My two cents on Ross Levine’s expert report: deep thinking on incentives, dynamics, and moral hazard.

Last week, the RBNZ released the three independent international experts’ assessments of the New Zealand Capital Review. The RBNZ commissioned three international experts in May as part of the bank capital review.

The reports by Cummings, Miles and Levine offer interesting views on bank capital. The reports are different in substance and style. However, they agree that the analyses of Reserve Bank are adequate, appropriate, balanced, well-reasoned, and thorough.

I will not dwell on the excellent reports of Cummings and Miles. It’s Levine’s report that you should to pay attention to. It contains lots of valuable and relevant insights. Levine clearly understands banking and bank capital. More interestingly, however, is the way he has written his report. He carefully respects the terms of reference, while at the same time his report is quite critical and outspoken.

A first-rate team that ignored important studies

For example, he stresses that the analyses of the RBNZ are unbiased. He also praises the RBNZ team working on the capital review for being first-rate analytically.

At the same time, however, Levine notices that the team missed out on important areas of literature. Specifically (but not exclusively) the lack of attention paid to incentives and dynamics. According to Levine this “limits the analyses of how capital requirements influence bank stability and efficiency.” These observations are not insignificant. I mean, suppose your General Practitioner admits to not having read more than half of the literature relevant for his work. Wouldn’t that worry you?

A plumbing view of banks

It is hard to not get the impression that Levine was amazed, perhaps irritated, by the lack of attention to behavioral consequences of the Reserve Bank’s analyses. Levine even uses the unflattering term “plumbing view of banks” to comment on the strong focus of the RBNZ team on the cost of bank funding. According to Levine, the allocation of credit deserves more attention.

You may get what you asked for

Another weakness of the RBNZ analysis, according to Levine, is that it does not really examine the wider consequences of the capital plans. For example, the higher capital requirements will make bank borrowing more expensive. This will attract non-bank competitors. Perhaps insurance companies or “shadow banks.” The question then is whether the Reserve Bank and other institutions will be able to deal with these newcomers. Looking at the behavior of New Zealand insurance companies, the RBNZ may want to think again. Do we want more competitors of the type that treat customers poorly, or are we happy-ish with the status quo where the Reserve Bank can concentrate its supervision resources on a few players?

Incentives are all too important

Levine criticizes the RBNZ for being overly concerned about large shocks that deteriorate asset values. He argues that crises can also be the result of excessive risk taking driven by moral hazard behavior.

Levine has a point. Since the Global Financial Crisis (poor) conduct has become an important problem for the banking industry. The Royal Commission into Misconduct in the Banking industry highlights just that: misconduct is a problem for Australian banks, four of which are the owners of subsidiaries operating in our country.

A more recent example is Dutch (state-owned) bank ABN Amro, which, despite a CET1 capital ratio of 18 percent, got into trouble because of money laundering. Similar problems affected Dankse Bank with a CET1 ratio of 17 percent and Swedbank, which until recently reported a CET1 ratio of 24.6 percent. These banks show that a focus on capital and credit risk won’t cut it. Conduct, behavior, and incentives are all too important.

The problem is that the Reserve Bank appears to struggle with conduct risk. Last year’s joint report with the FMA on Bank Conduct and Culture confirm this: “Conduct risk and customer outcomes are less easily quantified than credit risk or market risk, and are not easily controlled by standard compliance tools.” A recent examination of conduct in the insurance industry prompted the Reserve Bank to express disappointment with insurers’ response to the conduct and culture review – but not take important measures to improve that conduct.

On CoCo’s

On the question of what counts as capital, Levine initially appears to support the Reserve Bank in disallowing Contingent Convertible securities as Tier 1 capital. Having read the last pages of his report, I am not so sure. Throughout his analysis, Levine holds the view that investors in banks’ debt capital can discipline bank managers. This debtholder-monitoring and debt-holder disciplining view is common in academic bank literature. But that literature, however, does not discriminate between Tier 1 debt and Tier 2 debt.

Recent cases, e.g. Banco Popular, demonstrate that regulators also do not necessarily view Tier 1 and Tier 2 capital as separate.

As long as capital instruments count as debt for accounting and tax purposes and not as equity for prudential purposes, these securities can absorb losses and thus could count as regulatory capital.

On the last page of his analysis, Levine writes “while it is well-established that diffuse shareholders in widely-held companies typically have strong incentives to increase corporate risk after obtaining funds from debtholders, uninsured debtholders have incentives to limit this risk-shifting. Thus, well-designed Tier 2 capital can enhance incentives ... .” This reasoning could easily apply to Tier 1 Contingent Convertible debt.

Why wait for the supervisor?

Moreover, and this is the point made in my paper with Annelies Renders (and overlooked by many, including Levine), banks do not have to sit idle until a tardy supervisor decides to trigger a CoCo. Banks have the right to call these capital instruments (Tier 1 and Tier 2) whenever they want. This allows them to absorb losses in going concern. The results of our study then show that Tier 1 capital instruments, or CoCos, absorb losses more effectively than Tier 2 instruments.

Lastly, it is often mentioned that Contingent Convertible instruments are complex and onerous to supervise. Perhaps, but a Coasian solution would be to allow banks to hire an expert to determine the capital eligibility of convertible capital instruments. That is a far more efficient and effective solution than employing a platoon of capital experts to vet and monitor capital instruments. In fact, some of the New Zealand banks offered to pay experts for the vetting of capital instruments themselves. Apparently to no avail.

The politics of capital buffers: the IMF

Though the three experts discuss the countercyclical capital buffer, little attention goes out to the conservation buffer. That is a shortcoming. For example, the IMF in a recent report advised the Reserve Bank to align the conservation buffer with international standards. The IMF wants to double the buffer for domestic systemically important banks (D-SIB buffer) to 2 percent with a corresponding decrease of the capital conservation buffer for the other banks “since much of the structural systemic financial risks identified in the FSAP arise from the D-SIBs.”

According to the IMF, there is also “scope for a somewhat smaller increase of the general conservation buffer, given that other regulatory changes since the global financial crisis likely also lowered the probability of bank failure (e.g., higher liquidity requirements).”

I agree with the IMF, the proposed conservation buffer is large. With 7.5 percent it occupies about half of the proposed capital requirements. That’s a lot.

I am also worried about the location of the conservation buffer and the behaviors that it will elicit. The buffer resides at the top of the proposed capital stack. That makes me feel uncomfortable, because banks will likely hold a very small management buffer over the required amount of capital. See the picture below, which shows a negative relation between capital requirements and management buffers for European banks over the years 2012-2018.

Capital buffers are no cheese dips

Don’t be surprised to see our banks holding just sufficient capital to meet the requirements. The risk is that, at times, banks will start dipping into the conservation buffer.

What will the Reserve Bank do when a bank dips into the conservation buffer? Ross Levine offers some clues in his discussion of the countercyclical capital buffer: there is a risk that the conservation buffer will be mismanaged or abused for political purposes. To illustrate my point, this line from the 4th Capital Review paper is particularly worrying: “banks … would not be in breach of … conditions if they were to operate inside the prudential capital buffer …, for example following a large loss event.

For the same reasons that make Levine worry about the RBNZ’s ability to supervise the countercyclical buffer, I am worried that the supervisor would have an imperfect understanding of the nature and persistence of the reasons why a bank dips into that buffer.

A possible consequence of the current setup of the conservation buffer is that it may lack credibility once banks learn how to circumvent the proposed automatic restrictions. This can be done, for example, by timing the recognition of net earnings. The consequence is that banks can freely dip in to the buffer, as long as these dips are assumed to be temporary. That, to me, is a worry.

My suggestion therefore would be to follow the IMF and Basel III: keep the conservation buffer at 2.5 percent and locate it near the bottom of the stack. This will give the supervisor more leverage when it comes to enforcing restrictions on payments, bonuses, etc.