The case for low mortgage risk weights

I have touched on residential mortgage risk weights a couple of times in this blog, most recently in a post on the Dutch proposal to increase residential mortgage RW. This post explores the question of why residential mortgage RW under the Internal Rating Based (IRB) approach can be so low. More importantly, can we trust these very low risk weights (and the banks generating them) or is this yet more evidence that the IRB approach is an unreliable foundation for measuring bank capital requirements? It also touches on some of the issues we encounter in cross border comparisons of capital strength.

It has to be said at this point that IRB modelling is not an area where I claim deep expertise and I would welcome comments and input from people who do have this subject matter expertise. However, it is an important issue given the role that residential mortgage lending plays both in the economy as a whole. If nothing else, the post will at least help me get my thoughts on these questions into some kind of order and potentially invite comments that set me straight if I have got anything wrong. Notwithstanding the importance of the issue, this post is pretty technical so likely only of interest if you want to dig into the detailed mechanics of the IRB approach.

Recapping on the Dutch proposal to increase mortgage risk weights

First a recap on what the Dutch bank supervisor proposes to do. Residential mortgage RW in the Netherlands are amongst the lowest observed in Europe

DNB:Financial Stability Report Autumn 2019


The Dutch banks can of course cite reasons why this is justified but, in order to improve the resilience of the banking system, the Dutch banking supervisor proposes to introduce a floor set at 12% on how low the RW can be. The 12% floor applies to loans with a dynamic Loan to Value (LTV) of 55% or lower. The RW floor increases progressively as the LTV increases reaching a maximum of 45% for loans with a LTV of 100% or more. DNB expects the application of the measure to increase the average risk weights of Dutch IRB banks by 3-4 percentage points (from 11% to 14%-15%).

What drives the low end of the IRB Mortgage RW?

None of the discussion set out below is in any way intended to challenge bank supervisors seeking to apply limits to the low mortgage risk weights we observe being generated by the internal models developed by IRB banks. That is a whole separate discussion but the move to higher RW on these exposures broadly makes sense to me, not only for reasons of systemic resilience, but also with regard to the way that it reduces the disparity between IRB RW and those the standardised banks are required to operate against. It is however useful to understand what is driving the model outcomes before citing them as evidence of banks gaming the system.

This extract from Westpac’s September 2019 Pillar 3 Report shows a weighted average RW of 24% with individual segments ranging from 6% to 137%. The CBA Pillar 3 shows a similar pattern (RW range from 4.4% to 173.5%). I won’t get too much into the technical detail here but the effective IRB RW is higher when you factor in Regulatory Expected Loss. The impact on the RW in the table below is roughly 16% on average (I divide REL by .08 to translate it to an RWA equivalent and then divide by RWA) but this effect only becomes material for the 26% and higher RW bands).

Source: Westpac Pillar 3 Report – Sep 2019

I am very happy to stand corrected on the facts but my understanding is that the 6% and 14% RW bands in the table above capture “seasoned” portions of the loan portfolio where the Loan to Valuation (LVR) ratio has declined substantially from the circa 80% plus typically observed in newly originated loans. The declining LVR is of course a natural outcome for Principal and Interest loans which is the kind traditional prudent banking prefers.

What at face value looks like an incredibly thin capital requirement starts to make more sense when you consider the fact that the borrowers in these segments have demonstrated their capacity to service their loans and, perhaps more importantly, have built up a substantial pool of their own equity in the property that will absorb very substantial declines in property prices before the bank is likely to face a loss.

Australian owner occupied borrowers have an incentive to repay as fast as possible because their interest is not tax deductible (making the mortgage repayment one of the best applications of surplus cash) and they typically borrow on a floating rate basis. The natural amortisation of loan principal is also likely to have been accelerated by the progressive decline in interest rates in recent years which has seen a large share of borrowers apply the interest saving to higher principal repayments.

Comparing Dutch and Australian Mortgage Risk Weights

Looking at the Dutch RW provides some perspective on the mortgage RW of the Australian IRB banks and the initiatives APRA has implemented to increase them. I will only scratch the surface of this topic but it is interesting none the less to compare the 14-15% average RW the Dutch IRB banks will be required to hold with the 25% average RW that Australian IRB banks must hold.

The Dutch banks cite a favourable legal system that supports low LGD by allowing them to quickly realise their security on defaulted loans. That is a sound argument when you are comparing to a jurisdiction where it can take up to 3 years for a bank to gain access to the security underpinning a defaulted loan. That said, the Australian banks can make a similar argument so that does not look like a definitive factor in favour of lower Dutch RW.

The Australian LTV is based on the amortised value of the loan compared to the value of the property at the time the loan was originated. The Dutch LTV as I understand it seem to includes the updated value of the property as the loan ages. Again I don’t see anything in the Dutch system that renders their residential mortgage lending fundamentally less risky than the Australian residential mortgage.

The other positive factor cited by the Dutch banks is the tax deductibility of mortgage interest which applies even where the property is owner occupied. In Australia, interest on loans for owner occupied property is not tax deductible. The Dutch banks argue that the tax deduction on interest enhances the capacity of the borrower to service a loan but my guess is that this advantage is highly likely to be translated into higher borrowing capacity and hence higher property prices so it is not clear that there is a net improvement in the capacity to repay the loan.

I obviously only have a very rudimentary understanding of Dutch tax rules but my understanding is that tax deductibility of interest expense in some European jurisdictions is quid pro quo for including the implied value of rental on the property in the owner’s taxable income. If that is the case then it looks like tax deductibility of interest is a zero sum game from the lending bank perspective. Qualified by the caveats above, I will provisionally take the side of the Australian mortgage in this comparison. It seems equally likely to me that the the absence of a tax deduction creates an incentive for Australian borrowers to repay their loan as quickly as possible and hence for a greater proportion of loans outstanding to move into the low LVR bands that insulate the bank from the risk of loss. There does not seem to be the same incentive in the Dutch system, especially where the loans are fixed rate.

Summing up

The purpose of this post was mostly to help me think through the questions posed in the introduction. If you are still reading at this point then I fear you (like I) take bank capital questions way too seriously.

There are two main points I have attempted to explore and stake out a position on:

  • What at face value looks like an incredibly thin capital requirement for some parts of the residential mortgage portfolio start to make more sense when you consider the fact that the borrowers in these segments have demonstrated their capacity to service their loans and have built up a substantial pool of their own equity in the property that will absorb very substantial declines in property prices before the bank is likely to face a loss.
  • Cross border comparisons of capital are complicated but mortgages are a big part of the Australian bank risk profile and I still feel like they stack up relatively well in comparison to other jurisdictions that cite structural reasons why theirs are low risk.

If you have some evidence that contradicts what I have outlined above then by all means please let me know what I am missing.

Tony

Capital geeks take note – the IRB scaling factor strikes again

Capital

Another reminder on the importance of paying attention to the detail courtesy of a story that I picked up reading Matt Levine’s “Money Stuff” column on Bloomberg.

This extract from Matt’s column captures the essential facts:

Here, from Johannes Borgen, is a great little story about bank capital. Yesterday Coventry Building Society, a U.K. bank, announced “a correction to its calculation of risk weighted assets” that will lower its common equity Tier 1 capital ratio from 34.2% to 32.6%. That’s still well over regulatory requirements, so this is not a big deal. But the way Coventry messed up is funny:

“The Society uses Internal Ratings Based (“IRB”) models to calculate its Risk Weighted Assets (“RWAs”) and is seeking to update these models to ensure compliance with upcoming Basel III  reforms. During the process of transitioning models, the Society has identified an omission in connection with its historic calculation of its RWAs. Specifically, the necessary 6% scalar was not applied to the core IRB model outputs. The core IRB models themselves are not impacted.”

For banks that use Internal Ratings Based models, the way the Basel capital rules work is that you apply a complicated formula to calculate the risk weights of your assets, and then at the end of the formula you multiply everything by 1.06. That’s kind of weird. (The Basel capital regime for banks using IRB models “applies a scaling factor in order to broadly maintain the aggregate level of minimum capital requirements, while also providing incentives to adopt the more advanced risk-sensitive approaches.”) It’s weird enough that in the “upcoming Basel III reforms” regulators plan to get rid of it: The 1.06 multiplier is a kludge, and if you measure your risk-weighted assets a bit more accurately and conservatively, you shouldn’t have to multiply them by 1.06 at the end. 

Matt Levine, “Money Stuff”, Bloomberg

For anyone new to this game who wants to dig a bit deeper into how the advanced capital requirements are calculated, the Explanatory Note published by the BCBS in July 2005 is still a good place to start. I published a note on that paper on my blog here. The RBNZ also produced a useful note on how they used the IRB function in the portfolio modelling work they used to support their recent changes to NZ capital requirements.

It should be noted however that none of these documents discuss the 6% scaling factor. I open to alternative perspectives on this but my recollection is that the 6% scaling factor was introduced post July 2005 in one of the multiple recalibration exercises the BCBS employed to ensure that the IRB function did not reduce capital requirements too much relative to the status quo operating under Basel I. It is effectively a “fudge” factor designed to produce a number the BCBS was comfortable with (at that time).

Tony

Mortgage Risk Weights – revisited

I post on a range of topics in banking but residential mortgage risk weights is one that seems to generate the most attention. I first posted on the topic back in Sep 2018 and have revisited the topic a few times (Dec 2018, June 2019#1, June 2019#2, and Nov 2019) .

The posts have tended to generate a reasonable number of views but limited direct engagement with the arguments I have advanced. Persistence pays off however because the last post did get some specific and very useful feedback on the points I had raised to argue that the difference in capital requirements between IRB and Standardised Banks was not as big as it was claimed to be.

My posts were a response to the discussion of this topic I observed in the financial press which just focussed on the nominal difference in the risk weights (i.e. 25% versus 39%) without any of the qualifications. I identified 5 problems with the simplistic comparison cited in the popular press and by some regulators:

  • Problem 1 – Capital adequacy ratios differ
  • Problem 2 – You have to include capital deductions
  • Problem 3 – The standardised risk weights for residential mortgages seems set to change
  • Problem 4 – The risk of a mortgage depends on the portfolio not the individual loan
  • Problem 5 – You have to include the capital required for Interest Rate Risk in the Banking Book

With the benefit of hindsight and the feedback I have received, I would concede that I have probably paid insufficient attention to the disparity between risk weights (RW) at the higher quality end of the mortgage risk spectrum. IRB banks can be seen to writing a substantial share of their loan book at very low RWs (circa 6%) whereas the best case scenario for standardised banks is a 20% RW. The IRB banks are constrained by the requirement that their average RW should be at least 25% and I thought that this RW Floor was sufficient to just focus on the comparison of average RW. I also thought that the revisions to the standardised approach that introduced the 20% RW might make more of a difference. Now I am not so sure. I need to do a bit more work to resolve the question so for the moment I just want to go on record with this being an issue that needs more thought than I have given it to date.

Regarding the other 4 issues that I identified in my first post, I stand by them for the most part. That does not mean I am right of course but I will briefly recap on my arguments, some of the push back that I have received and areas where we may have to just agree to disagree.

Target capital adequacy ratios differ materially. The big IRB banks are targeting CET1 ratios based on the 10.5% Unquestionably Strong Benchmark and will typically have a bit of a buffer over that threshold. Smaller banks like Bendigo and Suncorp appear to operate with much lower CET1 targets (8.5 to 9.0%). This does not completely offset the nominal RW difference (25 versus 39%) but it is material (circa 20% difference) in my opinion so it seem fair to me that the discussion include this fact. I have to say that not all of my correspondents accepted this argument so it seems that we will have to agree to disagree.

You have to include capital deductions. In particular, the IRB banks are required to hold CET1 capital for the shortfall between their loan loss provision and a regulatory capital value called “Regulatory Expected Loss”. There did not appear to be a great awareness of this requirement and a tendency to dismiss it but my understanding is that it can increase the effective capital requirement by 10-12% which corresponds to an effective IRB RW closer to 28% than 25%.

The risk of a mortgage depends on the overall portfolio not the individual loan. My point here has been that small banks will typically be less diversified than big banks and so that justifies a difference in the capital requirements. I have come to recognise that the difference in portfolio risk may be accentuated to the extent that capital requirements applied to standardised banks impede their ability to capture a fair share of the higher quality end of the residential mortgage book. So I think my core point stands but there is more work to do here to fully understand this aspect of the residential mortgage capital requirements. In particular, I would love get more insight into how APRA thought about this issue when it was calibrating the IRB and standardised capital requirements. If they have spelled out their position somewhere, I have not been able to locate it.

You have to include the capital required for Interest Rate Risk in the Banking Book (IRRBB). I did not attempt to quantify how significant this was but simply argued that it was a requirement that IRB banks faced that standardised banks did not and hence it did reduce the benefit of lower RW. The push back I received was that the IRRBB capital requirement was solely a function of IRB banks “punting” their capital and hence completely unrelated to their residential mortgage loans. I doubt that I will resolve this question here and I do concede that the way in which banks choose to invest their capital has an impact on the size of the IRRBB capital requirement. That said, a bank has to hold capital to underwrite the risk in its residential mortgage book and, all other thinks being equal, an IRB bank has to hold more capital for the IRRBB requirement flowing from the capital that it invests on behalf of the residential mortgage book. So it still seems intuitively reasonable to me to make the connection. Other people clearly disagree so we may have to agree to disagree on this aspect.

Summing up, I had never intended to say that there was no difference in capital requirements. My point was simply that the difference is not as big as is claimed and I was yet to see any analysis that considered all of the issues relevant to properly understand what the net difference in capital requirements is. The issue of how to achieve a more level playing field between IRB and Standardised Banks is of course about much more than differences in capital requirements but it is an important question and one that should be based on a firmer set of facts that a simplistic comparison of the 39% standardised versus 25% IRB RW that is regularly thrown around in the discussion of this question.

I hope I have given a fair representation here of the counter arguments people have raised against my original thesis but apologies in advance if I have not. My understanding of the issues has definitely been improved by the challenges posted on the blog so thanks to everyone who took the time to engage.

Tony

Mortgage risk weights fact check revisited – again

The somewhat arcane topic of mortgage risk weights is back in the news. It gets popular attention to the extent they impact the ability of small banks subject to standardised risk weights to compete with bigger banks which are endorsed to use the more risk sensitive version based on the Internal Ratings Based (IRB) approach. APRA released a Discussion Paper (DP) in February 2018 titled “Revisions to the capital framework for authorised deposit-taking institutions”. There are reports that APRA is close to finalising these revisions and that this will address the competitive disadvantage that small banks suffer under the current regulation.

This sounds like a pretty simple good news story – a victory for borrowers and the smaller banks – and my response to the discussion paper when it was released was that there was a lot to like in what APRA proposed to do. I suspect however that it is a bit more complicated than the story you read in the press.

The difference in capital requirements is overstated

Let’s start with the claimed extent of the competitive disadvantage under current rules. The ACCC’s Final Report on its “Residential Mortgage Price Inquiry” described the challenge with APRA’s current regulatory capital requirements as follows:

“For otherwise identical ADIs, the advantage of a 25% average risk weight (APRA’s minimum for IRB banks) compared to the 39% average risk weight of standardised ADIs is a reduction of approximately 0.14 percentage points in the cost of funding the loan portfolio. This difference translates into an annual funding cost advantage of almost $750 on a residential mortgage of $500 000, or about $15 000 over the 30 year life of a residential mortgage (assuming an average interest rate of 7% over that period).”

You could be forgiven for concluding that this differential (small banks apparently required to hold 56% more capital for the same risk) is outrageous and unfair.

Just comparing risk weights is less than half the story

I am very much in favour of a level playing field and, as stated above, I am mostly in favour of the changes to mortgage risk weights APRA outlined in its discussion paper but I also like fact based debates.

While the risk weights for big banks are certainly lower on average than those required of small banks, the difference in capital requirements is not as large as the comparison of risk weights suggests. To understand why the simple comparison of risk weights is misleading, it will be helpful to start with a quick primer on bank capital requirements.

The topic can be hugely complex but, reduced to its essence, there are three elements that drive the amount of capital a bank holds:

  1. The risk weights applied to its assets
  2. The target capital ratio applied to those risk weighted assets
  3. Any capital deductions required when calculating the capital ratio

I have looked at this question a couple of times (most recently here) and identified a number of problems with the story that the higher risk weights applied to residential mortgages originated by small bank places them at a severe competitive disadvantage:

Target capital ratios – The target capital adequacy ratios applied to their higher standardised risk weighted assets are in some cases lower than the IRB banks and higher in others (i.e. risk weights alone do not determine how much capital a bank is required to hold).

Portfolio risk – The risk of a mortgage depends on the portfolio not the individual loan. The statement that a loan is the same risk irrespective of whether it is written by a big bank or small bank sounds intuitively logical but is not correct. The risk of a loan can only be understood when it is considered as part of the portfolio the bank holds. All other things being equal, small banks will typically be less diversified and hence riskier than a big bank.

Capital deductions – You also have to include capital deductions and the big banks are required to hold capital for a capital deduction linked to the difference between their loan loss provisions and a regulatory capital value called “Regulatory Expected Loss”. The exact amount varies from bank to bank but I believe it increases the effective capital requirement by 10-12% (i.e. an effective RW closer to 28% for the IRB banks).

IRRBB capital requirement – IRB banks must hold capital for Interest Rate Risk in the Banking Book (IRRBB) while the small standardised banks do not face an explicit requirement for this risk. I don’t have sufficient data to assess how significant this is, but intuitively I would expect that the capital that the major banks are required to hold for IRRBB will further narrow the effective difference between the risk weights applied to residential mortgages.

How much does reducing the risk weight differential impact competition in the residential mortgage market?

None of the above is meant to suggest that the small banks operating under the standardised approach don’t have a case for getting a lower risk weight for their higher quality lower risk loans. If the news reports are right then it seems that this is being addressed and that the gap will be narrower. However, it is important to remember that:

  • The capital requirement that the IRB banks are required to maintain is materially higher than a simplistic application of the 25% average risk weight (i.e. the IRB bank advantage is not as large as it is claimed to be).
  • The standardised risk weight does not seem to be the binding constraint so reducing it may not help the small banks much if the market looks through the change in regulatory risk measurement and concludes that nothing has changed in substance.

One way to change the portfolio quality status quo is for small banks to increase their share of low LVR loans with a 20% RW. Residential mortgages do not, for the most part, get originated at LVR of sub 50% but there is an opportunity for small banks to try to refinance seasoned loans where the dynamic LVR has declined. This brings us to the argument that IRB banks are taking the “cream” of the high quality low risk lending opportunities.

The “cream skimming” argument

A report commissioned by COBA argued that:

“While average risk weights for the major banks initially rose following the imposition of average risk weight on IRB banks by APRA, two of the major banks have since dramatically reduced their risk weights on residential mortgages with the lowest risk of default. The average risk weights on such loans is now currently on average less than 6 per cent across the major banks.”

“Despite the imposition of an average risk weight on residential home loans, it appears some of the major banks have decided to engage in cream skimming by targeting home loans with the lowest risk of default. Cream skimming occurs when the competitive pressure focuses on the high-demand customers (the cream) and not on low- demand ones (the skimmed milk) (Laffont & Tirole, 1990, p. 1042). Cream skimming has adverse consequences as it skews the level of risk in house lending away from the major banks and towards other ADIs who have to deal with an adversely selected and far riskier group of home loan applicants.”

“Reconciling Prudential Regulation with Competition” prepared by Pegasus Economics May 2019 (page 43)

It is entirely possible that I am missing something here but, from a pure capital requirement perspective, it is not clear that IRB banks have a material advantage in writing these low risk loans relative to the small bank competition. The overall IRB portfolio must still meet the 25% risk weight floor so any loans with 6% risk weights must be offset by risk weights (and hence riskier loans) that are materially higher than the 25% average requirement. I suspect that the focus on higher quality low risk borrowers by the IRB banks was more a response to the constraints on capacity to lend than something that was driven by the low risk weights themselves.

Under the proposed revised requirements, small banks in fact will probably have the advantage in writing sub 50% LVR loans given that they can do this at a 20% risk weight without the 25% floor on their average risk weights and without the additional capital requirements the IRB banks face.

I recognise there are not many loans originated at this LVR band but there is an opportunity in refinancing seasoned loans where the combined impact of principal reduction and increased property value reduces the LVR. In practice the capacity of small banks to do this profitably will be constrained by their relative expense and funding cost disadvantage. That looks to me to be a bigger issue impacting the ability of small banks to compete but that lies outside the domain of regulatory capital requirements.

Maybe this potential arbitrage does not matter in practice but APRA could quite reasonably impose a similar minimum average RW on Standardised Banks if the level playing field argument works both ways. This should be at least 25% but arguably higher once you factor in the fact that the small banks do not face the other capital requirements that IRB banks do. Even if APRA did not do this, I would expect the market to start looking more closely at the target CET1 for any small bank that accumulated a material share of these lower risk weight loans.

Implications

Nothing in this post is meant to suggest that increasing the risk sensitivity of the standardised risk weights is a bad idea. It seems doubtful however that this change alone will see small banks aggressively under cutting large bank competition. It is possible that small bank shareholders may benefit from improved returns on equity but even that depends on the extent to which the wholesale markets do not simply look through the change and require smaller banks to maintain the status quo capital commitment to residential mortgage lending.

What am I missing …

How much capital is enough? – The NZ perspective

The RBNZ has delivered the 4th instalment in a Capital Review process that was initiated in March 2017 and has a way to run yet. The latest consultation paper addresses the question “How much capital is enough?”.  The banking industry has until 29 March 2019 to respond with their views but the RBNZ proposed answer is:

  • A Tier 1 capital requirement of 16% of RWA for systemically important banks and 15% of RWA for all other banks
  • The Tier 1 minimum requirement to remain unchanged at 6% (with AT1 capital continuing to be eligible to contribute a maximum of 1.5 percentage points)
  • The proposed increased capital requirement to be implemented via an overall prudential capital buffer of 9-10% of RWA comprised entirely of CET1 capital;
    • Capital Conservation Buffer 7.5% (currently 2.5%)
    • D-SIB Buffer 1.0% (no change)
    • Counter-cyclical buffer 1.5% (currently 0%)

The increase in the capital ratio requirement is proposed to be supplemented with a series of initiatives that will increase the RWA of IRB banks:

  • The RBNZ proposes to 1) remove the option to apply IRB RW to sovereign and bank exposures,  2) increase the IRB scalar (from 1.06 to 1.20) and 3) to introduce an output floor set at 85% of the Standardised RWA on an aggregate portfolio basis
  • As at March 2018, RWA’s produced by the IRB approach averaged 76% of the Standardised Approach and the RBNZ estimate that the overall impact will be to increase the aggregate RWA to 90% of the outcome generated by the Standardised approach (i.e. the IRB changes, not the output floor, drive the increase in RWA)
  • Aggregate RWA across the four IRB banks therefore increases by approximately 16%, or $39bn, compared to March 2018 but the exact impact will depend on how IRB banks respond to the higher capital requirements

The RBNZ has also posed the question whether a Tier 2 capital requirement continues to be relevant given the substantial increase in Tier 1 capital.

Some preliminary thoughts …

There is a lot to unpack in this paper so this post will only scratch the surface of the issues it raises …

  • The overall number that the RBNZ proposes (16%) is not surprising.It looks to be at the lower end of what other prudential regulators are proposing in nominal terms
  • But is in the same ball park once you allow for the substantial increase in IRB RWA and the fact that it is pretty much entirely CET1 capital
  • What is really interesting is the fundamentally different approach that the RBNZ has adopted to Tier 2 capital and bail-in versus what APRA (and arguably the rest of the world) has adopted
    • The RBNZ proposal that the increased capital requirement take the form of CET1 capital reflects its belief that “contingent convertible instruments” should be excluded from what counts as capital
    • Exactly why the RBNZ has adopted this position is a complex post in itself (their paper on the topic can be found here) but the short version (as I understand it) is that they think bail-in capital instruments triggered by non-viability are too complex and probably won’t work anyway.
    • Their suggestion that Tier 2 probably does not have a role in the capital structure they have proposed is logical if you accept their premise that Point of Non-Viability (PONV) triggers and bail-in do not work.
  • The RBNZ highlight a significantly enhanced role for prudential capital buffersI am generally in favour of bigger, more dynamic, capital buffers rather than higher fixed minimum requirements and I have argued previously in favour of the base rate for the counter-cyclical being a positive value (the RBNZ propose 1.5%)
    • But the overall size of the total CET1 capital buffer requirement requires some more considered thought about 1) the role of bail-in  structures and PONV triggers in the capital regulation toolkit (as noted above) and 2) whether the impacts of the higher common equity requirement will be as benign as the RBNZ analysis suggests
  • I am also not sure that the indicative capital conservation responses they have outlined (i.e. discretionary distributions limited to 60% of net earnings in the first 250bp of the buffer, falling to 30% in the next 250bp and no distributions thereafter) make sense in practice.
    • This is because I doubt there will be any net earnings to distribute if losses are sufficient to reduce CET1 capital by 250bp so the increasing capital conservation requirement is irrelevant.
  • Last, but possibly most importantly, we need to consider the impact on the Australian parents of the NZ D-SIB banks and how APRA responds. The increase in CET1 capital proposed for the NZ subsidiaries implies that, for any given amount of CET1 capital held by the Level 2 Banking Group, the increased strength of the NZ subsidiaries will be achieved at the expense of the Australian banking entities
    • Note however that the impact of the higher capital requirement in NZ will tend to be masked by the technicalities of how bank capital ratios are calculated.
      • It probably won’t impact the Level 2 capital ratios at all since these are a consolidated view of the combined banking group operations of the Group as a whole
      • The Level 1 capital ratios for the Australian banks also treat investments in bank subsidiaries relatively generously (capital invested in unlisted subsidiaries is treated as a 400% risk weighted asset rather than a capital deduction).

Conclusion

Overall, I believe that the RBNZ is well within its rights to expect the banks it supervises to maintain a total level of loss absorbing capital of 16% or more. The enhanced role for capital buffers is also a welcome move.

The issue is whether relying almost entirely on CET1 capital is the right way to achieve this objective. This is however an issue that has been debated for many decades with no clear resolution. It will take some time to fully unpack the RBNZ argument and figure out how best to articulate why I disagree. In the interim, any feedback on the issues I have outlined above would be most welcome.

Tony

Distinguishing luck and skill

Quantifying Luck’s Role in the Success Equation

“… we vastly underestimate the role of luck in what we see happening around us”

This post is inspired by a recent read of Michael Mauboussin’s book “The Success Equation: Untangling Skill and Luck in Business, Sports and Investing”. Mauboussin focuses on the fact that much of what we experience is a combination of skill and luck but we tend to be quite bad at distinguishing the two. It may not unlock the secret to success but, if you want to get better at untangling the contributions that skill and luck play in predicting or managing future outcomes, then this book still has much to offer.

“The argument here is not that you can precisely measure the contributions of skill and luck to any success or failure. But if you take concrete steps toward attempting to measure those relative contributions, you will make better decisions than people who think improperly about those issues or who don’t think about them at all.”

Structure wise, Mauboussin:

  • Starts with the conceptual foundations for thinking about the problem of distinguishing skill and luck,
  • Explores the analytical tools we can use to figure out the extent to which luck contributes to our achievements, successes and failures,
  • Finishes with some concrete suggestions about how to put the conceptual foundations and analytical tools to work in dealing with luck in decisions.

Conceptual foundations

It is always good to start by defining your terms; Mauboussin defines luck and skill as follows:

“Luck is a chance occurrence that affects a person or a group.. [and] can be good or bad [it] is out of one’s control and unpredictable”

Skill is defined as the “ability to use one’s knowledge effectively and readily in execution or performance.”

Applying the process that Mauboussin proposes requires that we first roughly distinguish where a specific activity or prediction fits on the continuum bookended by skill and luck. Mauboussin also clarifies that:

  • Luck and randomness are related but not the same: He distinguishes luck as operating at the level of the individual or small group while randomness operates at the level of the system where more persistent and reliable statistical patterns can be observed.
  • Expertise does not necessarily accumulate with experience: It is often assumed that doing something for a long time is sufficient to be an expert but Mauboussin argues that in activities that depend on skill, real expertise only comes about via deliberate practice based on improving performance in response to feedback on the ways in which the input generates the predicted outcome.

Mauboussin is not necessarily introducing anything new in his analysis of why we tend to bad at distinguishing skill and luck. The fact that people tend to struggle with statistics is well-known. The value for me in this book lies largely in his discussion of the psychological dimension of the problem which he highlights as exerting the most profound influence. The quote below captures an important insight that I wish I understood forty years ago.

“The mechanisms that our minds use to make sense of the world are not well suited to accounting for the relative roles that skill and luck play in the events we see taking shape around us.”

The role of ideas, beliefs and narratives is a recurring theme in Mauboussin’s analysis of the problem of distinguishing skill and luck. Mauboussin notes that people seem to be pre-programmed to want to fit events into a narrative based on cause and effect. The fact that things sometimes just happen for no reason is not a satisfying narrative. We are particularly susceptible to attributing successful outcomes to skill, preferably our own, but we seem to be willing to extend the same presumption to other individuals who have been successful in an endeavour. It is a good story and we love stories so we suppress other explanations and come to see what happened as inevitable.

Some of the evidence we use to create these narratives will be drawn from what happened in specific examples of the activity, while we may also have access to data averaged over a larger sample of similar events. Irrespective, we seem to be predisposed to weigh the specific evidence more heavily in our intuitive judgement than we do the base rate averaged over many events (most likely based on statistics we don’t really understand). That said, statistical evidence can still be “useful” if it “proves” something we already believe; we seem to have an intuitive bias to seek evidence that supports what we believe. Not only do we fail to look for evidence that disproves our narrative, we tend to actively suppress any contrary evidence we encounter.

Analytical tools for navigating the skill luck continuum

We need tools and processes to help manage the tendency for our intuitive judgements to lead us astray and to avoid being misled by arguments that fall into the same trap or, worse, deliberately exploit these known weaknesses in our decision-making process.

One process proposed by Mauboussin for distinguishing skill from luck is to:

  • First form a generic judgement on what the expected accuracy of our prediction is likely to be (i.e. make a judgement on where the activity sits on the skill-luck continuum)
  • Next look at the available empirical or anecdotal evidence, distinguishing between the base rate for this type of activity (if it exists) and any specific evidence to hand
  • Then employ the following rule:
    • if the expected accuracy of the prediction is low (i.e. luck is likely to be a significant factor), you should place most of the weight on the base rate
    • if the expected accuracy is high (i.e. there is evidence that skill plays the prime role in determining the outcome of what you are attempting to predict), you can rely more on the specific case.
  • use the data to test if the activity conforms to your original judgement of how skill and luck combine to generate the outcomes

Figuring out where the activity sits on the skill-luck continuum is the critical first step and Mauboussin offers three methods for undertaking this part of the process: 1) The “Three Question” approach, 2) Simulation and 3) True Score Theory. I will focus here on the first method which involves

  1. First ask if you can easily assign a cause to the effect you are seeking to predict. In some instances the relationship will be relatively stable and linear (and hence relatively easy to predict) whereas the results of other activities are shaped by complex dependencies such as cumulative advantage and social preference. Skill can play a part in both activities but luck is likely to be a more significant factor in the latter group.
  2. Determining the rate of reversion to the mean: Slow reversion is consistent with activities dominated by skill, while rapid reversion comes from luck being the more dominant influence. Note however that complex activities where cumulative advantage and social preference shape the outcome may not have a well-defined mean to revert to. The distribution of outcomes for these activities frequently conform to a power law (i.e. there are lots of small values and relatively few large values).
  3. Is there evidence that expert prediction is useful? When experts have wide disagreement and predict poorly, that is evidence that luck is a prime factor shaping outcomes.

One of the challenges with this process is to figure out how large a sample size you need to determine if there is a reliable relationship between actions and outcome that evidences skill.  Another problem is that a reliable base rate may not always be available. That may be because the data has just not been collected but also because a reliable base rate simply may not even exist.

The absence of a reliable base rate to guide decisions is a feature of activities that do not have simple linear relationships between cause and effect. These activities also tend to fall into Nassim Taleb’s “black swan” domain. The fundamental lesson in this domain of decision making is to be aware of the risks associated with naively applying statistical probability based methods to the problem. Paul Wilmott and David Orrell use the idea of a “zone of validity” to make the same point in “The Money Formula”.

The need to understand power laws and the mechanisms that generate them also stands out in Mauboussin’s discussion of untangling skill and luck.

The presence of a power law depends in part on whether events are dependent on, or independent of, one another. In dependent systems, initial conditions matter and come to matter more and more as time goes on. The final outcomes are (sometimes surprisingly) sensitive to both minor variations in the initial conditions and to the path taken over time. Mauboussin notes that a number of mechanisms are responsible for this phenomenon including preferential attachment, critical points and phase transitions are also crucial.

“In some realms, independence and bell-shaped distributions of luck can explain much of what we see. But in activities such as the entertainment industry, success depends on social interaction. Whenever people can judge the quality of an item by several different criteria and are allowed to influence one another’s choices, luck will play a huge role in determining success or failure.”

“For example, if one song happens to be slightly more popular than another at just the right time, it will tend to become even more popular as people influence one another. Because of that effect, known as cumulative advantage, two songs of equal quality, or skill, will sell in substantially different numbers. …  skill does play a role in success and failure, but it can be overwhelmed by the influence of luck. In the jar model, the range of numbers in the luck jar is vastly greater than the range of numbers in the skill jar.”

“The process of social influence and cumulative advantage frequently generates a distribution that is best described by a power law.”

“The term power law comes from the fact that an exponent (or power) determines the slope of the line. One of the key features of distributions that follow a power law is that there are very few large values and lots of small values. As a result, the idea of an “average” has no meaning.”

Mauboussin’s discussion of power laws does not offer this specific example but the idea that the average is meaningless is also true of loan losses when you are trying to measure expected loss over a full loan loss cycle. What we tend to observe is lots of relatively small values when economic conditions are benign and a few very large losses when the cycle turns down, probably amplified by endogenous factors embedded in bank balance sheets or business models. This has interesting and important implications for the concept of Expected Loss which is a fundamental component of the advanced Internal Rating Based approach to bank capital adequacy measurement.

Mauboussin concludes with a list of ten suggestions for untangling and navigating the divide between luck and skill:

  1. Understand where you are on the luck skill continuum
  2. Assess sample size, significance and swans
  3. Always consider a null hypothesis – is there some evidence that proves that my base  belief is wrong
  4. Think carefully about feedback and rewards; High quality feedback is key to high performance. Where skill is more important, then deliberate practice is essential to improving performance. Where luck plays a strong role, the focus must be on process
  5. Make use of counterfactuals; To maintain an open mind about the future, it is very useful to keep an open mind about the past. History is a narrative of cause and effect but it is useful to reflect on how outcomes might have been different.
  6. Develop aids to guide and improve your skill; On the luck side of the continuum, skill is still relevant but luck makes the outcomes more probabilistic. So the focus must be on good process – especially one that takes account of behavioural biases. In the middle of the spectrum, the procedural is combined with the novel. Checklists can be useful here – especially when decisions must be made under stress. Where skill matters, the key is deliberate practice and being open to feedback
  7. Have a plan for strategic interactions. Where your opponent is more skilful or just stronger, then try to inject more luck into the interaction
  8. Make reversion to the mean work for you; Understand why reversion to the mean happens, to what degree it happens, what exactly the mean is. Note that extreme events are unlikely to be repeated and most importantly, recognise that the rate of reversion to the mean relates to the coefficient of correlation
  9. Develop useful statistics (i.e.stats that are persistent and predictive)
  10. Know your limitations; we can do better at untangling skill and luck but also must recognise how much we don’t know. We must recognise that the realm may change such that old rules don’t apply and there are places where statistics don’t apply

All in all, I found Maubossin’s book very rewarding and can recommend it highly. Hopefully the above post does the book justice. I have also made some more detailed notes on the book here.

Tony

Looking under the hood – The IRB formula

This post is irredeemably technical so stop here if that is not your interest. If you need to understand some of the mechanics of the formula used to calculate credit risk weighted assets under the advanced Internal Ratings Based (IRB) approach then the BCBS published a paper in 2005 which offers an explanation:

  • describing the economic foundations
  • as well as the underlying mathematical model and its input parameters.

While a lot has changed as a result of Basel III, the models underlying the calculation of Internal Rating Based Capital (IRB) requirements are still based on the core principles agreed under Basel II that are explained in this BCBS paper.

The notes in the linked page below mostly summarise the July 2005 paper with some emphasis (bolded text) and comments (in italics) that I have added. The paper is a bit technical but worth reading if you want to understand the original thinking behind the Basel II risk weights for credit risk.

I initially found the paper useful for revisiting the foundation assumptions of the IRB framework as background to considering the regulatory treatment of Expected Loss as banks transition to IFRS9. The background on how the RW was initially intended to cover both Expected and Unexpected Loss, but was revised such that capital was only required to cover Unexpected Loss, is especially useful when considering the interaction of loan loss provisioning with capital requirements.

Reading the BCBS paper has also been useful for thinking through a range of related issues including:

  • The rationale for, and impact of, prudential conservatism in setting the risk parameters used in the IRB formula
  • The cyclicality of a risk sensitive capital requirement (and potential for pro cyclicality) and what might be done to mitigate the risk of pro-cyclical impacts on the economy

If you have read this far then my summary of the BCBS paper and my comments /observations can be found here (and thank you).

I am not a credit risk model expert, so the summary of the paper and my comments must be read with that in mind. I did this to help me think through some of the issues with bank capital adequacy. Hopefully others will find the notes useful. If you see something wrong or something you disagree with then let me know.

APRA’s proposed revisions to capital requirements for residential mortgages

… there is a lot to like in what APRA have proposed but also some issues that would benefit from further thought

Many readers will be aware that APRA released a Discussion Paper (DP) last week titled “Revisions to the capital framework for authorised deposit-taking institutions”.   The paper sets out APRA’s proposed changes to ADI capital requirements defined by the Internal Ratings Based Approach (IRB) and Standardised Approach to Credit Risk, Interest Rate Risk in the Banking Book (IRRBB) and Operational Risk. The focus of this post will be the proposals impacting credit risk capital requirements for residential mortgage lending. This post presupposes that the reader is familiar with the detail of what APRA has proposed. For those of you who have not yet got around to reading the whole paper I have added a short summary of the proposals below (see “APRA’s proposals – in more detail”).

My gut reaction is that there is a lot to like in what APRA have proposed but there are also issues that deserve further consideration in order to address the risk of unintended consequence and to better deliver on the objectives of consistency, transparency and competitive neutrality.

Proposals which make sense to me:

  • The increased risk sensitivity of the proposed standardised RWs for residential mortgages is, I believe, a material enhancement of the capital adequacy framework
  • There are arguments (and indeed evidence) for why investor property loans can be as low risk as owner occupier loans (most of the  time) but APRA’s desire to address the systemic tail risk of this form of lending is I think an understandable policy objective for a prudential regulator to pursue
  • Continuing to pursue higher IRB RW via changes to the correlation factor also looks to be a better approach than the 20% floor on LGD currently applied and thankfully also up for revision
  • Applying a higher correlation factor to low PD loans also makes intuitive sense, especially if your primary concern is the systemic risk associated with the residential mortgage lending that dominates the balance sheets of your banking system
  • In addition, the potential for the correlation adjustment to reduce the sensitivity of residential mortgage RWA to the economic cycle (and hence reduce the risk of pro-cyclical stress on capital ratios) is particularly welcome though I believe there is much more to do on this general issue
  • The support for Lender’s Mortgage Insurance (LMI) is also welcome

Areas where I believe the proposed revised capital framework could be improved (or at least benefit from some more thought):

  • The discussion of relative standardised and IRB RW does not address the fact IRB banks are required to hold additional capital to cover any shortfall between loan loss provisions and Regulatory Expected Loss (REL)
  • Residential mortgage portfolios subject to the standardised approach should be subject to a minimum average RW in the same way that IRB portfolios are currently constrained by the 25% floor
  • Applying a fixed scalar to Credit RWA can be problematic as the composition of the loan portfolio continues to evolve

The discussion of comparative IRB and Standardised RW you typically encounter seems to assume that the two approaches are identical in every aspect bar the RW but people working at the coal face know that the nominal RW advantage the IRB banks have has been partly offset by a higher exposure measure the RW are applied to. It appears that APRA’s proposed revisions will partly address this inconsistency by requiring banks using the Standardised Approach to apply a 100% Credit Conversion Factor (CCF) to undrawn loan limits.  IRB banks are also required to take a Common Equity Tier 1 deductions for the shortfall between their loan loss provisions and REL. The proposed revisions do nothing to address this area of inconsistency and in fact the Discussion Paper does not even acknowledge the issue.

Residential mortgage portfolios subject to the standardised approach should be subject to a minimum average RW in the same way that IRB portfolios are constrained. The majority of new residential mortgages are originated at relatively high LVR (most at 70% plus and a significant share at 80% plus), but the average LVR will be much lower as principal is repaid (and even more so if you allow for the appreciation of property values).  The introduction of a 20% RW bucket for standardised banks poses the question whether these banks will have an advantage in targeting the refinancing of seasoned loans with low LVR’s. The IRB banks would seek to retain these customers but they will still be constrained by the 25% average RW mandated by the FSI while the standardised banks face no comparable constraint.

This is unlikely to be an issue in the short term but one of the enduring lessons learned during my time “on the inside” is that banks (not just the big ones) are very good at identifying arbitrages and responding to incentives. It is widely recognised that housing loans have become the largest asset on Australian bank balance sheets (The Royal Commission issued a background paper that cited 42% of assets as at September 2017) but the share was significantly less when I started in banking. There has been a collection of complex drivers at play here (a topic for another post) but the relatively low RW has not harmed the growth of this kind of lending. Consequently, it is dangerous to assume that the status quo will persist if incentives exist to drive a different outcome.

This competitive imbalance could be addressed quite simply if the standardised banks were also subject to a requirement that their average RW was also no lower than 25% (or some alternative floor ratio that adjusted for the differences in exposure and REL noted above).

Another lesson learned “on the inside” is that fixed scalars look simple but are often not. They work fine when the portfolio of assets they are scaling up is stable but will gradually generate a different outcome to what was intended as the composition of the loan book evolves over time. I don’t have an easy solution to this problem but, if you must use them, it helps to recognise the potential for unintended consequence at the start.

Read on below if you have not read the Discussion Paper or want more detail on the revisions APRA has proposed and how these changes are proposed to be reconciled with the FSI recommendation. This is my first real post so feedback would be much appreciated.

Above all, tell me what I am missing … 

Tony

Note: The original version of this post published 22 February 2018 stated that inconsistent measurement of the exposures at default between the standardised and IRB approaches  was not addressed by APRA’s proposed revisions. I believe now that the proposed application of a 100% CCF in the Standardised Approach would in fact address one of the areas of inconsistency. The treatment of Regulatory Expected Loss remains an issue however. The post was revised on 24 February to clarify these points.

APRA’s proposals – in more detail

Good quality loans fully secured by mortgages on occupied residential property (either rented or occupied by the borrower) have been assigned concessionary risk weights (RW) ever since risk weighted capital adequacy ratios were introduced under Basel I (1988). The most concessionary risk weight was initially set at 50% and reduced to 35% in the Basel II Standardised Approach (2006).

APRA currently applies the concessionary 35% RW to standard eligible mortgages with Loan Valuation Ratios (LVR) of 80% or better (or up to 90% LVR if covered by Lender’s Mortgage Insurance) while the best case scenario for a non-standard mortgage is a 50% RW. Progressively higher RW (50/75/100) are applied for higher risk residential mortgages.

Under the Standardised Approach, APRA proposes:

  • The classification of a Standard Eligible Mortgage will distinguish between lowest risk “Owner-occupied P&I” and a higher risk “Other residential mortgages” category which is intended to be conceptually similar to the “material dependence” concept employed by Basel III to distinguish loans where repayment depends materially on the cash flows generated by the property securing the loan
  • 6 RW bands for each of these two types of residential mortgage (compared to 5 bands currently)
  • Standard Eligible Mortgages with lower LVR loans to be assigned lower RW but these loans must also meet defined serviceability, marketability and valuation criteria to qualify for the concessionary RW
  • The higher RW applied to “Other residential mortgages” may take the form of a fixed risk-weight schedule (per the indicative RW in Table 3 of the Discussion Paper) but might also be implemented via a multiplier, applied to the RW for owner-occupied P&I loans, which might vary over time “… depending on prevailing prudential or financial stability objectives or concerns”
  • Relatively lower capital requirements to continue to apply where loans are covered by LMI but its preferred approach is to apply a RW loading to loans with LVR in excess of 80% that are not insured (i.e. the indicative RW in Table 3 assume that LMI covers the high LVR loans)
  • Non-Standard residential mortgages should no longer benefit from any RW concession and be assigned a flat 100% RW irrespective of LVR and LMI

While the IRB requirements impacting residential mortgages are largely unchanged under Basel III, APRA proposes the following changes to the Australian IRB Approach to reflect local requirements and conditions:

  • Increased capital requirements for investment and interest-only exposures; to be implemented via a higher correlation factor for these loans
  • The (currently fixed) correlation factor applied to residential mortgages to be amended to depend on probability of default (PD); reflecting empirical evidence that “… the default risk of lower PD exposures is more dependent on the economic cycle  and can consequently increase at a relatively higher rate in a downturn”
  • A reduction in the minimum Loss Given Default (LGD) from 20% to 10% (subject to APRA approval of the LGD model); in order to facilitate “… better alignment of LGD estimates to key drivers of loss such as LVR and LMI”
  • Capital requirements for non-standard mortgages use the standardised approach; increasing consistency between the IRB an standardised approaches

APRA’s proposals seek to strike a balance between risk sensitivity and simplicity but must also take account of the FSI recommendations that ADI capital levels be unquestionably strong while also narrowing the difference between standardised and IRB RWs for residential mortgages. APRA is undertaking a Quantitative Impact Study (QIS) to better understand the impact of its proposals but the DP flagged that APRA does not expect the changes to correlation factors to meet its objectives for increased capital for residential mortgage exposures.

APRA could just further ramp up the correlation factor to generate the target IRB RW (which I assume continues to be 25%) but the DP notes that this would create undesirable inconsistencies with the correlation factors applied to other asset classes. Consequently, the DP indicates that the target increase in IRB RWA will likely be pursued via

  • A fixed multiplier (scalar) applied to total Credit RWA (i.e. althoughBasel III removes the 1.06 Credit RWA scalar, APRA is considering retaining a scalar with a value yet to be determined); and
  • If necessary, by applying additional specific RWA scalars for residential (and commercial) property.

These scalars will be subject to consultation with the industry and APRA has committed to review the 10.5% CET1 benchmark for unquestionably strong capital should the net result of the proposed revisions result in an overall increase in RWA’s relative to current methodologies.