Everyone wants to know how to stop the next financial crisis.
One lesson from the last one was that reliance on complicated models created a crisis of confidence in bank balance sheets.
So how banks provide for losses is changing.
On 1 January 2018, a new accounting standard for how banks report on financial instruments, IFRS9, comes into force.
Financial reporting standards rarely sound exciting to non-accountants, but this one will have a real effect on banks and the economy. Impairment losses are the largest factor affecting bank profits, so changing how they are calculated will have a real effect. Yet is still poorly understood.
Financial policy is currently going through a period of change, but it can look more akin to playing regulatory whack-a-mole, since fixing one problem can often prompt another. In this case, changes to a little-known accounting rule could well make lending to the real economy look very different.
One of the problems from the crisis was that too many people tried to “learn lessons”. Bank regulators wanted their wards to be safer, so they implemented tough new standards to make them hold more rainy day money. Banks now have both far more capital and significantly more liquidity.
The trouble is that accountants went away and did exactly the same thing, and created their own new, higher standard.
Banks will now be required to estimate future losses on their lending. Come the new year, they will be gazing into their crystal balls and introducing new models to calculate “expected loss”. They will then have to hold larger credit provisions against this – in other words, even more rainy day money.
The idea is laudable. If all banks are looking at all the relevant forward-looking information, we should have a better idea of which banks are riskier and, overall, the system should be safer.
However, IFRS9 is no panacea. In fact it may lead to volatility, inconsistency, lack of comparability, and the exacerbation of financial instability.
An estimate of future losses is just that – an estimate – and a highly subjective one at that. If a recession is predicted, these expected losses will accelerate, even if the current economic situation is benign.
The requirement to hold more capital amplifies this, hence the increase in volatility.
But banks don’t like volatility, and their shareholders like it even less. This is therefore likely to mean banks change who they lend to and how they treat the customers they do lend to.
Any unsecured lending, for example, is likely to come under intense scrutiny. Banks won’t want to show erratic performance, so may reduce this type of business. There may also be particular sectors that show wide variations in loan losses. Banks will treat these industries less favourably too.
Comparisons between banks will be difficult, since their views of the future could be radically different.
The practicalities of considering several possible scenarios, calculating the probability of each occurring, and modelling the impact will be extremely challenging.
The sheer size of loan portfolios, variety of credit products sold, and diversity of customers, geographies, and sectors will all have to be factored in when assessing the expected losses for the entire business.
Because of the way the standard was drafted, there will be significant “cliff-edge” effects. Some of a bank’s profit numbers can change very rapidly if lending issues look like they are developing.
This may mean that banks intervene a lot sooner than businesses have been used to in the past. Businesses with some performance issues may find the bank manager knocking on the door sooner rather than later, perhaps to look at the prices of the loans.
In this way the standard could exacerbate financial instability, rather than countering it.
No accounting rule is perfect. But there is a misconception that IFRS9 will fix more than it can, and its shortcomings may become evident very soon.
Rather than creating technical issues over which accountants and analysts scratch their heads, this has the potential to influence how banks are perceived, with real knock-on effects to the economy and access to finance.
The more people understand some of the wrinkles in the new rules, the less likely we are to see businesses affected. This is why it is so important to increase understanding of what this change to the rules will mean.
Ultimately, the best chance of avoiding the next crisis is for investors, businesses and policymakers to better understand how what may sound like a dry technical accounting issue actually works.
Read more: How do you audit a robot?