It is generally acknowledged that high bank leverage was a key factor contributing to the severity of the Global Financial Crisis. For the UK, Bank of England data suggest that UK banks’ average simple leverage – the ratio of banks’ total assets to their shareholder equity – was 24.4 over 2006.[1]

Basel III then sought to counter high leverage by imposing a minimum required leverage ratio on regulated banks.[2] The ‘leverage ratio’ is the inverse of the leverage.

So how onerous are Basel III’s constraints on bank leverage? More precisely, what is the maximum permitted leverage under Basel III?

A clear expression of the Basel rules on this point is the following from the 2015 Prudential Regulatory Authority (PRA) Rulebook in the UK:

3 MINIMUM LEVERAGE RATIO

3.1 A firm must hold sufficient tier 1 capital to maintain, at all times, a minimum leverage ratio of 3%.[3]

3.2 For the purposes of complying with 3.1, at least 75% of the firm’s tier 1 capital must consist of common equity tier 1 capital [CET1].[4]

This minimum required leverage ratio is expressed in terms of the ratio of Tier 1 capital to the leverage exposure. Section 3.1 of the PRA Rulebook would suggest that the maximum permitted leverage is then 1/3 percent = 33.33.

Section 3.2 of the PRA Rulebook then states that at least 75 percent of the firm’s Tier 1 capital should consist of CET1 capital.

Now 25 percent of the 3 percent minimum Tier 1 leverage ratio is 0.75 percent, so the minimum required leverage ratio expressed in terms of CET1 capital = 3 percent minus 0.75 percent = 2.25 percent.

This implies that the maximum permitted CET1 leverage = 1/0.0225 = 44.44.

However, Basel III also allows banks to include a ‘sin bucket’ of non-CET1 capital items as part of their reported CET1.

So let’s distinguish between ‘reported’ CET1 (or CET1 including the sin bucket) and ‘clean’ CET1 (or CET1 purged of the sin bucket).

Under Basel III rules, the clean CET1 can be as low as 85 percent of reported CET1.[5]

Let’s also assume that bankers make maximum use of the sin bucket so the clean CET1 = 85 percent reported CET1.

This means that the Leverage Ratio using clean CET1 can be as low as 85 percent 2.25 percent = 1.9125 percent and still comply with the Basel III minimum required leverage ratio. Inverting this number gives the maximum permitted leverage using clean CET1, i.e., 1/1.9125 percent = 52.29.[6]

A loss of 2 percent of the leverage exposure would then be more than enough to wipe out a bank’s CET1 capital.

In plain English, the Basel III capital rules allow banks to maintain remarkably high leverage and still be Basel III-compliant. Indeed, the Basel capital rules would appear to allow banks to maintain considerably higher leverage than they had on the eve of the financial crisis!

I should also note some qualifications, which will serve to further loosen the impact of the Basel III maximum leverage constraint:

The first is that these calculations ignore sources of hidden leverage such as accounting standards that cause capital to be over-reported or the additional leverage in off-balance sheet positions.
The second is that the above calculations relate to book values, not to market values, and market-value leverage these days will typically be higher than book-value leverage.

In fact, since Basel III does not impose any restriction on market-value leverage, the maximum permitted market-value leverage under Basel III is theoretically unbounded.

End Notes

* I thank Anat Admati, Tim Bush, Gordon Kerr and Sir John Vickers for helpful discussions on the subject covered in this blog posting.

[1] Obtained from the Bank of England’s Financial Stability Report for November 2016, p. 58, where the 2006 average simple leverage ratio is reported as 4.1 percent. The simple leverage is then 1 ÷ 4.1 percent = 24.4.

[2] Critics however have argued that such requirements are onerous. These include many leading bankers (e.g., Jamie Dimon and former Deutsche Bank chairman Josef Ackermann) and even central bankers (e.g., Alan Greenspan), the American Bankers Association, and the British Bankers Association (see, e.g., the citations in A. Admati and M. Hellwig, The Bankers’ New Clothes: What’s Wrong with Banking and What to Do About it, Princeton University Press, 2013).

[3] As an aside, it is disappointing to see that the PRA’s own rulebook explicitly endorses the ‘capital is a rainy day fund’ fallacy by stating that banks “hold” capital. Banks do not “hold” capital. To suggest that they do is to suggest that capital is an asset to a bank and it is not. Read Admati and Hellwig.

[4] Prudential Regulation Authority (2015) PRA Rulebook: CRR Firms: Leverage Ratio Instrument 2015. London: PRA, p. 5.

[5] For more on the ‘sin bucket’ see T. F. Huertas (2014) Safe to Fail: How Resolution will Revolutionise Banking, Basingstoke: Palgrave Macmillan, p. 23, or Basel Committee on Banking Supervision (June 2011) “Basel III: A global regulatory framework for more resilient banks and banking systems,” pp. 21-6 and 65.

[6] If one took account of systemic or countercyclical buffers, however, the maximum permitted leverage would be somewhat lower, but not much.