David Lewis is General Manager, Supervision Framework at the Australian Prudential Regulation Authority (APRA)
GB: David, thank you for being with us. I would also like to thank you once again for appearing at the Edmund Rice Business Ethics Initiative Forum on “Banking for Good or Evil”. It was a very good night and your contribution was a very significant part of that.
DL: My pleasure. I was happy to do it.
GB: David, one of the important points you made, and that the government is also emphasising is that Australian Banking Institutions are in pretty good financial shape; that the four big banks in Australia are one of thirty that still have a AA credit rating internationally.
DL: Four of the remaining twenty, I think.
GB: There were a number of reasons why that’s the case. I think one that was mentioned at the Forum was the former Reserve Bank Governor McFarlane’s comments about the four pillar bank policy. But also, good regulation has helped to contribute. Could you perhaps just go over some of that reasoning, please?
DL: There have been a number of reasons put forward as to why Australia seems to have withstood the global financial crisis better than most other developed countries, ranging from very good management to just plain good luck – and good regulation for that matter. From APRA’s point of view, some of the explanations put forward are just superficial ex post rationalisation, rather than any genuine attempt to analyse cause and effect. In reality, we tend to get more credit than we probably deserve when things go well and we tend to get more blame than we really deserve when things go bad.
In my view, there is no single factor underpinning Australia’s relative success, but rather a combination of reasons:
- maintenance of sound fiscal and monetary policy settings;
- the structure of the market; and
- conservative prudential regulation.
Each of these factors played there part; without any one of these, Australia’s position could have been much worse.
Certainly, as Ian McFarlane alludes to, structural differences, not just the four pillars policy but a range of things, have really helped the Australian position vis-à-vis other markets. The dynamics of mortgage lending in Australia are fundamentally different from the United States. In the United States they have what they call an “originate-to-distribute” model. That means a separation of all the lending processes down to mortgage originators, packagers, securitizers, wholesalers and asset managers. It is a much more disaggregated model than that which prevails in Australia. The problem in that model is that ownership of risk gets lost, particularly, in an expanding market. –Under such a model, the people who make lending decisions do not really have any skin in the game: they do not have to live with the consequences of their decisions. Sure, the people who buy loan securities should enforce greater discipline on the sellers of those securities, but the transparency of the underlying asset quality is lost and so is the capacity to exercise market discipline – and that was a fundamental breakdown that precipitated the crisis in the United States.
The model in Australia is different. While our banks do securitize mortgage loans, the predominant model in Australia is an intermediation model in which banks retain the bulk of the credit risk on the loans they originate and, where loans are originated through mortgage brokers, the banks impose significant disciplines on mortgage originators to ensure that the bank’s credit standards are maintained.
Secondly, as Ian MacFarlane has pointed out, the banking oligopoly in Australia has served us well. It has afforded us institutions that are fundamentally sound and, generally, well-managed. They have been spared from a lot of the takeover pressures that has been evident in overseas markets due to the presence of the four pillars policy.
And, while I would not want to overstate it, effective prudential regulation has also played its part. At the Forum, I drew out a couple of examples of this. APRA has always erred on the conservative side when it comes to capital adequacy. We have always been wary of “sub-prime” lending (as have our banks), and in 2004 we introduced higher capital charges for “loc-doc” loans.
Also, it has always been an APRA requirement that, where banks originate loans through brokers, they must ensure that the brokers apply the bank’s own credit assessment standards (ADI Prudential Standard 112). We also require that banks monitor and audit all loans coming in through those channels to ensure that they continue to comply with the bank’s lending criteria. We would also note that best practice in this area is for the bank to pay a lagged or a risk-based commission to brokers, so as to ensure that their incentives are aligned to risk.
GB: There are some examples of Australian banks doing some – at least with the wisdom of hindsight – very strange business. I am thinking of the OPES Prime and Storm Financial as two recent examples.
DL: Well, while we expect our banks to maintain the highest risk management and ethical standards, we can’t expect them to be omnipotent. It’s in the nature of any risk management system that there will be failures. No system is ever going to be perfect. To me, a good risk management system is one that is constantly evolving, recognising that things are going to get through from time to time but, when they do, you learn from those mistakes and build a better system to improve your risk management capability going forward.
A concrete example: NAB FOREX trading
GB: When you say, “learn from your mistakes”, what does that mean in practice? Are there changed practices? Is there something you can point to that banks or you change in response to, say, a situation like the well-publicised NAB foreign exchange options losses of a couple of years’ ago? Were you involved in that investigation?
DL: Yes. I was.
GB: And NAB has subsequently published the report, so we can all see it.
DL: The situation at NAB in 2004 was highly unusual – not least because these sorts of things are not normally played out in public. Not that it had much choice in the matter, but I think that it was to NAB’s great credit that it was prepared to bare its soul in public and clearly stake out that it was prepared to change.
In one sense you could say that the problems at NAB were quite particular. They were related to manipulation of a trading book that occurred on just one desk in its trading room – the foreign currency options desk. There were just four people directly involved inside a huge trading room. Even the losses that emerged were relatively small. At no stage did they threaten the bank or its ability to meet its obligations to its beneficiaries. The losses themselves were not significant; it was their underlying cause that so exposed the bank.
The real issue was what these losses revealed about the culture of the organisation. It was a culture that was essentially self-reinforcing. It was a ‘can-do’ culture in which bad news didn’t travel up. It was a “good news” culture. Nobody wanted to deliver any bad news. In that sort of environment, you don’t learn from issues that emerge. Small issues become big ones because they’re not headed off early.
GB: So, how do they manage to stop the flow of bad news? How is that done?
DL: It all comes down to organisational culture. Organisations are very good at promoting certain types of behaviour – sometimes unwittingly. People soon learn that – irrespective of what is said – certain types of behaviour get rewarded and reinforced.
GB: So, based on the way bonuses are handed out, everyone picks up the message pretty quickly?
DL: Yes and by simple positive reinforcement. Who is it that gets rewarded? Who does well or why they are doing well? Rarely, do the plaudits go to people who stand up and point out things that are going wrong.
GB: Was there some effective stopping of risk-checking in that area of the bank?
DL: There were breakdowns on a number of levels. Within the desk itself, there was a breakdown in proper risk management practices. The head of the trading desk was a very dominant personality and he was able to manipulate the people around him and exert undue influence on back office control staff. The traders were able to manipulate the trading system by inputting ‘dummy’ trades which had the effect of falsifying reports that came out of the system. Various oversight committees failed to pick this up. Even though there were alarm bells ringing, these were summarily dismissed and not acted on. It was another manifestation of that “good news” culture.
NAB’s senior executives were genuinely shocked when they started to realise the consequences of all this and that’s why they accepted the need to make substantial changes – which they have done.
GB: In the chapter of the APRA report on the organisational culture, the culture was described as one where the central value was “Profit is King”. Did you understand that “Profit is King” is something that guided the structures or was it just a symptom of the way things were? Did it actually create things?
DL: Again, this was a manifestation of the culture. It was never an articulated goal; just something that took hold in practice. The irony is that expression was one that was used by the traders themselves. They realised that, as long as they were making money, they would be afforded a lot of lee-way.
GB: So, the actions would be things like the way they award bonuses; the way they might control questions about risky-looking figures?
DL: It’s not as if questions weren’t being raised from time to time by control areas such as internal audit, but they were quickly brushed aside.
GB: So, it sounds like some sort of bullying going on?
DL: Yes, sort of. As I say, at a management level, it was probably unconscious. It easy to dismiss audit reports as “nitpicking”.
Characteristics of a good culture
GB: So what would be a good way of describing a good organisational culture?
DL: That’s actually a very difficult thing to do. A lot of CEOs will tell you, “We’ve got a really good culture”. I’d say to them, “Well, what is that? How do you know that you’ve really got a good culture?” Because, I think it’s actually quite a difficult thing to tie down. I think the most important characteristics of a good culture would be: one that is capable of evolving; one that recognises that things can change so the organisation needs to be willing to question and challenge the existing mindset. These are the cultures that are best placed to deal with the unexpected, whether it’s just foreign exchange options trading or a global economic crisis. They’re the ones that are best placed to deal with the things that come up in any risk business.
GB: And by implication, they don’t shoot the bearers of bad tidings?
DL: Yes. You’ve got to give your people the chance to get their views up and properly consider a view that might be challenging to the status quo.
GB: You might look at your bonus-awarding procedures to not discourage the discordant voice?
DL: Financial institutions need to take a long-term focus when it comes to performance rewards because the decisions you take may yield a profit today but harbour risks that may not be revealed for some time into the future. They need to make sure the bonus structure is capable of capturing the full impact of the decision over time, so some form of staggered delivery of performance-based payments is important. It’s also important for financial institutions to have a culture that’s just open to questioning. A good “whistle-blowing” policy is an important component of this.
GB: Returning to your description of one of the failings of the mortgage market in the US about no “skin in the game” or implicit conflicts of interest, is there connection between that problem and the way incentives are awarded in Australia?
DL: The extent to which mis-aligned incentive structures have contributed to the current financial crisis is receiving a lot of attention at the moment. (APRA has recently published new draft prudential standards on this issue.) While much of the media attention is on the level of executive remuneration, the issue for a prudential regulator is: “What impact do remuneration practices have on the decisions that are made about risk?” As I have said, a longer-term approach to remuneration is important because the risks in the business that a financial institution undertakes can sometimes take a year or two to manifest themselves. Business conditions can change and not all potential risks are apparent at the point of decision. On the other hand, if lenders are rewarded solely on the basis of volume of business written (as was the case with most of the broker-originated lending in the United States), this will encourage unsound risk-seeking behaviour: loans will be written to yield a short-term bonus without regard to the risk being carried over the longer term.
GB: What would you say would be a proper long-term time scale? Are we talking about two years, three years, ten years?
DL: That depends on the type of risk involved. Incentive structures should align with the risk profile of the business being written. But, yes, generally, for a lending business we would expect to see a deferral of some two or three years. And there should be some capacity for the financial institution to reduce or eliminate performance-based remuneration in line with actual experience.
Characteristics of good regulation
GB: Could we move on to another matter you raised at the Forum, namely the difference between good and bad regulation? If I remember correctly, you talked about principles-based regulation: trying to get the people you regulate to buy into the principle, rather than trying to tie everything down as perhaps, to take an example, Sarbanes-Oxley might be seen to attempting to do.
DL: Yes, that’s right. At APRA, we are strong adherents of a ‘principles-based’ approach to regulation. The effectiveness of any regulation depends upon the underlying purpose and making sure the scheme of regulation is well-designed to meet that underlying purpose. The primary focus of prudential regulation is risk prevention. Our risk remediation role only comes into play where risk prevention has failed (which from time to time it will do). So, unlike most other regulators who act in response to observed breaches of the regulations, a prudential regulator tries to intervene before a problem can manifest itself. With prevention as the goal, a heavily rules-based system just won’t work. You just can’t prescribe every kind of risk that could ever eventuate in a set of rules.
So, while we too have ‘rules’, they are “principles-based”. In designing prudential standards, we seek to identify those principles which best describe sound risk management in the area and then tailor our regulatory interventions to these objectives. Effective regulation has very little to do with the volume of the regulations, and everything to do with the way in which these are implemented. As a crude illustration of this, you only have to look at the volume of banking regulation in Australia in comparison to Australia: in the United States, there are approximately 5,000 pages of banking regulations, whereas in Australia the figure is around 500. (And, if you compare relative font size, the differential is even greater!)
The point is that effective prudential regulation can never be totally imposed from the outside. We believe that prudential regulation works best when the goals of the regulation – broadly, the adoption of sound risk management practices – are internalised within regulated financial institutions. Some academics use the term “meta-regulation” to describe this model.
GB: How does the process of sharing those principles work? I imagine that there are a number of stakeholders in the process, and you are seeking good “buy-in” from all of them?
DL: Well, just like the financial institutions that we regulate, APRA has to maintain an open mind and be willing to engage in a constructive dialogue with regulated institutions about risk management priorities. First of all, we need to be clear about is the objective we’re trying to achieve. Under a principles-based approach, the focus is on the achievement of good risk management outcomes and less on the way in which that outcome is achieved. This means that, if the prudential outcome we have set out in our prudential standards can be achieved in a different way from the one we have prescribed, we are open to looking at that. The outcome is the important thing.
Secondly, a lot of our prudential oversight takes place outside the bounds of the formal prudential standards and laws. We receive a lot of data from the institutions that we supervise, as well as conducting a regular program of on-site prudential reviews which target specific risk areas for investigation. So, one way or another, we are monitoring the activities of regulated financial institutions most of the time. When we see potential risk issues that concern us, we act on them – usually by raising the issue with the institution concerned as part of our regular dialogue so that it can address the particular issue. Of course, if we think the financial institution’s response is inadequate or unsatisfactory, it is open to us to take some form of enforcement action. For example, we might form a conclusion that the institution is taking on a greater level of risk and that this might warrant the imposition of a higher capital charge. So, again, there is a feedback loop between risk and return.
GB: This requires a high level of constant engagement.
DL: That goes back to the cultural issue again. A stable, well-regulated financial system is not only for the benefit of consumers of financial services, but also in the interests of financial institutions themselves. In that sense, the prudential regulator is one of the key stakeholders that financial institutions need to engage with, so when we raise things we think are issues or problems, we expect to get heard and have those issues taken seriously. This needs to be a two-way process: it may be that the institution concerned will respond with an alternative proposition. The important thing is that there is a dialogue that leads to an appropriate risk management outcome. It’s all part of maintaining a questioning culture; we are just one of the avenues by which those questions can arise because we are bringing a broader risk perspective from outside the organisation.
How wide should stakeholder involvement be?
GB: I’ve asked a number of bankers about the importance they might have of community groups as serious stakeholders in that same vein. Not that they’re going to be experts on risk management, but they can be providing feedback about what the bank or other large businesses’ effects are on the community. For example, in the context of mortgages, more obvious noise about the difficulties people were facing might have helped the banks pull up faster.
DL: Given the role that banks play in our community, I think it’s important that they remain open to a broad church of opinion. When it comes to recouping debts via mortgage foreclosure, a bank needs to be conscious of the duties it owes to its depositors as well as to its borrowers. Foreclosing a mortgage is an expensive and difficult process for all concerned; it is not in the bank’s best interest if it can be avoided. That’s basically a last resort, and I think that banks recognise that. If they can do things to assist their customers to try and get through a short-term problem, that’s ultimately in their best interest.
GB: Another issue related to this that was also raised in the Forum, someone argued that people who really are on “struggle-street” are having credit cards pushed at them by the banks. Might this be a situation where the banks may need to listen at least to this very, perhaps, peculiar sector, but it’s quite important because it’s connected to very significant levels of misery?
DL: That’s more an issue for ASIC, I’d suggest. I think David Bell (CEO of the Australian Bankers Association) commented during the Forum that banks have models for whom they target for offering additional credit card limits and those models are based on people’s ability to repay based on their repayment history and so on. But, like any model, there are always gaps and I guess it’s incumbent on banks to go back and look at those models to see how effectively they’re generating these additional offers.
GB: The models can also be suffering from short-termism, as you implied earlier. For example, the bank might take the attitude that as long as it gets payment back for the first couple of years and then it doesn’t care what happens afterwards.
DL: Yes, possibly. But, again, that’s an issue relating to selling practices. It’s not something we deal with at APRA.
Focus on the future?
GB: Dr Jerry Jordan, ex-Chairman of the Cleveland Branch of the US Federal Reserve, who’s been on a speaking circuit, sponsored by US Embassies, both in New Zealand and here. I notice that in one of his talks, he was asked about better regulation for the banks – I suppose the question really being about the US – and his response, as I understood it, was that there are quite good rules and regulations in place in the US. The problem has been the lack of supervision. You’ve just been painting a picture of fairly close supervision, close ongoing engagement with the banks, but he was painting a picture of very large financial institutions with very complex financial products, like those CDOs and no one was really supervising what was going on. He says that if the organisation is too big to fail or to big to supervise, it is just too big.
DL: I didn’t manage to catch Jerry Jordan when he was here, but I have been corresponding with him. His view is that the crisis was first failure of corporate governance, then a failure of prudential supervision. While he agrees with many of the reforms proposed, he thinks more attention needs to be paid to the way in which prudential supervision is implemented. I think he is probably right about that. As I have said, effective prudential outcomes are more likely to come from the promotion – and acceptance – of sound risk management practices, than from mountains of rules.
In many ways, Jerry Jordan reinforces the approach we have taken in Australia. In fact, if you look at the list of potential reforms he advocates for the US, many of these are already in place in Australia – for example, the reforms in relation to broker-originated lending and sub-prime loans which I have mentioned. These were major sources of market failure in the US.
GB: And there was a whole industry of untruth fabrications, wasn’t there?
DL: Yes. There was a level of complacency born of good times in which many lenders lost sight of underlying market realities. In the United States, a whole market opened up in lending to people who really had very little possibility of repaying their loan without a significant appreciation in the value of the mortgage property. This was done on a very large scale – and sometimes even promoted by Government policy. But, eventually, that bubble burst. It was very short-sighted.
GB: One of the things you need to do better is smoothing out the cycles?
DL: Yes. Internationally, that is one of the main items on the regulatory reform agenda. There is a view that current provisioning practices place too much reliance on current market values. Proposals are being developed for a provisioning regime that measures provisioning requirements through the economic cycle, so that provisions are built up in good times to help cope with inevitable market declines.
GB: I think we can leave it there. Thank you very much David.
DL: My pleasure.