There is a huge difference between the implementation of the AMA and SMA.  The AMA, with its large number of moving parts and adjustable parameters, which could easily run into the hundreds, is so convoluted that local regulators placed much emphasis on trying to understand its configuration, making sure it is implemented correctly and not being misused.  

The SMA, on the other hand, is highly prescriptive, with all of its parameters clearly and transparently defined.  So there is little room for manoeuver when it comes to its implementation.  The local regulators’ role in the implementation of the SMA would, therefore, be more focused on checking if the correct data are being used in the capital calculation process, because this is the only place that the numbers can be played with.

 

 

Several research papers have already appeared on this subject, assessing the impact of the SMA on capital.  The main conclusion of most of these is that the SMA causes an uplift in the aggregate capital in comparison with the AMA capital across a consortium of banks, with some increasing and others decreasing.  This means that some banks will be over insured and others underinsured against operational risk in comparison to their AMA capital, but this assumes that the AMA capital is correct and the SMA is not.  The studies also show an apparent bias of the SMA in favour of the US against the European banks.

 

 

Focus will be on simplicity, consistency and transparency. Given the highly prescriptive and transparent nature of the SMA, it will be easier for local regulators to better understand the models and determine, on the spot, whether the numbers generated do or do not make sense.  This way, regulatory checks and feedback will be much swifter, with time scales spent on regulatory reviews significantly shorter.  

The SMA will also enable the risk managers to use the models independently, without having to rely on the op risk quants to generate the numbers.  The SMA will therefore make communications between management and the modelling teams more interactive and, certainly, much less intimidating due to its more simple and transparent nature.

 

 

Absolutely!  Recent research shows that the SMA has a number of flaws.  However, many of these flaws were found with a strong bias towards keeping the AMA and getting rid of the SMA altogether. 

If there is no choice but to bring in the SMA in its current, proposed form, changes will have to be made to it and 

one of them is to shift it more towards an LDA-type equation, albeit at the entity level and excluding all the details that have plagued the AMA and the complications that led to its downfall. The AMA route, with all the baggage it carries, is now very likely going out of the window and the way forward is to make simplicity, consistency and transparency a major focus in the next generation of op risk capital models, whether it’s the proposed SMA or a different version of it.

 

 

These conferences are the best places to meet people who share your interests, ideas and expertise.  So they are ideal for learning something new, getting up to speed and networking.