Banking on Basel – The case of Internal Rating Based approach

By
anjum
-
71
205

The mortgage crisis of 2008 demonstrated the vulnerability of financial intermediaries towards extreme events emanating mainly from fragile risk management practices. Technically, financial intermediation is a pseudo derivative transaction, and banks take both long (loans) and short (deposits) position simultaneously. Therefore, the inherent risks in banking firms are extreme warranting a prudent risk assessment and management. The basic ingredient of risk management for commercial banks – capital adequacy ratio – is based on a regulatory driven standardised approach for calculation of capital adequacy. The standardised approach was proposed by Basel Committee on Banking Supervision as an interim arrangement and banks were expected to develop their own risk assessment systems. However, almost seven years after the second Basel Accord, banks in Pakistan are still relying on the risk weights provided by the central bank to assess and mitigate their overall risk. This standardised approach is seriously flawed with multiple caveats that should be of concern to commercial banks.
Primarily, the risk weights assigned under standardised approach for corporate sector are classified into rated (by approved rating agencies) and unrated exposures with different risk weights for investment (low) and speculative (high) grade entities. This must be interesting to note that none of the default in the last decade was detected ex ante by either of the two rating agencies in Pakistan (despite their claims that ratings are forward looking). Adding insult to injury is that some of the defaulted entities were enjoying investment grade status by these agencies, couple of weeks before declaring default and were ex post subsequently downgraded to default. This raises questions on prudence of risk mitigating practices that are based on such expert opinions. Therefore, if banks develop internal rating based models, they will be in better position to ex ante assess the likelihood of default and provide adequate capital cushion.
Similarly, for unrated corporate exposures, banks should allocate 100% risk weight for capital adequacy purpose. In Pakistan less than hundred industrial (excluding instrument ratings) firms have entity ratings as per the website of two local rating agencies. That would imply a 100 per cent capital charge by commercial banks for most of their corporate clients. This could be alarming as banks with credit portfolios skewed towards bad borrowers will be allocating less capital than required, exposing them to high default risk. This could also leads to Moral Hazard as capital charge for all unrated exposures is same, and banks might seek high risk borrowers to earn premium for augmenting their interest rate spread by hiding the risk under the rug. On the contrary, banks with premium quality unrated borrowers would managing a non-existing risk putting constraints on their capital. An internal rating base model can address these issues by classifying unrated borrowers according to their repayment capacity and consequently assessing appropriate capital for the relevant risks.
Lastly, standardised approach provides no benefits for diversification in portfolios which is a conventional tool for risk management. At present, banks are required to allocate risk weights to individual exposures regardless of the situation of their overall portfolio. Again, this might lead to the situation where banks will increase concentration of their portfolios to yield higher spreads, as they find no economic benefit from diversification. It is also important to note that diversification is not only diversifying one particular risk (intra diversification) but also applies to diversify the possible interaction of various risks (intra diversification) as market risk could trigger credit risk and credit risk could ultimately lead to market risk. None of these types of diversification are considered in standardised approach and banks with high concentration and low concentration will be on a similar pitch to allocate capital against contingent risks. The diversification impact is incorporated in Internal Rating Based (IRB) approach by assuming correlation of exposures within the portfolio for credit risk and under advanced management approach for operational risk, while in market risk, all value at risk models (under IRB) account for diversification.
Adapting internal rating based approach is beneficial both for commercial banks and the central bank. Commercial banks could have more accurate estimate of their exposures and can hedge only the relevant risk. This will mitigate the excess risk of bad borrowers and will free up capital against good borrowers, which can be used in other ventures. If the inherent risks are mitigated the economic cost of supervision for the central bank would be less and this would lower frictions in the financial system. SBP introduced a road map in 2005 to adopt internal rating based approach by January 2010 but unfortunately later they allowed flexibility for transition to internal models. Given, the risk management benefits from IRB, Commercial Banks should consider developing their internal assessment methodologies and SBP should follow up on adapting Basel II in its essence because otherwise we have seen that “one size fits all” approach is only good tp manage risk in a Gaussian World but could not sustain even a smallest black swan.

The author holds a PhD in Quantitative Finance from Paris Dauphine. He is Associate Professor of Finance at Lahore School of Economics and provides consultancy on risk management through Synergistic Financial Advisors.

RELATED ARTICLESMORE FROM AUTHOR

71 COMMENTS

  2. Hey!!

    Sorry to intervene you. Can you exactly tell which Internal Based Rating Approach should be adopted? Please provide adequate reasons behind your stance as your above article depicts your lack of knowledge regrading Internal Based Rating Approach. As far as I know there are normally two approaches widely adopted by international banks. They are, a rating industry approach and a
    Merton-type model. Again, no offence, your article appears to Wikipedia based research.

    (Mr Mirza, I was also a BBA student of LSE)

    • MR. A
      Above article has raised a question, that you may able to reply. (1) What is the criteria of rating by rating agencies? Same company is rated different by different agencies at a same time. (2) What is the rating of these rating agencies? There should be some regulations for these rating agencies.

  3. very informative article. This has raised issues that must be considered by SBP to improve Banking risk management in Pakistan. Extremely well written

  4. excellent. Mr A I am sorry but if you are criticizing for the sake of criticism its ok, else your comment depicts ur lack of knowledge. IRB are not merton or industry, they are foundation and advanced – merton KMV is just one method out of the whole universe of statistical models in advanced IRB to estimate only one input ie pd (probability of default) and you are calling it a whole IRB approach, so plz do ur homework before jumping to conclusions – i blv this is no academic research and the details provided are adequate – news article are always “wikipedia” style as u may say

    • Fix your English man!! Can you tell me what do you mean by foundation and advanced? Have you ever entered in an International Bank? I am not calling Merton type Model as IRB approach. I have written two key approaches behind IRB. Many banks go for industry approach (in layman terms, it is quite similar to approach adopted by Credit Rating agencies).

  5. IRBs are very critical for risk management. I have been involved in their implementation here in canada for some time. Nicely presented – Mr A it seemed u google IRB and come up with argument based on first two results – please do ur hw before coming up with useless stuff especially when it is to criticise

    • Ms Shanzay!!
      Do you know any thing about IRB Approach? For sake of your knowledge, I don't need to google it because I am myself a postgraduate student at Stanford University.

  7. Hay A, you dont have to mention u r from LSE, ur imaturity and poor knwldge explains that u r from lahore school for everyone

    • O Man! I am not proud to be a LSE student because I am currently a postgraduate student at Stanford University (2nd after Harvard).

  8. Mr A,

    isnt it strange that you claim to know a lot about IRB approach and you dont know about "foundation" and "advanced"? Stanford LOL – now you need to google lol

    • I did not say him to define ‘foundation’ and ‘advanced’ approach but I was asking him which one should be adopted by Pakistani banks? And, what is the rationale behind his recommendation?
      Let me comment. For instance Barclays uses both techniques. Foundation IRB is employed for wholesale bank loans while Advanced IRB is adopted for retail, equity and non-credit obligations exposures.
      What you said about Stanford. Your absurd statements are not going to create any impact on me or Stanford.
      You should first assess your knowledge as you said KMV and Merton account for same. Working for Moody’s and so ignorant! Seems you haven’t taken a single course in finance specially econometrics.
      You people cannot bear positive criticism.

      • Damn – which stanford you are in LOL – Isnt it obvious that if Banks are not using IRB, they will start with foundation – and then move to foundation – but then only if you understand what IRB is – LOL hail to your "econometrics" skills when your IQ is so challenging –

        • Barclay's Board of Directors have gone mad. They should seek your advice because I have explained Barclay's is using both Advanced and Foundation IRB at same time. Man you are hilarious, join any english classes because 'Advanced' does not always mean at 'later stage'. Dude!! come on and have a cup of tea with me in Stanford. My address is

          Building 128, Escondido South, Stanford, CA 94305.

      • bachay, do you exactly know whats the difference b/w foundation and advanced? do you know the Pillar 2 of Basel? Do you know that once bank moves to advanced they cant have foundation IRB for any of portfolio. Read that pillar 2 and then speak out

  9. and about me knowing IRB – well i work for KMV (yea Moody's) so IRB is my bread and butter – and btw KMV is what you are crying out Merton Merton Merton – i guess thats enough wasting time for this – you please rest in peace with your "Stanford"

  10. guys guys guys ..

    KMV is a trademark of KMV Corporation that was founded in 1989. The KMV model calculates the Expected Default Frequency (EDF) based on the firm’s capital structure, the volatility of the assets returns and the current asset value.

    there is a difference between KMV model and Merton model . though the basis of KMV model is a paper written by Merton in 1974 .

    there are different approaches to make these model some are econometric approach and other are based on actuarial approach. I deal with econometric approach which is becoming popular and lots of new things have been done in that setup . i am not a Stanford guy as u see i wear dhot I can tell you but there is lots of scope in this field but to produce good models you need good data and econometric skills and SAS.

    once u have that send me your resume i always look for good modellers

  12. This is an interesting article, and the Dhoti wala raises some intriguing points. Let us not start imposing our own arrogance on other people’s work and well all of us have an opinion which you may or may not agree with. If Mr Stanford has such an issue, why does he not write an article on the same pages and allow it to come under scrutiny. All of us, will be very glad to read your pearls of Wisdom Mr A.

    • Okay sure, BTW I am publishing my paper in March in Journal of Financial Econometrics not here on Pakistan Today!! LOL. If you don’t have access to Journal of Financial Econometrics, tell me I will post or go to LUMS library to access .You people cannot bear positive criticism.

      • Dumb you dont understand the difference b/w an academic article and a newpaper one – hilarious – plz rest in peace – for your info (which like always is incomplete) in Pakistan JSTOR is accessible at all universities since 2005 and so is Blackwell, Scientdirect etc – why dont you share your publication acceptance here or else give us your SSRN page so we have the honor to read it now – LOL tch tch – grow up man "postgraduate" lol

        • LOL JSTOR will provide you access after 5 years. Science Direct, no one uses in USA. Check EBSCO Host or Wiley!! You look like illiterate students of LSE, science direct science direct …LOL

          • LOL A (i wonder what A stands for), i thankGod i do not have the honor to study from Lahore School for Everyone, and again even punjab university has updated JSTOR access – tch tch tch – wake up and get info where HEC has taken Pakistan to – why dont you post your web presence here so we can see you and your stanforddddddd – stanfordeee lol- some web link SSRN your class page something LOL

          • Dude!! don't twist the story around. Reply to my earlier comment. Man you are hilarious. You say first foundation and then advanced. You are considering detonative meanings of these words. Yea right, only Punjab University students can have such stupid knowledge. You want to know my name so that you can tell to Mirza. I know HEC has put universities in pit. LOL.

      • It seems JFE publishing a special issue in your "honor" in march – otherwise their next issues are due either in Jan or end April LOL

          • Asma,

            When was the last FJE issue, end of September. When will be of January? end of jan–.Just googling and saying that it is January. Do you know exact date of publication? It is normally available in libraries in March. Not pakistani universities libraries. LOL
            you are just showing jealousy

          • Library – who will care for library when it is available online. BTW publication date tht counts is when it is published not when it is recieved in library, basics of citations. Sure lets wait and i will repost the link of next issue here. After all we all must have the honor of reading it (there will not b many names of Pk origin in one issue) Still u can post the link to its working paper version to save us from waiting. But then you want to remain secret

  13. @dhoti wala – precisely – and what was missing in Merton was the volatility in assets, since it is not possible to observe "market value" of assets KMV proposed an iterative process for its estimation – Plus, PD is one input of that model and two others are Exposure at default and loss given default – under foundation IRB these two will be provided by the regulator while in advanced IRB they will be estimated internally – unfortunately our "Stanford post graduate" is asking – what do you mean by advanced and foundation

  14. thats ok sometime we don't know and there is no harm in asking .we all learn by discussing and reading I like to read while wearing dhoti :))) For those who wants to know more on this allow me to suggest a site that i often visit my place
    http://www.bis.org/

    complete document mention on BASEL ! BASEL !! and 111 accord and tonnes of scholarly articles how to measure and implement that. The new debate is IFRS and new accounting system and how this sometime collide with BASEL II .. keeping BASEL III a side.

    just a fruit for your thoughts :))

  15. A well written piece and quite informative too. I'll be looking forward to more articles by Dr. Mirza in the future.

  16. Guys
    u can't tolerate criticism. Mr. A is right . instead of giving proper answer to MR A question, ur criticizing him. as Dr Mirza has least information regarding implementation of IRB approach in Pakistan. I am working with State Bank of Pakistan. IRB approach is a sophisticated approach and it takes time to adopt it. Many banks were supposed to adopt it on 1st jan 2010 but some have started partial IRB approach. specific requirements have been set by SBP for enforcing IRB approach. there are several issues that were a hurdle like data collection across. soon IRB approach will be adopted by pakistani banks. Dr Mirza is completely wrong as State Bank has never ignored IRB approach and SBP has been working on it for last 6 years. it takes time to adopt IRB approach especially in a developing country like pakistan. it is not child's play as Dr Mirza thinks, SBP team has been working hard. Dr Mirza has inaccurate information and he should tell how many banks of India and China have adopted IRB approach?

    • I guess the point here was not about india or china and if i read correctly the circular issued by SBP after agreement with ceos of banks, pakistan must have adapted IRB from jan 2010 and there were no ifs and buts. Data, bank does not have their own data for risk analysis? Not even for a simple altman style statistics? Yup its not a childs play (i am a phd finance myself and knows what this IRB in and out is) but didnt SBP realized that in 2005? While world is moving to Basel iii we are still stuck with standardized. But then what can be expected where a governor will be fired on if he reduces to lower discount. and if u read the article it does not criticize SBP (which i wld have done if i have written) it simply suggests that SBP shld adapt for xyz reasons – so before giving bhashans read the article and then come back

      • Well said Haroon – a bunch of losers from sbp who knws nothing think they are smart **** before writing this comment i read annual reports of 10 largest banks in pk and let me tell u wht they mean by partial. They state we have now a dept tht will comply with Basel lol imagine this is partial IRB in 6 years. When i researched further all such “depts” are one man “depts” but yea SBP is doing the business lol

        • Keep in limits. don't use indecent language.u r saying loser to me. this is what u have been taught at home. u have no etiquettes.

  17. Well its quite easy to suggest IRB approach but in practical at the moment implementing IRB approach would be an uphill task for Pakistani banks.

    However, a good analysis on the whole but lacks some factual aspects.

    • You are absolutely right. I have also tried to explain the same but few people have started using indecent language.

      • Why dont u write an article to explain your point? With all the info u have as central banker. And provide factual knowledge. We will learn from as info is never complete, so please go ahead and share your accurate and factual response. Cause it is easy to say that xyz has inaccurate info etc, ok fine, then share yours, write 1000 words

      • especially on the state of Pk banks about IRB and “six years” efforts of SBP, that wld add to our knowledge

  18. very well written article. Probably this is max that can be done in a newspaper article. Very informative for me (even if it is not for othe Einstien’s around)

  19. nice article, its well composed. I wish central bankers should have brains rather than beans. retards on us that know how to print notes and thats it

  20. I was searching for IRB in emerging markets on google and i end up on this page. Very informative article and very energetic comments, typical of a third world mentality i suppose that instead of extending information just expressing baseless arguments. If there are issues Ms Hassan come up with your version and as someone mentioned with your accurate updated insider information, we would love to know how far regulator has reached in all these years. Although, as much as i understand, the article is abt advantages of IRB and not on if central bank is doing anything or not. But since you have implied that meaning, come on shoot, we are all waiting – cheers

    • I also ended up here searching some material for our working paper on implementation of Basel II. I am myself a researcher at Sheffield Hallam University. Arguments are not baseless. In academic circle, just giving advantages of some approach is considered to be a poor practice, one should give its limitations as well. Mirza says he is professor, unfortunately he is unaware of usual academic practice. It is not a third world mentality, it is always good to learn from others by pinpointing their mistakes.
      Vendy!! which part of the world you live in?? I always see people writing their names Wendy not Vendy.

      Lisa Bevere
      Accounting, Finance and Performance Measurement Research Centre
      Sheffield Hallam University

  21. My question is for so-called financial experts like Harun_Nust who knows IRB in and out, Shanzay, Asma and Dr Mirza.
    I think many people are exhibiting jealousy and providing irrelevant comments here. As most of you don’t understand IRB even Dr Mirza. My question was pretty simple and you cannot understand financial jargon, so I am again putting my question for all experts and Dr Mirza.
    Risk components in IRB approach are the probability of default (PD), the loss given default (LGD), the exposure at default (EAD) and the effective maturity (M).

    My question was and is how banks calculate default probabilities? Normally, banks employ two methods. One method to estimate probability of default is based on Merton Model approach (precisely saying, KMV model) and the other methodology to assign probabilities of default is more or less based on external ratings.
    Both methods will yield different results.
    1)Which one is better and why?
    2)What would be the likely impact on capital requirements of banks? (If they use Merton Model approach to estimate default probabilities or external ratings based approach)
    Please don't give irrelevant answers

        • and that is because you punjabis are hijras, if there were balls in any, bengalis must not have set u upside down, pathetic, none of them will have the guts to write and disclose their names because they are all ashamed of their existence and their bloodline. Punjabis will just bla bla here – nation of speakers – mourn ur life now

  22. mate since you are being multiplied by zero every time, why are you crying out loud. LOL – why dont you have peace or else write an article here – just for your information these are not the only two methods. a whole list of methods is available at (and even they are not exhaustive)
    http://www.springer.com/business+%26+management/f…

    i am sure at "stanaford" you can access springer LOL

    jaa apni hasratoon per ansoo baha kar soo jaaa

    • Haahaa!! Lady, you are funny. Just searching an article is not a big deal. Try to read paper at page number 79. You even did not care to read it and copy pasted the link here. They are suggesting casual methods of estimating probabilities of default (what I earlier mentioned and I referred to it as Merton Model, or KMV, etc), other methods to estimate probabilities of default are heuristic models and statistical models (but they are irrelevant in our approach for internal ratings based)
      And for your kind information, international banks estimate PD's using generally methods mentioned in earlier comment.
      In article you are referring above, they are constructing confidence intervals around estimated probabilities of default (PD). In IRB approach probabilities of default are calculated using one year of data.
      Take it in IRB perspective.
      What about answers to my questions??
      Now should I not say this "jaa apni hasratoon per ansoo baha kar soo jaaa"

      I have access to all in my Stanford accommodation, now you are witness. I accessed springer link at 11:40 PM..LOL

  23. Hey M nice to see your continuous passion on IRBs, i recall that from our days at BNP ParisBas (some nine years back) as derivative strategists that how enthusiastic you were about Basel Basel Basel. Surprisingly you are thinking at a place where probably cumulative investment portfolio of banking sector is less than the the investment portfolio our team used to manage. But still good luck, though after browsing down this page, i still stand by my comments to you some years back, dude do not end at wrong part of the globe, sadly you have

    Avot

  24. hahah … I was enjoying your discussion and was avoiding comments on author. I read his article twice and came to this conclusion that its a Charba . Information has been take from many sources and known articles and massaged it up. I don't disagree with agree with Mr. A. that its not a very coho rant piece of writing does not show an elegant flow of information

    and I tell you why I said so

    Its not clear if author is writing about IRR. internal risk rating of banks or some business entities (each financial have its risk raking and within financial institution each borrowers is also having a risk rating depending on banking product and amount of loan.) For retail banking risk rating for borrowers can be achieved using 3rd party PD models or internal PD models but when it comes to small business or cooperate loan method of estimating risk rating are different and its the choice of bank whether they rate in house or use third party score like SBSS6.0. For the very large commercial loans banks also need to have rating system based on the borrowing company's financials . While a commercial bank can have general loan loss estimates based on Basel .. its the responsibility of state bank to see if banks are adopting Basel or not and following its methods or not . it is also possible that state bank can do internal rating of all commercial banks and see what is the ranking of these banks in terms of risk rating and the risky loans they are carrying . Form Mirza's article its not clear what he is interesting in Is he talking about business companies or banks ? at point this division is blurred. look at the sentence "for unrated corporate exposures, banks should allocate 100% risk weight for capital adequacy purpose"

    Mirza sb.. this not true and i tell u why even if the business companies are not rated the probability that all of the unrated companies are going to default at the same time or in a certain fixed time is o,ooooooooooooooo1 percent so banks dont need 100 more capital to ensure that they will remain viable

    you are infact mixing estimation of economic capital under extreme stress testing situation with generalized loan loss allowance required under BASEL . these two things are different though both fall under the category of BASEL.

    with this note . I am going to have some HOOKA now while wearing Dhoti

  25. we are a pathetic nation. From showoff to indecent language everything is here. Personal grudges and i am only right attitude. As someone said if there are issues why the einstiens who are challenging or arguing here share their composed thoughts thru an article or share with us their research findings. Isnt it strange that majority who are criticizing or applauding are under fake/hidden names (may be they are one may be they are many) We dont have the guts to stand up and argue with identity (same goes for me), we are just a nation of morners and fake identities, pathetic r us.

    • – mate sure ur ideas thru a planned article so we get some benefit, knwldge sharing is something we badly need in Pakistan, Or else lets drown in dust – hail to our fake existence for mud slugging

  27. I also ended up here searching some material for our working paper on implementation of Basel II. I am myself a researcher at Sheffield Hallam University. Arguments are not baseless. In academic circle, just giving advantages of some approach is considered to be a poor practice, one should give its limitations as well. Mirza says he is professor, unfortunately he is unaware of usual academic practice. It is not a third world mentality, it is always good to learn from others by pinpointing their mistakes.

    Lisa Bevere
    Accounting, Finance and Performance Measurement Research Centre
    Sheffield Hallam University

    • Is this an academic article? Is this article submitted for journal of xyz research? The problem with academicians are they are too myopic to understand beyond their myopic brains. No hypothesis was tested here. Lastly all academic research has proven to be useless (i dont say that read Nassim Taleb on that) Again as many suggested why dont you just write a newspaper style article in response. At least you also as an identity that you disclosed. You can do that unlike many who are just hiding under cover. I guess third world mentality was about hiding and crying and that is obvious here

    • so whoever wants to highlight limitations of basel should go ahead and explain to the world not merely expressing personal grudges or showoff or so called challenges as they knw everything and no one else does. Challenges, who is listening, except for few of us who are here to make “fun”

  28. According to recent consumer research four out of ten consumers have bought a non-life insurance policy through a comparison website – where does that leave the role of the professional financial intermediary? The emergence of aggregators and price comparison websites has had a huge impact on consumer behavior and the distribution of general insurance products. These sites are now visited by increasing numbers of consumers who use them to identify insurance, mortgage products and other products that may be available to them before being sent to the providers' website to complete their purchase.

    (stata statistical analysis)

Comments are closed.