Bank stress tests
Author
Discussion

thinfourth2

Original Poster:

32,414 posts

235 months

Friday 4th November 2011
quotequote all
I heard about bank stress tests on the radio this morning while they were talking to Bob from barcleys. Bob seemed to be talking sense.

Anyway these bank stress test probably don't involve Bob, duct tape and a Justin Bieber CD.

So what is a stress test and how is it done?

ATG

23,834 posts

303 months

Friday 4th November 2011
quotequote all
thinfourth2 said:
I heard about bank stress tests on the radio this morning while they were talking to Bob from barcleys. Bob seemed to be talking sense.

Anyway these bank stress test probably don't involve Bob, duct tape and a Justin Bieber CD.

So what is a stress test and how is it done?
It's exactly the same process you might apply to see how your own finances would stand up to some mayhem. For example, let's say you've got some cash on deposit, a fixed rate mortgage and pension savings. You might ask yourself what position you'd be put in if the stock market dropped 30%, and interest rates doubled. Rates going up would be good for your money on deposit, the fixed mortgage would be OK for a while and your pension pot would have probably devalued quite a bit. And you could dream up other fairly extreme but not completely unrealistic scenarios to span all the kind of mayhem you can imagine. If your personal finances would survive all of the scenarios, then you're probably in fairly good shape right now.

Stress testing banks is the same idea. You look at the assets and liabilities on their balance sheets and you revalue them under various scenarios ... e.g. Greek defaults and pays out 50%. If that means the bank goes bust, they need to start reducing their exposure right now or put money aside right now to provide a cushion to absorb the potential loss.

Puggit

49,825 posts

279 months

Friday 4th November 2011
quotequote all
I think it involves Unicorns?

munky

5,328 posts

279 months

Friday 4th November 2011
quotequote all
Stress testing defined and implemented properly is probably THE most useful risk management tool. However, the EBA (regulatory) stress tests are not defined properly, they are way too mild. Dexia passed the latest stress test with flying colours... then immediately had to be rescued by France and Belgium.

We don't have to do the EBA stress test, but we do several of our own and they are much, much more severe (and varied) than the EBA one.

elster

17,517 posts

241 months

Friday 4th November 2011
quotequote all
I thought it was a spinning wheel with numbers from -20% to +20% and wherever it lands is what the banks rating is.

Scientific as much as the EU is democratic.

munky

5,328 posts

279 months

Friday 4th November 2011
quotequote all
elster said:
I thought it was a spinning wheel with numbers from -20% to +20% and wherever it lands is what the banks rating is.

Scientific as much as the EU is democratic.
..would be an incorrect assumption. We have 7 full time staff on stress testing in London alone, one of whom has a PhD in nuclear physics for example, from the French equivalent of Oxbridge. That's not counting the IT support staff assigned to it. Traded products - bonds, stocks, derivatives - are all re-priced under different assumptions of price shocks in all markets. Counterparty risk is recalculated under stressed conditions using simulations of PFE (potential future exposure) in addition to more simple downgrade (stressed transition matrix) assumptions for RWA stressing, to predict regulatory capital impacts. Funding costs and liquidity stress impacts are calculated under various scenarios. And so on. We combine the lot, take the worst scenario loss, and add that to our capital buffer.

Not much wheel spinning goes on.

elster

17,517 posts

241 months

Friday 4th November 2011
quotequote all
munky said:
elster said:
I thought it was a spinning wheel with numbers from -20% to +20% and wherever it lands is what the banks rating is.

Scientific as much as the EU is democratic.
..would be an incorrect assumption. We have 7 full time staff on stress testing in London alone, one of whom has a PhD in nuclear physics for example, from the French equivalent of Oxbridge. That's not counting the IT support staff assigned to it. Traded products - bonds, stocks, derivatives - are all re-priced under different assumptions of price shocks in all markets. Counterparty risk is recalculated under stressed conditions using simulations of PFE (potential future exposure) in addition to more simple downgrade (stressed transition matrix) assumptions for RWA stressing, to predict regulatory capital impacts. Funding costs and liquidity stress impacts are calculated under various scenarios. And so on. We combine the lot, take the worst scenario loss, and add that to our capital buffer.

Not much wheel spinning goes on.
How come so many banks that the general public knew just from common sense were going to fail, passed the Euro Bank Stress Tests with flying colours?

munky

5,328 posts

279 months

Friday 4th November 2011
quotequote all
elster said:
munky said:
elster said:
I thought it was a spinning wheel with numbers from -20% to +20% and wherever it lands is what the banks rating is.

Scientific as much as the EU is democratic.
..would be an incorrect assumption. We have 7 full time staff on stress testing in London alone, one of whom has a PhD in nuclear physics for example, from the French equivalent of Oxbridge. That's not counting the IT support staff assigned to it. Traded products - bonds, stocks, derivatives - are all re-priced under different assumptions of price shocks in all markets. Counterparty risk is recalculated under stressed conditions using simulations of PFE (potential future exposure) in addition to more simple downgrade (stressed transition matrix) assumptions for RWA stressing, to predict regulatory capital impacts. Funding costs and liquidity stress impacts are calculated under various scenarios. And so on. We combine the lot, take the worst scenario loss, and add that to our capital buffer.

Not much wheel spinning goes on.
How come so many banks that the general public knew just from common sense were going to fail, passed the Euro Bank Stress Tests with flying colours?
Because they were a farce... and the EBA took so long to devise them and gave the banks so long to calculate the results, that they were already well out of date (the latest one didn't even include a Greek default, even though we all knew a Greek default was imminent, FFS). I suspect the EBA didn't want to suggest a Greek default was possible for political reasons, being a European regulator and all.