Page 41 - Monocle Quarterly Journal Vol 1 Issue 1 Q4

lIquIdIty At RISK: A meASuRemeNt AppRoAch wIthIN BANKING INStItutIoNS
and principal components for simulations of Net Interest Income (NII) and other measures of bank pro tability.
ere is, however, no world-wide standard on the contractual terms or conditions which may be imposed on a bank’s products. Cash ow functions must be customised to cater for the unique product types and payment structures a bank o ers its customers.
e risk, across many product types, is that the cash ows expected under the contractual pro le are di erent from actual, experienced cash ows. is is because of the options customers have to deviate from the initial terms and conditions of their account; Pre-Payment Risk (the risk that a customer will repay an asset before the contractual maturity date); Early-Redemption Risk (the risk that a client will withdraw a deposit before the contractual maturity date); or Rollover Risk (the risk that a liability will reach a maturity date, and a higher funding cost is demanded by the depositor to roll it over). But it could also be because of behavioural market impacts, such as credit risk, resulting in the cessation of a particular set of cash ows.
Behavioural modelling, using econometric techniques and statistical methods, can estimate future client behaviour, and transform the results of these models into predicted cash ows. It also includes statistical techniques used to predict the behaviouralised cash ows of portfolios of uctuating products, such as savings and current accounts. Simply, it adjusts contractual cash ow calculations to re ect the most likely client behaviour in the future.
As a rule, behavioural models used by banks have typically been long- run averages of past behaviour, so they’re not sensitive to the prevailing economic environment, or potential future economic developments. But it is possible to directly relate the level of each behavioural risk to an economic factor. Let’s look at pre-payments, for example. We can create a statistical model which relates levels of pre-payments in housing loans to selected interest rates or interest rate changes. Logically, as interest rates decrease, banks and their competitors are able to reduce the level of interest charged to products such as housing loans. And, of course, o ers of lower rates entice customers to re nance.
So we can estimate the level of prepayment on sub-portfolios, given the prevailing level and changes in key interest rates. Behavioural cash ow models address the full range of possible behavioural adjustments for each and every product, for diverse client types.
“ e risk, across many product types, is that the cash ows expected under the contractual pro le are di erent from actual, experienced cash ows.”
39