The Goldilocks zone
29 March 2016
Glenn Horner of State Street reviews the newly proposed securities financing transaction standardised calculation, finding it to be just right
Image: Shutterstock
Following the financial crisis in 2008, global regulators were presented with an unenviable task of rewriting banking regulations in a sensible manner while adhering to a number of core principles.
Among the guiding principles, measures should:
Be simple enough to be easily understood and implemented across banking organisations;
Be conservative enough to capture potential risks in a stressed market environment;
Reflect the actual potential future exposure a banking entity might incur;
Be risk sensitive; and
Across similar economic exposures, be similar in magnitude so as to avoid regulatory arbitrage.
Historically, for securities financing transactions (SFTs), many of the largest agent lenders utilised a simple value-at-risk (VaR) measure to calculate the potential exposure-at-default (EAD) for each counterparty. As a result of the financial crisis, however, the use of such simple VaR measures came under greater scrutiny from global regulators. The primary objections to the use of a simple VaR measure were: (i) the relative complexity of the models utilised; and (ii) the fact that different banks reported different EADs for similar loan and collateral portfolios.
Additionally, despite agent lenders reporting no losses due to their loan indemnities, many regulators believed that the simple VaR models resulted in EADs that did not fully reflect the potential risks of a stressed market.
As a result of concerns that the various internal models for simple VaR did not reflect stressed market conditions and the lack of comparability across firms, regulators have explored utilising a standardised floor for banks that calculate risk-weighted assets (RWA) under an advanced approach.
Under the Collins Amendment to the Dodd-Frank Act, the largest US banks are required to calculate RWA under both the advanced and standardised methods and to manage to the more conservative of the two at the firm-wide level. The Basel Committee on Banking Supervision is exploring a similar floor on a global basis. Additionally, the standardised method has been proposed as the method to calculate exposures under both the single counterparty concentration limits (SCCL) in the US and Basel’s large exposure limits. As such, the importance of the standardised method has increased significantly, even for those advanced approach banks that have historically utilised a simple VaR measure for calculating SFT exposures.
The existing standardised approach for SFTs meets regulators’ goals of simplicity, consistency and conservativeness. However, it fails on the other key criteria, namely, risk sensitivity and reflection of actual exposures that may be incurred even under stressed market conditions.
Further, the resulting EAD for SFTs varies substantially from the EAD that would be calculated for derivative transactions with identical or very similar economic exposures. These deficiencies are driven primarily by the standardised approach’s lack of consideration for correlation and diversification benefits within a portfolio of loans and collateral. The existing measure relies on a set of additive haircuts for both loan and collateral positions even if these positions are largely offsetting.
As a simplified example, suppose a lender lends Apple shares worth $100 and receives Intel shares worth $105 as collateral. The current model assumes that, in the event of a counterparty default, the Apple shares would increase in value by 10.6 percent over the five-day buy-in horizon while the Intel shares would decrease in value by 10.6 percent during the same period.
Example EAD using existing standardised approach, equity versus equity loan:
= (Assumed higher value of Apple shares) – (assumed lower value of Intel)
= ($100 *(1 + 10.6 percent)) – ($105 *(1 – 10.6 percent)) = $16.73
= $16.73 / $100 (original loan value) = 16.73 percent
This hardly seems plausible and becomes even more unrealistic when one considers that many loan and collateral netting sets include hundreds or even thousands of positions. In fact, estimates from market participants indicate that the EADs from the existing standardised approach are 20 to 30 times higher than EADs calculated using a simple VaR model.
With these considerations in mind, the Basel Committee on Banking Supervision has proposed a new standardised methodology for SFTs utilising the following formula:
EAD = |(0.4 * net exposure)|+ (0.6 * gross exposure/?N)
Where:
Net exposure would allow for the offsetting of collateral and loan positions, reflecting the systemic risk of the portfolio;
Gross exposure would be cumulative across loans and collateral, reflecting the idiosyncratic risk of the portfolio positions; and
N represents the number of securities in the portfolio (with some relative size requirements).
In addition to the proposed new formula the Basel Committee on Banking Supervision proposed raising certain haircuts.
The new haircuts for a five-day holding period for main index equities would be 14.4 percent. If we used our prior example, the new EAD would be as follows:
EAD using proposed new standardised approach, equity versus equity loan:
= |0.4 *($100*14.4 percent - $105*14.4 percent)| + (0.6 * ($100*14.4 percent + $105*14.4 percent)/?2)
= 0.28 + 12.52 = $12.80
= $12.80 / $100 = 12.8 percent
We would see a marginal benefit under the example. However, the benefits would be further magnified as the portfolio of loans and collateral increased. Early estimates by market participants indicate that average EADs would decrease by approximately 70 percent using the proposed new formula.
While markedly lower relative to the existing standardised approach, EADs under the proposal would still be very conservative at approximately five to seven times higher than the EADs calculated under a simple VaR method.
Overall, the new proposal by the Basel Committee on Banking Supervision is quite remarkable in that it meets all requirements for a sound regulatory measure.
It maintains a conservative estimate of EAD that is relatively simple to understand and implement across different organisations, and it represents the potential impact of a stressed scenario.
Further, it addresses key shortcomings of the existing standardised measure since it is also risk sensitive, represents actual potential future exposures and eliminates the significant difference in measurement for like economic exposures under derivative and cash-based trades.
With the new standardised approach proposal for SFTs, the Basel Committee on Banking Supervision has truly identified a Goldilocks solution: one that is not too complex for most users, but which represents a conservative yet reasonable calculation of EAD.
While this solution, if implemented, would result in a calculation that meets the key regulatory objectives it should be noted that it is only a proposal from Basel.
Until such time as it is adopted in the US, large banks will still be required to calculate their risk-based capital calculations on both the advanced and current standardised approaches due to the Collins Amendment of the Dodd-Frank Act.
Once finalised by the Basel Committee on Banking Supervision, the proposal would have to go through a similar process within the US whereby regulators would put out a proposal, receive comments and then ultimately finalise the new calculation.
Obviously, with the multiple steps for adoption there may be resistance at some levels and it will likely be a multi-year process.
Among the guiding principles, measures should:
Be simple enough to be easily understood and implemented across banking organisations;
Be conservative enough to capture potential risks in a stressed market environment;
Reflect the actual potential future exposure a banking entity might incur;
Be risk sensitive; and
Across similar economic exposures, be similar in magnitude so as to avoid regulatory arbitrage.
Historically, for securities financing transactions (SFTs), many of the largest agent lenders utilised a simple value-at-risk (VaR) measure to calculate the potential exposure-at-default (EAD) for each counterparty. As a result of the financial crisis, however, the use of such simple VaR measures came under greater scrutiny from global regulators. The primary objections to the use of a simple VaR measure were: (i) the relative complexity of the models utilised; and (ii) the fact that different banks reported different EADs for similar loan and collateral portfolios.
Additionally, despite agent lenders reporting no losses due to their loan indemnities, many regulators believed that the simple VaR models resulted in EADs that did not fully reflect the potential risks of a stressed market.
As a result of concerns that the various internal models for simple VaR did not reflect stressed market conditions and the lack of comparability across firms, regulators have explored utilising a standardised floor for banks that calculate risk-weighted assets (RWA) under an advanced approach.
Under the Collins Amendment to the Dodd-Frank Act, the largest US banks are required to calculate RWA under both the advanced and standardised methods and to manage to the more conservative of the two at the firm-wide level. The Basel Committee on Banking Supervision is exploring a similar floor on a global basis. Additionally, the standardised method has been proposed as the method to calculate exposures under both the single counterparty concentration limits (SCCL) in the US and Basel’s large exposure limits. As such, the importance of the standardised method has increased significantly, even for those advanced approach banks that have historically utilised a simple VaR measure for calculating SFT exposures.
The existing standardised approach for SFTs meets regulators’ goals of simplicity, consistency and conservativeness. However, it fails on the other key criteria, namely, risk sensitivity and reflection of actual exposures that may be incurred even under stressed market conditions.
Further, the resulting EAD for SFTs varies substantially from the EAD that would be calculated for derivative transactions with identical or very similar economic exposures. These deficiencies are driven primarily by the standardised approach’s lack of consideration for correlation and diversification benefits within a portfolio of loans and collateral. The existing measure relies on a set of additive haircuts for both loan and collateral positions even if these positions are largely offsetting.
As a simplified example, suppose a lender lends Apple shares worth $100 and receives Intel shares worth $105 as collateral. The current model assumes that, in the event of a counterparty default, the Apple shares would increase in value by 10.6 percent over the five-day buy-in horizon while the Intel shares would decrease in value by 10.6 percent during the same period.
Example EAD using existing standardised approach, equity versus equity loan:
= (Assumed higher value of Apple shares) – (assumed lower value of Intel)
= ($100 *(1 + 10.6 percent)) – ($105 *(1 – 10.6 percent)) = $16.73
= $16.73 / $100 (original loan value) = 16.73 percent
This hardly seems plausible and becomes even more unrealistic when one considers that many loan and collateral netting sets include hundreds or even thousands of positions. In fact, estimates from market participants indicate that the EADs from the existing standardised approach are 20 to 30 times higher than EADs calculated using a simple VaR model.
With these considerations in mind, the Basel Committee on Banking Supervision has proposed a new standardised methodology for SFTs utilising the following formula:
EAD = |(0.4 * net exposure)|+ (0.6 * gross exposure/?N)
Where:
Net exposure would allow for the offsetting of collateral and loan positions, reflecting the systemic risk of the portfolio;
Gross exposure would be cumulative across loans and collateral, reflecting the idiosyncratic risk of the portfolio positions; and
N represents the number of securities in the portfolio (with some relative size requirements).
In addition to the proposed new formula the Basel Committee on Banking Supervision proposed raising certain haircuts.
The new haircuts for a five-day holding period for main index equities would be 14.4 percent. If we used our prior example, the new EAD would be as follows:
EAD using proposed new standardised approach, equity versus equity loan:
= |0.4 *($100*14.4 percent - $105*14.4 percent)| + (0.6 * ($100*14.4 percent + $105*14.4 percent)/?2)
= 0.28 + 12.52 = $12.80
= $12.80 / $100 = 12.8 percent
We would see a marginal benefit under the example. However, the benefits would be further magnified as the portfolio of loans and collateral increased. Early estimates by market participants indicate that average EADs would decrease by approximately 70 percent using the proposed new formula.
While markedly lower relative to the existing standardised approach, EADs under the proposal would still be very conservative at approximately five to seven times higher than the EADs calculated under a simple VaR method.
Overall, the new proposal by the Basel Committee on Banking Supervision is quite remarkable in that it meets all requirements for a sound regulatory measure.
It maintains a conservative estimate of EAD that is relatively simple to understand and implement across different organisations, and it represents the potential impact of a stressed scenario.
Further, it addresses key shortcomings of the existing standardised measure since it is also risk sensitive, represents actual potential future exposures and eliminates the significant difference in measurement for like economic exposures under derivative and cash-based trades.
With the new standardised approach proposal for SFTs, the Basel Committee on Banking Supervision has truly identified a Goldilocks solution: one that is not too complex for most users, but which represents a conservative yet reasonable calculation of EAD.
While this solution, if implemented, would result in a calculation that meets the key regulatory objectives it should be noted that it is only a proposal from Basel.
Until such time as it is adopted in the US, large banks will still be required to calculate their risk-based capital calculations on both the advanced and current standardised approaches due to the Collins Amendment of the Dodd-Frank Act.
Once finalised by the Basel Committee on Banking Supervision, the proposal would have to go through a similar process within the US whereby regulators would put out a proposal, receive comments and then ultimately finalise the new calculation.
Obviously, with the multiple steps for adoption there may be resistance at some levels and it will likely be a multi-year process.
NO FEE, NO RISK
100% ON RETURNS If you invest in only one securities finance news source this year, make sure it is your free subscription to Securities Finance Times
100% ON RETURNS If you invest in only one securities finance news source this year, make sure it is your free subscription to Securities Finance Times
