"Disaster Risk through Investors’ Eyes: a Yield Curve Analysis" [pdf]
This paper presents a model for extracting disaster risk from yield curve data. I employ an asset pricing model with disasters, which influence bond prices through changes in consumption growth, inflation, and sovereign default risk. I decompose bond prices into a non-disaster theoretical price and a disaster wedge. Using high-frequency yield curve data, I estimate disaster probabilities of around 50 countries over two decades. The model can identify key stylized facts such as the non-anticipation of the Russian-Ukrainian war, the impact of Mario Draghi's "whatever it takes" speech, and the effect of the COVID-19 pandemic.
"Caught in a Trap: Simulating the Economic Consequences of
Internal Armed Conflict" with Hannes Mueller
[pdf]
This study proposes a statistical model to capture the economic impact of the "conflict trap" phenomenon - a period of recurring outbreaks of internal armed conflict. The framework captures conflict dynamics through a discrete-time Markov process. We estimate the transition matrix and link the states to GDP per capita growth distributions through country fixed effects regressions. This allows for simulating the distribution of developmental effects of the conflict trap. We find that the trap has a large detrimental effect on long-term economic development, reaching a relative decline of GDP per capita of over 50% in the most affected countries.
- Media: BSE Focus
"Coordination Effects among Global Games: An Application on Tax Havens"
[pdf].
Regime change global games are coordination games with incomplete information in which an entity's regime changes if a sufficiently large number of agents take a certain action. This paper extends the game to multiple entities to account for the possible coordination effects among them. To analyze this, I design a model where multiple regime change global games take place simultaneously, and in an ex-ante stage, agents decide which one they play. Then, I compare the effects of altering the public information on the overall coordination. The whole model is conducted using a tax evasion application. My results show that worsening the public information of just one tax haven can increase (ease) or decrease (hinder) evasion (coordination), depending on the relative perception of each one. When the tax haven with the best public perception for evading is threatened, it leads to less evasion. However, if the tax haven with the worst public perception is threatened too harshly, it leads to more evasion due to a Crowding-in effect. Whereas a symmetric worsening always hinders coordination. Therefore, modeling a single entity global game when, in fact, players could choose among several of them, might be missing notorious coordination effects. Indeed, these effects can explain the inefficacy of the international policies to undermine tax evasion. Yet, the oncoming Minimum Global Tax Rate will reduce evasion.
Work in Progress
"On the Measurement of Fragility"