By Jon Gregory, Independent Consultant
Regulation such as Basel III has increased counterparty risk related capital requirements through components such as the CVA capital charge. Regulatory capital for counterparty risk has become increasingly costly and is one reason why some banks have seen return on capital for OTC activities return at best low single digit returns. It is therefore not surprising that the growing rigour around valuation adjustments is being extended to the lifetime cost of holding regulatory capital though KVA (capital value adjustment). Incoming regulation such as the leverage ratio and the use of capital floors has only enforced the significant capital costs that must be borne for OTC transactions, especially long-dated ones.
Banks have for many years set capital hurdles when determining prices to quote in transactions. In some sense, therefore, KVA is the oldest VA of them all. However, traditionally capital hurdles were set loosely and not necessarily with direct reference to actual required regulatory capital. KVA can be seen as formalising such a practice, in line with the relatively sophisticated quantitative calculations that are typically used for other components, such as CVA and FVA. In short, KVA is an up-front amount that would generate a specific return on capital via rolling (post-tax) profits over the lifetime of a transaction. As shown in Figure 1, a proper calculation of KVA therefore requires calculating the expected regulatory capital over the full term of a transaction. Note that KVA will only produce the correct return on capital in expectation and therefore, like other VA terms, hedging considerations arise.

The accurate calculation of KVA is complicated by a number of aspects:
- capital charges change as a function of credit and market risk factors. A correct calculation of KVA should ideally deal with this properly rather than taking a single projected scenario in the future. Since regulatory capital calculations may be Monte Carlo based (at least for banks with internal model method approval), this represents a computationally challenging problem.
- Changes and additions to regulatory rules such as the leverage ratio and the SA-CCR make future capital charges impossible to predict accurately. Recent proposals by the Basel Committee on Banking Supervision suggest that the CVA capital framework may also have a significant overhaul in the coming years.
- Future changes to transactions such as restructurings or unwinds or different underlying collateral terms cannot be predicted with certainty and yet will affect future capital requirements.
- Regulatory capital is generally a portfolio level calculation and therefore the pricing of KVA for a new transaction should ideally account for such portfolio effects.
- The return on capital specified is a rather qualitative parameter not directly linked to any fundamental observable economic quantity. Corrections for effects such as efficiency and tax are also required.
Regulatory capital for counterparty risk has become increasingly costly and is one reason why some banks have seen return on capital for OTC activities return at best low single digit returns.
For the above reasons, KVA is often calculated in a relatively simplistic framework with simplifying assumptions. Whilst techniques are likely to become more advanced and align the KVA calculation with the actual future regulatory capital requirements, behavioural aspects (e.g. restructurings and regulatory change) will always create problems.
KVA is being more commonly seen in market prices, although it does not yet have the ubiquity of other components such as CVA and FVA. Banks can take substantially different approaches to reflecting capital costs in transactions. Nevertheless, it seems likely that KVA will become another relatively standard valuation adjustment with time. This raises the prospect of whether KVA will also become an accounting adjustment, driven by the exit price concept attached to fair value. In one sense this might seem inevitable but there are large hurdles: for example, banks would have to report substantial losses to reflect capital costs not previously part of financial statements. Furthermore, XVA desks in banks would be incentivised to transfer-price and risk manage KVA, leading to even larger P&L swings and hedges than such desks already require.
A well-known aspect of banking is that profits on transactions are often recognised immediately and paid out in bonuses and dividends. This leaves nothing left to show as a return on a transaction in future years. Since OTC derivative transactions can exist for many years, or even decades, this has long been viewed as a problem and was seen as contributing to a reckless bonus culture in banks. A fascinating feature of KVA is that it has the ability to redress this imbalance. A bank reporting KVA adjustments would accrue the profit over the life of a transaction rather than upfront. Almost a decade since the start of the financial crisis began, KVA is more than another valuation adjustment and may change the entire derivatives business model.
- www.bis.org/bcbs/publ/d325.htm