There is an article in the Financial Times today that says:
New accounting rules could cripple parts of the banking sector by forcing earlier recognition of loan losses, as the coronavirus pandemic threatens to push the world into recession.
“There is a problem with the new accounting rules,” which would “increase provisioning in a dramatic way”, one member of the European Central Bank's governing council told the Financial Times, referring to sweeping changes to the way loans are accounted for.
It has to be said that this is complete and utter nonsense and comes from the school of accounting called ‘Let's pretend everything is okay when it isn't, and hope no one notices'. Please excuse my cynicism, but this is not the time for backtracking on IFRS9 reforms.
Let me explain why. There are, in essence, two methods of accounting for losses in accounts. They are called the arising loss method and the anticipated loss method. An example would help. Suppose a bank has lent £5,000 to be repaid over five years in equal instalments (a little simplistic: it will do). Almost immediately after the loan has been given the borrower, due to unforeseen circumstances (for example, their business closing down due to a pandemic) advises the bank that they cannot make loan repayments and that without government support their business will fail. The bank faces the immediate prospect that the £5,000 loan will not be repaid.
Under the arising loss method of accounting the £5,000 will be written off as a loss by the bank over the entire loan period i.e. as each instalment that was due fails to be paid. So, in the first year the loss that would be recorded would be £1,000, and in the second year the same loss would be shown in the accounts, and so on until at the end of year five the entire loss would have been recognised. The problem is that throughout this period there then remains an asset on the balance sheet which is knowingly false: The loan is not going to be repaid.
Alternatively, under the anticipated loss method the entire loss is recognised at the time that it becomes clear that no repayment is to take place. This, of course, means that in the first year a loss of £5,000 would be recorded.
Until 2005 under the then rules of UK accounting the anticipated loss method of accounting was always in use. This was considered prudent. What it meant was that there was no asset recorded on the balance sheet of the bank which its directors knew was entirely artificial because the prospect of any loan repayment taking place was actually, or close to, zero. The logic was quite straightforward: there should be no false representation to the shareholders that an asset existed when that was not true.
Under the rules of the International Accounting Standards Board, introduced in 2005, this changed. Losses were then recognised on an arising basis. The consequence was that many of the losses that should have been provided between 2005 and 2008 were left on the balance sheet of banks and we all know the consequences: those banks then failed and had to be bailed out. False accounting assisted this process, which is exactly why pressure was brought to bear by a quite limited number of accountants to change the rules so that the anticipated loss method of accounting was re-introduced.
Now we are seeing the backlash: just at the moment when we need an anticipated loss method of accounting for banks some are claiming that they want to move back to an arising loss basis and as a result let banks misrepresent the true natures of their balance sheets. We cannot afford this and must not afford this. It would be an outright disaster: lies never help good management.
Let me add one thing though. I have been suggesting that banks allow their borrowers repayment holidays. There is good reason for this. It is not just that this would mean that losses on an arising basis would not be recognised because repayments would not be due; it also means that, presuming that at least some of the borrowers can reorganise their affairs so that they can then reschedule their debts and pay the banks what they owe, albeit over longer time periods, then losses to be recognised on an anticipated basis will also be reduced. This suggestion does, then, relfect a completely practical relationship between granting bank loan repayment holidays and preserving the value of bank assets without ever misrepresenting the truth to anyone.
It is deeply disappointing that some are wishing to change accounting rules to provide false representation of the true state of the banks at this time of crisis. What we do, instead, need is truthful accounting backed up by policy that is to the benefit of both the bank and its customers, which is what happens when loans can be carefully extended to give greatest chance of recovery of debts owing.
Thanks for reading this post.
You can share this post on social media of your choice by clicking these icons:
You can subscribe to this blog's daily email here.
And if you would like to support this blog you can, here:
Richard, first up, thank you for all you do as a critical voice and your commitment to raising and discussing important issues. However, I just want to query your analysis of IFRS 9 in the interest of accuracy.
Under the previous rules your example would (or at least should) have led to the recognition of the loan as impaired and a partial or full provision for the total amount due depending on the assessment of likely future cash flows. The trigger for impairment testing is clear and you couldn’t (or shouldn’t) just provide for the next repayment if future repayments are also compromised.
The real difference is that under the previous model you had to wait for a triggering event for impairment testing (in respect of individual loans, not for portfolios of similar assets which were already considered collectively). Now you have to assess and provision probable losses over the life of the loan from the moment of initiating it based on credit risk and keep that up to date.
You have a higher level of provisioning because some loans always go bad. You just don’t know which ones in advance. So the difference in the current environment only should only affect provisions on loans which don’t yet show some individual trigger for impairment testing.
This is the theory. In practice the focus has changed from discussions about what constitutes triggering events for impairment and the impact on future cash flows to discussion about the extent of expected future losses based on changes in credit risk.
There is however no question that there is a general problem about when and how provisions are recognized, financial or non-financial. The approach to climate change is the clearest example as you have highlighted. Why do you now have to provide up front for credit risk but not for non-financial risks? Best regards, Stephen.
I’m going to be honest and say I am simply ou of energy to debate that one now
Sorry….the spirit is willing but I’ve done my bit for now
No worries, Richard. It’s not important at the moment. All the best, Stephen.
[…] Second, they want to continue use of the false accounting rules on losses that helped precipitate that crisis and which we have only just got rid of, which I discussed here. […]