In October’s GT Alert on non-bank lending, we focussed on whether there was a justification in banks being subject to regulatory capital requirements when non-bank financial institutions (NBFIs) were not. Professor Simon Gleeson’s evidence – that banks are permitted to take deposits from the public, whereas NBFIs are not – was an important justification for a differentiated regulatory approach.
In this GT Alert, we provide an overview of how the amount of regulatory capital that a UK bank is required to hold is quantified and whether there is a case for reconsidering the current approach. This is based largely on the evidence Lord King of Lothbury, the former governor of the Bank of England, provided to the Financial Services Regulation Committee (the Committee) of the House of Lords. We also consider certain other observations Lord King made in relation to the regulation of banks and NBFIs.
The Calculation of Regulatory Capital
The system of calculating regulatory capital requirements imposed on banks is largely based on the works of the Basel Committee on Banking Supervision (the BCBS), which is headquartered at the Bank for International Settlements – the so-called central bank for central banks – in Basel, Switzerland. The BCBS provides guidance on regulatory capital, liquidity, and financial stability in relation to banks and, while its pronouncements do not have the force of law, countries who participate in its work are expected to implement its recommendations into their local law. Its most prominent work is the Basel Capital Accords, which require banks to hold sufficient capital and manage risk.
While a detailed examination of the principles on which regulatory capital is quantified is beyond the scope of this GT Alert, in essence, it is largely based on a concept known as “risk weighted assets.” Conventionally, the primary assets of a bank are the loans that it makes and holds. The loans generate interest, providing the bank with income. However, these assets are also a source of risk because if the borrowers fail to pay interest or repay principal as they are required to do, the bank may suffer a loss. Regulatory capital is intended to absorb losses so that creditors of the bank, critically depositors, are not adversely affected by these losses and that their claims against a bank will be honoured, but it is not meant to protect banks against risk. It follows that a bank that engages in higher-risk lending should be required to hold more regulatory capital than one whose lending involves less risk, as it would be more vulnerable to losses and has a greater need for loss absorbing capital.
The formula for determining the amount of regulatory capital that a bank is required to maintain for a loan is as follows: the amount of the loan (for example, £100) multiplied by a minimum capital ratio that applies to all loans – 8% of the loan amount (£8 in this example) multiplied by the risk weight assigned to that loan.
- If the loan was, under the BCBS framework, perceived to be without risk of default, the risk weight would be 0%. This means that the regulatory capital related to that loan would also be zero (£100 x 8% x 0%). A loan made to a multilateral development organisation may have a risk weight of 0%.
- If the loan was, under the BCBS framework, perceived to have a low risk of default, the regulatory capital a bank would require may be determined based on a risk weight of, for example 35%. This would result in a requirement of £2.80 (£100 x 8% x 35%).
- If a loan was, under the BCBS framework, perceived to be riskier still, the risk weight ascribed might be 100%, resulting in a regulatory capital requirement of £8 (£100 x 8% x 100%). Loans made to unrated or sub investment grade corporates would typically have a risk weight of 100% to reflect their higher probability of default and their uncertain value. Some loans may have an even higher risk weight. While it may feel counterintuitive to require a bank to give a parcel of loans a risk weighting of more than 100%, this is because a 100% risk weighting would normally result in substantially less than 100% capital being required to be held against it. Capital requirements absorb unexpected losses rather than cover the possibility of default on the entire loan.
As indicated above, this approach makes intuitive sense – the riskier a loan, the greater the risk of loss and the greater the need for loss absorbing capital to prevent creditors of the bank being adversely impacted by the loss.1
Banks are, in turn, required to hold capital (normally in the form of Common Equity Tier 1 (CET1) capital, predominantly the bank’s share capital and retained earnings) against their risk weighted assets. Currently, UK systemically important banks hold such CET1 capital in the region of 14% to 15% of their risk weighted assets.
The Difficulty with Risk – Weights
In Lord King’s view, determining regulatory capital requirements based on differentiated risk weights pre-supposes that it is possible to formulate an accurate, forward-looking assessment of the risk of different kinds of lending:
…[the system of risk weights] presumes knowledge that we can accurately assess the riskiness of different kinds of lending. The risk weights that are put into the Basel III and other frameworks tend to reflect people’s quantitative estimates of riskbased on normal times, but the purpose of having the capital to absorb losses is for when there is a crisis. At that point, risk weights are a very bad indicator of the riskiness of different elements on the balance sheet. The best example is that, before 2008, it was assumed—and the risk weights reflected this—that mortgage lending was the safest kind of lending. That turned out to be completely false when it came to 2008…. It is just too difficult to assess the riskiness of different kinds of lending… (emphasis added)
This assessment may be worth considering. The riskiness of any financial instrument may not be correctly assessed at the time that it is made (as was the case with residential mortgage lending in the years immediately before the Global Financial Crisis (GFC), given the deterioration of origination standards). Even if it was, the riskiness of that financial instrument may change over time for any number of reasons, both of a systemic or an idiosyncratic nature. Any attempt at achieving mathematical precision in measuring riskiness gives rise to complexity, evidenced by the tens of thousands of pages of regulations that have been implemented by regulators but which individuals responsible for making lending decisions cannot have a detailed understanding of, as well as the adoption of a bureaucratic, rather than judgement based, approach. Reverting to his time at the Bank of England, Lord King commented:
…When we started work at the Bank [of England] on the financial stability report or producing concepts of risk, every month I would get a list of 75 risks. This was not helpful. I would have preferred to have a much smaller group of people, most of whom had years of experience and remembered the previous crisis, at least, who could go out and come back, and say, “This doesn’t feel right.” They could use their judgment to say, “This is the one risk that we should worry about,” rather than trying to pretend that there are 75 risks…. (emphasis added)
In other words, Lord King recognised the merits of a qualitative approach to risk assessment rather than a purely quantitative one.
What Is the Alternative?
Lord King’s preference, based on his evidence to the Committee, is to base regulatory capital requirements not on a system of differentiated risk weights but rather the amount of leverage that an individual bank uses to fund its operations. In his evidence, he stated:
“…I would much rather have a robust and much simpler system that focuses on leverage and by which banks in trouble can have access to the central bank liquidity facility….”
He referred to his experience during the GFC in substantiating this view:
“…In 2008, what really went badlywrong was that the banking sector rapidly expanded its balance sheet, not by issuing loss-absorbing equity or other similar instruments, but by borrowing itself. Its own leverage rose to very high levels, and it had almost negligible liquid financial assets. That was the big risk. It did not matter what the exposure was….” (emphasis added)
To be fair to the BCBS, leverage is already an element in its framework. The leverage ratio, which supplements the differentiated risk weight-based approach, requires that a bank maintains a certain amount of “Tier 1 Capital” (in broad terms, equity, being the best form of capital) relative to its “Total Exposures” (in broad terms, its liabilities both, on and off balance sheet). It is not based on a bank’s assets and their perceived riskiness. The minimum leverage ratio prescribed by the BCBS is 3%, meaning that a bank is required to have an amount of Tier 1 Capital that is at least 3% of its total exposures, though for certain banks that have systemic importance, the requirement may be up to 6%. A bank may manage its leverage ratio by either increasing its Tier I Capital or, alternatively, reducing the amount of its liabilities. This is a less complex concept than differentiated risk weights.
In addition, Lord King focused on the importance of relationship-based knowledge, as opposed to credit scoring, in banks forming a view on the risk of making a loan, again emphasising the importance of qualitative factors.
What of NBFIs?
While Lord King’s view on the merits of a leverage-based approach to determining regulatory capital requirements for banks is clear, his view on the regulation of NBFIs are equally noteworthy. The starting point of his analysis was to question what the purpose of financial regulation is. In broad terms, it is to promote financial stability, ensuring that the financial system supports normal economic activity, so that individuals and firms can save, spend, pay money to each other, and have access to credit without disruption.
In his evidence to the Committee, Lord King stated:
…The non-bank sector comprises a multitude of different kinds [of institution]. The word “ecosystem” has been used to describe it, but it is much more. It is like different life forms across the entire planet – insurance companies, pension funds, bond funds, private equity, hedge funds, venture capital. All of these are completely different animals.
If an individual insurance company were to fail, that is not, in itself, a systemic risk, and there is protection to protect individuals who may be suffering from it. If an insurance company were to fail, and that led the entire insurance industry to find itself in a position where it could not offer insurance to people, that would be systemic.
If a pension fund failed, we have mechanisms for insuring the individuals in that fund. If the entire industry ran into trouble, would we be concerned? ….. I do not see why, if a hedge fund, or several hedge funds, were to fail that constitutes a systemic issue…. (emphasis added)
Based on this and other statements made to the Committee, it seems to us that Lord King’s view is that:
- Banks have a systemic importance that NBFIs do not necessarily have, at least at the current time. This is based upon taking deposits and providing access to payment systems.
- Banks can access central bank liquidity in a way that NBFIs cannot, and this is also an advantage that justifies more stringent regulation.
- NBFIs are not homogeneous, and some may be more systematically important than others.
- Some NBFIs, such as insurance and pension providers, already operate within a framework of prudential regulation and those that do not, such as hedge funds, do not have sufficient systemic significance to warrant it.
- Regulating NBFIs less stringently than banks may promote financial stability by directing more risky activities to NBFIs, thus preventing banks from getting into trouble.
However, there is at least one example of a hedge fund whose losses required regulatory intervention in order to maintain financial stability – Long Term Capital Management (LTCM).2 It seems that much depends on the size of the hedge fund and its interconnectedness with the regulated financial sector.
Conclusion
Lord King’s experience as both an eminent academic and a governor of the Bank of England, particularly during the GFC, gives him a unique perspective on financial stability. Whether the system of differentiated risk weights is to be replaced by a leverage ratio-based approach as he suggests is a wider question and would require a departure from a well-entrenched approach, but there may be merit in the view that banks and NBFIs play different roles in a financial system and so a differentiated approach to regulation between them has a logic to it. Indeed, the differentiated regulatory approach may promote financial stability by limiting the lending activity of banks. However, the example of LTCM suggests that not all NBFIs should be treated in the same way from a prudential regulation perspective.
The Committee’s work continues. Greenberg Traurig will continue to monitor the developments.
1 As indicated, this description is simplification. It does not recognise the distinction between the “standardised approach” to differentiated risk weights, which is the approach that the BCBS prescribes as a default position and the “internal risk-based approach,” which sophisticated banks may adopt and which may result in different results from the standardised approach. This is based on sophisticated banks being in a better position to assess risks than banking regulators.
2 Long Term Capital Management (LTCM) was a hedge fund established in 1994 and that at one point managed $3.5 billion of investor capital. As a result of the Russian debt default, it sustained several losses and the U.S. government had to intervene, facilitating a bail out by 14 banks and financial institutions in order to prevent a wider financial crisis.