Home page of Tony Bellotti

UK retail banks – commission report

agbellotti — Wed, 13 Apr 2011 07:57:54 +0000

UK news story: Retail banks should be ring-fenced, says Independent Commission on Banking:
http://www.bbc.co.uk/news/business-13032403

Loss given default models incorporating macroeconomic variables for credit cards

agbellotti — Fri, 04 Mar 2011 12:18:59 +0000

New paper accepted for publication:

T. Bellotti and J. Crook. Loss given default models incorporating macroeconomic variables for credit cards, International Journal of Forecasting, Article in Press 2011, Accepted Manuscript (available online).

Abstract

Loss Given Default is an important measure of credit loss used by financial institutions to compute risk within credit portfolios, expected loss on individual loans and capital requirements. The Basel II Capital Accord gives banks the opportunity to calculate their own estimates of Loss Given Default. Based on UK data for major retail credit cards, we build several models of Loss Given Default based on account level data including Tobit, a decision tree model, a Beta and fractional logit transformation. However, we find that Ordinary Least Squares models with macroeconomic variables perform best for forecast of Loss Given Default at account and portfolio level on independent hold-out data sets. The inclusion of macroeconomic conditions in the model is important since it provides a means to model Loss Given Default in downturn conditions as required by Basel II and enables stress testing.

agbellotti — Thu, 02 Dec 2010 12:51:28 +0000

Center for Operations Research and Econometrics

CORE – ECORE

Econometrics seminar/LSM Finance seminar

Speaker:

Tony BELLOTTI, Imperial College London

Title:

Support vector machines in finance: application to the prediction of bank ratings

Date and place:

Wednesday, December 8, 2010 at 14h30

CORE – room b_135

Abstract:

We compare the ability of ordinal choice models and support vector machines to model and predict international bank ratings. Although support vector machines can identify the significant determinants of ratings we argue that ordered choice models are more reliable for this purpose. Our findings suggest that ratings reflect a bank’s financial position, the timing of when the rating was made and a bank’s country of origin. Accounting for country effects in the model was found to be particularly important not least because they substantially improve the predictive performance of the models. We find that support vector machines can produce considerably better in- sample predictions of international bank ratings than the standard method currently used for this purpose, ordered choice models. This appears to be due to the support vector machine’s ability to estimate a large number of country dummies unrestrictedly, which was not possible with the ordered choice models due to the small sample size. Given that the primary purpose of modelling ratings is prediction this is an important result.

(This is joint work with R. Matousek and C. Stewart.)

A paper is available for consultation – Please consult web page:

http://www.uclouvain.be/en-43677.html

Recession and recovery – hard times or smoke and mirrors?

agbellotti — Wed, 06 Oct 2010 12:54:42 +0000

Quick plug for this Royal Statistical Society sub-conference: Recession and recovery – hard times or smoke and mirrors?
27 October 2010 10:00 – 16:30 Includes lunch
Savoy Place, London WC2R 0BL

One week after details of widely anticipated major cuts in public spending are set out in the coalition government”s comprehensive spending review, this year”s Statistics User Forum conference will probe some of the fundamental statistical issues relating to both the recession and recovery. Our keynote speakers will review and comment on these issues in plenary sessions. In two parallel sessions, conference delegates will then be able to explore them further from either an economic or social perspective respectively. Among the questions being addressed are: > What lessons can be learned in considering the role of recession in long-term economic and social change? > Do we measure ”recession” in the ”right” way and are there data gaps that need to be filled? > To what extent did lack of the right data contribute to us being unprepared for the recession or not taking preventative action in advance? > In what ways has this recession differed from previous ones? > Is the public sector deficit being measured properly? > Do we have the right data and analytic tools to enable us to measure the impact of policies to reduce the deficit? > How will we know when recovery has arrived?

Lectureship announcement

agbellotti — Mon, 27 Sep 2010 08:24:52 +0000

I look forward to joining the Department of Mathematics at Imperial College London next week as a lecturer in statistics with special interest in Credit Scoring.

Are Rating Agencies’ Assignments Opaque? Evidence from International Banks

agbellotti — Mon, 27 Sep 2010 08:12:31 +0000

My latest paper, Are Rating Agencies’ Assignments Opaque? Evidence from International Banks, has been accepted for publication in Expert Systems with Applications. This is a joint work with Roman Matousek and Chris Stewart at London Metropolitan Business School whom I would like to thank for involving me in this interesting project.

Abstract

We compare the ability of ordered choice models and support vector machines to model and predict international bank ratings. Although support vector machines can identify significant determinants we argue that ordered choice models are more reliable for this. Our findings suggest that ratings reflect a bank’s financial position, the timing of rating assignment and a bank’s country of origin. Accounting for country effects substantially improves predictive performance. We find that support vector machines can produce considerably better predictions of international bank ratings than ordered choice models due to the formers ability to estimate a large number of country dummies unrestrictedly.

Flash crash

agbellotti — Fri, 20 Aug 2010 08:43:43 +0000

On 6 May 2010 world stock markets becoming overloaded with trades and markets went hay-wire. This is what can happen now the world markets are linked by high-frequency automated trading. Have a look at this article in the Wall Street Journal:
http://online.wsj.com/article/NA_WSJ_PUB:SB10001424052748704545004575353443450790402.html

Thanks to Pierre Giot for forwarding the article to me.

A simulation study of Basel II expected loss distributions for a portfolio of credit cards

agbellotti — Thu, 04 Mar 2010 08:54:55 +0000

The Journal of Financial Services Marketing has published my latest research paper “A simulation study of Basel II expected loss distributions for a portfolio of credit cards“.

Abstract

Credit scoring models have been used traditionally as the basis of decisions to reject or accept credit applications. They are also used to categorize applicants or existing accounts into risk groups. Based on estimates of probability of default (PD), the risk groups may seem well separated. However, by considering distributions on risk elements such as model estimation uncertainty, exposure at default and loss given default, a simulation approach is used to compute Basel II expected loss distributions for a portfolio of credit cards. These show that discrimination between risk groups is not as clear as is immediately suggested simply by PD estimates. Based on these distributions, we also show that measuring extreme credit risk with Value at Risk can lead to considerable underestimation if distributions on these risk elements are not entered into the computation.

LGD workshop

agbellotti — Thu, 10 Sep 2009 10:05:17 +0000

I will be attending a LGD workshop organized by Marfintel in Belgium on 8 October to present practical work and methods for estimating LGD with macroeconomic conditions.

Building models of consumer LGD with macroeconomic data to improve estimation and enable stress testing: (1) How to include macroeconomic variables in models of LGD, (2)Evidence showing improved estimation of LGD at account and aggregate levels, (3) Using the models for stress testing and estimation of downturn LGD.

Other speakers are Prof Bart Baesens and Dr David Martens.

Computational and Financial Econometrics conference 2009

agbellotti — Wed, 29 Jul 2009 09:21:33 +0000

I will be attending CFE 2009 in Cyprus and will present the following paper.

An exercise in stress testing for retail credit cards
Tony Bellotti and Jonathan Crook

Abstract

Stress tests are becoming increasingly important for evaluating consumer credit risk and they are recognised as a key tool in helping financial institutions make business strategy, risk management and capital planning decisions. We present a simulation-based method for stress testing using discrete survival analysis to build a dynamic model of default at the account level which includes macroeconomic conditions as risk factors. We discuss methods to generate plausible but extreme economic simulations within this framework and apply Monte Carlo simulation to compute empirical distributions of estimated default rates. We present experimental results for a large data set of UK credit card accounts that show that bank interest rates, production index, retails sales index and the FTSE 100 index are all statistically significant systemic risk factors for default. Stress tests using this model generate right skewed distributions of default rates and yield plausible extreme values of risk as measured by Value at Risk and expected shortfall. Finally we discuss statistical validation of stress tests through back-test of the loss distribution.