Forecasting recovery rates on non-performing loans with machine learning

We compare the performance of a wide set of regression techniques and machine-learning algorithms for predicting recovery rates on non-performing loans, using a private database from a European debt collection agency. We find that rule-based algorithms such as Cubist, boosted trees, and random forests perform significantly better than other approaches. In addition to loan contract specificities, predictors that refer to the bank recovery process — prior to the portfolio’s sale to a debt collector — are also shown to enhance forecasting performance. These variables, derived from the time series of contacts to defaulted clients and client reimbursements to the bank, help all algorithms better identify debtors with different repayment ability and/or commitment, and in general those with different recovery potential.     

Optimal Environmental Policy Differentials in Open Economies under Emissions Constraints

Is there a case for preferential treatment of the exposed sector in an economy when compliance to an aggregate emissions constraint induced by an international environmental agreement is mandatory? This question is being debated in many countries in the context of the implementation of the Kyoto Protocol. We address the issue in a general equilibrium framework and theoretically cover several market structures, including perfect competition, the large country case and oligopoly. We identify the conditions under which preferential treatment of the exposed sector is not warranted from the point of view of maximizing social welfare. In addition, we demonstrate that in the case of oligopoly, instituting a more stringent environmental policy on the exposed sector might be profit-enhancing for this sector. This finding lends theoretical support to a specific interpretation of the Porter hypothesis.     

General equilibrium impact of an energy-saving policy in the public sector

We analyse a disregarded environmental policy instrument: a switch in government expenditure away from energy (or other natural resources) and toward a composite good which includes energy-saving expenditure. We first develop two variants of an analytical general equilibrium model. A composite good is produced with constant returns to scale, and energy is imported or produced domestically with diminishing returns, yielding a differential rent to its owners. The government purchases energy and composite goods from private firms. Such a policy unambiguously increases employment. It also raises private consumption and welfare under two conditions: (i) it is not too costly and (ii) the initial share of the resource is smaller in public spending than in private consumption, or the difference is small enough. We then run numerically a model featuring both importation and domestic production of energy (oil, gas and electricity), for the OECD as a whole. Simulations show that employment, welfare and private consumption rise. We provide magnitudes for different parameter values. Earlier versions of this paper have benefited from conference participants at the European Council for an Energy-Efficient Economy, International Society for Ecological Economics World Congress, CIRED seminar and EUREQua environmental economics seminar. We especially thank Michèle Sadoun and two anonymous referees. The usual disclaimer applies.     

Threshold models for trended time series

This paper presents the theoretical development of a new threshold autoregressive model based on trended time series. The theoretical arguments underlying the model are outlined and a nonlinear economic model is used to derive the specification of the empirical econometric model. Estimation and testing issues are considered and analysed. Additionally we apply the model to the empirical investigation of U.S. GDP.This paper is the result of work carried out for the author's Ph.D. thesis. I would like to thank Hashem Pesaran for his help, encouragement and insights during the preparation of this paper. I also thank Gary Koop and Sean Holly for helpful comments. Financial assistance from the Economic and Social Research Council is gratefully acknowledged.     

Noisy chaotic dynamics in commodity markets

The nonlinear testing and modeling of economic and financial time series has increased substantially in recent years, enabling us to better understand market and price behavior, risk and the formation of expectations. Such tests have also been applied to commodity market behavior, providing evidence of heteroskedasticity, chaos, long memory, cyclicity, etc. The present evaluation of futures price behavior confirms that the resulting price movements can be random, suggesting noisy chaotic behavior. Prices could thus follow a mean process that is dynamic chaotic, coupled with a variance that follows a GARCH process. Our conclusion is that models of this type could be constructed to assist in forecasting prices in the short run but not over long run time periods.First version received: June 2001/Final version received: March 2003     

Carbon Taxes and Joint Implementation. An Applied General Equilibrium Analysis for Germany and India

Germany has committed itself toreducing its carbon emissions by 25 percent in2005 as compared to 1990 emission levels. Toachieve this goal, the government has recentlylaunched an environmental tax reform whichentails a continuous increase in energy taxesin conjunction with a revenue-neutral cut innon-wage labor costs. This policy is supposedto yield a double dividend, reducing both, theproblem of global warming and high unemploymentrates. In addition to domestic actions,international treaties on climate protectionallow for the supplementary use of flexibleinstruments to exploit cheaper emissionreduction possibilities elsewhere. One concreteoption for Germany would be to enter jointimplementation (JI) with developing countriessuch as India where Germany pays emissionreduction abroad rather than meeting itsreduction target solely by domestic action. Inthis paper, we investigate whether anenvironmental tax reform cum JI providesemployment and overall efficiency gains ascompared to an environmental tax reformstand-alone. We address this question in theframework of a large-scale general equilibriummodel for Germany and India where Germany mayundertake JI with the Indian electricitysector. Our main finding is that JI offsetslargely the adverse effects of carbon emissionconstraints on the German economy. JIsignificantly lowers the level of carbon taxesand thus reduces the total costs of abatementas well as negative effects on labor demand. Inaddition, JI triggers direct investment demandfor energy efficient power plants produced inGermany. This provides positive employmenteffects and additional income for Germany. ForIndia, joint implementation equips itselectricity industry with scarce capital goodsleading to a more efficient power productionwith lower electricity prices for the economyand substantial welfare gains.     

Overlapping Generations or Infinitely-Lived Agents: Intergenerational Altruism and the Economics of Global Warming

Do we need an overlapping generations model for the economics of global warming? To answer this question, an infinitely-lived agent (ILA) approach and an overlapping generations (OLG) model are contrasted. ILA and OLG can be viewed as polar representations of intergenerational altruism. With ILA an immortal agent acts through his investment/savings decisions as trustee on the behalf of the future generations. With OLG, agents need not behave altruistic. They simply save during working years and dissave completely during retirement. Nevertheless, ILA and OLG must not differ in their implication for greenhouse policy. Greenhouse gas abatement is a straightforward alternative to physical capital formation and, even without altruism, each age cohort has an incentive to provide current abatement in order to reduce future damages attributable to climate change. Indeed, under reasonable assumptions and parameter values, our simulations reveal such an invariance result. Provided carbon taxes are the only policy tool and tax revenues are recycled through socially mandated rules, projections of economic growth, climate change and energy consumption are only insignificantly affected by the choice of approach.     

Hedonic models and air pollution: Twenty-five years and counting

This paper reports a meta analysis of how effectively hedonic property models have detected the influence of air pollution on housing prices. Probit estimates are reported describing how data, model specification, and local property market conditions in cities represented in thirty-seven studies influence the ability of hedonic models to uncover negative, statistically significant relationships between housing prices and air pollution measures.Partial support for this research was provided by the Russell Sage Foundation. Thanks are due David Cordray, Heidi Hartman, and Larry Hedges of the Foundation's Meta-Analysis Panel for constructive comments, to Ray Palmquist for suggestions and assistance in assembling the results from his studies, to Rick Freeman and Tom Tietenberg and two anonymous referees for comments on the research, and to Barbara Scott for constructive editing of earlier drafts of this paper.     

The Economics Of Partner Out Trading in Sexual Markets

This paper derives a model of participation in what is commonly known as ‘adultery’. Arguably the best sex survey in the world is used to produce estimates of participation functions. The results show a great deal of support for bioeconomic models and reveal some interesting similarities and differences between the male and female equations. This revised version was published online in July 2006 with corrections to the Cover Date.     

Self-enforcing international environmental agreements with a stock pollutant

In this paper the stability of an International Environmental Agreement (IEA) among N identical countries that emit a pollutant are studied using a two-stage game. In the first stage each country decides noncooperatively whether or not to join an IEA, and in the second stage signatories jointly against nonsignatories determine their emissions in a dynamic setting defined in continuous time. A numerical simulation shows that a bilateral coalition is the unique self-enforcing IEA independently of the gains coming from cooperation and the kind of strategies played by the agents (open-loop or feedback strategies). We have also studied the effects of a minimum participation clause finding that for this case a self-enforcing IEA just consists of the number of countries established in the clause.JEL Classification: C73, D62, Q28 Corresponding author : Santiago J. RubioThis paper is based on chapter four of Begoña Casinos Ph. Dissertation. Financial support from the Instituto Valenciano de Investigaciones Económicas, the Ministerio de Ciencia y Tecnología under grant BEC2000-1432 and Fundación BBVAis gratefully acknowledged. We also appreciate the helpful comments of three anonymous referees, whose suggestions improved the paper. Regarding any remaining inadequacies, the usual caveat applies.