Page 44 - Monocle Quarterly Journal Vol 1 Issue 1 Q4

BANKING
“LaR simulates the progression of a portfolio’s value through time – how cash ows should behave, how balance sheet values should change, and ultimately the cash in ows and out ows that result from these changes.”
will use simulated market interest rates as a driving economic factor, because interest rates usually show the strongest predictive power in behavioural modelling. e interest rate simulation produces 10 000 to 100 000 observations of possible future interest rate paths, along with paths for other market factors.
ese interest rate paths, and other market factor paths, are used to drive the contractual cash ow models, the behavioural cash ow models, growth models, and interest rate re-pricing models. ese models interact to provide a picture of how the balance sheet ‘evolves’ over time. New ‘synthetic’ accounts are created to compensate for predicted growth value, and cash ows are, in turn, calculated for these new accounts. e replication of the balance sheet is an image of how the future balance sheet may look.
One method we use is to hold interbank activity conducted by the bank constant over the funding horizon, to measure the bank’s reliance on the interbank market in times of heightened demand for liquidity. With the foregoing assumptions and processes in place, the simulation is run. At each month during the horizon, it’s possible to measure the liquidity gap on a point-in-time basis. Simply, the cash in ows and out ows are considered in isolation, to assess whether that month’s liquidity gap is positive or negative.
At the end of the one year horizon, for example, banks can calculate a distribution of possible one year Net Cumulative Cash Flows (NCCF). is is the sum of all twelve point-in-time liquidity gaps over the year, representing an accumulation of cash ow shortages, or excesses, from the beginning to the end of the year. At the 99.97 percent con dence level for example (consistent with a AA rating), LaR will be the 31st worst observation in this distribution.
LaR simulates the progression of a portfolio’s value through time – how cash ows should behave, how balance sheet values should change, and ultimately the cash in ows and out ows that result from these changes. It is known as a dynamic portfolio approach as the methodology assumes that the portfolio is constantly changing as loans mature and new loans are created. e goal of LaR is to produce 10 000 to 100 000 values for the NCCF at the desired horizon, which could be overnight (1 day), 7 days or monthly, from 1 month to 12 months.
42