The elusive leverage ratio

October 5, 2009

By Peter Thal Larsen and Rolfe Winkler

LONDON/NEW YORK, Oct 5 (Reuters) – Of all the reforms proposed by global financial regulators over the past 12 months, none looks as appealingly straightforward as the leverage ratio. What could be simpler than linking the total amount of assets a bank can hold to the amount of capital it has to absorb losses it makes on them?

Alas, such a task is more difficult than it appears. There is little international agreement about how to calculate banks’ assets or capital, let alone what the ratio between the two should be.

A few years ago, even the idea of such a simple measure seemed hopelessly out of date. Banks were busy building sophisticated computer models to measure the risks they faced. They were allowed to tweak the amount of capital they held against assets depending on how risky the computer thought those assets to be. Then those models failed.

Now bank regulators want a blunt measure that will cap banks’ expansion, regardless of what their models say. Despite the complexities, they are right to try.

Still, finding a consistent way to measure banks’ assets is a daunting challenge. For instance, U.S. accounting rules allow banks to report their net derivatives exposure. International Financial Reporting Standards used by most European banks don’t. Take Deutsche Bank: At the end of 2008 its gross derivatives book accounted for nearly half its total assets of 2.2 trillion euros. Using IFRS, Deutsche’s leverage ratio is 1.3 percent but under U.S. GAAP it’s 4.2 percent.

(Click on chart to enlarge in new window)


Regulators agree that any calculation of leverage should adjust for these differences to discourage regulatory arbitrage. But barring an unlikely shift to global accounting standards, this will undermine the leverage ratio’s simplicity, which is central to its appeal.

Measuring capital is also a thorny issue. U.S. regulators believe that only tangible equity — cash raised from shareholders or retained through earnings — should count as capital. That is sensible: Only this capital occupies the true first loss position. Without it, investors higher up the capital structure tend to panic and run for the exits when losses mount.

But the emphasis on tangible equity has gone down badly with European banks, which have historically stuffed preferred shares and other forms of hybrid capital into their capital structures. If these are excluded from capital measures, European banks will have to find tens of billions of additional equity — or shrink their balance sheets even further. Another objection is that U.S. banks are currently allowed to count deferred tax assets as capital, even though these are worthless if the institution cannot make a profit.

Even if these problems can be overcome, the leverage ratio is not foolproof. After all, U.S. regulators have long imposed a 4 percent leverage ratio on banks. But this did not include off-balance sheet assets. More recently, Swiss regulators introduced a leverage ratio for UBS and Credit Suisse that explicitly excludes the banks’ Swiss loan books from their asset calculations. Some bankers argue that, because the ratio does not adjust for balance sheet risks, it may even encourage banks to load up on risky assets.

Placing an appropriate cap on the expansion of banks’ balance sheets is crucial to the future of banking regulation. If it’s too loose, banks may rush into another crisis. If it’s not applied consistently, banks will arbitrage differences between regimes. Some will argue that if it’s too tight, banks will constrict lending or will dump safe assets in favour of risky ones.

But this misses the point. The leverage ratio is not designed to replace risk-based measures of capital: it is a safety net in case those measures fail. Despite the difficulties involved, regulators should not be deterred from introducing it.

(e-mail: peter.thal.larsen at; rolfe.winkler at


We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see

Perhaps creation of an international rules body with the power to impose standards would help. In other fields, such approaches help. Examples include international building codes and standards that promote safe construction.

Posted by Marvin McConoughey | Report as abusive

Consensus on simplification, liberalization and unification of personal bankruptcy laws could be achieved easier, made credit riskier and immediately lead to reduction of overcomplex credit risk accounting worldwide.

And that would be treaty on freedom, not another bureaucratic installment.

Posted by anarchistas | Report as abusive

[…] The elusive leverage ratio […]

Posted by Monday Afternoon Reading | The Big Picture | Report as abusive

[…] The elusive leverage ratio […]

Posted by Monday Afternoon Reading | | Report as abusive

RE: Comment at 3:42 GMT

Correct me if I’m wrong but aren’t the Basel Accords and subsequent modifications, i.e., Basel II doing what is proposed here?

Posted by Mike | Report as abusive

[…] The Elusive Leverage Ratio Rolfe Winkler […]

Posted by Links 10/6/09 « naked capitalism | Report as abusive

Good article. Just to clarify, my understanding of ‘leveraging’ is ‘gearing’, that is the debt to equity ratio, which doesn’t seem to be the case here. I suppose in essence one wants to avoid a ‘run’ on the banks, as a basis for the debate. I understand the latter as being ‘minimum capital requirements’. Interesting the approach that Swiss Banks have, and don’t discount Belgium banks, after all, this is where a lot of ‘clearing or ‘movement’ of large monies takes place. Brazil seems to have escaped the crisis by using +- 12%, and is now very (over-) confident in the markets. The problem remains which tiers/hybrids to include and conversely, what assets to include and how to value them all. Correct me if I am wrong, but there is a convergence of US GAAP and IFRS at the moment and remember, the World has banks outside of the US, Britain and Europe. That also leaves the Federal Banks and Treasuries, as intermediaries, open to discussion. The key here is ‘value’. On a more flippant approach, any % x zero value = zero, that would include insurance products, junk bonds, derivatives/futures, options and off-balance sheet horrors that simply can’t or refuse to unwind, for whatever reason. So let’s exclude those for now and focus on ‘tangible’ items valued on the basis of bond, stock/property, forex, money and capital markets’ ‘weighted parity exchanges index’ ? That would place us on both sides of the balance sheet. Don’t forget about Amex, Diners Club, Master and VISA markets, they tend not to forget about you. Otherwise, take an average/mean of all the banks, study which are deviant and deal with it. This could be dangerous if the sample is not representative. Dean Baker elsewhere today suggests a sliding scale. Otherwise, change the measurement to a % relative to related turnover, which is more difficult to fudge, but that could open another can of weighting worms altogether.

Mike, Basle II has a driveway and pillars, but lacks a foundation, pillars for HR and legal risk, walls, windows, (back-) doors, a roof, basic services, a fence, a family, an alarm system, a security company that responds at all and a tornado bunker. That’s my opinion.

Posted by aNON | Report as abusive

I think we’re all missing the point here, if the cost of borrowing was set at 10% banks wouldn’t be able to leverage to these ratios anyway. Money is too cheap; if it wasn’t we wouldn’t have any subprime or any of these regulatory failings or any of this “growth” which is now being destroyed; we would have had that elusive stability that we crave.

Interest rates need to rise and quickly before the damage is way beyond repair, if it isn’t already.

Posted by Harlequin002 | Report as abusive

Harlequin002: We might be debating at cross-purposes here. Even if you were correct, the weighted average cost of capital (WACC) of, e.g. bank, is much more complex and dynamic than a flat rate of 10%.

Posted by Gaspard | Report as abusive

[…] Contingent Capital discusses an easy ratio for determining the health of banks. […]

Posted by Weekend Recap – October 18, 2009 | Report as abusive