Measuring Credit Risk of a Bank's Corporate Loan Portfolio Using Advanced Internal Ratings Base Approach

Abstract

The use of technical and advanced approaches in the measurement of credit risk of banks' portfolios has nowadays become a very hot issue. The most recent technical report issued by the Basel Committee in May 2003 has concentrated heavily on the measurement of credit risk using either foundation or advanced Internal Ratings Base (IRB) approaches. This empirical research study attempts to measure credit risk of a bank's corporate loan portfolio, including firms from 10 different Turkish sectors. The monthly observations of the total amount of corporate loans and the total amount of corporate loans at default across various sectors are downloaded from the web page of Central Bank of Turkey (CBT) in a period of 1999-2002. This period covers 47 monthly observations since CBT has captured sectoral corporate loans beginning of 1999. Therefore, the observed sectoral default rates are needed to be simulated to obtain a nicely shaped distribution. Monte Carlo simulation is applied for 1,000 times. Based on the simulated default rates, the expected sectoral default rates are computed. Next, a credit quality rating scale is fitted into sectoral default rates distributions. Finally, the sectoral weights in the whole loan portfolio are multiplied by the expected sectoral default rates matrix, considering cross-sectoral correlations to get the total amount of the bank's credit risk and capital requirement. It is assumed that sectoral monthly default rates are a good representative of the default risk of a sample bank's corporate loan portfolio since no publicly available data on any particular bank's corporate loan portfolio composition exists. Nevertheless, this research may be a good application for measuring the credit risk of banks' corporate loan portfolios using advanced IRB approach.     

Forecasting and stress testing credit card default using dynamic models

We present discrete time survival models of borrower default for credit cards that include behavioural data about credit card holders and macroeconomic conditions across the credit card lifetime. We find that dynamic models which include these behavioural and macroeconomic variables provide statistically significant improvements in model fit, which translate into better forecasts of default at both account and portfolio levels when applied to an out-of-sample data set. By simulating extreme economic conditions, we show how these models can be used to stress test credit card portfolios.     

Comparisons of control schemes for monitoring the means of processes subject to drifts

Although statistical process control (SPC) techniques have been focused mostly on detecting step (constant) mean shift, drift which is a time-varying change frequently occurs in industrial applications. In this research, for monitoring drift change, the following five control schemes are compared: the exponentially weighted moving average (EWMA) chart and the cumulative sum (CUSUM) charts which are recommended detecting drift change in the literature; the generalized EWMA (GEWMA) chart proposed by Han and Tsung (2004) and two generalized likelihood ratio based schemes, GLR-S and GLR-L charts which are respectively under the assumption of step and linear trend shifts. Both the asymptotic estimation and the numerical simulation of the average run length (ARL) are presented. We show that when the in-control (IC) ARL is large (goes to infinity), the GLR-L chart has the best overall performance among the considered charts in detecting linear trend shift. From the viewpoint of practical IC ARL, based on the simulation results, we show that besides the GLR-L chart, the GEWMA chart offers a good balanced protection against drifts of different size. Some computational issues are also addressed.     

Comparisons of linear regression and survival analysis using single and mixture distributions approaches in modelling LGD

Estimating the recovery rate and recovery amount has become important in consumer credit due to the new Basel Accord regulation and the increase in the number of defaulters as a result of the recession. We compare linear regression and survival analysis models for modelling recovery rates and recovery amounts, in order to predict the loss given default (LGD) for unsecured consumer loans or credit cards. We also look at the advantages and disadvantages of using single and mixture distribution models for estimating these quantities.     

Optimization tools for design and marketing of new/improved products using the house of quality

Quality function deployment (QFD) has helped many firms realize significant reduction in product design costs and development time. The QFD process includes ranking customer preferences, rating the competitors, and parts deployment for the new/improved product. Prior to this research, such activities have been performed based on expert opinion, or the “best-in-class” approach. We develop and solve optimization models for the identification of consensus rankings and ratings, that take into account the priorities and perceptions of all the customers in a target market. Then, based on the consensus rankings, we identify a parts mix for the new/improved product that satisfies a budget constraint and matches or exceeds the performance expectations of all customers surveyed in the target market. Finally, we show how the QFD charts can be used to identify competitors that are falsely perceived as superior, as well as areas where the firm's marketing strategies have had the desired effects. Such insights are useful in developing the future marketing strategy of the firm.     

The economic value of patent portfolios

Patent holders may choose to protect innovations with single patents or to develop portfolios of multiple, related patents. We propose a decision‐making model in which patent holders allocate resources to either expanding the number of related patents or investing in higher value of patents in the portfolio. We estimate the derived value equation using portfolio value data from an inventor survey at the level of individual inventions rather than the firm as a whole. We find that investments in individual inventions exhibit diminishing returns, and that a good part of the value of a portfolio depends on adding new patented inventions. Also, while diminishing returns to individual inventions are stable across subsamples, the returns to portfolio size vary between complex and discrete industries, and between inventions that are science‐based or driven by customer information. When firms seek to strengthen appropriability, the returns to an increase of portfolio size are not different from the sample average. Thus, a higher number of inventions in a portfolio may reflect both stronger appropriability via patents and genuine creation of value.     

Sample size and the Binomial CUSUM Control Chart: the case of 100% inspection

The Binomial CUSUM is used to monitor the fraction defective (p) of a repetitive process, particularly for detecting small to moderate shifts. The number of defectives from each sample is used to update the monitoring CUSUM. When 100% inspection is in progress, the question arises as to how many sequential observations should be grouped together in forming successive samples. The tabular form of the CUSUM has three parameters: the sample size n, the reference value k, and the decision interval h, and these parameters are usually chosen using statistical or economic-statistical criteria, which are based on Average Run Length (ARL). Unlike earlier studies, this investigation uses steady-state ARL rather than zero-state ARL, and the occurrence of the shift can be anywhere within a sample. The principal finding is that there is a significant gain in the performance of the CUSUM when the sample size (n) is set at one, and this CUSUM might be termed the Bernoulli CUSUM. The advantage of using n=1 is greater for larger shifts and for smaller values of in-control ARL. First version: September 1998/Third revision: September 2000     

商业银行的企业违约概率度量方法发展沿革及比较研究

违约概率度量是指对可能引起信用风险的因素进行定性分析、定量计算,以测量借款人的违约概率,为贷款决策提供依据。国际上违约概率度量领域的研究和实际应用,有从主观判断分析、财务比率分析、统计分析转向人工智能、以资本市场理论和信息科学为支撑的方法等动态计量分析方法为主的发展趋势。本文对商业银行的企业违约概率度量方法发展沿革进行了比较研究,并对违约概率度量方法的国内研究作了综合评述。     

The long-run performance of initial public offerings: Stochastic dominance criteria

We examine the long-run performance of initial public offerings (IPOs) using the idea of stochastic dominance. The analysis is a first attempt using a non-event study methodology to evaluate long-horizon performance. We find that there is no first-order stochastic dominance relation between the IPO portfolio and the benchmark of a broad index or a portfolio including either small size or low book-to-market stocks. However, those benchmarks second-order stochastically dominate the IPO portfolio. When using a portfolio including both small size and low book-to-market stocks as benchmark, there is a clear dominance of the IPO portfolio over the benchmark for both orders. Our findings generally imply that the question of assessing portfolio performance between IPO firms and benchmark portfolios depends critically on the specific construction or the cumulative distribution function of the benchmark portfolios. The empirical results also potentially explain the extent of sample dependent results in the literature.     

Dominant tracking signals

The objective of this research is to find the best methods of automatically monitoring an exponentially smoothed forecast. Three conclusions are drawn. First, previously used performance measures are inadequate. As a consequence, currently available control limits can give false alarm rates that are substantially different than advertised. Second, two commonly used tracking signals can be substantially improved by choice of smoothing parameters. Finally, when measured by the new criteria proposed in this paper, the smoothed error tracking signal is substantially better than the unweighted sum of errors (CUSUM) method.     

An asset protection scheme for banks exposed to troubled loan portfolios

We examine a specific portfolio credit derivative, an Asset Protection Scheme (APS), and its applicability as a discretionary regulatory tool to reduce asymmetric information and help restore the capital base of troubled banks. The APS can be a fair-valued contract with an appropriate structure of incentives. We apply two alternative multivariate structural default risk models: the classical Gaussian Merton model and a model based on Normal Inverse Gaussian processes. Using a data set on annual farm level data from 1996 to 2009, we use the Danish agricultural sector as a case study and price an APS on an agricultural loan portfolio. We compute the economic capital for this loan portfolio with and without an APS. Moreover, we illustrate how model risk in the form of parameter uncertainty is reduced when an APS is attached to the loan portfolio.     

Default models based on scale mixtures of Marshall-Olkin copulas: properties and applications

We present a unification of the Archimedean and the Lévy-frailty copula model for portfolio default models. The new default model exhibits a copula known as scale mixture of Marshall-Olkin copulas and an investigation of the dependence structure reveals that desirable properties of both original models are combined. This allows for a wider range of dependence patterns, while the analytical tractability is retained. Furthermore, simultaneous defaults and default clustering are incorporated. In addition, a hierarchical extension is presented which allows for a heterogeneous dependence structure. Finally, the model is applied to the pricing of CDO contracts. For this purpose, an efficient Laplace transform inversion approach is developed. Supporting a separation of marginal default probabilities and dependence structure, the model can be calibrated to CDS contracts in a first step. In a second step, the calibration of several parametric families to CDO contracts demonstrates a good fitting quality, which further emphasizes the suitability of the approach.     

Performance of cumulative sum schemes for monitoring low count-level processes

In many instances attributes data must be used to monitor a manufacturing (or other) process that, in normal conditions, should operate at very low count levels for defects. Lucas (1989) has directed attention to this problem, and has investigated a new control scheme for low count-level processes. An alternative scheme is proposed, based on a Cumulative Sum (CUSUM) of the number (termed Run-Length) of successive samples having zero count-levels between samples having at least one count. Using the criterion of Average Run Length (the average number of samples until a signal is generated) comparisons of the Lucas scheme and the Run-Length CUSUM scheme indicate that ARL values for the Run-Length CUSUM can be up to 50% lower.     

Surveillance of the mean behavior of multivariate time series

In this paper several cumulative sum (CUSUM) charts for the mean of a multivariate time series are introduced. We extend the control schemes for independent multivariate observations of crosier [ Technometrics (1988) Vol. 30, pp. 187–194], pignatiello and runger [ Journal of Quality Technology (1990) Vol. 22, pp. 173–186], and ngai and zhang [ Statistica Sinica (2001) Vol. 11, pp. 747–766] to multivariate time series by taking into account the probability structure of the underlying stochastic process. We consider modified charts and residual schemes as well. It is analyzed under which conditions these charts are directionally invariant. In an extensive Monte Carlo study these charts are compared with the CUSUM scheme of theodossiu [ Journal of the American Statistical Association (1993) Vol. 88, pp. 441–448], the multivariate exponentially weighted moving-average (EWMA) chart of kramer and schmid [ Sequential Analysis (1997) Vol. 16, pp. 131–154], and the control procedures of bodnar and schmid [ Frontiers of Statistical Process Control (2006) Physica, Heidelberg]. As a measure of the performance, the maximum expected delay is used.     

Measuring connectedness of euro area sovereign risk

We introduce a method for measuring the default risk connectedness of euro zone sovereign states using credit default swap (CDS) and bond data. The connectedness measure is based on an out-of-sample variance decomposition of model forecast errors. Due to its predictive nature, it can respond to crisis occurrences more quickly than common in-sample techniques. We determine the sovereign default risk connectedness using both CDS and bond data in order to obtain a more comprehensive picture of the system. We find evidence that there are several observable factors that drive the difference between CDS and bonds, but both data sources still contain specific information for connectedness spill-overs. In general, we can identify countries that impose risk on the system and the respective spill-over channels. Our empirical analysis covers the years 2009–2014, such that the recovery paths of countries exiting EU and IMF financial assistance schemes and the responses to the ECB’s unconventional policy measures can be analyzed.     

AFT regression-adjusted monitoring of reliability data in cascade processes

Today’s competitive market has witnessed a growing interest in improving the reliability of products in both service and industrial operations. A large number of monitoring schemes have been introduced to effectively control the reliability-related quality characteristics. These methods have focused on single-stage processes or considered quality variables which are independent. However, the main feature of multistage processes is the cascade property which needs to be justified for the sake of optimal process monitoring. The problem becomes complicated when the presence of censored observations is pronounced. Therefore, both the effects of influential covariates and censored data must be taken into account while presenting a monitoring scheme. In this paper, the accelerated failure time models are used and two regression-adjusted control schemes based on Cox-Snell residuals are devised. Two different scenarios with censored and non-censored data are considered respectively. The competing control charts are compared in terms of zero-state and steady-state average run length criteria using Markov chain approach. The comparison study reveals that the cumulative sum based monitoring procedure is superior and more effective. It should be noted that the application of the proposed monitoring schemes are not restricted to manufacturing processes and thus service operations such as healthcare systems can benefit from them.     

Default estimation,correlated defaults,and expert information

Capital allocation decisions are made on the basis of an assessment of creditworthiness. Default is a rare event for most segments of a bank's portfolio and data information can be minimal. Inference about default rates is essential for efficient capital allocation, for risk management and for compliance with the requirements of the Basel II rules on capital standards for banks. Expert information is crucial in inference about defaults. A Bayesian approach is proposed and illustrated using prior distributions assessed from industry experts. A maximum entropy approach is used to represent expert information. The binomial model, most common in applications, is extended to allow correlated defaults yet remain consistent with Basel II. The application shows that probabilistic information can be elicited from experts and econometric methods can be useful even when data information is sparse. Copyright © 2009 John Wiley & Sons, Ltd.     

How useful is an Asian Currency Unit (ACU) index for surveillance in East Asia?

An Asian Currency Unit (ACU) index is constructed using an alternative procedure which minimizes a basket or portfolio of assets expressed in terms of national currencies. Using this estimated ACU index and an ACU deviation indicator, the main finding of this study based on the current trajectory of East Asian currencies relative to this regional ACU benchmark is that there is a formation of two contrasting groups of countries in the region—one group of strong currencies and one group of weak currencies. We emphasize that this contrasting trajectory in East Asian intra-regional exchange rates implies disturbed competitive trading relationships in the region, which may result in wasteful beggar-thy-neighbor policies. As emphasized in other recent studies (e.g., Kawai and Takagi, 2012), the region needs a framework for exchange rate policy coordination that will promote intra-regional exchange rate stability. We suggest two important ways in which the region can capitalize on using an ACU index for surveillance purposes in the immediate term. One way is to assess “over- and undervaluation” of individual currencies from the regional ACU average. The other is to use it as a monitoring device for excessive flows of international capital within the region.     

Career survival strategies for materiel managers

The skill sets and attitudes required for materiel management professional survival continue to change. Just as the industry has evolved for calculators and three-part forms, so too must today's managers evolve into facilitators and information-oriented management. While these new skills are being perfected, three focused strategies initiated right now can enhance your chance for survival. First, get to know your customers intimately. Go to their meetings. Tour their departments. Second, assume responsibility and report materiel cost data that has meaning to senior management and customers. Fill rates and inventory turns offer little value to other managers, but supply cost per discharge can be of value. Third, be proactive in competing with outsourcing competitors. Do not wait until they are on your doorstep before developing a strategy to compete.     

Explaining incubators using firm analogy

Incubators are initiated to accelerate the development of new technology-based firms. Policy actors see them as a tool to initiate or revive innovativeness in regions and universities as a way to commercialize research results. However, even though the intended results of the incubator are good it is not known how the incubator should be managed and organized in order to achieve this end. When faced with a new type of organizations analogies can provide insights gathered from other contexts. To contribute to the further understanding of incubators this paper discuss the implications, in terms of highlighted dimensions and further clarifications needed when using the analogy of a firm. The paper uses empirical findings from six incubators. The discussion shows it is not clear who is the actual customer of the incubator. For example, can the policy actors that provide the funding to the incubator be seen as a customer paying for the service of regional revival and the NTBFs customers when their fees are not in relation to the services they are provided? In the discussion it is suggested that the incubator can have many customers with different value creation processes or no customers depending on the viewpoint taken.