held at the time was unaffected when risk factor such as default probabilities changed. This resulted with trades being overvalued as the valuation didn’t account for the likelihood of default. In the current regulatory environment, trades are re-valued frequently reflecting their fair value. Not only does this lead to losses being continuously accounted for, but this also makes the balance sheet of banks reflect a much more accurate value of their assets and liabilities.
6.3 Incentives and incremental CVA
The findings in section 5.4.2 demonstrated how incremental CVA can be both negative and positive, indicating that entering into a risky transaction can actually be risk reducing for banks in some cases.
Furthermore, the analysis demonstrated how banks with different existing exposure will face different incremental CVA from the same trade. Implying how the profitability of entering a trade can vary from bank to bank.
These findings demonstrate how OTC derivatives deviate from other financial instruments. As doing business with stocks and bonds does not lead to different P/L results for different banks. If two different banks would buy the same shares of the same company, they should be subject to the same P/L as well as the same fair value of those assets. This is not a direct result of the transaction between the bank and its client but rather an adjustment made on the banks books to comply with regulatory requirements.
CVA as a balance sheet adjustment and P/L contributor is applied on total netted and aggregated portfolios of banks (Hull, 2012). And as the analysis indicates, the CVA P/L can be a critical factor affecting the profit of banks from year-to-year. Therefore, it shouldn’t come as a surprise that banks are nowadays coming up with various measures to try to manage CVA, attempting to limit the losses suffered from it and increase the profit exploited from it. As discussed in section 3.3.3, banks have in recent years initiated frameworks to incorporate CVA in the incentives for front office employees. The results highlight how incremental CVA can vary from trade to trade, and how it can deviate from its independent CVA. Consequently, incorporating the incremental CVA in dealer’s compensation frameworks seems as a good risk management tool.
6.4 Further potential research
The bonus incentives setup in banking without incremental CVA leads to an adverse selection problem as described in section 2.2. Adverse selection in this case rises since the bank does not have the material knowledge (of the expected incremental CVA) which the employee is likely to have. But by calculating, monitoring and allocating incremental CVA on trade level, banks can see the incremental CVA
contribution from each trade and incorporate it into the bonus system. How this affects the behavior of derivatives sales managers and thereby the overall risk of banks is a highly relevant but an unexplored topic within CVA research.
In this thesis, DVA was completely neglected. This means that the total effect of counterparty risk presented in the analysis did not account for the possibility of own default. In practice, CVA and DVA are often calculated simultaneously and presented as [CVA – DVA] as DVA works as a natural hedge against CVA. Potential future research could include similar analysis including DVA expecting that the total [CVA – DVA] adjustment would be lower than the CVA adjustment only.
Perhaps the most important aspect of DVA is that banks make a profit when their own credit quality weakens. But as already discussed, there are multiple other factors associated with banks credit quality which weakens profitability as credit quality decreases. Comparing the DVA profit associated with a weakened credit worthiness spreads to other negative effects and thereby assessing the net impact of weakened credit quality, is a highly interesting potential research area within the xVA area.
7 Conclusion
The aim of this thesis was to explore CVA and to build and calibrate a CVA model for interest rate swaps and to assess the dynamics in CVA by sequentially adding trades to a netting set. A special focus was set on analyzing incremental CVA and the effect of credit quality on CVA. To meet these objectives, two counterparties were selected to compare different credit qualities and how they affect CVA. The relationship between the exposure from three interest rate swaps and marginal default probabilities derived from CDS quotes of the counterparties were analyzed. Furthermore, the dynamics of CVA were explored via incremental CVA calculations indicating how banks are affected by CVA from new trades.
The results presented in this thesis demonstrate a few main findings. They show how credit spreads affect CVA as higher spreads represent a higher probability of default. They also highlight the importance of netting for CVA calculations. The contrast presented by applying the same EE along with two different credit profiles represented by different CDS curves demonstrated, to no surprise, the effect of default risk on CVA. It is important to keep in mind that both counterparties had, like banks in general, a strong credit profile. However, Deutsche Bank has struggled in recent years resulting with a relatively weak credit profile for an international bank. The CVA difference for the entities was quite high which indicates how counterparties’ credit quality can affect the attractiveness of doing business with them and lead to different pricing due to CVA charges.
Incremental CVA figures from the calculations demonstrate how independent and incremental CVA are different. These findings show how fair value adjustments can bet both negative and positive depending on the existing portfolio and its correlation with new trades. Implications of these effects were discussed considering the information asymmetries between dealers and banks. How banks react to CVA by allocating incremental CVA on trade level affecting the incentive compensation of employees was also discussed in relation with the findings.
Whether or not regulators expected banks to dive so deep in reaction to the CVA regulation is questionable. But the practice of allocating the impact of CVA on trade level seems to be not only harmless but also a neat tool for mitigating counterparty risk on trade level as it should incentivize dealers to consider the counterparty risk associated with their deals.