What was the root cause of the financial crisis? Ask any economist or banker and undoubtedly they will at some point mention leverage (see e.g. here, here and here). Yet when a capital requirement based on leverage — the leverage ratio requirement — was introduced, fierce criticism followed (see e.g. here and here). Drawing on the insights from a working paper, and thinking about the main criticism — that a leverage ratio requirement could cause excessive risk-taking — this seems not to have been the case.
The chief criticism against the ‘leverage ratio’ was actually that it targeted leverage too directly — at the expense of targeting risk — since it is defined simply as a percentage of a bank’s total exposures. This contrasts to the pre-existing risk-based capital requirement, which as you might guess, is defined in relation to a bank’s risk-weighted (RWA) assets. But the leverage ratio was never designed to replace the risk-based requirements — it operates alongside them — so it adds as a complimentary feature to the capital framework. This complimentary feature is important because as the paper shows it adds a natural limit on risk-taking. So while the paper illustrates that a leverage ratio requirement can incentivise banks to take on more risk, it’s not too much when you compare this additional risk-taking to the additional loss-absorbing capacity the leverage ratio brings. In other words, the paper suggests it’s an overall net benefit.
Why a leverage ratio?
If you’ve ever seen the movie ‘The Big Short’, you might remember the opening scene which begins with a quote they actually miscredit to Mark Twain: ‘It ain’t what you don’t know that gets you into trouble. It’s what you know for sure that just ain’t so’. Despite the misaccreditation, it sets the film up to address a key message at the heart of finance, and for that matter the leverage ratio; and that is the message of uncertainty.
As the saying goes, ‘uncertainty is the only certainty’. Even when we think we know or understand something, often, the truth is that we really don’t. It is not just a point about hubris, it is much more fundamental. In psychology, this phenomenon is called the ‘illusion of explanatory depth’. There are just many things we don’t know that we don’t know — unknown unknowns — and there is nothing we can do about it.
But when it comes to safeguarding the financial stability of the banking system, it might be prudent to try and guard against this. The leverage ratio is a direct response; a guardrail against these unknown unknowns. While the established risk-based capital framework hinges on knowledge of the underlying probability distribution of asset returns, the idea behind the leverage ratio is to add something a little simpler; something that doesn’t depend on understanding, or even knowing this distribution.
This is where the paper begins. It identifies a key issue with the risk-based framework that creates a friction the leverage ratio can solve. While theoretically risk-based requirements are an ideal way to map risk to capital requirements — the more risk a bank takes, the more capital it must hold — underlying this mapping is a model, and if you’ve ever done any modelling yourself, you’ll know they’re far from perfect.
The financial crisis brought these problems to the forefront of our attention. Assets which we previously thought were safe turned out to be the complete opposite; they then blew up. Losses mounted, and to make things worse, because these supposedly safe assets attracted low capital requirements, banks had been allowed to leverage massively; multiplying these losses.
So what’s not to like? Theoretical findings
The paper builds out these issues in a theoretical micro banking model that considers the trade-offs imposing a leverage ratio entails. The fact is that the leverage ratio’s key strength is also its weakness. Because it ignores the riskiness of individual assets — a fact that enables it to cover these unknown unknowns — banks can simply load up on the riskiest assets. I mean, why not? Riskier assets offer a higher return, and if your binding requirement is not risk-based, you can achieve all this for the same level of capital.
But this is a trade-off remember, so let’s take a step back for a minute to think this through. The above is simply a one-dimensional argument about marginal cost. While a leverage ratio requirement might on the one hand incentive greater risk-taking — in technical terms, because the marginal cost of risk-taking declines — it’s not the complete story. There are two mitigating factors on the other side.
First, consider who is bearing this risk. Ultimately the bank will. And that makes a difference. To put this into perspective, take the case of an extremely leveraged bank with an infinitesimal capital ratio. While the bank benefits from all the upside of risk-taking, it bears none of the downside. If it makes a loss, it just goes bankrupt, and due to limited liability, passes these losses onto its creditors. As a result, it has an incentive to risk-up: to load up on the riskiest assets.
Now, suppose we force this bank to increase its capital-to-asset ratio — this is what the leverage ratio is doing. While the bank still bears all the upside from risk-taking, it now also bears more of the downside. This subdues the incentive to risk-up; it can even have an incentive to de-risk.
The point is that raising capital increases banks’ skin-in-the-game, so banks internalise more of the consequences from additional risk-taking. So instead of the leverage ratio endlessly incentivising banks to increase risk-taking, there is in fact an offsetting effect.
Second, the risk-based framework still exists. The leverage ratio hits only a subset of banks, i.e. those that are highly leveraged. Because of this, there is a natural limit to how much additional risk a bank can take. If it takes too much, the bank will merely revert to having a risk-adjusted capital requirement as its binding requirement; and this does involve raising further capital.
These two effects subdue the extent to which banks want to risk-up. Furthermore, this is just one side of the coin. Banks may take a bit more risk, but what about all this additional capital they now hold. How does this trade-off manifest itself? Well, actually pretty well. The paper shows that you would expect better probabilities of default and lower expected losses on depositor funds.
Empirical evidence: more risk-taking, but more stable banks
And what does the data say? We can check this since banks have known they will need to satisfy a leverage ratio requirement since late 2009. In the paper, using data on more than 650 EU banks over the years 2005-2014, a difference-in-difference type analysis is used to isolate the leverage ratio effect. Using a 3% leverage ratio threshold to separate banks, and 2010 as the announcement date in respect to the Basel consultative document, the paper finds that banks are incentivised to take on riskier assets, but not much. Using various estimation techniques and samples, the paper finds that banks bound by the leverage ratio requirement increase their RWA to total assets ratio by around 1.5-2.5 percentage points more than non-bounded banks. So, while there is an increase in risk-taking, it appears rather small. This is especially true considering banks raised their capital ratios at the same time because of the leverage ratio — see Chart 1.
Chart 1: Percentage of banks with a leverage ratio below 3%
Source: Acosta-Smith, Grill and Lang (2018).
And how does this increased capital position affect the picture? We find two interesting findings. First, using a logit model to quantify the risk-stability trade-off associated with a leverage ratio, we find that in determining distress, the leverage ratio is an order of magnitude more important than any slight increase in the riskiness of banks’ assets. A 1 percentage point increase in the leverage ratio would reduce the odds ratio (on banks in distress versus safe) by around 40-50%. This is much larger than the equivalent effect from a 1 percentage point increase in the RWA ratio, which raises the odds ratio by only around 1-3.5%. Second, using a counterfactual simulation in which we can simulate banks raising both their capital ratios and their risk-taking, the paper finds significant reductions in the probability of distress with a leverage ratio. The empirical evidence therefore supports the conceptual thinking: while there may be some additional risk-taking, it is offset enough such that banks should become safer.
So what can we conclude?
It has been said that a leverage ratio requirement is akin to setting the same speed limit on every single road. This, however, is incorrect. It’s actually more like imposing a national maximum speed limit, and there’s a lot of evidence this increases road safety. The authorities can always impose different speed limits on other roads depending on risk, and the same is true through the combination of the leverage ratio and the risk-based requirement.
While there may be some increased risk-taking at the margin, this needs to be put into perspective: the evidence would suggest it’s outweighed by the benefit of higher loss-absorbing capacity.
Jonathan Smith works in the Bank’s Policy Strategy and Implementation Division.
If you want to get in touch, please email us at firstname.lastname@example.org or leave a comment below.
Comments will only appear once approved by a moderator, and are only published where a full name is supplied. Bank Underground is a blog for Bank of England staff to share views that challenge – or support – prevailing policy orthodoxies. The views expressed here are those of the authors, and are not necessarily those of the Bank of England, or its policy committees.