In this section
The comment stream
- 1 of 31728
- 1 of 434
The news stream
- RBNZ to go back to its toolbox? 30
- While you were sleeping: Oil sinks as OPEC holds 27
- IRD targets property speculators 17
- Bernard's Top 10 15
- $512k gets you 56sq m in downtown Auckland 10
- Housing costs rising faster than incomes 8
- Smith and Auckland Housing Accord praised and damned 8
- What happened Thursday 6
- Friday's guest Top 10 5
- Disappointing US data 5
Alfred Duncan suggests a possible way of better aligning bankers’ incentives with financial-system safety and capital-allocation efficiency. You agree?
By Alfred Duncan*
The Reserve Bank of New Zealand (RBNZ) regulates bank leverage in accordance with the Basel II capital adequacy framework adopted in 2008.
These regulations are designed to ensure that banks’ liabilities include a significant share of loss-absorbing capital, ideally reducing the risk of bank failures and lessening the incentive to rely too much on on-demand deposit funding.
This recognises the ‘moral hazard’ problem that occurs when implicit or explicit government guarantees on bank deposits separate bank risk from the interest rates demanded by depositors.
Under Basel II, broad capital requirements are prescribed against different types of assets. Each asset is assigned a risk weight and the bank’s equity share of funding must exceed 8% of aggregate risk-weighted assets.
In this way, rather than being a constant limit on bank leverage, the leverage limits prescribed by the Basel II framework are dependent on the risk of the assets held by the bank, at least as adjudicated by the regulations. In lieu of accepting prescribed risk weights, banks can apply weights determined by their own internal risk models (providing these models and their inputs are approved by the RBNZ).
In June 2011, the RBNZ adjusted its estimates of risk parameters for the determination of rural-lending risk weights. Banks using internal risk models are required to adopt the RBNZ’s estimates for the expected losses in the event of default for all farm lending.2 This will increase the risk weights associated with rural loans and therefore the amount of regulatory capital those banks are required to hold against these loans.
Banks that do not use their own risk models, including TSB and SBS, are unaffected by this change.
Equity capital is an expensive source of funding for banks. The increased capital requirements will force higher interest rates on farm borrowers and will divert bank lending towards other uses including residential mortgages.
From the perspective of farm businesses, there will be less bank credit available; this will encourage borrowing from non-bank financial institutions such as finance companies and from foreign lenders.
Family farming businesses relying greatly on local bank finance will see their costs of borrowing increase and may struggle to compete for funding with corporate farm businesses (which have access to cheaper, perhaps foreign, capital).
Vertically integrated banks can often operate more efficiently than other finance companies in markets such as farm lending. This is because the relationships built between the bank and the borrower can be developed over years and through a range of products. Over time, the bank increases its knowledge about the borrower and is able to better judge the risk associated with the borrower’s loans.
Banks with personal relationships may be able to offer loans at lower interest rates, and more adroitly renegotiate or modify loans when covenants are broken or payments are missed.
As bank-intermediated funding is replaced by direct non-bank financing in this and other markets, bank risks associated with rural loan performance may subside. But these new unregulated and more opaque capital flows may bring with them new systemic risks.
Simply shifting capital flows from banks to other financial institutions is unlikely to reduce the risks of future financial crises.
Moreover, the RBNZ’s main levers for supporting the financial system in times of crisis (the target interbank lending rate or ‘official cash rate’, and direct lending to banks) will likely be far less effective in combating crises arising from disruptions to non-bank financial institutions.
Heading off bank runs
The on-demand nature of bank deposits leaves them reliant on depositor sentiment. At any time, depositors could withdraw their funds: an ‘en masse’ withdrawal creates a so-called bank run.
Banks not holding sufficient cash to meet withdrawals may need to liquidate assets at any price, and low prices could lead to bank insolvency.
In order to tackle the risks associated with bank runs, the RBNZ has introduced a limit on the extent to which banks can rely on short term or on-demand debt. The core funding ratio places a lower bound on retail deposits, long-term debt and equity funding as a share of the total liabilities of New Zealand banks. The contribution of retail deposits to regulatory core funding is dependent on the concentration of those deposits.
For example, ten depositors each with $100,000 deposited contribute more regulatory core funding than one depositor with $1,000,000 in their account.
These rules affect markets in a number of ways. Most importantly, the requirement increases the costs associated with wholesale and (to a lesser extent) retail deposit funding. While retail deposit funding may be less volatile than wholesale funding, it is no less subject to shifts in sentiment.
With electronic banking, household depositors can withdraw their funds very quickly. The dramatic queues of depositors outside failed UK bank Northern Rock in September 2007 also provide a stark reminder of the spectre of bank runs.
No more ‘them versus us’
The RBNZ regulates bank activity directly through capital and liquidity controls. In this way, bankers have incentives to circumvent RBNZ rules.
A better approach would align the incentives of bankers with the safety of the financial system and the efficient allocation of capital.
Specifically, regulating liability rather than activity would encourage banks to reduce risk while allowing them to choose the most efficient way of doing so.
Most large banks are incorporated as single liability companies: shareholders stand to lose only their initial investment if the bank becomes bankrupt. This contrasts with extended liability for bank shareholders, which was popular in the UK in the 19th century: if the bank failed, shareholders were jointly and severally liable for losses over and above the amount invested.
One common form of extended liability was multiple liability, where shareholders were liable up to a predetermined multiple of the face value of their shares (a system which was in place in New Zealand until the 1950s).
Extended-liability regimes force depositors to monitor the shareholders as well as the managers of the bank.
Shareholders also must consider their counterparts’ ability to pay their share if bankruptcy occurs. Where it is more costly to monitor shareholders than managers, single liability is likely to be more efficient than extended liability.
In many large OECD countries some banks issue shares publicly; these are freely traded in stock exchanges and private transactions. Under extended liability, depositors would need to monitor shareholders and shareholders would need to monitor each other. The added complexity of the contract could also decrease liquidity in the market for such shares. These factors could make it very difficult for banks to raise the same amounts of capital that they could raise under a single-liability regime.
In New Zealand, however, the five major banks (ANZ,3 Westpac, ASB, BNZ and Kiwibank) are wholly owned by single shareholders. Their shares are not publicly traded so would not suffer from the aforementioned costs associated with decreased share liquidity and cross-shareholder monitoring.
Reducing moral hazard
One potential form of extended liability might take the form of a guarantee of a given percentage of the retail deposits of the bank.
Retail deposits are most likely to be guaranteed by the government in the case of bankruptcy because of their importance to trade and to households.
The expectation of government guarantees causes moral hazard: depositors don’t consider the health of their bank, as they expect to be guaranteed by the government even if the bank fails.
An extended-liability scheme forcing bank shareholders to guarantee a significant share of retail deposits would shift the costs of this moral hazard from the government back to the banks’ shareholders.
Limited liability can encourage bank shareholders to extract deposits from the bank if they suspect that their bank is insolvent. They stand to lose only the deposits that are tied up in the bank. Under extended liability, shareholder deposits invested in the bank are independent of the depositors’ claim on the shareholders in case of insolvency. Shareholders are more likely to keep funds in the bank, potentially leading to a faster and smoother resolution or sale of the bank.
Regulating liability rather than activity thus offers new options for consideration in the New Zealand regulatory debate.
1. This project was suggested by Lewis Evans, who also provided comments on earlier drafts. A formal analysis of extended liability regimes, and a description of the Scottish banking market of the 19th century (where banks of varying liability regimes competed) can be found in: L Evans & N Quigley (1995) ‘Shareholder Liability Regimes, Principal-Agent Relationships, and Banking Industry Performance’ The Journal of Law and Economics 38(2) pp497-520.
2. I Harrison & K Hoskin (2011) ‘Bank farm capital: does it cost the earth?’ Reserve Bank of New Zealand Bulletin 74(2) pp5- 14.
3. The National Bank is now fully subsumed into ANZ.
Alfred Duncan is a PhD student in financial economics at the University of Glasgow and a former research assistant at ISCR.
This article was first published by the ISCR in the November 2012 issue of their "Competition and Regulation Times". It is republished here with permission.