Working papers

















































































































Covid 19: A New Challenge for the European Monetary Union, joint with Alexis Guillaume

There was a risk of another euro crisis in March 2020: the pandemic triggered portfolio rebalancing across the Euro Area reflecting a loss of financial markets’ confidence into the capacity of several member sovereigns to repay their debt. But spreads narrowed down quickly after policy interventions. We document this episode by providing a detailed account of the patterns of government bond spreads during the pandemic and the crisis management period. We first quantify the resilience capacity of E.A. members and we find that Italy and Greece were the most vulnerable, far ahead of Spain and Portugal. Second, we find that contagion was significantly higher in 2020 than during the 2011-2012 crisis episode. Third,we assess the contribution of different policy announcements by the European polity during the pandemic: ECB speeches were a game changer as it had been in 2012; more unusual: the policy announcements by the European Council also helped to significantly calm down the market stress. Unfortunately, their effect was partly offset by Eurogroup’s lack of coordination. In sum, crisis management by European institutions/ body contributed to absorb the initial heterogeneity among members but it could have been more efficient as our results suggest that markets sanctioned uncertainty due to inter-governmental disagreements.

CEPR DP 14848

Companion website with Data Vizualization


Grey Zones in International Finance, joint with Amélie Guilin and Vincent Vicard

Tax avoidance schemes generate artificially complex cross-border financial structures inflating measured international investment stocks in tax havens. Using a standard gravity framework, we estimate that about 40\% of global assets (FDI, portfolio equity and debt) are 'abnormal' - unexplained - stocks. Abnormal stocks are increasing over time and concentrated in a limited number of jurisdictions. Six jurisdictions including three European countries are the largest contributors: Cayman, Bermuda, Luxembourg, Hong Kong, Ireland and the Netherlands. Interestingly, the Luxleaks in 2014 do not appear to have diverted cross-border investments away.



Credit under influence: Evidence from France, with A. Matray and N. Pinardon-Touati


Formally independent private banks change their supply of credit to the corporate sector for the constituencies of contested political incumbents in order to improve their reelection prospects. In return, politicians grant such banks access to the profitable market for loans to local public entities among their constituencies. We examine French credit registry data for 2007--2017 and find that credit granted to the private sector increases by 9\%--14\% in the year during which a powerful incumbent faces a contested election. In line with politicians returning the favor, banks that grant more credit to private firms in election years gain market share in the local public entity debt market after the election is held. Thus we establish that, if politicians can control the allocation of rents, then \it{formal} independence does not ensure the private sector's \it{effective} independence from politically motivated distortions.

Media coverage: Le Monde, Alter Eco, Les Echos, Libération

CEPR DP 14409

The transmission channels of unconventional monetary policy, Evidence from a change in collateral requirement in France, with Pranav Garg and Jean Imbs

Using a bank-firm level credit registry combined with firm-level balance sheet data we establish the presence of heterogeneity in the effects of unconventional monetary policy transmission. We examine the consequences of a loosening in the collateral eligibility requirement for credit refinancing in France. The policy was designed to affect bank lending positively. We expect a linear increase in lending and an additional increase in loans to firms with newly acceptable rating. We find a large heterogeneity of the monetary policy transmission including the unexpected reduction of lending by the banks benefiting the most from the policy. These are small, risk-averse banks whose foremost concern after the recession was to strengthen their balance sheets. Banks least affected by the policy respond with a reduction in credit to low risk borrowers in reaction tothe change in the market structure. Last we document heterogenous effects of the policy on firmsdepending on their size

CEPR DP 13693, voxeu column, online appendix


Banks defy Gravity, joint with V. Bouvatier and G. Capelle-Blancard


This paper provides the first quantitative assessment of the contribution of global banks inintermediating tax evasion. Applying gravity equations on a unique regulatory dataset based oncomprehensive individual country-by-country reporting from all the Systemically ImportantBanks the European Union, we find that: 1) Tax havens generate a threefold extra presence of foreign banks; 2) The favorite destinations of tax evasion intermediated by European banks are Luxembourg and Monaco 3) British and German banks display the most aggressive strategies in tax havens ; 4) New transparency requirements imposed in 2015 have not changed European banks commercial presence in tax havens; 5) Banks intermediate e550 billion ofoffshore deposits, that is 5% of their origin countries' GDP.

CEPR DP 12222, voxeu column


This figure displays the contribution of policy announcements on the government bond spread of E.A. members during the pandemic (in b.p.). Red squares represent the mean of the spreads over the period and white diamonds represent the individual vulnerability to exoegnous shocks (the lower, the least vulenrable) computed as the contribution of time-invariant individual factors to spreads returns (in b.p.). The period of estimate is Jan-July 2020.

This figure shows the spread of local public entity debt relative to Treasury bonds of similar maturities in the timeseries (left panel) and in the cross section (right panel)