Select Page


Job Market Paper

 Capital Gains Taxation, Learning and Bubbles                                                              Download!

Abstract. Why have there been more asset price boom-bust cycles since the 1980s despite the drop in macroeconomic risk? This paper argues that the fall of the Capital Gains Tax (CGT) is one of the reasons. In a model of learning about prices, I show that a lower CGT make prices more responsive to changes in investors’ beliefs, thereby elevating the likelihood of self-fulfilling booms and busts. This novel mechanism dominates the more traditional lock-in effect -according to which a lower CGT reduce price fluctuations- whenever the tax elasticity of realization is not too negative. A structural estimation of the model focusing on the US stock market suggests that CGT cuts account for at least 25% of the increase in Price-Dividend fluctuations, while also generating a rise in stock market valuations and a sizable equity premium. Empirical estimates using survey beliefs support the model’s prediction of an increase in the sensitivity of prices to subjective expectations due to lower taxes. Finally, I show that optimal policy prescribes a CGT that leans against market expectations, preventing belief-driven business cycles and enhancing the autonomy of monetary policy regarding financial stability concerns.

Presented at: University of Oxford, Bocconi University, University of Mannheim, UC3M, Universitat de Barcelona, Universitat Autònoma de Barcelona, Barcelona School of Economics, ENTER, Universitat de les Illes Balears, Winter Meeting of the Econometric Society, SAEe2020 and 2022, EconMod2021. 

Work in progress

The Fiscal Channel of Quantitative Easing (with Eddie Gerba and Luis E. Rojas)                                [Draft]

Abstract. This paper is a theoretical examination of the role of fiscal distortions in shaping the effects of Quantitative Easing (QE). The presence of deadweight losses from taxation breaks Wallace’s neutrality since QE influences the level and volatility of such losses. Under some conditions, QE can stimulate demand by removing tax distortions, but it increases the risk premium. This differs from the standard view that QE stimulates demand precisely by lowering risk premiums due to the relaxation of financial frictions. A Central Bank must strike the right balance between the efficiency gain of more QE against the additional risks it entails. By exploiting the risk premium from capital ownership, QE emerges as an alternative to costly taxation, suggesting an efficiency-risk trade-off for public finances.

Presented at: Bank of England, T2M Conference London 2022, UAB MacroClub, 53rd  MMF conference, SAEe2022, 2023 ASSA meeting, CESC 2023.

Dematerialization in part explained by the burst of the housing bubble (with Marina Requena[Draft]

Abstract. We document that a set of countries grew their GDP while decreasing their Material Footprint (MF) over the 2007-2017 decade, breaking the previous trend. This paper analyzes the drivers of this absolute dematerialization. In accounting terms, it simply reflects a reduction in Material Intensity (MI) typically associated with technology. Nonetheless, we show that a wide range of variables related to technology can only explain between 2% and 10% of the MI variation. Alternatively, we hypothesize that the observed dematerialization is in part a cyclical phenomenon resulting from the housing prices bust that depresses construction activity and, under some conditions, hits MF harder than GDP. Indeed, the data analysis reveals that housing prices and construction explain between 19% and 46% of the dematerialization variance. Besides, we show that the absence of the housing boom would have accelerated the dematerialization, although being insufficient to bring MF within their sustainable limits.

Heterogeneous Expectations and Stock Market Cycles (with Adrian Ifrim and Janko Heineken)              [Draft]

Abstract. We present a model of heterogeneous expectations. In the short run, agents learn about prices with different intensities due to their distinct levels of confidence regarding the signal-to-noise content of price news. Beliefs fluctuate around idiosyncratic means, which set agents’ different views about the asset’s long-run value. The model micro-founds the heterogeneous extrapolation and the persistent and procyclical disagreement present in survey data. The subjective belief system is embedded in an otherwise standard asset pricing framework, which can then quantitatively account for the dynamics of prices and trading. In the model, learning from prices leads to disagreement and trading, which reshuffles the distribution of wealth between lower- and higher-propensity-to-invest agents, affecting aggregate demand and prices. This feedback loop complements the expectations-price spiral typical of models with extrapolation, placing heterogeneity and trading as key drivers of price cycles.

Solving asset pricing models with learning through the Parameterized Expectations Algorithm