UK: IFRS 9 And The "Is It A Bird, Or Is It A Bomber?" Problem

Last Updated: 23 February 2017
Article by Alexander Marianski and Ian Wilson

With the adoption of the IFRS 9 accounting standard into EU law, it is full steam ahead for banks to deploy credit models that estimate Expected Credit Loss (ECL) accounting values. The standard requires firms to account for lifetime ECL on loans that have experienced a "significant increase in credit risk" (SICR), but allows firms to reach their own conclusions as to just how much credit risk ought to be viewed as "significant".

Most firms have developed their first suite of ECL models, and are able to estimate 12-month and lifetime ECL under various input assumptions such as forward economic scenarios. However, many firms have held off finalising their definition of SICR until additional forward economic scenarios have been developed. The wait-and-see approach also allows some degree of consensus in interpretation to emerge.

Credit risk, and movements in credit risk, are of course quantifiable using well-established techniques. But what should count as a "significant" increase? The IFRS 9 standard strongly suggests in paragraph 5.5.9 that SICR should be a function of movements in remaining lifetime PD as well as other qualitative factors, but gives little detail on how or where to calibrate the PD threshold.

In this article, we explore binary classifier techniques borrowed from machine learning applications, and explain how they can be used to select the optimal definition of SICR.

Receiver Operating Characteristic

Pioneers of Radar faced a similar problem in choosing the threshold at which to call the blip on the screen a big flock of birds, or a bomber. The Receiver Operating Characteristic (ROC) curve, which measures the discriminatory power of a binary classifier and is familiar to scorecard modellers, has its roots in radar engineering.

With the application of machine learning techniques for automated decision-making, it became necessary to set a decision rule which optimises the trade-off between identifying all bombers (the true positive result), all birds (the true negative result) and falsely calling bombers birds (false negative) or birds bombers (false positive).

Other examples of automatic classifiers include sorting of post, recognising hand-written numbers on cheques, and medical imaging. Many applications introduce the concept of a "cost function" – in the postal example the cost of misclassification is empirically measurable in terms of additional time spent resolving errors as well as compensation to customers.

IFRS 9 features and cost

The science of machine learning places considerable emphasis on first identifying which characteristics are of interest – termed the "feature vector" – as well as the relative cost of correct and incorrect classification. The IFRS 9 standard allows us to circumvent this first step by providing a handy (and concise) non-exhaustive list of factors which may be included in the feature vector (paragraph B5.5.17) including changes in credit spread, credit rating, economic conditions, borrower's financial results, and pricing.

In addition, the IFRS 9 standard disallows losses (i.e. the inclusion of collateral cash flows) as an input into stage allocation (paragraph B5.5.9), leading us to conclude that expected credit losses (i.e. the obvious choice of cost function) should not be introduced to weight the classification to stage 1 or 2. Based on this information, standard machine learning techniques can be used to set the optimal boundary for classifying accounts into those likely to default (which should be classed as stage 2) and those not (which belong in stage 1).

Linear Discriminant Functions

In the case where the classes are linearly separable and assumed multivariate Normal, then the straight line through the origin which gives maximum class separation can be found with relative ease to be a linear combination of feature vectors on the margin or, equivalently, a function of the class means and standard deviations. The use of a discriminant function handily avoids the problem of having to estimate probability densities, and the approach can be extended to:

  • Situations where the classes are not separable; and
  • Non-linear boundaries between classes, as alluded to by the IFRS 9 standard (paragraph B5.5.9) permitting a larger absolute movement for the best-rated customers at initial recognition.

The four possible results of a binary classifier are:

  • True positive;
  • True negative;
  • False positive; and
  • False negative.

These can be described using a 2x2 matrix, often referred to as a confusion matrix. Note that when weighted by a cost function, the term "cost matrix" is also used. Many market participants have adopted approaches which evaluate variations in true positive rate (TPR), true negative rate (TNR), false positive rate (FPR) and false negative rate (FNR). Definitions of "positive" and "negative" are typically articulated using business rules which often include progression to arrears and default from stages 1 and 2, as well as rules around accounts' propensity to flip-flop between stages.

Unsurprisingly, stages 1 and 2 are not separable (an expected result – with perfect separation, banks could choose to lend money only to borrowers who are known never to default) and some trade-off between the four metrics described above is therefore required. In the sciences, Matthews Correlation Coefficient and Mean Square Error are often used to combine the metrics into a single criterion to be minimised. However, few banks have combined the four metrics, and have instead relied on judgement to select their preferred trade-off between business rules, i.e. a semi-informed guess at the optimal boundary between stage 1 and stage 2.

Perceptron and Multi-Layer Networks

With a linear decision boundary, this approach is consistent with analysis of linear discriminants. However, the decision boundary is often non-linear and the Normality assumption mentioned is usually violated. Linear discriminant analysis can be extended to cope with non-linear boundaries by transforming explanatory variables before taking the weighted combination. This is the "perceptron". The perceptron permits a much wider range of decision boundaries, and is trained using gradient descent. Since the confusion matrix may have regions of constant performance, perceptron training instead introduces an error function which always has non-zero gradient.

The perceptron is, however, not able to deal with any general separation problem such as SICR. A perhaps extreme example of a problem which cannot be solved using a single perceptron is the "exclusive or" problem. A generalised decision boundary can, however, be achieved by introducing successive transformations or "layers". This is the "multi-layer perceptron", and is the cornerstone of modern artificial intelligence. A two-layer network is sufficient to approximate any continuous decision boundary.


The general problem of how to find the boundary between stages 1 and 2 can be addressed by introducing "features" described in the IFRS 9 standard as a vector of information, and training a multi-layer perceptron to discriminate between accounts which subsequently default, and accounts which do not.

It is, however, important to recognise that:

  • When a machine "learns" to mimic observed patterns of behaviour, it has not learned to articulate the rationale for its results. Therefore, the outputs of a multi-layer perceptron (and non-linear decision boundaries) can be difficult or impossible for management to explain.
  • Monotonicity seems likely to be a desirable feature of decision boundaries in credit risk applications. Therefore, it is likely that sufficient SICR classification performance can be achieved using a single layer perceptron.
  • It may be desirable to enforce an intuitive functional form on the decision boundary, such a "PD multiple" (straight line through the origin) or "rating notch difference" (probit or logit). This can in fact be achieved without the perceptron: Linear discriminant techniques also work if features are pre-mapped into a higher-dimensional space using the desired functional form.

A final word of caution: This kind of analysis is data-intensive and should be performed across the economic cycle. After all, the radar operator who calibrates the receiver in December may be in for a shock when birds return from their winter migration.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.

To print this article, all you need is to be registered on

Click to Login as an existing user or Register so you can print this article.

In association with
Related Topics
Related Articles
Related Video
Up-coming Events Search
Font Size:
Mondaq on Twitter
Mondaq Free Registration
Gain access to Mondaq global archive of over 375,000 articles covering 200 countries with a personalised News Alert and automatic login on this device.
Mondaq News Alert (some suggested topics and region)
Select Topics
Registration (please scroll down to set your data preferences)

Mondaq Ltd requires you to register and provide information that personally identifies you, including your content preferences, for three primary purposes (full details of Mondaq’s use of your personal data can be found in our Privacy and Cookies Notice):

  • To allow you to personalize the Mondaq websites you are visiting to show content ("Content") relevant to your interests.
  • To enable features such as password reminder, news alerts, email a colleague, and linking from Mondaq (and its affiliate sites) to your website.
  • To produce demographic feedback for our content providers ("Contributors") who contribute Content for free for your use.

Mondaq hopes that our registered users will support us in maintaining our free to view business model by consenting to our use of your personal data as described below.

Mondaq has a "free to view" business model. Our services are paid for by Contributors in exchange for Mondaq providing them with access to information about who accesses their content. Once personal data is transferred to our Contributors they become a data controller of this personal data. They use it to measure the response that their articles are receiving, as a form of market research. They may also use it to provide Mondaq users with information about their products and services.

Details of each Contributor to which your personal data will be transferred is clearly stated within the Content that you access. For full details of how this Contributor will use your personal data, you should review the Contributor’s own Privacy Notice.

Please indicate your preference below:

Yes, I am happy to support Mondaq in maintaining its free to view business model by agreeing to allow Mondaq to share my personal data with Contributors whose Content I access
No, I do not want Mondaq to share my personal data with Contributors

Also please let us know whether you are happy to receive communications promoting products and services offered by Mondaq:

Yes, I am happy to received promotional communications from Mondaq
No, please do not send me promotional communications from Mondaq
Terms & Conditions (the Website) is owned and managed by Mondaq Ltd (Mondaq). Mondaq grants you a non-exclusive, revocable licence to access the Website and associated services, such as the Mondaq News Alerts (Services), subject to and in consideration of your compliance with the following terms and conditions of use (Terms). Your use of the Website and/or Services constitutes your agreement to the Terms. Mondaq may terminate your use of the Website and Services if you are in breach of these Terms or if Mondaq decides to terminate the licence granted hereunder for any reason whatsoever.

Use of

To Use you must be: eighteen (18) years old or over; legally capable of entering into binding contracts; and not in any way prohibited by the applicable law to enter into these Terms in the jurisdiction which you are currently located.

You may use the Website as an unregistered user, however, you are required to register as a user if you wish to read the full text of the Content or to receive the Services.

You may not modify, publish, transmit, transfer or sell, reproduce, create derivative works from, distribute, perform, link, display, or in any way exploit any of the Content, in whole or in part, except as expressly permitted in these Terms or with the prior written consent of Mondaq. You may not use electronic or other means to extract details or information from the Content. Nor shall you extract information about users or Contributors in order to offer them any services or products.

In your use of the Website and/or Services you shall: comply with all applicable laws, regulations, directives and legislations which apply to your Use of the Website and/or Services in whatever country you are physically located including without limitation any and all consumer law, export control laws and regulations; provide to us true, correct and accurate information and promptly inform us in the event that any information that you have provided to us changes or becomes inaccurate; notify Mondaq immediately of any circumstances where you have reason to believe that any Intellectual Property Rights or any other rights of any third party may have been infringed; co-operate with reasonable security or other checks or requests for information made by Mondaq from time to time; and at all times be fully liable for the breach of any of these Terms by a third party using your login details to access the Website and/or Services

however, you shall not: do anything likely to impair, interfere with or damage or cause harm or distress to any persons, or the network; do anything that will infringe any Intellectual Property Rights or other rights of Mondaq or any third party; or use the Website, Services and/or Content otherwise than in accordance with these Terms; use any trade marks or service marks of Mondaq or the Contributors, or do anything which may be seen to take unfair advantage of the reputation and goodwill of Mondaq or the Contributors, or the Website, Services and/or Content.

Mondaq reserves the right, in its sole discretion, to take any action that it deems necessary and appropriate in the event it considers that there is a breach or threatened breach of the Terms.

Mondaq’s Rights and Obligations

Unless otherwise expressly set out to the contrary, nothing in these Terms shall serve to transfer from Mondaq to you, any Intellectual Property Rights owned by and/or licensed to Mondaq and all rights, title and interest in and to such Intellectual Property Rights will remain exclusively with Mondaq and/or its licensors.

Mondaq shall use its reasonable endeavours to make the Website and Services available to you at all times, but we cannot guarantee an uninterrupted and fault free service.

Mondaq reserves the right to make changes to the services and/or the Website or part thereof, from time to time, and we may add, remove, modify and/or vary any elements of features and functionalities of the Website or the services.

Mondaq also reserves the right from time to time to monitor your Use of the Website and/or services.


The Content is general information only. It is not intended to constitute legal advice or seek to be the complete and comprehensive statement of the law, nor is it intended to address your specific requirements or provide advice on which reliance should be placed. Mondaq and/or its Contributors and other suppliers make no representations about the suitability of the information contained in the Content for any purpose. All Content provided "as is" without warranty of any kind. Mondaq and/or its Contributors and other suppliers hereby exclude and disclaim all representations, warranties or guarantees with regard to the Content, including all implied warranties and conditions of merchantability, fitness for a particular purpose, title and non-infringement. To the maximum extent permitted by law, Mondaq expressly excludes all representations, warranties, obligations, and liabilities arising out of or in connection with all Content. In no event shall Mondaq and/or its respective suppliers be liable for any special, indirect or consequential damages or any damages whatsoever resulting from loss of use, data or profits, whether in an action of contract, negligence or other tortious action, arising out of or in connection with the use of the Content or performance of Mondaq’s Services.


Mondaq may alter or amend these Terms by amending them on the Website. By continuing to Use the Services and/or the Website after such amendment, you will be deemed to have accepted any amendment to these Terms.

These Terms shall be governed by and construed in accordance with the laws of England and Wales and you irrevocably submit to the exclusive jurisdiction of the courts of England and Wales to settle any dispute which may arise out of or in connection with these Terms. If you live outside the United Kingdom, English law shall apply only to the extent that English law shall not deprive you of any legal protection accorded in accordance with the law of the place where you are habitually resident ("Local Law"). In the event English law deprives you of any legal protection which is accorded to you under Local Law, then these terms shall be governed by Local Law and any dispute or claim arising out of or in connection with these Terms shall be subject to the non-exclusive jurisdiction of the courts where you are habitually resident.

You may print and keep a copy of these Terms, which form the entire agreement between you and Mondaq and supersede any other communications or advertising in respect of the Service and/or the Website.

No delay in exercising or non-exercise by you and/or Mondaq of any of its rights under or in connection with these Terms shall operate as a waiver or release of each of your or Mondaq’s right. Rather, any such waiver or release must be specifically granted in writing signed by the party granting it.

If any part of these Terms is held unenforceable, that part shall be enforced to the maximum extent permissible so as to give effect to the intent of the parties, and the Terms shall continue in full force and effect.

Mondaq shall not incur any liability to you on account of any loss or damage resulting from any delay or failure to perform all or any part of these Terms if such delay or failure is caused, in whole or in part, by events, occurrences, or causes beyond the control of Mondaq. Such events, occurrences or causes will include, without limitation, acts of God, strikes, lockouts, server and network failure, riots, acts of war, earthquakes, fire and explosions.

By clicking Register you state you have read and agree to our Terms and Conditions