One of the key shortcomings of the first two Basel Accords is they approached the solvency of each institution independently. The recent crisis highlighted the additional ‘systemic’ risk that the failure of one large institution could cause the failure of one or more of its counterparties, which could trigger a chain reaction
Basel III addresses this issue in two ways – 1) by significantly increasing capital buffers for risks related to the interconnectedness of the major dealers and 2) incentivising institutions to reduce counterparty risk through clearing and active management (hedging). Since Basel III may not explicitly state how some of the new provisions address systemic risk, some analysis is necessary.
Basel III Provisions
Basel III substantially raises the amount and quality of core Tier 1 capital from 2% to 7%, plus an additional countercyclical buffer of up to 2.5% and a discretionary surcharge for ‘systemically important’ institutions, i.e., the big dealers. It also fixes known mispricing of securitisation risks, which is very important given the fundamental role of securitisation in the global banking system. Another key innovation is the inherent recognition that the risk-weighted capital ratio alone is not sufficient. Basel III supplements the capital model with a leverage ratio and liquidity requirements. Each of these enhancements has a systemic risk management objective.
Restricting the leverage of major dealers is clearly important from a systemic risk perspective. Basel III adds a minimum Tier 1 balance sheet leverage ratio of 3%, subject to further calibration. There are two reasons for this addition. First, countries that imposed a leverage ratio, e.g., Canada, seemed to fare better during the crisis. Second, the leverage ratio serves as a form of safety net for the capital ratio, which is vulnerable to arbitrage in both the numerator (capital) and denominator (risk-weighted assets).
In addition to the leverage ratio, Basel III introduces a short-term liquidity coverage ratio and a longer term net stable funding ratio, designed to address the duration mismatches in bank assets and liabilities. Banks fund a substantial portion of assets in the repo markets and when these markets froze due to declining mark-to-market collateral values, inter-bank lending also dried up causing systemic shocks. The link between liquidity and leverage amplified these shocks. This linkage comes from widening haircuts on repo collateral, which banks must fund with their own capital. Ultimately, these liquidity requirements are intended to prevent another ‘run’ on the shadow banking system and global seize-up of credit.
One of the critical sources of liquidity risk came from short-term funding of securitised assets in the repo markets, a practice that banks had ramped up to take advantage of regulatory arbitrages. Basel I and II under-priced risk weights for securitisations allowing banks to increase leverage (and returns). They further increased leverage by manufacturing additional super senior collateral through re-securitisation (e.g. CDO-squared). The fact that Basel made no distinction for re-securitisations encouraged this activity. Banks also moved securitised assets from the banking book to the trading book to access the more favourable capital treatment. Basel III (II ½) firmly addresses all of these regulatory arbitrages while providing a detailed ‘carve out’ for dealers with sufficiently robust risk management processes.
Along with the supplemental leverage and liquidity measures, the core capital model has been enhanced to address systemic risks more effectively. Capital models typically involve Monte Carlo simulations of future market scenarios using historical volatilities for the relevant market factors. An obvious weakness is that volatility tends to go down in normal (stable) times, resulting in lower capital reserves. Correlations also tend to be under-estimated during normal times. Conversely, when volatilities and correlations spike during a crisis, banks are forced to raise capital and deleverage as credit markets tighten. Basel III attempts to mitigate this ‘procyclicality’ through new capital charges for ‘stressed’ CVA VaR and correlation between financial intermediaries, which are expected to more than triple counterparty risk capital.
With the dramatic capital increases, Basel III incentivises banks to actively manage (hedge) counterparty risk. Many larger banks already hedge a significant portion of counterparty risk through central CVA desks and there appears to be general consensus and movement towards this model, accelerated by Basel III (and the desire to reduce earnings volatility). However, there are immense operational and practical challenges in setting up a CVA desk. The main operational challenges involve gathering position and market data and implementing scalable technology with robust CVA analytics. Some of the practical issues are illiquidity of many names, managing residual correlation and basis risks, and handling of DVA. DVA represents a gain (that can never be realised) based on the credit quality of the trader’s own institution and can’t be hedged with CDS.
Clearly, the best hedge for counterparty risk is collateral. While dealers typically have margin agreements between them, central clearing standardises the process and enforces tighter controls around collateral risks and rehypothecation. Clearing also helps immunise the system from the failure of any one big bank. Basel III assigns a minimal (1-3%) risk weight for cleared transactions, thereby fostering central clearing and the systemic benefits.
Conclusions
Whereas Basel III represents progress, there are several ongoing challenges. The first set of challenges has to do with the regulation itself. The timeline provides for a phased implementation period extending out to 2019. Another crisis could certainly occur within that time. While quantitative studies have shown limited impact of the higher capital requirements on the real economy, banks may choose to curtail or exit certain lending businesses if the returns are too low. A consequence could be the expansion of the unregulated and relatively opaque sector of the shadow banking system to fill the credit gap.
The second set of challenges is structural. Banks are moving toward active management of counterparty risk. However, there is limited or no liquidity in CDS contracts needed to hedge a significant number of counterparties and institutions will continue to manage a substantial portion of counterparty credit risk through traditional reserves and exposure limits. The residual counterparty risk portfolio is essentially a pool of loans and therefore fraught with the complexities of CDO structures. These complexities include model specification and configuration, manipulating large and diverse sets of position and market data, and managing unhedgeable correlation and basis risks. Therefore, counterparty risk portfolios will continue to be susceptible to large unexpected losses.
Another structural issue is related to clearing. While the near zero risk weight encourages dealers to clear CDS and other hedge transactions, not all products will be cleared, which means a critical mass of bilateral counterparty risk will likely remain in the system. Clearinghouses may also specialise in specific products, potentially increasing net counterparty risk. Finally, a clearinghouse could conceivably fail and there is no evidence that the 1-3% risk weighting will provide an adequate capital cushion to contain the systemic fallout.