Bank of England stress tests are getting tougher – but do they still add value?
The Bank of England 2016 stress test was the toughest to date, with a more severe scenario which incorporated a combination of both a UK and global recession. Combined with a higher pass mark, this created challenges for several institutions.
All of the banks started the exercise in more robust shape with a lower risk profile than in previous years, but the severity of the stresses meant the results produced the highest average reduction in capital over the three year history of the test.
The Royal Bank of Scotland (RBS) saw a shortfall versus its systemic reference point and its Tier 1 leverage ratio hurdle rate. While this failure grabs headlines, these results exclude the capital raised in 2016 and the subsequent changes to the bank’s capital plan which will improve its resilience. All else being equal, this will provide a stronger starting point for next year’s exercise.
Our analysis shows that, once again, misconduct costs weigh heavy. Over the stress tested period, there were projected incremental losses of around £40bn. While extreme, this sum is less than the actual losses experienced over the period 2011 to 2015 (estimated at £55bn based on the KPMG Bank Benchmark Report). The Bank of England stress test only reflects projections for known misconduct issues – the likelihood of no new issues arising is arguably questionable.
Read more: RBS on renewed drive to cut costs after stress test flop
It’s clear that the Bank of England is walking a tightrope between making tests too difficult and seeing banks fail and making them so easy that people question their value.
Our recent stress testing benchmark survey found that a quarter of firms globally estimate stress tests cost them in excess of $100m a year and over three months to complete. With such a huge resourcing effort being put into these tests, and such public pressure on banks to be robust, firms need to get better at running stress testing and putting the results into practice.
Regulators, including the Bank of England, also have a role to play in terms of maintaining a stable framework and providing as much guidance as possible to ensure testing continues to add value.
Next year banks will have to comply with two separate tests, the annual cyclical exercise and for the first time an exploratory stress test. There have been some indications as to what this experimental test will involve, namely that it will look seven years ahead and will focus on business model not the macroeconomic environment.
This poses some questions. Most banks are working to three or five year plans. Even putting aside the elements of change we don’t yet know about, within seven years companies will have implemented Basel 4 and IFRS9. Both will fundamentally change banks’ capital structures – and maybe even the make-up of the firm itself, as staffing needs change and automation is increasingly introduced.
Furthermore, in a world where six months involves events like Leicester City, Brexit and Trump, it is difficult to predict the next few months, let alone seven years!
Read more: The OBR shouldn't be expected to forecast so far into the future
Having withstood 2008, 2009 and 2016, banks have proved their ability to survive shocks in real life as well as in regulation. Interestingly, the currency swings included in this year’s stress test were actually less severe than we experienced after Brexit.
The Bank of England is clear that stress tests plan against hypothetical shocks rather than making a prediction of what’s ahead. But it does leave many asking: what else is there to test against?