Seeing the unseeable – credit risk in a world without defaults

15 March, 2021

When I was younger, my friend bought his first house. There were cracks in the brickwork and the roof had started to sag, but he didn’t plan to fix it. His thinking? If it hadn’t fallen in yet, it likely wouldn’t.

What this logic missed was that if the roof did collapse, it would be more due to factors like rain, snow and wind storms than it would with the passage of time alone. Estimating credit losses in the current economy reminds me of that roof. Some countries have started to emerge from the pandemic with defaults yet to transpire. Still, the reckoning might be ahead. Borrowing from my friend’s experience...

Maybe it only feels better because it's actually worse

Businesses and individuals haven’t suffered as much as we might’ve expected financially, but it isn’t because there haven’t been losses. Government spending has filled the gaps. It’s hard not to be optimistic if savings are up, markets are at all-time highs and indicators like GDP and unemployment continue to perform better than expected. Except that’s due to support being greater and longer than any of us thought, and that’s only because economic paralysis has lasted much longer too. This raises multiple issues like:

  • What are the potential consequences of stimulus that’s funded by government debt?
  • When will stimulus stop?
  • What happens when it does?
  • What will the new business environment look like?

Until we know…

Looking backwards to look forwards still might not work

Models are calibrated to the past as it’s usually our best indicator of the future when it comes to credit losses. But previously the past hasn’t included a global pandemic, economic shut-downs, ‘shelter in place’ orders and massive government stimulus. That’s still true. Some are now arguing that the absence of significant defaults indicates that previous overlays might need to be reversed, retreating to historically informed models instead. However – and this is really important – past models can only predict the future if the impacts of the four issues above are insignificant. And we’re now also seeing…

Sovereign and other risks are beginning to emerge

In order to shield us from losses, governments have been picking up the tab on our behalf. Albeit necessary, this isn’t without risk of replacing problems today with new ones tomorrow (think Greece and Venezuela). While we all hope for growth, governments might also escape indebtedness through:

  • Taxation.
  • Inflation.
  • Sustained low (possibly negative) interest rates.
  • Default.

Any of these could have profound effects on credit losses in the future. Then there’s potential for things like additional waves, hybrid variants, permanent sectoral impacts from changes in behaviour (like travel and retail). The list goes on. And with all this in mind...

‘It’s better to be roughly right than precisely wrong’

After a year developing robust estimates in a challenging environment (to say the least), maybe now is a good time to step back and assess how such factors are being related to macroeconomics. Which includes whether more or different scenarios and possible adjustments to probabilities might be warranted. As previous overlays unwind, we might find that others become necessary too - and that’s okay.

And in case you’re wondering, my friend did hedge his bets in the end. Like the Flaming Lips said, he couldn’t ‘see the unseeable’. But some risks are just too great to ignore.

All views expressed are the authors

Contact us

Follow us