A Macro-finance View on Stress Testing

Transcription

1 ECONOMIC & CONSUMER CREDIT ANALYTICS June, 2012 MOODY S ANALYTICS A Macro-finance View on Stress Testing Prepared by Andrea Appeddu Economist Juan M. Licari Senior Director José Suárez-Lledó Associate Director

2 A Macro-finance View on Stress Testing BY ANDREA APPEDDU, JUAN M. LICARI, AND JOSÉ SUÁREZ-LLEDÓ For most financial practitioners, stress testing is a must-do activity, even if it is not a regulatory requirement. Such stress testing encompasses a wide range of sophisticated and quantitative exercises, including assessments of market, credit and liquidity risks. This article discusses several approaches and outlines a foundation for a robust and consistent stress-testing framework. Risk-neutral valuation is widely used in academic and applied finance to model and simulate the behavior of security prices. However, this methodology has significant limitations. It is not well-suited to forecast out-of-sample data, and its model parameters lack clear economic interpretation. We argue below that a macro-finance approach may be more appropriate for forecasting and stress testing exercises, even though such models may generate larger pricing errors than others (such as option theory formula models). More important, option pricing models built as partial equilibrium modules cannot incorporate cross-asset correlation in a robust and transparent fashion. Additionally, they suffer from an assumption of market completeness. In contrast, macro-finance models clarify the role of real business cycles in asset price determination and can anchor expectations based on multiple sources of information. Critically, because they are based on general equilibrium, macro-finance models better capture feedback effects and are thus more suited for comprehensive and consistent stress-testing. Among the more commonly used models, three classes can be identified: large scale structural macro models, dynamic stochastic general equilibrium models, and vector autoregressive models. Structural macro models, developed by Laurence Klein (Nobel 1980), consist of large systems of aggregate demand and supply equations with explicit modeling of industries and macro sectors. They are not connected to economic theories of consumer decision and production. DSGE models, developed by Tom Sargent (Nobel 2011), Edward Prescott (Nobel 2004), and Robert Lucas (Nobel 1995), are modern macroeconomic models with micro foundations. They account for the intertemporal optimizing behavior of agents such as consumers, firms and government with rational expectations. These models also incorporate inter-temporal consistency and market-clearing conditions for both complete and incomplete markets. Central banks and other institutions often employ DSGE models, although they may limit them to key macroeconomic series. Vector autoregressive models and structural VARs were pioneered by Chris Sims (Nobel 2011). The systems of equations reflect data-driven models and are generally easier to implement and maintain than other models discussed. They relate economic variables of interest to their specific history and to exogenous variables. Although they do not seamlessly connect to economic theory, they often serve as workhorse models for short-term forecasting. They are also suitable for satellite models that incorporate financial or other variables with an effect on the macroeconomy. Each of these models provides baseline forecasts, simulations, system responses to one-variable shocks and other relevant post-estimation outputs. They also maintain internal consistency based on the system s equilibrium conditions. An advantage of the macro-finance approach versus others is that the macroeconomic variables can be explicitly modeled (as opposed to being inferred from simulations on asset pricing models). This is particularly important for practitioners who want to test their portfolios under particular events, something not easily performed with other models. They can build specific event-driven and data-driven scenarios reflecting plausible economic developments such as a double-dip recession or crisis in the Euro zone. These can be precisely calibrated such that a specific drop in GDP growth will translate to relevant unemployment, interest rates, housing prices and so on. These forecasts and scenarios can then be viewed in the context of a distribution of economic outcomes that are consistent with a particular narrative of risks facing the economy. (Generally the baseline forecast will be the mean forecast interpreted as the most likely outcome of the model.) Data-driven scenarios can be generated by drawing simulations around estimated DSGE or SVAR models constrained to follow relevant distributions, Monte Carlo simulations from normal distributions, or by drawing from the distribution of estimation errors in the historical sample (non-parametric bootstrapping). The latter captures richer dynamics in the data and should generate more interesting and accurate scenarios than some other approaches. MOODY S ANALYTICS / Copyright

3 ANALYSIS A Macro-finance View on Stress Testing Real GDP Simulations Annual Growth Rates, % over to 12 9 to to 9 6 to to 6 3 to to 3 0 to to 0-3 to -1.5 less than -3 Source: Moody s Analytics 2016Q1 2014Q1 2012Q1 Historic U.S. 10-Year Swap Rates % Baseline Moderate double-dip recession Fed stress scenario 0 Jan-00 Jan-02 Jan-04 Jan-06 Jan-08 Jan-10 Jan-12 Jan-14 Source: Moody s Analytics Stress-Testing Key Risk Parameters: Interest Rates, CDS Spreads and Ratings Migrations The methodology for stress-testing in a general equilibrium framework uses satellite models with explicit links to macro variables.»» Term Structure of Interest Rates The macro-finance literature surveyed in Rudebusch (2010) models interest rates for different maturities as a function of two latent components extracted with factor analysis techniques: the so-called level and slope of the curve. These components are modeled as autoregressive processes related to the macroeconomy through a term capturing expectations about the medium-term inflation target (for the level) and through terms capturing the reaction of monetary policy to short-run fluctuations in GDP and inflation (for the slope).»» Corporate CDS Spreads The empirical behavior of corporate CDS indexes has seldom been investigated in economic literature. A multivariate, parametric, semi-structural approach illustrated below uses explanatory variables based on modern portfolio theory. A principal component, the corporate credit factor, extracts relevant information from the covariance matrix of different CDS indexes, which are modeled as autoregressive processes exogenously driven by the CCF. Since CDS traders consider the information on corporate creditworthiness embedded in equity indexes, the CCF is specified as a function of a riskappetite factor that captures the systemic corporate risk. ITraxx Investment-grade CDS Index»» Ratings Migrations The evolution of ratings migration probabilities can be analyzed as a time series with three distinct groups being defined: (1) ratings that ITraxx Investment-Grade CDS Index keep their status (diagonal of a transition Index 250 matrix); (2) ratings Baseline that are downgraded 200 or upgraded by one Euro zone crisis notch; (3) ratings 150 that are downgraded 100 or upgraded by two or more notches. 50 Ratings in the diagonal are fairly stable 0 and can thus be Jun-04 Jun-06 Jun-08 Jun-10 Jun-12 Jun-14 Jun-16 Jun-18 Jun-20 modeled as the residual elements Source: Moody s Analytics from the one- and two (or more)-notch movements. One-notch movements are interrelated to a high degree and are modeled as a constrained VAR process. A set of exogenous variables may be included to account for the economic cycle and used to model shocks to the system based on various scenarios. The key feature of movements of two (or more) notches is that their observations are either zero or small positive values, which suggests the use of discrete-choice models. This dual behavior is captured by means of a two-part model. The first is designed to predict the occurrence of zeroes; the second is conditional on positive observations. The final model for these transitions is the joint probability of the two parts. Concluding Remarks Maintaining consistent scenarios and accounting for cross-correlation across assets requires a fully specified general equilibrium model. Since stress testing exercises can suffer from correlation biases, macro-financial models may provide a more appropriate framework. To facilitate forecasting and stress testing, we propose using consistent and comprehensive macro models and linking their tail distributional outcomes to relevant risk parameters. Depending on the nature of the asset class being analyzed, additional satellite models may be relevant. Examples include equity indexes, FX rates, and historical and option-implied asset volatilities. MOODY S ANALYTICS / Copyright

4 AUTHOR BIO About the Authors Andrea Appeddu Andrea Appeddu is an associate economist in the London office of Moody s Analytics. He covers economic analysis for several European countries. Before joining Moody s Analytics, Andrea worked as an associate at IHS Global Insight, where he analyzed and forecast for ferrous and nonferrous metals and the capital goods sector. Andrea holds an MSc in economics and finance from Warwick Business School. Juan M. Licari Juan M. Licari is a director at Moody s Analytics. Dr. Licari is a member of the Credit Analytics group and specializes in financial economics. Juan leads consulting projects with major industry players, builds econometric tools to model credit phenomena, and has implemented several stresstesting platforms to quantify portfolio risk exposure. He has a leading role in the development and implementation of credit solutions and is actively involved in communicating these to the market. Dr. Licari holds a PhD and an MA in economics from the University of Pennsylvania and graduated summa cum laude from the National University of Cordoba in Argentina. José Suárez-Lledó Jose Suarez-Lledo is an associate director at Moody s Analytics based in London. As part of the Credit Analytics team he designs retail and corporate credit models as well as macro-econometric models for key economic and financial variables. Before joining Moody s Analytics, Jose held a research position at the Universidad Autonoma de Barcelona, where he developed models for illiquid financial markets and the dynamics of asset prices and credit. Jose holds a PhD and an MA in Economics from the University of Pennsylvania. More on Global Economies from Moody s Analytics MOODY S CREDITCYCLE TM Integrating regional economics with industry-leading consumer credit forecasting & stress testing. GLOBAL ALTERNATIVE SCENARIOS Meet regulatory requirements, evaluate the impact of shocks, expose vulnerabilities, and develop strategic business plans. U.S./CANADA EMEA +44 (0) ASIA/PACIFIC OTHER LOCATIONS help@economy.com MOODY S ANALYTICS / Copyright

5 About Moody s Analytics Economic & Consumer Credit Analytics Moody s Analytics helps capital markets and credit risk management professionals worldwide respond to an evolving marketplace with confidence. Through its team of economists, Moody s Analytics is a leading independent provider of data, analysis, modeling and forecasts on national and regional economies, financial markets, and credit risk. Moody s Analytics tracks and analyzes trends in consumer credit and spending, output and income, mortgage activity, population, central bank behavior, and prices. Our customized models, concise and timely reports, and one of the largest assembled financial, economic and demographic databases support firms and policymakers in strategic planning, product and sales forecasting, credit risk and sensitivity management, and investment research. Our customers include multinational corporations, governments at all levels, central banks and financial regulators, retailers, mutual funds, financial institutions, utilities, residential and commercial real estate firms, insurance companies, and professional investors. Our web and print periodicals and special publications cover every U.S. state and metropolitan area; countries throughout Europe, Asia and the Americas; and the world s major cities, plus the U.S. housing market and other industries. From our offices in the U.S., the United Kingdom, and Australia, we provide up-to-the-minute reporting and analysis on the world s major economies. Moody s Analytics added Economy.com to its portfolio in Its economics and consumer credit analytics arm is based in West Chester PA, a suburb of Philadelphia, with offices in London and Sydney. More information is available at , Moody s Analytics, Inc. and/or its licensors and affiliates (together, Moody s ). All rights reserved. ALL INFORMATION CONTAINED HEREIN IS PROTECTED BY COPYRIGHT LAW AND NONE OF SUCH INFORMATION MAY BE COPIED OR OTHERWISE REPRODUCED, REPACKAGED, FURTHER TRANSMITTED, TRANSFERRED, DISSEMINATED, REDISTRIBUTED OR RESOLD, OR STORED FOR SUBSEQUENT USE FOR ANY PURPOSE, IN WHOLE OR IN PART, IN ANY FORM OR MANNER OR BY ANY MEANS WHATSOEVER, BY ANY PERSON WITHOUT MOODY S PRIOR WRITTEN CONSENT. All information contained herein is obtained by Moody s from sources believed by it to be accurate and reliable. Because of the possibility of human and mechanical error as well as other factors, however, all information contained herein is provided AS IS without warranty of any kind. Under no circumstances shall Moody s have any liability to any person or entity for (a) any loss or damage in whole or in part caused by, resulting from, or relating to, any error (negligent or otherwise) or other circumstance or contingency within or outside the control of Moody s or any of its directors, officers, employees or agents in connection with the procurement, collection, compilation, analysis, interpretation, communication, publication or delivery of any such information, or (b) any direct, indirect, special, consequential, compensatory or incidental damages whatsoever (including without limitation, lost profits), even if Moody s is advised in advance of the possibility of such damages, resulting from the use of or inability to use, any such information. The financial reporting, analysis, projections, observations, and other information contained herein are, and must be construed solely as, statements of opinion and not statements of fact or recommendations to purchase, sell, or hold any securities. NO WARRANTY, EXPRESS OR IMPLIED, AS TO THE ACCURACY, TIMELINESS, COMPLETENESS, MERCHANTABILITY OR FITNESS FOR ANY PARTICULAR PURPOSE OF ANY SUCH OPINION OR INFORMATION IS GIVEN OR MADE BY MOODY S IN ANY FORM OR MANNER WHATSOEVER. Each opinion must be weighed solely as one factor in any investment decision made by or on behalf of any user of the information contained herein, and each such user must accordingly make its own study and evaluation prior to investing. MOODY S ANALYTICS / Copyright

6 CONTACT US For further information contact us at a location below: U.S./CANADA EMEA ASIA/PACIFIC OTHER LOCATIONS us: help@economy.com Or visit us: Copyright 2012, Moody s Analytics, Inc. All Rights Reserved.