Commercial and industrial lending in the US has been flat for the past six months, the longest stretch of sluggishness in six years. The static trend is weighing on the year-over-year change: business loans increased by just 2.0% in May, the softest increase since 2011, according to Federal Reserve data.

Nonetheless, recession risk remains low, based on data published through May, and near-term projections suggest that slow-to-moderate growth for the economy will prevail. But the stronger headwind in business lending deserves attention as a possible warning sign for the US economic recovery, which marked its eighth birthday last month – the third-longest expansion on record, based on NBER data.

History suggests that business and commercial loans are a lagging indicator for monitoring recession risk. In the last downturn, for instance, the year-over-year growth in lending remained strong in the first few months of the recession that started in January 2008. Even as late as April that year, the pace of lending was accelerating, picking up to a robust 21.5% increase vs. the year-earlier level – a 35-year year high!

The history business lending in previous decades was similarly slow to signal rising recession risk. The one exception: the 1990-91 downturn, when the growth rate for loans fell sharply in the months leading up to the recession.

Overall, the record for business lending isn’t encouraging if you’re looking for timely recession-risk signals. Is it different this time?

Perhaps, although the jury’s still out. The broad trend for the US economy remains positive, albeit at a moderate pace. Nonetheless, the weakness in business lending is a risk factor that warrants close attention in the months ahead.

Researchers have long documented a pro-cyclical relationship between the supply of bank loans and the business cycle, both in the US and abroad. It’s clear that lending dries up once a recession trend is established. The debate is whether this is primarily due to supply (banks tighten lending standards and offer fewer loans), demand (companies reduce borrowing), or a mix of both.

As an added twist, new sources of lending may be affecting the reliability of the traditional bank-lending data as a business-cycle indicator. As one example, Inc. Magazine recently noted that Amazon

fairly easily entered the lucrative small business lending marketplace. Currently, the company only offers short-term business loans ranging from $1,000 to $750,000 for up to 12 months to micro, small and medium businesses that sell on Amazon. It has the potential to expand into retail banking. Amazon could begin setting up deposit accounts or lending to small business borrowers outside of its own network. This could help Amazon grow its core business of online sales as well as establish the company as one of the leading online lenders to small companies.

It’s unclear if alternative lenders have rendered conventional bank loan trends obsolete as a business-cycle indicator. Even if the established data for commercial and business lending is still the leading source for credit modeling, the data suffers from a sketchy history as a timely indicator of business-cycle risk.

No one indicator is a silver bullet for analyzing economic risk, particularly when that indicator’s history has been uneven at best. But at a time of rising geopolitical risk, weak-to-moderate US economic growth, and a Federal Reserve that’s intent on tightening, if only modestly, monetary policy, the sight of softer loan growth isn’t easily dismissed. All the more so if business lending’s pace slips into negative terrain in the months ahead.

It’s still prudent (and always will be) to use a broad set of indicators to reliably estimate recession risk. But it’s possible that business lending could be an early warning sign this time. For the moment, there’s minimal corroboration that a new downturn has started and so it would be hasty to conclude that the recovery has stalled. Indeed, a majority of key datasets suggest otherwise, at least for now.

The question is whether business lending knows something that the rest of the economy doesn’t?