How to think about risky lending?
In today's Finshots, we talk about US banks and their latest approach to lending.
The Story
Banking is a risky business. You loan money to individuals only to find out some of them won’t pay you back. So it’s imperative to set aside some extra money so that you have enough financial leeway to absorb losses when these events do materialize i.e. You need to set aside funds for when some of your customers default on their repayment obligation.
But wait — How do you decide on the quantum of allocation?
I mean really, how do you know what kind of money you’ll need to set aside for these unforeseen events?
Well, for the longest time, bankers followed a relatively simple method. They created an allowance based on what they called — the Incurred Loss Model.
The assumption here is simple. There is no reason to set aside money until there is evidence that the borrower might not repay in full.
And to this end, the implementation is rather straightforward. You lend money to an individual and you go about your day without thinking too much about it. However, in the event that the borrower delays payment when it's due, you start thinking about future losses. In the off-chance, the delays breach a certain threshold, say — 60 days, then you set aside some money based on your past experience with such people. For all you know, the borrower might still repay you in full someday in the future, but you’d want to be prudent in your assessment. So you do your provisioning when you’re fairly certain that the borrower might default. And for the most part, this formulae works quite well.
Except during a recession.
For instance, an economic downturn forces borrowers to fall short of their loan payments. Banks, in turn, will be forced to set aside more money. And as they set aside more money, they won’t have enough money to lend. And if they don’t lend, corporates and individuals desperately looking to borrow new money during a recession will have to look elsewhere. It will only exacerbate the crisis. It’s a vicious cycle this.
In fact, this is precisely what happened the last time around when the US housing bubble popped. So in a bid to avoid a similar crisis, the Financial Accounting Standards Board mandated US banks to adopt a slightly different approach.
So now we have another model called The Expected Credit Loss Model.
This time around, you don’t pretend every borrower is going to pay you back. After all, an individual’s ability to repay loans depends on a long list of variables — including his/her credit history, the duration of the loan, other economic factors, etc. So you take all of this into account and you build a model that will help you peer into the future. Once you have a good idea of the potential losses that could accrue, you set aside a small portion of the loan the moment you disburse it. So in effect, you are provisioning based on future expectation. And experts believe this might be a more prudent alternative.
For instance, look at the events that preceded the US financial crisis. First, you had big banks loan large amounts of money to prospective home buyers. However, the banks had very little idea of their repayment ability. Once people were made aware they can invest in real estate using borrowed money, there was an explosion in demand. US housing prices kept appreciating beyond all reasonable levels and bankers kept doling out new housing loans pushing housing prices even further. Eventually — prices crashed. People defaulted on their repayments. Banks were forced to set aside more money based on the incurred loss model. And as the recession gained momentum, banks realised they didn’t have money they could lend to individuals and corporates desperately looking to borrow. This only accentuated the crisis.
But what if they had adopted the expected credit loss model? How would things have panned out if we were living in an alternate universe?
So let’s take it from the top once again. Big banks start doling out large amounts of money to prospective home buyers. Soon enough they are disbursing loans to people with no real credit history. But at this moment they are forced to account for future losses arising out of potential defaults. So they start setting aside some money based on the expected credit loss model. And soon enough, they realise they can’t be reckless with their lending standards. They are already setting so much money they can’t put to good use. They can’t keep doing this. They have to scale back. And as home-buyers realise they no longer have access to cheap debt, they drop their aspirations to buy new homes. Eventually, demand fizzles out. Housing prices stabilize. And the bubble never pops — because hey, there is no bubble now.
What a wonderful world, no?
But alas, things are more complicated.
For instance, consider J.P. Morgan. The bank has already adopted this new approach. And right now, they are being forced to set aside billions of dollars in provisions because they are lending money to people with no jobs (considering the lockdown and all that). Meaning, they can’t be lending with complete flexibility at a time when new loans might be the only recourse for most individuals and corporates.
In fact, the RBI has been engaging with Indian banks hoping to convince them to switch and adopt the Expected Credit Loss model. However considering most banks in India are tentative about lending as is, do you think forcing them to set aside more money (under the new approach) would be a wise thing to do? Especially in the midst of a pandemic!!!
You tell us…
Until then…