For the last few years, Basel Committee for Banking Supervision was trying to finish the largest regulatory overhaul in living memory – Basel 3, as a replacement of Basel 2 that was found wanting in the crisis of 2008.
The last remaining major piece of the new regulatory picture was market risk capital charges, issued in January this year. This delay was not for the lack of trying – the first consultative document to what is termed “Fundamental Review of Trading Book” (FRTB) was issued in May 2012, and since then BCBS issued an updated version of the proposed rules every year.
While the basic ideas of FRTB are simple, the number of issues that needed to be solved makes the whole picture complex. For any implementer, the situation was not made any simpler by BCBS changing their approach with almost every iteration – although luckily the final version was not substantially different from the last version issued in summer 2015.
So what did we get after all those years?
FRTB is first and foremost concerned capitalising banks for market risk. That is the risk arising from instruments that are reasonably tradable – have a two way price in reasonably liquid markets.
As the name implies, instrument like these are mostly kept in the “trading books” of financial institutions, as opposed to the banking books. Distinction between these was one of the issues that failed under the last regulation, and FRTB addresses.
Under Basel 2, the definitions of a trading and a banking book were vague and the decision which is which rested with the financial institution. The boundary was very permeable. As a result, some institutions moved risk between trading and banking books almost at will, in effect arbitraging the regulations, and using whatever capitalization was more beneficial.
One of the major goals of FRTB is to stop this behavior. To this effect, the definition of a trading book is much more prescriptive, even enumerating products (such as options, listed equities, instruments accounted for using fair value) as automatically presumed to be in a trading book. In addition, moving a position from trading to banking books was made much harder, and cannot improve capital charge.
FRTB also addresses some of the perverse incentives and outcomes created by Basel 2.5. One of the most visible is the so called split-hedge. Hedges meant to manage CVA exposure risk looked to the previous regulation as a naked risk taking, and thus attracted capital charges. Institutions could then choose between prudent hedging penalized by extra capital charges, or more risky behavior with smaller capital requirements. In effect, a competitive advantage for being less prudent.
FRTB now splits all the trading activity into two categories. CVA hedges, which include any trades used to manage CVA (and are presumed to sit in a CVA desk) – both the credit side and exposure, and everything else. While this is not an optimal solution, BCBS assumed that creating a consistent and workable inclusion of CVA into the main market risk framework was beyond the skill (and needs) of a number of smaller institutions.
On the other hand, this split means that there are in fact two market risk frameworks. The first one is often called FRTB-Trading Book (FRTB-TB), captured in BCBS paper 352. The second one deals with CVA, and has its own consultative paper, 325.
Broadly speaking, FRTB framework allows for two methods of calculating the capital charge. So called Standardised Approach (SA), and Internal Model Approach (IMA). SA charge is also sometimes called SBA – Sensitivity Based Approach, as a part of the charge is based on risk sensitivities of the trading portfolio. While the original CVA proposal included both SA and IMA approaches, the new consultation paper BCBS 369 removes the IMA approach for CVA entirely.
BCBS recognizes that calculating CVA, and its sensitivities, requires significant degree of sophistication and computational capacity. Even the sensitivity based approach can be too much for smaller institutions, with very small trading portfolios.
As a result, the CVA market risk capital charge introduces a third method, so called Basic Approach-CVA (BA-CVA).
Taking a step back, the regulatory picture once both 352 and 325 are implemented will look like this:
The original post was published at Vlad Ender’s Linked In page. This is a slightly updated version.
 Note that FRTB does not calculate Risk Weighted Assets – it calculates the direct capital charge, which of course can be used to create nominal RWA.
 Which is not entirely correct, as SA is made of several charges, and only one of them is based on sensitivities, but nevertheless the naming stuck.
 While strictly speaking BA-CVA is not required to be implemented by an FI with FRTB-CVA accreditation, it is conceivable that the regulator may withdraw the accreditation and the FI would be forced to implement BA-CVA.