News / Big Data /

Using analytics to prevent another credit crisis

IFRS 9 may sound like a sci-fi movie sequel, but it’s a new international financial reporting standard that will require banks to provide more timely recognition of expected credit losses based on future expectations. That’s where predictive analytics will come into play.

Share this article:

This is a story about how the 2008 global credit crunch is changing the way Canadian banks use technology to crunch their numbers today.

Back in 2008, the credit crisis (remember sub-prime U.S. mortgages, the Lehman Bros. bankruptcy and the Bear Stearns government bailout?) rattled the global economy. It also rattled regulators.

According to a recent report by SAS, financial industry regulators analyzed the crisis and “identified the delayed recognition of credit losses on loans and other financial instruments as a weakness in existing accounting standards.”

In other words, banks and other financial institutions were lending way too much money to customers who couldn’t really afford to pay it back on time; mass loan defaults ensued. To prevent a repeat of that debacle, regulators wanted to make banks disclose these types of credit risks to investors and financial markets much sooner.

So one of the biggest regulators for bean counters, the International Accounting Standards Board, is bringing in IFRS 9. It sounds like a sci-fi movie sequel but stands for International Financial Reporting Standard. As explained in the SAS report, IFRS 9 “requires banks to provide more timely recognition of expected credit losses based on future expectations as opposed to the current ‘incurred loss’ model.”

Right now, banks must disclose the actual credit losses they’ve suffered from bad loans. Although they’ve always had to provide some sort of outlook about risks to credit quality going forward, the new standard is much more forward-looking. Under IFRS 9, banks must calculate potential credit losses they can expect based on various risk factors.

That’s where predictive analytics comes into play.

To comply with IRFS 9, banks “have to look at millions of customers with hundreds of data points,” said Darryl Ivan, national lead of risk management at SAS Canada.

“The banks have millions of clients and any of them could have a home, a line of credit, credit cards and so forth,” he told me in an interview.

“Now lay out those individual data points against millions of loans and start applying predictive analytics to say ‘where do we think this loan is going to go under various economic conditions?’”

Since file folders and Post-It notes aren’t up to the task, “IT plays a critical role because you’ll need a very complex infrastructure to support this,” said Ivan. “You want an automated process. The IT people don’t want to kill themselves maintaining and modifying and enhancing a complex system in such a compressed timeframe.”

Although IFRS 9 doesn’t come into effect worldwide until Jan. 1, 2018, most Canadian banks will start including it in their earnings reports on Nov. 1, 2017. That’s because the fiscal reporting year for most banks here ends Oct. 31 instead of Dec. 31. We’ve gotta be different, eh?

As Ivan noted, “globally, Canadian banks will be the first to start reporting under IFRS 9.”

Being the first to dive into uncharted regulatory waters, our financial institutions aren’t taking any chances.

“They want to start the process, ideally, a year ahead to test multiple scenarios before going into full production,” said Ivan.

A year seems prudent, given that banks traditionally have a fair amount of legacy IT baggage. They’ll also have to make sure all that data stays secure. On the integration side, banks that have gone through mergers and acquisitions may “have to cobble together various data sources,” Ivan suggested.

Quite daunting. Yet Ivan said some banks are using IFRS 9 as an opportunity to modernize their accounting and risk management platforms with the latest technology. That includes giving some of their line-of-business managers simple, self-serve access to these predictive data insights.

“Banks are trying to leverage these regulatory initiatives and converge and really support business decisions,” he said.

For complex institutions like banks, doesn’t that sound more forward-looking than any predictive data model they’ll ever put out?

Photo courtesy of Free Digital Photos

Share this article:
Comments are closed.