Research Papers

Built from decades of experience utilizing extensive datasets and proven models, EDF measures have been validated on defaults and credit spreads and have become the de facto standard for lenders and investors worldwide.

The RiskCalc™ Banks v4.0 Model is intended for assessing the probability of default (PD) for banks across different geographies and regulatory environments. The model provides a unified framework to assess bank risk across different countries and regions, as well as different economic cycles. The one-year model is based upon a set of well-defined and ready-to-calculate financial ratios that effectively measure bank profitability, leverage, liquidity, growth, and asset quality. The five-year model combines these ratios with a measure derived from an economic capital framework based upon portfolio theory. Specifically, this measure captures the unexpected loss of a bank’s loan portfolio relative to its loss-absorbing capital.

Authors: Yanruo Wang, Douglas Dwyer, Janet Yinqing Zhao
Date: July 1, 2016
Download PDF
With this insight, we constructed the RiskCalc™ Banks v4.0 Model, intended for assessing the probability of default (PD) for banks across different geographies and regulatory environments. The model provides a unified framework to assess bank risk across different countries and regions, as well as different economic cycles. The one-year model is based upon a set of well-defined and ready-to-calculate financial ratios that effectively measure bank profitability, leverage, liquidity, growth, and asset quality. The five-year model combines these ratios with a measure derived from an economic capital framework based upon portfolio theory. Specifically, this measure captures the unexpected loss of a bank’s loan portfolio relative to its loss-absorbing capital.

Authors: Yanruo Wang, Douglas Dwyer, Janet Yinqing Zhao
Date: July 25, 2014
Download PDF
To help our clients build benchmark commercial and industrial (C&I) loss models for the Federal Reserve’s Comprehensive Capital Analysis and Review (“CCAR”)/DFAST exercises, we have developed an approach designed specifically to calculate provisions for losses of C&I portfolios. Our approach utilizes Moody’s Analytics probability of default (PD), loss given default (LGD), and exposure at default (EAD) econometric models, which are intuitive, parsimonious, make economic sense, and have good statistical fit. We construct these models using our public EDF credit measures, RiskCalc™ private firm EDF credit measures, and Moody’s Default & Recovery Database and Credit Research Database.

Authors: Nan Chen, Jian Du, Heather Russell, Douglas Dwyer, Jing Zhang, Zhong Zhuang
Date: July 1, 2014
Download PDF
In this paper, we detail a RiskCalc™ Stress Testing Model (ratio-based approach), based upon economic and accounting principles. Our simple, yet fundamental, model assumptions make the framework adaptable to many uses, including: loss forecasting, pro forma analysis, stress testing, as a challenger or benchmark model, and for customized scenario analysis.

Authors: Douglas Dwyer, Yinqing (Janet) Zhao, Monalisa Sen
Date: July 25, 2014

Download PDF

Using a unique data set pooled from multiple US financial institutions, we empirically study the credit line usage of middle-market corporate borrowers. We find that defaulted borrowers draw down more of their lines than non-defaulted borrowers. They also increase their usage when approaching default. Riskier borrowers tend to utilize a higher percentage of their credit lines. We find that firms rated as “pass” grade by the lender draw down the credit lines more than those rated below “pass” grade. Usage ratios also vary by collateral type, commitment size, loan purpose type, and prior quarter’s usage. Further, we find evidence that usage ratios are higher during economic downturns. The evidence is stronger for non-defaulted firms than for defaulted firms. Taken together, these results suggest that credit line usage is a function of both borrowers’ characteristics and banks’ monitoring and control of these lines.  

Author: Janet Yinqing Zhao, Douglas W. Dwyer, Jing Zhang
Date: June 23, 2014

Download PDF
On any given day, credit analysts monitor multiple names. Some names have been reviewed recently, but not all. Some names represent large exposures, while others are small. Some are known high credit risks, while others are low credit risks. The risk profile of some exposures may have changed recently, while others remain unchanged. Exposures to some names can be meaningfully reduced, while some names are more difficult to hedge. Which names should the analyst review first? Next? Take action on? Ultimately, how does the institution value a credit review?

Information’s value lies in its potential to change actions. This paper focuses on measuring the economic value of a review and deriving an optimal review policy. We develop a framework for constructing a review strategy as well as its associated costs and benefits. For specific assumptions, our methodology derives an optimal monitoring strategy. The actual value of the information depends upon the risk return profile of the loan, the cost of obtaining the information, and the ability of the lender to reduce their exposure to the borrower if the review is negative. Timely information regarding credit risk is especially valuable when credit risk is elevated (the return may or may not justify the risk), and the bank can meaningfully reduce their exposure to the borrower. 

Author: Douglas W. Dwyer
Date: September 30, 2013
Download PDF
In this article, we assess our December 2012 predictions for expected C&I loss rates for the 2013 CCAR in light of the Fed’s recently announced results. Moody’s Analytics used Stressed EDF metrics based on the Fed’s supervisory severely adverse scenario as the PD in expected loss calculations to predict the projected loss rate for C&I loans under the 2013 CCAR. Our projection of 6.7% for 17 BHCs in the aggregate compares favorably to the Fed’s estimated post-stress loss rate of 6.9% for the same set of BHCs (or 6.8% for all 18 participating BHCs).

Author: Danielle H. Ferry
Date: March 27, 2013
Download PDF

Olam’s probability of default has jumped significantly since the start of year, from 0.2% to its current level of 1.17%, suggesting heightened risk of a credit event. The firm’s low margins, increasing debt levels to fund agricultural investments, liquidity concerns, and a deteriorating market capitalization, all indicate that the firm’s probability of a credit event has increased. The firm’s EDF measure has underperformed its industry peer group, which according to Moody’s Analytics’ research is an early warning signal of default risk.

Author: Irina Makarova
Date: December 19, 2012

Download PDF

Earlier this year Sharp’s EDF measure began to trend in a range suggesting very heightened risk of default, rising from 1.21% In January 2012, to 20.85% as of November 15, 2012. The firm’s weak liquidity, substantial operating losses, and heightened EDF measure – equivalent to a Ca implied rating – indicates that the likelihood of a credit event in the near future remains high.

Author: Irina Makarova
Date: November 11, 2012

Download PDF

Stressed EDFTM (Expected Default Frequency) credit measures are one-year, firm-level default probabilities conditioned on a range of macroeconomic scenarios. Stressed EDF measures can substitute for probability of default measures whenever it is necessary to assess credit risk in alternative macroeconomic situations. Two examples of such applications are Basel II/III capital and loan loss provision calculations. Stressed EDF measures bring together macroeconomic scenarios from Moody’s Analytics’ economic forecasting unit and the public firm EDF model, the industry-leading structural credit risk model for default probability. Our unique approach affords users a rigorous means to evaluate the impact of plausible macro-financial events on credit risk at both the firm and portfolio levels. Stressed EDF metrics – for a baseline, one upside, and three downside economic scenarios – are available at a monthly frequency, over a five-year forecast horizon, and are updated each month. This paper reviews the fundamental methodology of the Stressed EDF model for Western Europe, which includes firms in Austria, Belgium, Denmark, Finland, France, Germany, Greece, Italy, Ireland, Luxembourg, the Netherlands, Norway, Portugal, Spain, Sweden, Switzerland, and the United Kingdom.

Authors: Danielle H. Ferry, Tony Hughes, Min Ding
Date: October 29, 2012
Download PDF

In our second Alpha Factor Portfolio Insight report in which we address the question of liquidity constrained investing we impose a significant constraint on the US dollar and euro investment grade and dollar high yield model portfolios in the form of only “buying” liquid issues. Even with this limitation the Alpha Factor framework still produces healthy portfolio excess returns with high Sharpe Ratios. We believe that this reinforces that conclusion that Alpha Factors can play a valuable role in fund managers’ investment and portfolio surveillance processes. We extend the analysis by also constraining our euro and dollar investment grade model portfolios by sector, so they match the breakdown of their respective indices between financial institutions and industrials/utilities on a market weight basis. Regardless of the combination of investment restraints, the Alpha Factor-based portfolios all exhibit significantly better risk and return characteristics than their respective indices.
 
Authors: David Munves & Yukyung Choi 
Date: October 24, 2012
Download PDF

Through much of its history Best Buy was considered one of the most successful retail stores in the US. However, since 2010 the electronics retailer has faced business and financial challenges that are placing increasing pressure on its credit quality.
 
Author: Irina  Makarova 
Date: October 2, 2012
Download PDF

Since Groupon reported a fourth-quarter 2011 loss of USD 9.8 million on an adjusted basis, Moody’s Analytics’ public EDF measure had increased from 0.04% to 2.71% as of August 16, 2012.
 
Author: Irina  Makarova 
Date: August 18, 2012
Download PDF

Peugeot S.A.’s EDFTM (Expected Default Frequency) credit metrics have deteriorated dramatically over the past 12 months. The company’s one-year EDF measure, for example, increased from 0.6% in July 2011 to 8.57% today.
 
Author: Irina  Makarova 
Date: August 9, 2012
Download PDF

RIM does not have traded bonds or CDS from which to observe credit spreads, and is not rated by Moody’s Investors Service. However, Moody’s Analytics’ public EDF measure effectively captures and quantifies changes in the company’s credit risk.
 
Author: Irina  Makarova 
Date: June 22, 2012
Download PDF

The EDF measure for Shanghai Zendai has largely tracked its peer group, China Real Estate Group, prior to March 2011. But since then its EDF has risen at a much faster pace than its peers.
 
Author: Irina  Makarova 
Date: June 13, 2012
Download PDF

The EDF measure for Shandong Helon Co.’s has signaled a high level of default risk since the time of the financial crisis in 2008. In 2010 its EDF measure began to trend in a range suggesting heightened risk of default, and in June 2011 its EDF jumped from 2.6% to over 7%. Its EDF measure jumped again in April 2012 to over 10%.
 
Author: Irina  Makarova 
Date: May 31, 2012
Download PDF

Residential Capital Llc is one of the last subprime mortgage lenders of the early 2000s to file Chapter 11. The heightened level of its CDS-implied EDF measure reflects the company’s inability to repay debt taken on to finance the issuance of home mortgages and continue its operations in the wake of the global financial crisis.
 
Author: Irina  Makarova 
Date: May 22, 2012
Download PDF

Bankia SA’s one-year probability of default jumped sharply in May, from 0.45% at the start of the month to 2.24% as of May 24.
 
Authors:Irina  Makarova 
Date: May 25, 2012
Download PDF

Elpida Memory Inc. does not have traded bonds or CDS from which to observe credit spreads, and is not rated by Moody’s Investors Service. However, Moody’s Analytics’ public EDF measure effectively captured and quantified changes in the company’s risk of default.
 
Authors:Irina  Makarova 
Date: May 21, 2012
Download PDF

The GAME Group plc does not have any credit risk measures like bond or CDS spreads available, and it is not rated by Moody’s Investors Service. However, the public EDF measure is able to capture and quantify changes in the company’s risk of default.
 
Author: Irina Makarova
Date: April 25, 2012
Download PDF

The paper presents an updated and expanded version of the public firm EDF methodology on both an intuitive and a rigorous quantitative basis. We also review the historical performance of the model, and discuss recently developed EDF extensions, namely CDS-implied EDF measures, Through-the-Cycle EDF measures, and Stressed EDF measures, that were developed for applications such as calculating regulatory capital and economic stress testing.

Author: Zhao Sun; David Munves; David T. Hamilton, PhD
Date: June 28, 2012

Download PDF

In this Viewpoints, we briefly recount the methodology used to construct Stressed EDF measures and then highlight some of their strengths for macroeconomic stress testing. History shows that Stressed EDF measures are capable of accurately predicting credit risk under severe economic onditions. The degree of granularity afforded by these firm-level PDs increases flexibility and improves precision in credit analytics where portfolio composition is important. We also show that Stressed EDF measures can be used to simulate the macroeconomic stress testing exercises of supervisory authorities, such as the Federal Reserve’s Comprehensive Capital Analysis and Review (CCAR).

Author: Danielle Ferry
Date: June 18, 2012

Download PDF

Econometricians need to build models for forecasting or prediction, strucutural analysis or hypothesis testing, and policy or shock analysis. Each of these applications has an underlying loss or risk function that governs how the model should be built and the properties the preferred model specification should retain. In this paper, we condiser stress-testing, which we view as somewhat distinct from other types of econometric analysis.

Author: Tony Hughes
Date: June 7, 2012

Download PDF

In this paper, we describe the modeling methodology Moody’s Analytics employs to generate Stressed EDF measures for North American firms. The methodology can be best described as a macroeconomic-based approach to projecting default probabilities which are themselves the product of an asset value model that utilizes equity market information and financial statement data.

Author: Danielle H. Ferry, Tony Hughes, and Min Ding
Date: May 25, 2012

Download PDF

We develop a model-based approach to constructing investment grade and high yield corporate bond portfolios that both outperform their respective prevalent benchmark indices and popular Exchange Traded Funds (ETFs) with better risk-return profiles, i.e., higher returns with lower or similar risk. More importantly, the outperformance is obtained after controlling for credit risk, duration risk, and downside risk. The outperformance is robust across a number of different specifications of the strategy and over time. We also achieve outperformance using relatively smaller, more realistic portfolios. These model portfolios can be potentially converted into new fixed income indices or ETFs. Our model-based approach utilizes Moody’s Analytics’ EDF™ (Expected Default Frequency) credit measures and Fair-value Spread (FVS) valuation framework as powerful tools to control for credit risk and to exploit relative value in the bond market.

Author: Zan Li, Jing Zhang, Christopher Crossen
Date: April 19, 2012

Download PDF

The recovery amount from a defaulted financial instrument can be measured by either ultimate recovery (UR) or post-default value (PDV). While these two measures have the same expected value, UR variance is usually greater than PDV variance. In most cases, the appropriate measure of variance to use in an economic capital framework is somewhere between UR and PDV variance, but is much closer to PDV variance. This paper proposes two approaches to estimating recovery variance based on UR variance and PDV variance, respectively. We find that the impact of recovery variance parameterization on economic capital can be non-trivial, especially when PD-LGD correlation (PLC) is accounted for. In this paper we demonstrate that two common parameterizations can result in capital differences of over 15% in a realistic setting.

Authors: Amnon Levy, Qiang Meng, Douglas Dwyer and Irina Korablev
Date: November 17, 2011

Download PDF

In this paper, we validate the performance of the Moody’s Analytics Public EDF™ (Expected Default Frequency) model for global financial firms during the last decade, including the recent credit crisis and its recovery period. We divide the decade into two sub-periods: an early period, 2001–2007, and a later one, 2008–2010, and then compare the model’s performance during these two periods. We focus on the model’s ability to prospectively differentiate between defaulters and non-defaulters, its comparison to agency credit ratings, the timeliness of its default prediction, and its accuracy of levels.

Author: Christopher Crossen and Sue Zhang
Date: October 2011

Download PDF

We validate the performance of the Moody’s Analytics Public EDF™ (Expected Default Frequency) model for Asian-Pacific and Japanese corporate firms during the last decade, including the recent credit crisis and its recovery period. We divide the decade into two sub-periods: an early period, 2001–2007, and a later one, 2008–2010, and then compare the model’s performance during these two periods. We focus on the model’s ability to prospectively differentiate between defaulters and non-defaulters, the timeliness of its default prediction, and its accuracy of levels.

Author: Christopher Crossen and Xu Zhang
Date: October 2011

Download PDF

We validate the performance of the Moody’s Analytics Public EDF™ (Expected Default Frequency) model for European corporate firms during the last decade, including the recent credit crisis and its recovery period. We divide the decade into two sub-periods: an early period, 2001–2007, and a later one, 2008–2010, and then compare the model’s performance during these two periods. We focus on the model’s ability to prospectively differentiate between defaulters and non-defaulters, the timeliness of its default prediction, and its accuracy of levels.

Author: Christopher Crossen and Sue Zhang
Date: October 2011

Download PDF
Print