Banks Offer a Window on the Future of Local Economies

We tend to view booms and busts and national phenomena, but economic conditions can vary wildly from state to state, with implications for business, regulators and individuals. New research suggests publicly available bank date could help to get a better grasp on where local economies are headed.

Print this page
Based on research by Urooj Khan and N. Bugra Ozel

At the end of 2012, as the rest of the country was still hobbling towards recovery from the 2008 financial crisis, North Dakota was flying high. Buoyed by high oil and gas prices, the Rough Rider state posted 19.5 percent annual growth, while the US as a whole managed to eke out a mere 2.5 percent.

In the United States, economic conditions are like weather patterns: They can vary wildly from place to place. When it’s raining money in one state, there can be droughts of investment and employment in others. Accurate local forecasting matters a great deal not only to individuals whose prospects are defined by local economic conditions, but to corporations, policymakers, and candidates for political office as well.

According to Urooj Khan, when it comes to forecasting where state economies are headed, banks can be particularly sensitive barometers. Because they’re able to collect private financial data from the thousands of individuals and companies they loan to, the disclosures commercial banks make about their loan portfolios — and in particular, their provision for loan and lease losses — can significantly improve predictions of economic growth on the state level.

Using aggregated disclosures about loan portfolios between 1990 and 2013, Khan and N. Bugra Ozel, of the University of Texas at Dallas, show a significant negative correlation between banks’ provision for loan and lease losses and the coincident index, a measure of state economic activity produced monthly by the Federal Reserve Bank of Philadelphia (FRB).

“Banks make decisions based on the information they have about their customers,” Khan says. “And based on the decisions they're making, we can infer their expectations of what the future of their local economy will look like.”

These connections make sense: When banks in a given state collectively become more pessimistic about the prospect of people paying them back, it’s a bad sign for the local economy. Khan and Ozel’s data shows that the provision for loan and lease losses is a powerful predictor of the coincident index up to four quarters into the future—which is why the authors argue that it should be incorporated into the FRB’s leading index, the top predictor of state-level economic growth.

Economic forecasters aren’t the only ones interested in bolstering predictions about state-level growth. When corporations make decisions about where to increase investments or build new facilities, they’re liable to favor states that aren’t headed for downturns. “If I’m working for a corporation, and I'm thinking of relocating our facilities, local economic conditions are definitely in play,” Khan explains.

And the same goes for workers. “Similarly, if I were offered an opportunity relocate, and had a spouse who would need to find a job in the same area, would I be willing to move to a place that’s economically going downhill?” Khan says.

There are also political implications. A number of studies have indicated that flagging state economies lead voters to turn against incumbent governors. This contradicts the prevailing belief that voters are primarily concerned with national trends. It seems that all economics, like all politics, is local.

“It can help in terms of your point of view and how you come up with your message in the election — both for the incumbent and other people contesting,” Khan says, explaining how stronger predictions of economic growth might influence state-level elections. It could also affect how national parties disburse funds to local races.

Regardless of whether they’re facing electoral challengers, state policymakers could also use enhanced growth projections to more accurately predict annual tax revenues and plan budgets. And since these disclosures can be used to predict recessions, incorporating the provision for loan and lease losses into the FRB’s leading index could help states figure out how and when to implement countercyclical measures.

“The leading index is one of the timeliest metrics out there,” Ozel says. “Being able to improve the leading index would improve the timing of federal help to local governments in terms of overcoming recessions.” Critically, the provisions are available on a regular and frequent basis, allowing states to act quickly to minimize the fallout from economic downturns. And even during months when the FRB must release its leading index numbers before banks have filed their quarterly call reports, Khan and Ozel find the provision can be useful, since it remains a robust predictor even with a one-month lag.

For the researchers, their results prove the under-tapped value of much publicly available accounting data. “To many people, accounting summarizes what happened in the past,” Khan says. “What we're showing is a certain aspect of accounting that's actually based on future expectations and is forward looking.”

About the researcher

Urooj Khan

articles by Topic