France has a long and unique experience of public intervention in household debt restructuring. When facing financial distress, households can file a case to a “households’ overindebtedness commission” (HOC). These HOCs either grant a delay of payment or impose a partial reimbursement of the secured or unsecured debt. This paper evaluates the ex post impact of this decision on the creditors’ recovery rate, the household’s re-default rate and the net benefit of the treatment, defined as the amount recovered by the creditors minus the public cost of treatment of the file. The random allocation of the households over file managers with different pro-household friendlinesses is used to correct for endogeneity. Sixty percent of households are ordered to repay part of their debt. Over a two-year horizon, the possibility to grant a delay of payment decreases the average redefault rate and the average repayment rate respectively from 13 to 7 percentage points and from 14 to 12 percentage points. The net benefit over a small fraction of low distressed households offsets the loss observed over a large fraction of more distressed households. Our results highlight a substantial impact of the severity of the case manager. JEL : D1, G2, K35 During the last financial crisis, the numbers of personal bankruptcies and household indebtedness have reached levels never experienced before on a worldwide scale. According to Eurostat figures taken from national accounts, the household debt-to-income ratio exceeded 200% in Ireland, the Netherlands and Denmark in 2009. Meanwhile, according to statistics from the US courts, personal bankruptcies hit a record high of 1.531 million people in the US in 2010. As a result, the policy debate seems to have shifted from the set up of an ex ante “optimal” bankruptcy regime to the implementation of “ex post” special policy programs restructuring the debt of households in financial distress. Put more simply, the urgent question has increasingly shifted towards how to restructure the huge amount of outstanding debt rather than how to prevent new indebtedness. For example, a program was implemented in Pennsylvania in 2010 to provide help for unemployed borrowers. The program (Homeowners’ Emergency Mortgage Assistance Program) enables borrowers to obtain an interest-free loan to pay arrearages plus a portion of their mortgage in case of unexpected losses of income. 1 In 2009, the US launched a federal program – the Home Affordable Modification Programin order to facilitate the modifications of loans granted to homeowners at risk of foreclosure. There are ongoing discussions on tackling the debt restructuring issue in Spain. Among the existing systems, French and German bankruptcy laws are theoretically considered close to optimum (White, 2006 and Kilborn, 2007). Nevertheless, this statement stands mainly for the ex ante analysis of the legal process. Indeed, compared to the US system, French and German laws possess screening mechanisms designed for avoiding opportunistic use of the law. An overindebted household must be considered in “good faith”, otherwise its file is simply rejected. Strategic filings are thus not of great concern in these two countries. However, these systems do not prevent ex post problems, such as re-default, from occurring. In this respect, very few studies are available, especially because of a lack of comprehensive data available. In this study, we use a unique dataset made up of the population of households that filed for bankruptcy in France from mid-2007 to 2008 and tracked until 2010. This unique dataset provides the entirety of the debt obligations for each household. It also provides all of the information concerning the procedure, including the outcomes of the procedure, i.e., recovery rates and re-defaults. The French procedure presents an interesting and unique feature. Since 1990, it possesses a compulsory ex ante alternative method of dispute resolution (following the denomination of Shavell, 1995) in order to restructure the secured and (mostly) unsecured debt of the financial distressed household. Households have to file to a conciliation board (commission de surendettement), i.e., the households’ over-indebtedness commissions (denoted HOC hereafter). The purpose of HOC is to make creditors and debtors agree on a settlement when households cannot meet their commitments because of excessive debt. Inside of the HOC, one key actor is the file manager. Indeed, she orients a procedure issue by assessing the file and proposing the appropriate outcomes. Hence, file allocation 1 For a first assessment of the program, one can read Orr et al., 2011. between case managers may have an important influence on the efficiency of the procedure, according to the personal inclination of the chosen case managers. Using a two-year horizon, a case manager must basically make the choice between asking the household to repay a part of its debt or granting it a moratorium period (or “observation period”), during which it would have nothing to pay. The HOC examines the financial situation of the household again at the end of the observation period (“OP” hereafter). The household can then benefit from a total discharge against a liquidation of his assets if he is still genuinely in dire conditions. 2 Otherwise, a modification plan is set up. In this paper, our contribution is to explore a piece of literature that has received little attention so far –public programs of household debt restructuring. The public program under review does not focus solely on mortgage debt but on all types of debt. We use an original source of identification, a measure of the pro-creditor friendliness of case managers in charge of bankruptcy files, to identify the effects of debt repayment on ex post re-default and recovery rates. We propose a welfare measure of the program by comparing the public cost of treating a file with the debt recovered by creditors. Previous works have mainly focused on the trade-off faced by any personal bankruptcy system. On the one hand, personal bankruptcy law aims at insuring households against unanticipated negative shocks of income. On the other hand, it aims at limiting strategic behaviors from the part of households who could over-borrow in anticipation of debt relief (White, 2005). From an international comparison perspective, the main result is that different legal systems governing personal bankruptcy laws have different balances between the objective of creditor protection and debtor protection. For instance, the US system, even after adopting the Bankruptcy Abuse Prevention and Consumer Protection Act 3 (BAPCPA) in 2005, can be qualified as being debtor-friendly, whereas France and Germany have more creditor-protective systems. One strand of the literature has tried to exhibit the main features of efficient bankruptcy law systems. These works have explored microeconomic theory, such as Wang and White (2000), and macroeconomics, such as Athreya (2002) and Livshits, MacGee and Tertilt (2006). These works also focus on ex ante perspectives and confirm the relative nature of adopted solutions to solve the trade-off between creditor and debtor protection. As underlined by Han and Li (2011), who stress ex post bankruptcy borrowing, little is known about households’ behaviors after bankruptcy. This paper complements a few of the recent empirical papers dealing with the issue of mortgage loan modification in the US. Quercia, Ding and Ratcliffe (2009) assess the impact of different types of mortgage loan modifications on the likelihood of a re-default and find that only a strong payment reduction makes modified loans sustainable. Using a sample of a dataset that covered 60 percent of the mortgage market and an indirect measure of loan modification, Adelino, Gerardi and Willen (2009) show that, from 2007 to 2008, loan servicers performed only approximately 3 percent 2 As documented by Blazy et al. (2011), 80% of households benefiting from total discharge do not have any valuable assets. 3 A comprehensive economic analysis of the reform can be found in White (2007). of payment-reducing modification. They explain that re-default and self-cure risks make investors reluctant to renegotiate. Basing their analysis on loan servicers’ data, Agarwhal et al. (2011) find that securitization substantially explains the difficulty to renegotiate. They also find that a decrease in 100 percentage points of the mortgage interest rate reduces the re-default probability within 6 months by 4 percentage points. Mayer, Morrison, Piskorski and Gupta (2010) exploit a natural experiment impacting a key actor of the subprime market to show that creditors have limited incentive to renegotiate because they fear the adoption of strategic behaviors by households. As the closest to our work, Agarwhal et al. (2012) carefully evaluate the impact of the home affordable modification program both on the intensity of the renegotiation of mortgages within and outside the program and on subsequent outcomes (e.g., foreclosure rate, house price). By contrast, we “internally” evaluate the French procedure, comparing the impacts of the decision of the case manager whether to order a repayment on the subsequent outcomes. The scope of our dataset enables us to widen the perspective by addressing any cause of personal bankruptcy and any type of debt, bearing in mind that defaulting on a mortgage is a very rare event in France. This paper uses a measure of the pro-creditor friendliness of case managers in charge of bankruptcy files to identify causal effects of debt repayment on ex post re-default and recovery rates. One could indeed suspect that the assessment of these rates is plagued by some endogeneity issues. Despite having at our disposal unusually rich sets of controls, it is very likely that case managers have pieces of information on households that are unobservable to econometricians, e.g., family background or finan

France has a long and unique experience of public intervention in household debt restructuring. When facing financial distress, households can file a case to a “households’ over-indebtedness commission” (HOC). These HOCs either grant a delay of payment or impose a partial reimbursement of the secured or unsecured debt. This paper evaluates the ex post impact of this decision on the creditors’ recovery rate, the household’s re-default rate and the net benefit of the treatment, defined as the amount recovered by the creditors minus the public cost of treatment of the file. The random allocation of the households over file managers with different pro-household friendlinesses is used to correct for endogeneity. Sixty percent of households are ordered to repay part of their debt. Over a two-year horizon, the possibility to grant a delay of payment decreases the average redefault rate and the average repayment rate respectively from 13 to 7 percentage points and from 14 to 12 percentage points. The net benefit over a small fraction of low distressed households offsets the loss observed over a large fraction of more distressed households. Our results highlight a substantial impact of the severity of the case manager.

This paper provides empirical evidence of the heterogeneous borrowing behaviours of French regions, despite a common accountability constraint that forces them to balance their budget and to borrow only to finance investment expenditure (golden rule). To this end, we use a quantile regression analysis covering the period from 1999 to 2007. The heterogeneity is very pronounced when the regions face a negative shock on debt, for instance a tightening of financial conditions. We explain our findings as a consequence of the fact that the Golden rule can be thought of as a “soft” rule if some local administrations believe that a financial rescue from the central government is automatic (as the regions receive transfers from the later). In this case, some regions find it advantageous to consider borrowing as an adjustment variable when taking their budgetary decisions.

In this paper we examine whether the weights of the forecasted macroeconomic variables in the valuation of sovereign risks has changed after the adoption of the Bale 2 new regulation framework (around the years 2005/2006 when the new rules begun to be implemented in Europe). For purpose of illustration, we examine the case of France, Germany and Greece. We estimate a time-varying probability Markov Switching model in which the break date is endogenous. Our results show differences across countries.

Résumé Dans cet article, nous examinons si le poids accordé aux anticipations des variables macroéconomiques pour évaluer le risque des obligations souveraines a été plus important après l’adoption des nouvelles règles prudentielles de Bâle 2, c’est-à-dire autour des années 2005-2006 (années à partir desquelles ces règles ont commencé à être appliquées dans les pays d’Europe). À titre illustratif, nous comparons trois pays de la zone euro, à savoir l’Allemagne, la France et la Grèce. Pour ce faire, nous estimons un modèle Markov-switching avec probabilités de transition variables. Nos résultats révèlent des comportements différents entre pays, au moins à court terme.

Résumé Cet article montre qu’en dépit de contraintes comptables communes les obligeant à équilibrer leur budget et à emprunter uniquement pour financer l’investissement (règle d’or), les comportements d’emprunt des régions françaises sont hétérogènes. Ce résultat est établi par des régressions quantiles. L’hétérogénéité est très prononcée lorsque les régions subissent un choc d’endettement négatif, durcissement des conditions financières par exemple. Ceci pourrait résulter du fait que la règle d’or soit considérée comme une « soft rule » par des collectivités anticipant un sauvetage financier automatique de la part du gouvernement. Ainsi, dans le cas français, la plausibilité de l’hypothèse de bailing-out n’est pas invalidée.

This paper provides empirical evidence of the heterogeneous borrowing behaviours of French regions, despite a common accountability constraint that forces them to balance their budget and to borrow only to finance investment expenditure (golden rule). To this end, we conduct a quantile regression analysis. The heterogeneity is very pronounced when the regions face a negative shock on debt, for instance a tightening of financial conditions. Our findings may be due to the fact that the Golden rule can be thought of as a “soft” rule if some local administrations believe that a financial rescue from the central government is automatic. Hence, in the French case the bailing-out hypothesis cannot be rejected. Classification JEL: H74, E62, K34, R5

On 16-17 September 2010, the Banque de France’s Directorate General Economics and International Relations and the Bureau d’economie theorique et appliquee (BETA) of Strasbourg University jointly hosted a conference on the topic “New challenges for public debt in advanced economies” that brought together 70 economists from French and foreign universities, ESCB and other central banks, and European and international institutions.

Les 16 et 17 septembre 2010, la Direction generale des Etudes et des Relations internationales a organise en collaboration avec le Bureau d’economie theorique appliquee (BETA) de l’universite de Strasbourg une conference reunissant 70 economistes issus du monde academique francais et etranger, de banques centrales du SEBC ou d’ailleurs et d'institutions europeennes et internationales sur le theme « Nouveaux defis poses par la dette publique dans les pays developpes ».

Using second generation Panel Unit Root Tests (PURT), panel cointegration tests and panel Granger causality tests we find that although the financial crisis may have increased risk aversion for investors, it did not make disappearing speculative behaviours on structured credit markets. On the contrary, support measures to the banking sector and fiscal stimulus packages have given the opportunity for some investors to speculate on sovereign debt through Credit Default Swap (CDS) vehicles. Thus this article supports ongoing initiatives to strengthen the prudential regulation undertaken in these markets under the Group of 20 (G20) or the Financial Stability Forum (FSF).

Cet article présente l’une des causes déjà identifiées de la crise du crédit, le rôle des agences de notation, à travers un aspect souvent méconnu : l’analyse de l’influence du droit sur l’activité de titrisation et, partant, sur les phénomènes déclencheurs de la crise. Les agences de notation sont en effet critiquées pour la qualité de l’information qu’elles donnent lors de l’évaluation des produits structurés. La question se pose aujourd’hui avec d’autant plus d’acuité que les agences se sont vues conférer, dans le secteur bancaire, un rôle de quasi-régulateur - ou au moins de participation à la régulation - par la convention de Bâle II sur la régulation bancaire. Il convient ainsi d’apporter un éclairage sur la part des facteurs juridiques dans les causes de la crise et, notamment, de la complexité des montages juridiques de titrisation qui en sont à l’origine. Cet article présente également quelques remèdes envisageables pour éviter qu’une telle crise ne se reproduise dans l’avenir.

Although many works support creditor friendly bankruptcy laws, an evolution towards debtor friendly systems is at work. This paper proposes a theoretical ground to meet this paradox. It reconsiders the economic role of bankruptcy law by stressing on the courts’ production of information. It reveals that the transmission of a lenient signal by judges makes it possible to reduce the hazard that bad risks seek to avoid going on trial. Thus, it shows that debtor friendly bankruptcy laws are not systematically opposed to creditors’ interests. They reduce the risk of the economy and contribute to the improvement of the global efficiency.

Financial rating agencies are key actors of financial markets. Their different ratings are usually pre-conditions for issuing securities or bonds. If they assess the financial viability of the financial instruments issued, they also assess the soundness of the legal arrangements used. Thus, taking the law into account is a key element of their assessment. At a macro level, their ratings may express an opinion of the economic attractiveness of a territory. The same may apply to the legal framework used by issuers. Surprisingly, the literature on this topic reveals to be scarce. This paper aims at fulfilling this gap.