Working Papers:
Securitization, Non-Recourse Loans and House Prices
We study how securitization and non-recourse (limited liability) laws jointly affect house prices. Our model has the novel prediction that a combination of securitization and non-recourse laws increases house prices during booms. Risk averse originators stop screening borrowers when loans are securitized, unable to credibly signal loan quality; speculators subsequently sell houses, and their nonrecourse put-option increases prices. We test this prediction for US states/MSAs during the 00s boom. We .find a positive relationship between securitization and house prices which is roughly doubled in non-recourse states. This mechanism may explain 75% of the average price diff.erence between recourse and non-recourse states.
Civility and Hostility in Parliamentary Politics (joint with Kuniaki Nemoto)
This paper creates a new measure of hostility in speeches to investigate whether a shift from majoritarianism to consensualism due to electoral reform affects the way a parliament operates. We use a sentiment analysis approach to create a lexicon of words associated with anger and assign anger scores to all speeches made by New Zealand MPs between 1987 and 2016. Confirming our methodology’s internal and external validity through different means, evidence suggests that the electoral reform is associated with a decrease in hostility, particularly in the upper percentiles, and that the decrease in hostility is driven by government members’ use of neutral words in speaking to potential coalition partners, suggesting that politicians use hostility strategically. Letter to the FT
Work in Progress:
Dynamics of Regulation of Strategically Complex Financial Products (joint with Michel Azulai)
Banks often design complex new financial products, which are costly to screen by regulators. Given that many of these products are good, banks can strategically hide some bad products by making them complex. These bad financial products can in turn lead to moral hazard issues, as banks are bailed out in case of adverse shocks. What is the optimal regulation given that bailouts prevent financial punishments as an incentive mechanism? We find the optimal regime when regulators can make binding commitments. Initially, regulators do not screen and banks do not make complex bad products. This is sustained as after enough time has passed, regulators stop screening and allow banks to make use of bad, complex products. This is a reward for their previous discipline. We explore how results change under different modelling assumptions. We surmise that our base results will not hold when regulators cannot commit: the optimal regime in this case likely follows a cyclical path of alternating phases of lax and harsh screening.
Untitled Project on International Trade and Physics (joint with Johann Caro Burnett)
Untitled Project on Pollution in Brazil
Other Papers:
Contamination between Experiments: playing social preference games in sequence (March 2009, Master Thesis, PUC-Rio)
This paper deals primarily with the issue of contamination between experiments, which we define as altered behaviour by participants in an experiment due to experience in a previous experiment. We study three channels for this contamination, expectations update, a wounded pride model and through fatigue. We test these hypotheses by having participants play either an Ultimatum or a Dictator game or participate in a test, followed by a novel game design, an Altered Public Goods (APG) game, and compare to a control group. We cannot reject the hypothesis of contamination via an expectations update. We find mixed evidence that contamination occurs due to players behaving in a wounded pride model. We find fatigue as possible contaminant to be non-significant. (final draft available at the PUC-Rio website)