I hold a PhD in Economics from the University of Zürich, specializing in industrial organization, econometrics and competition policy.

I believe in knowledge sharing, and I try to contribute by sharing my code, tutorials and lecture notes. I write on Medium for Towards Data Science.

DOWNLOAD CV

Posts

Should we ever use DiD?
Assume parallel trends. DiD nails the ATT. Would a double-machine learning (DML) estimator also be able to recover the ATT? And the ATE and ATC? When should we use DiD if we have DML?
Nov 21, 2023 13 min read
Instrumental Variables
How to estimate causal effects when you cannot randomize treatment. A/B tests are the golden standard of causal inference because they allow us to make valid causal statements under minimal assumptions, thanks to randomization.
Nov 12, 2023 14 min read
Instrumental Variables
Beyond Churn Prediction and Churn Uplift
A very common task in data science is churn prediction. However, predicting churn is often just an intermediate step and rarely the final objective. Usually, what we actually care about is reducing churn, which is a separate objective, not necessarily related.
Jul 25, 2023 14 min read
Beyond Churn Prediction and Churn Uplift
Evaluating Uplift Models
One of the most widespread applications of causal inference in the industry is uplift modeling, a.k.a. the estimation of Conditional Average Treatment Effects. When estimating the causal effect of a treatment (a drug, ad, product, …) on an outcome of interest (a disease, firm revenue, customer satisfaction, …), we are often not only interested in understanding whether the treatment works on average, but we would like to know for which subjects (patients, users, customers, …) it works better or worse.
Jul 12, 2023 20 min read
Evaluating Uplift Models
Bayesian AB Testing
An introduction to the Bayesian approach to randomized experiments. Randomized experiments, a.k.a. AB tests, are now the established gold standard in the industry to estimate causal effects. Randomly assigning the treatment (new product, feature, UI, …) to a subset of the population (users, patients, customers, …) we ensure that, on average, the difference in outcomes (revenue, visits, clicks, …) can be attributed to the treatment.
Jul 11, 2023 15 min read
Bayesian AB Testing
See all posts

Newsletter

XYZ#5
A causal inference newsletter News The Causal AI Conference 2023 Two weeks ago, CausalLens hosted one of the most important causal inference conferences, especially in the industry. The full conference is available on YouTube.
Oct 9, 2023
XYZ#5
XYZ#4
A causal inference newsletter News Encouragement Designs and Instrumental Variables for A/B Testing Spotify researchers introduce IV for industry applications, when randomizing the treatment is not possible or desirable, for example because the feature is common knowledge.
Aug 27, 2023
XYZ#4
XYZ#3
A causal inference newsletter News Speeding up HelloFresh’s Bayesian AB Testing PyMC researchers tested different ideas to speed up thousands of concurrent AB tests. First, they decreased model parametrization, and sampling length, leading to marginal speed increases.
Aug 20, 2023
XYZ#3
See all

Projects

Algorithmic Collusion on Online Marketplaces
I show that online marketplaces, by algorithmically controlling consumers’ attention, have the incentives and ability to either facilitate or prevent algorithmic collusion, depending on their business model. I also explore platforms’ dual role.
PDF Slides
Algorithmic Collusion on Online Marketplaces
Foreclosure Complementarities
We use a computational model of market dynamics to investigate the interplay between exclusionary bundling and predatory pricing. We show that the two foreclosure practices are complementary and we investigate potential policy interventions.
PDF Slides
Foreclosure Complementarities
Algorithmic Collusion Detection
I show that algorithms can learn reward-punishment schemes that are fully independent from the rival’s actions and I propose a model-free test for algorithmic collusion based on historical data.
PDF Slides
Algorithmic Collusion Detection
Approximation Methods for Large Dynamic Stochastic Games
I compare existing approximation methods to compute Markow Perfect Equilibrium in dynamic stochastic games with large state spaces. I also propose a new approximation method called “Games with Random Order”.
PDF
Approximation Methods for Large Dynamic Stochastic Games
Ownership and Product Similarity
I generate a time-varying measure of S&P500 firm similarity using a zero-shot clustering model. The model takes as input BERT embeddings of product descriptions and is trained on market definitions from the EU commission. The objective is to estimate the causal effect of common ownership on product similarity.
Ownership and Product Similarity

Contact