In my earlier writings (here, here, here and here, for a start) I may have given the impression that I am not a great fan of the Bank of England’s so-called stress tests, mainly because the Bank’s stress scenarios barely break into a sweat.
My skepticism was heightened further when a little birdie from Moorgate suggested that the stress tests were really designed to ensure that the big institutions passed, ‘cos otherwise there’d be problems.
I had long known that the exercises had zero credibility, but even so, it comes as a bit of a shock to learn that people on the stress test team were more cynical about them than I was.
So now we know what I had long suspected: the stress test project is one big con job PR exercise to sell the Bank’s ‘Great Capital Rebuild’ fairy story to the public.
Here are some highlights from the December 2019 Financial Stability Report. I present them in ascending order of absurdity.
Exhibit #1
Consider the following chart.
The chart gives the banks’ projected cumulative profits before tax under the baseline scenario and the stress scenario.
On p. 8 we are given a little more ‘hard’ information: a loss of £77 billion after 1 year, which is the peak of the stress; projected impairment charges of £151 billion after 5 years, but which the chart suggests are basically borne after 1 year; and the projected impairment rate of 4.5%.
One is immediately struck by the thought that 4.5% is an awfully low impairment rate for a decent stress.
A good impairment rate for a reasonable crisis might be 10% and the impairment rate for a good crisis might be 15% or more.
If you had a 10% impairment rate – so you are talking about a reasonable but not too severe crisis – you would be looking at a loss of £336 billion after 1 year and the capital of the banking system would be wiped out pretty much however you measure it.
If you had a 15% impairment rate, you would be looking at losses of about £500 billion, which is on much the same scale as the losses the banks experienced from the last financial crisis.
Or look at it this way. If we extrapolate the Bank’s impairment charges for the Big Five banks, subtract those from their current market cap and then plot the stressed market cap against the projected impairment rate, we obtain the following chart:
Big Five UK Banks: Stressed Market Cap vs. Stressed Impairment Rate
The banks’ market cap falls sharply as the projected impairment rate increases. Indeed, it goes negative before the impairment rate hits 5%. A 10% impairment rate is enough to wipe out their capital more than twice over and a 15% impairment rate is enough to wipe it out thrice over.
So you can see what the Man from the Moorgate was getting at.
Exhibit #2
Then there are projected impairment charges (i.e., losses) on the banks’ real estate portfolios given in the first and third columns:
You call that a stress?
These numbers give no idea of how much the banks are exposed to a real estate downturn.
More on that later.
Exhibit #3
Finally consider the UK banks’ exposures to Hong Kong and China:
To quote the Bank’s Financial Stability Report (p. 27):
UK banks have significant exposure to Hong Kong, representing around 160% of their common equity Tier 1 (CET1) capital. The recent political protests in Hong Kong have been accompanied by a sharp slowdown in growth and falling asset prices. GDP growth contracted by 3.2% in Q3 — the weakest quarterly growth rate since the peak of the financial crisis in 2009 …
So according to the Bank, UK banks have 160% of their CET1 capital exposed to HK. 160% of the banks’ then CET1 is £410 billion.
Consider also that HK is in absolute turmoil. Its GDP is plunging, its real estate is obviously overvalued and real estate prices are starting to fall, its political system is being trashed by China and people are fleeing abroad.
Then the Bank of England runs a ‘stress’ exercise in which it projects that UK banks stand to lose only £22 billion in Hong Kong and China, a loss rate of 5.4% of its exposure.