How Machine Learning Can Save Economics

According to a friend, the wide embrace of non-theoretical machine learning methods is indicative of fundamental economic problems.

That might sound bad, but … isn’t that how science (albeit a social science) is supposed to work? A conversation between theory and empirics?

In the past, pivots in econometric methodology have led to better economic models. It’s taken for granted that VARs, DSGEs, dynamic factor models, etc. “saved” economics (at least empirical macro) by displacing the old regime of large Cowles Commission-style models.

Now, it’s machine learning’s turn to save economics.

My enthusiasm for the Macro ML revolution is uniquely excited by Phillipe Goulet Coloumbe’s (PGC) work on the Macro Random Forest (MRF) and now the Hemisphere Neural Network (HNN).

The new HNN paper strikes with an informative example at the heart of economic theory: the Phillips Curve (PC).

Walking through the Covid era recursively (to ensure the model does not overfit the Covid data ex-post), PGC finds, among other things, that today’s output gap is not near zero, as existing approaches imply. He also finds that the PC coefficient of the output gap has increased reliably since 1990, contradicting the popular narrative that the PC is dead.

More generally, HNN is a new way to estimate latent economic variables that we don’t actually observe (e.g., the output gap, neutral interest rates, term premia). HNN improves upon existing methods in a number of ways.

  1. It dispenses with restrictive law of motion assumptions on model parameters and latent states. State-space approaches often impose that factors or coefficients evolve according to a random walk or an arbitrary AR process.
  2. It produces a linear output layer. So, despite all the non-linear ML, there is a final, fully interpretable end-product.
  3. Important predictors of the latent states can be identified, allowing for more interpretability over and above both traditional machine learning and factor-based econometric methods.
  4. It allows for a novel sense of volatility that addresses known weaknesses in the widely-used GARCH and SV approaches.
  5. As an extension of 4, the model can predict its own demise. It will explicitly tell you in real-time if its predictions are uncertain.

I believe PGC’s recent work is an important step forward in the ML-meets-macroeconometrics literature. MRF and HNN directly, intuitively, and effectively address many concerns about macro models head-on using clever modifications of widely-known ML approaches. Stay tuned to this literature.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: