HOME > OUR THINKING > Credit Analysis > BLOG

Assessing Recoveries After Default: The Other Side Of The Coin

Risk managers and credit analysts at banks, asset managers, as well as analysts at non-financial corporations, all need to “look at both sides of the coin”: not only shall they assess risk of default, but also estimate what fraction of an exposure can be likely recovered in case of default, because both elements are important drivers of expected losses and there is empirical evidence that recovery rates decrease, when aggregate default rates increase in certain regions.1

This is especially true in the economic conditions that we are facing these days.

During the latest global economic crisis, central banks across the world have adopted accommodative monetary policies to sustain businesses and financial markets, by reducing interest rates to record lows (and in some instances extending them to negative territory) and/or embarking on multiple cycles of quantitative easing. As a result of these “ultra-favorable conditions”, corporate bonds (and loans) debt issuance has boomed to progressive highs, while default rates have been artificially stabilized or partially reduced, despite the incumbent economic downturn.2

As the United States and the European Union economies are slowly normalizing,3 central banks on both sides of the ocean have started (or are considering) to slowly increase interest rates and remove extraordinary credit support to businesses, potentially leading to a progressive increase of corporate default rates.

To complete the picture, let us not forget the imminent kick-off of new accounting standards that require calculation of future expected credit losses of performing loans and bonds, again making use of recovery rates! 4

Looking at “both sides of the coin” using S&P Global Market Intelligence’s tools and analytics

Our Credit Analytics suite offers statistical models that help to look at both sides of the coin in a simple, quick and scalable way: probability of default and credit scoring models (PD Model Fundamentals and CreditModelTM) and recovery models (LossStatsTM Model).

LossStats Model is a statistical tool that uses a compact set of highly-predictive, exposure-level and macro-economic factors to estimate the whole range of possible recovery outcomes in case of default, along with their likelihood. This is particularly important, since an average estimate is often not sufficient without understanding the spread of potential outcomes around that estimate.5 For example:

  • Using an average value can lead to underestimating or overestimating the actual expected losses and making overly aggressive or conservative decisions;
  • Ignoring the dispersion of the recovery outcomes can lead to incorrect risk pricing.

Figure 1: LossStats Model interface on Credit Analytics.

Probability distribution for loss given fault

Loss stats

Source: S&P Capital IQ platform (as of June 2017). For illustrative purposes only.

The model was trained on the historical database of observed recoveries, collected by S&P Global Ratings over several decades, for corporate bonds and bank loans in the US and Europe.6

Learn More About Credit Analysis
One of the challenges that we often hear from risk managers at large non-financial corporations is that “there are too many exposures, too few analysts and not enough time to review the whole portfolio of exposures more than once per year”.

Here, we propose a simplified, yet effective framework7 that will come in handy and speed up the portfolio review workflow: the idea is to organize all exposures into a “risk map”, looking at three components: default risk, recovery rate and exposure size.

Figure 2: Example of risk map for a simplified portfolio; the marker size is proportional to the exposure amount.

Example of risk map of a simplified portfolio updated
Source: S&P Global Market Intelligence (as of July 2017). For illustrative purposes only.8

By fine-tuning the risk tolerance thresholds according to their risk appetite (see dashed lines in Figure 2), risk managers can immediately identify the problematic exposures (highlighted in red, within the top-right quadrant in this example), and prioritize them for more frequent, targeted reviews. This may also serve as a starting point before including further risk/reward considerations and rebalancing the portfolio.  

Learn more about S&P Global Market Intelligence’s Credit Analytics models.

Copyright © 2017 S&P Global Market Intelligence, a division of S&P Global. All rights reserved. Legal Disclaimers.

1 Altman, E., B. Brady, A. Resti and A. Sironi (2003), “The Link between Default and Recovery Rates: Theory, Empirical Evidence and Implications,” Journal of Business (2005, vol. 78, issue 6, 2203-2228).

2 See for example: US Bond Market Issuance and Outstanding (xls) - annual, quarterly, or monthly issuance to February 2017 (issuance) and from 1980 to 2016 Q3 (outstanding), updated 03/07/17, from and “Addressing Market Liquidity: A broader perspective on today's Euro corporate bond market”, August 2016, from

3 Click here for more details.

4 The new accounting standards, IFRS9 and CECL, will become effective between 2018 and 2020, in countries following the International Financial Reporting Standard and in the United States of America, respectively.

5 The Loss Given Default (LGD) is equal to 1 - recovery rate. In some instances, LGD is negative, because the creditor can recover more than the initial exposure.

6 The historical recovery rates for bonds and loans in a selection of European countries was made available by S&P Global Ratings on an anonymized basis, and its use is restricted to modelling purposes only.

7 This is an evolution of the framework originally proposed by S&P Global Market Intelligence (2013).

8 S&P Global Ratings does not contribute to or participate in the creation of credit scores generated by S&P Global Market Intelligence. Lowercase nomenclature is used to differentiate S&P Global Market Intelligence credit scores from the credit ratings issued by S&P Global Ratings.

Subscribe to Insights
Oct 05, 2017
Jun 20, 2017
Jun 14, 2017
Credit Markets
Jun 05, 2017