In credit risk management many banks/institutions are now using the
"expected loss" methodology to calculate the level of provisions they
need to hold against their portfolio. I believe the idea is that you
set aside provisions to cover you for expected losses and the bank's
own capital is meant to address what can be called "unexpected loss".
The Basel II guidelines address this later point. My question relates
to the expected loss portion. If you calculate the expected loss on a
specific loan (ie likelihood of default times outstanding balance on
the loan less any risk mitigants - security/collateral etc) how do you
calculate from this the level of provisions that you need to hold
(must be related to time value of money and the risk margin you are
pricing). And secondly (the important bit)can you assume from a
theoretical standpoint that each loan's provisions should be
sufficient to address that particular loan. In other words if you do
your expected loss calculation correctly am I safe to assume that you
should not be relying upon any unused provisions on one loan to cover
you for greater than expected losses on a second or third loan (these
will be covered be those loan's own expected loss provisions, plus any
unexpected loss capital cover, plus your general portfolio risk
management strategy)