No, it’s not a simplified double-helix. It’s a graphic representation of dynamic provisioning — the new darling of accounting methods. As JP Morgan notes in its 184-page opus on financial reform, regulators around the world are looking at the method as a way to better prepare banks against losses.
But there’s at least one place where the system is already in use — Spain.
Dynamic provisioning essentially combines two accounting methodologies; `expected loss,’ and `through the cycle’ methodology. Expected loss means forecast credit losses are taken before evidence of any impairment. It has the effect of smoothing losses through economic cycles, as you can see in the chart at left.
‘Through the cycle’ means making provisions for the entire lifetime of a loan, usually based on long-run historical data. It’s in contrast to `point in time’ provisioning, which would be based on where the bank currently is in the economic cycle (i.e. the upswing or downswing in the first chart).
Dynamic provisioning combines the two. Banks would basically build up a buffer of extra provisions in the good times, to cushion against the bad times. Here’s JP Morgan:
A dynamic provision (also sometimes called a statistical provision or generic provision) is effectively an ‘expected loss’ ‘through the cycle’ methodology. This can be used in combination with a point in time provision, based on an incurred loss or expected loss methodology. In the Spanish system there is a ‘specific’ provision which is point in time, and a ‘generic’ provision which is through the cycle.
Great — and the Spanish method of accounting has often been cited as one of the reasons Spanish banks have so far weathered the financial crisis reasonably well. But there are some potential problems.
For a start, the expected loss method is almost entirely dependent on (you guessed it) banks’ expectations of future losses. To make those forecasts they often use financial models based on historical data. If the financial crisis has taught us anything, it’s that these sorts of systems are prone to collapse or stochastic-schism when defaults start rising above a certain threshold or correlation.
The worry then is that the buffer built up via dynamic provisioning may be overwhelmed by actual losses, and banks (Spanish or otherwise) might have to up their through-the-cycle provisioning.
Back to JPM:
. . . the visibility of the data can be a limiting factor, and for the Spanish banks the scale of the generic provision has been seen to be too small, and may be revised upwards in the future. Whilst there is a huge amount of debate about the direction provisioning requirements should take, we would observe that the Spanish system of ‘dynamic provisioning’ has helped the Spanish banks significantly through this downturn, and so we expect there will be pressure to adopt some of the features on a Europe wide basis.
Related links:
A smooth IASB and an impairment change – FT Alphaville
BBVA, an exercise in Spanish banking losses – FT Alphaville
Spanish banking crises, then and now – FT Alphaville
Spanish catastrophe, datapoint del dia – FT Alphaville