When a company wants to borrow $1 million, creditors typically ask for something they can seize and keep in the event of a default, such as equipment, real estate, inventory, or receivables. However, this does not apply to all businesses: a company that develops software, for example, does not have much to pledge to creditors. Recognizing this, lenders allow businesses to borrow against their operating cash flows.
According to an analysis by Chen Lian of the University of California at Berkeley and Yueran Ma of Chicago Booth, while less traditional, this method of corporate borrowing has actually become dominant. They estimate that 80 percent of US business borrowing by value is based on cash flows from operations, with only 20 percent tied to physical and other separable assets that lenders can seize.
Lian and Ma compiled a database of corporate debt that did not include financial institutions. They gathered aggregate debt data from a variety of sources, then zeroed in on individual company debt using data from S&P Capital IQ, Thomson Reuters’ DealScan, and other databases from 2002 to 2018.
Borrowing patterns varied across industries, according to the researchers. Airlines, for example, borrowed heavily against physical assets, as did small businesses.
Meanwhile, for companies that borrowed against their cash flows, total debt was typically limited not by asset count, but by a multiple of recent operating earnings. “Among large, nonfinancial firms, around 60% have explicit earnings-based covenants in their debt contracts,” the researchers write.
Every dollar increase in operating earnings as measured by EBITDA (earnings before interest, taxes, depreciation, and amortization) was associated with an additional 28 cents in net debt issuance for companies that used cash flow–based credit. This link was strongest at companies with contractually limited borrowing amounts, and it was also strongest when the economy was weak and lenders were more likely to check in on borrowers more frequently.
The researchers also gathered real-estate values, but found only a shaky link between real-estate prices and debt levels. Borrowing increased by only 2–3 cents for every additional dollar value of property at large nonfinancial companies.
“The good news from this research is that we’ve established a good institutional environment in the United States that allows firms to pledge their business value,” says Ma. Even if a company does not have machinery to use as collateral, it can usually borrow money if it can demonstrate that its operations are profitable. In the United States, legal structures such as Chapter 11 bankruptcy laws facilitate cash flow–based lending, and similar institutions can do the same in other countries.
On the other hand, she observes, businesses in countries with legal systems that are more oriented toward liquidating companies in bankruptcy, or that have less robust and reliable accounting and auditing systems, are at a disadvantage.
The study cites pre-2000 Japan as an example, where companies typically offered assets, particularly real estate, as collateral. Japan experienced a real-estate boom and bust in the 1990s, and when prices fell, a lot of corporate debt went bad, limiting companies’ ability to borrow. According to the researchers, the United States experienced a similar boom and bust in the 2000s, but its businesses were not as badly harmed. Because large US corporations did not rely on real estate to borrow, the crash in real-estate values had less of an impact on their debt capacity.
In contrast, US households relied heavily on borrowing against real estate, and their debt contributed to the Great Recession.
The study has implications for economic models, and Ma believes the findings are especially significant in the coronavirus era, when asset-heavy industries such as airlines, hotels, and oil and gas have been hit the hardest. Older macroeconomic models may need to be reconsidered to ensure that policies effectively support an economic recovery.