those stemming from heterogenous consumption baskets between rich and poor. In a macroeconomic model with heterogenous households (HANK) that matches the empirical distribution of nominal positions, the wealth redistribution generated by the current inflation shock is found to be expansionary for aggregate consumption, as the households who lose (wealthy middle-age and elderly) adjust their consumption by less than the households who gain (young middle-class mortgagors).
In the second stream, we look at macroeconomic models with heterogenous agents (HANK) which typically feature households immediately adjusting to macroeconomic shocks (e.g. a rise in interest rates). Once the responses of all individual households are aggregated, the IRFs to macro shocks implied by the HANK models therefore always exhibit a spike on impact and a gradual return to the steady state in the following quarters. This contrasts with most of the empirical evidence obtained e.g. through VARs using macro time series, which tend to display some inertia (a gradual increase peaking after some quarters). To reconcile the two IRFs, the literature has proposed introducing some behavioural frictions (e.g. sticky expectations of households) in HANK models. However, these frictions lack a clear empirical foundation. Moreover, some recent papers have been uncovering some immediate response to macroeconomic shocks also in aggregated time series, once the estimation is done on daily data. Yodlee's data would allow us to simultaneously uncover aggregate and individual responses to macro shocks, allowing us to estimate a full-fledged HANK model.
How will the Stone Centre grant help your research?
The Stone Centre grant will help us with the purchase of a dataset provided by a fintech company (Yodlee) which contains every flow in and out of the bank accounts of a representative sample consisting in around 4-5 million users (typically households) in the US.
What will you produce as part of your research?
We'll produce working papers for both research streams.
In addition, we could also produce CORE Econ education materials, such as some data visualisations like the CORE's Skyscrapers, which would show the distributional effects of the current inflation shock in the US. We could also produce some animation showing which households typically react first to MP shocks, or how this reaction transmits to aggregate demand more broadly.
About this grant
Title of the project: High frequency data for heterogeneity in macro
Value of the grant: £25,000
Duration: September 2023 – ongoing
