Custodians could face higher Basel G-Sib surcharges

Global banks with large asset custodian businesses may face higher capital requirements if the Basel Committee’s systemic bank assessment methodology, currently under a routine review, is changed in an effort to reflect systemic risk more accurately.

BNY Mellon, Citigroup, JP Morgan and State Street would be the most affected if the cap on one of five categories in the methodology – ‘substitutability’ – is removed, say bank lobbyists and academics.

“It is very important to take seriously this big difference among these five variables that they focus on in the methodology. The variables have very different behaviours across banks; it has a very skewed distribution. In terms of substitutability, we have banks with super high values and banks with super low values. The other four sources of systemic risk are much more concentrated, so substitutability ends up running the show,” says Christophe Perignon, associate professor of finance at HEC Paris.

The Basel Committee on Banking Supervision first developed its assessment methodology for global systemically important banks (G-Sibs) in 2011, leading to the publication of the first list of G-Sibs by the Financial Stability Board in November of that year. The list is divided into buckets, each one assigned a Basel risk-weighted capital surcharge ranging from 1% to 3.5%.

In the most recent list (November 2016), Citigroup and JP Morgan were in the highest populated bucket, at 2.5%, with BNY Mellon and State Street at 1%. The methodology employs five categories, each weighted at 20%, to score banks on their contribution to systemic risk.

The Basel Committee released a consultation paper in March as part of its regular review of the G-Sib assessment methodology. The proposals included removing the 500 basis point substitutability cap that was added in 2013. Substitutability measures how replaceable a bank is in the event of failure, using three indicators: payments, asset custody and securities underwriting.

The cap was adopted, as the committee explained in 2013, because data “revealed the substitutability category had a greater impact on the assessment of systemic importance than was intended”. This effect is due to the volatility of this category relative to the other categories: a high score in substitutability has a disproportionate effect on a bank’s final score.

Scoring bias 

In March, Perignon and two other French academics published an updated version of a paper examining the “pitfalls” of the Basel G-Sib methodology. “An unintended consequence of the scoring method is to bias scores towards the categories that are most volatile in the cross-section,” the paper says. “Indeed, when variables are aggregated in [the] absence of any form of standardisation, they are effectively weighted by their standard deviation.”

With the cap in place, this effect is reduced; however, the cap also has the effect of decreasing the overall scores of custodian banks. And so with this consultation, the Basel Committee is considering the possibility that the cap could remove the incentive for these banks to reduce their systemic footprint.

The potential loss of custody, payments and underwriting services was one of the reasons why governments felt obliged to step in during the 2008 financial crisis. As the 2017 consultation states: “The provision of payments, custody and underwriting services can be disrupted and difficult to substitute in the event of a bank’s failure, thereby undermining financial markets in a manner that does not necessarily diminish concentration in the provision of these services.”

To cap or not to cap?

Removing the cap would affect the composition of the G-Sib list and thus have a real impact on the businesses of these banks, says Perignon. He and his co-authors have calculated the 2014, 2015 and 2016 scores without the cap. Their findings – from a unique dataset they assembled of 119 banks in 22 jurisdictions over this three-year period – show the four banks with large custody businesses are the most affected.

BNY Mellon ($30.6 trillion assets under custody (AUC)) and State Street ($22.5 trillion AUC) are specialist custodians, the largest and second-largest in the world by AUC, respectively. JP Morgan’s custody and fund services business has $21.4 trillion AUC. Citi’s institutional clients group also features fairly prominently, with some $15.9 trillion AUC.

In some cases, the scores change to the extent that the bank moves up a bucket (see table A). Citigroup, which rose into the 2.5% bucket in 2016, would already have moved there in 2015 without the substitutability cap. BNY Mellon would have become the bank with the highest score in the 1% bucket in 2016, just a few points below the lowest bank in the 1.5% bucket. Since the scoring is relative, this means a fairly small reduction in the scores of other banks could push BNY Mellon into the 1.5% bucket as well.

The most striking impact, however, can be seen with JP Morgan. In both 2015 and 2016, the US bank would have been pushed into the top bucket. This involves a sharp increase in the capital surcharge to 3.5%.

According to the academic research paper: “Given the risk-weighted assets, as of year-end 2014, of JP Morgan Chase (€1,213 billion), this means that JP Morgan Chase [was] able to reduce its regulatory capital by €12.13 billion, 8.94% of its Tier 1 capital or 29% of its systemic-risk charge.”

The cap, then, reduces the systemic risk score of the custodian banks, but removing it is problematic too, because of the wild standard deviation of the substitutability category. Although none of the four banks would comment publicly, Risk.net understands there is real alarm about the proposal to remove the cap. The comment period for the consultation closes on June 30.

This is Statistics 101. It’s what everybody does when they aggregate scores – everybody except the Basel Committee. It’s not rocket science, it’s not fancy econometrics; it’s simple stats

Christophe Perignon, Ecole des Hautes Etudes Commerciales

Industry groups have long complained the Basel methodology seems to focus heavily on size, and uncapping the substitutability score would evidently amplify that effect for banks holding large custodial assets. And yet, the custodian business itself is inherently low risk.

Andrés Portilla, managing director of regulatory affairs at the Institute of International Finance, says Basel should follow the lead being set by asset management and insurance regulators, and begin to focus more on specific activities that create systemic risk, rather than purely on size.

“We have seen policymakers’ philosophy about Sifis [systemically important financial institutions] evolve since 2009. There has been a lot of effort to develop a framework for understanding whether individual firms could be deemed systemic. Now the debate is at a place where you can see a growing understanding that designating asset management firms individually, for example, is not the right approach. You can clearly see policymakers stepping back from the designation of individual asset managers as Sifis and looking more at activities than at individual firms. This has started to happen also in the insurance space,” says Portilla.

According to Perignon, there is another very simple solution to the substitutability problem: standardising each category by its volatility before the overall score is calculated. He submitted this recommendation to the Basel Committee in 2013. 

“Instead of just aggregating the five raw indicators, you aggregate them and scale each indicator by volatility,” says Perignon. “This is Statistics 101. It’s what everybody does when they aggregate scores – everybody except the Basel Committee. It’s not rocket science, it’s not fancy econometrics; it’s simple stats.”