Misinformed Speculators And Mispricing In The Housing Market (with Chris Mayer) [Paper, NBER Digest]
Review Of Financial Studies (2016) 29(2): 486-522.
When you buy a home, you have to tell the county clerk where to mail your property tax bill. We use this information to identify both local and out-of-town second-house buyers in US cities from 2000 to 2008. We find that demand from out-of-town (but not local) buyers predicts future house-price booms. And, we introduce a novel identification strategy to show that this effect isn’t being driven by reverse causation.
Sparse Signals In The Cross-Section Of Returns (with Adam Clark-Joseph and Mao Ye) [Paper]
Journal Of Finance (2019) 74(1): 449-492.
We use the LASSO to make -minute-ahead return forecasts for each stock. The LASSO boosts out-of-sample predictive power by choosing predictors that trace out the consequences of unexpected news announcements. What’s more, the size of the LASSO’s best-fit penalty parameter, per month, shows that this success is not just the result of loading on traditional monthly predictors only faster.
Estimating The Anomaly Baserate (with Andy Neuhierl and Michael Weber) [Paper]
Journal of Financial Economics, R&R.
The academic literature contains hundreds of statistically significant cross-sectional predictors, causing many to question whether we’re using the right statistical tests. But, here’s the thing: even if a researcher does use the right tests, he will still draw the wrong conclusions if he starts out with the wrong priors. So, which priors should a researcher use? We propose a new statistical approach to answer this question.
The Sound Of Many Funds Rebalancing (with Slava Fos) [Paper]
FRA Best-Paper Prize (2017).
Noise makes financial markets possible. (Black, 1986) We propose a new source of this all-important ingredient: complexity. Modern financial markets contain all sorts of different traders following a wide variety of different trading rules. So, even if you are fully rational, a stock’s demand might still appear random because the interactions of these various trading rules are too computationally complex to predict.
The Madness Of Crowds And The Likelihood Of Bubbles [Paper]
FRA Best-Paper Prize (2018).
Market participants are constantly swimming in a sea of psychological biases and trading constraints. Yet, in spite of all these biases and constraints, large pricing errors such as speculative bubbles are rare. Why is this? How often should we expect the limits of arbitrage to bind? This paper proposes a new kind of model to answer this question. The model makes out-of-sample predictions about where bubbles are most likely.
Trading On Coincidences [Paper]
Management Science, R&R.
Feature-Selection Risk [Paper]