A Ratio at One Date Can Hide a Year of Borrowing
This series examines hidden leverage in publicly traded nonfinancial corporations: borrowing exposure that can matter economically even when it is missing from the balance-sheet snapshot used to calculate reported leverage. This opening chapter focuses on the basic question behind the evidence: what can a corporation owe, use, or pay for during a reporting interval that the balance sheet does not show at the reporting date? The answer supports a practical judgment, not an accusation. A low leverage ratio may be useful and still incomplete, and the difference matters because investors, boards, lenders, and auditors may not all see the same exposure.
The balance sheet answers a narrower question than it seems
A balance sheet is a picture taken at a particular date. It records assets and liabilities that qualify for recognition and remain outstanding then. A leverage ratio built from that picture—debt divided by assets, for example—answers a defined question: how much recognized debt is present relative to the asset base at this moment? It does not automatically answer how much financing the corporation used during the quarter, what future payments it has committed to, or how much borrowing cost it incurred to keep operations moving.
Consider a simple illustration. A company needs cash for a temporary working-capital shortfall, borrows on Monday, and repays the loan before the quarter closes. On the reporting date, the debt balance can be zero even though the company used borrowed funds during the quarter and paid interest for doing so. The snapshot is not necessarily numerically wrong. It is simply silent about a transaction that no longer exists in the balance-sheet photograph. The income statement may still contain the interest expense, and the timing of cash flows may tell a different story from the period-end liability figure.
This is a stock-versus-activity distinction. The liability balance is a stock: an amount outstanding at a moment. Interest expense and lease payments describe activity or commitments over time. A stock can be a sensible measure for questions about solvency on that date, but it becomes incomplete when a reader uses it as a proxy for every financing exposure the corporation carried during the period. The Federal Reserve study makes that distinction operational by comparing what is reported at the date with signals that accumulate across the interval.
The same mismatch can arise without a loan being repaid. A corporation may control equipment, offices, vehicles, or other productive assets through a lease. If the accounting treatment leaves the related obligation outside recognized debt, the corporation still has a schedule of future payments, but the leverage ratio does not treat that schedule as debt at the date. Economic exposure continues across time even when its accounting presentation is separated from the liabilities that appear beside it.
That distinction is the starting point for reading the Federal Reserve study behind this series. The paper does not claim that every balance sheet is deceptive or that every unrecognized obligation is improper. It asks whether a reporting system can make some corporations look less leveraged than their financing activity would suggest, and what happens when previously separate information is brought into the common view.
A reporting-date number can also matter precisely because people make decisions from it. A lender may assess borrowing capacity, a board may evaluate management, and an investor may compare companies using a ratio that appears standardized across firms. If one corporation has a large payment commitment or repeatedly borrows and repays inside the period while another does not, the same reported leverage can conceal different paths to the same endpoint. The question is therefore not whether the snapshot has value. It is whether the snapshot is sufficient for the decision being made.
Two ways a corporation can fall outside the snapshot
The study concentrates on two forms of hidden leverage in nonfinancial corporations. The first is the treatment of operating leases before the adoption of ASC 842. Under the older distinction, capital leases appeared on both the asset and liability sides of the balance sheet, while operating leases were left off the balance sheet and disclosed in notes. That presentation made operating-lease commitments available to a careful reader without making them part of the headline debt ratio.
There is nothing inherently suspicious about choosing a lease. Leasing can provide flexibility, tax advantages, risk sharing, or a highly collateralized way to obtain the use of an asset. A corporation may prefer those features even if all audiences understand the commitment. The information problem appears when the economic obligation is important to one decision-maker but is easy to overlook for another. A sophisticated lender or rating analyst may reconstruct the present value of future lease payments; a board member, investor, or compensation formula may rely more heavily on recognized leverage. The same contract can therefore be economically ordinary and informationally consequential at the same time.
The second form is intra-period borrowing: short-term debt issued and repaid within a reporting period, before the balance sheet is measured. This is a timing gap rather than a classification gap. The debt can be absent at quarter-end because it has been repaid, while the interest expense remains. The study uses that surviving expense, relative to reported debt, to look for unusually high implied borrowing costs. At the broader market level, the paper also shows that nonfinancial commercial paper outstanding at quarter-end has understated within-quarter amounts by about 10 percent since 2005. That pattern is evidence that a reporting-date number can differ materially from activity during the interval; it is not, by itself, proof that every corporation used the gap to mislead anyone.
The two mechanisms should not be merged into a single vague category. An operating lease creates a continuing payment commitment that was historically disclosed but not placed in recognized debt. Intra-period borrowing creates a temporary liability that can vanish before the reporting date and must be inferred from another statement or from repeated observations. One concerns what the accounting model recognizes; the other concerns when the measurement is taken. Both can lower reported leverage, but they leave different trails and require different tests.
They can also meet in the same corporation. A firm that benefits from a lease's flexibility may also need short-term financing for timing mismatches, or it may find that keeping several exposures outside the headline ratio produces a more favorable reported profile. The paper later tests whether the overlap is greater than chance would predict. At this stage, the important point is more modest: a single ratio may hide more than one mechanism, so finding one missing piece does not establish that the full financing picture has been recovered.
Calling these exposures “hidden” also requires care. In the study, hidden means absent from the balance-sheet measure at the relevant time. It does not mean that the information was impossible to find, that a legal disclosure was necessarily violated, or that management intent has been established. The central issue is an information differential: what is visible in the financial statements, what a reader can reconstruct with effort, and what a decision-maker actually incorporates into a judgment.
The problem is neither universal nor negligible
The paper's evidence suggests a minority pattern with substantial exposure among the firms affected. From 1995 through 2018, 29.2 percent of firms in the annual sample were flagged for substantial operating-lease leverage under the study's criteria. In the quarterly sample covering 1984 through 2025, 12.0 percent were flagged for suspiciously high interest expense relative to reported debt. Over the common 1995–2018 period, 8.2 percent were flagged for both forms, a larger overlap than independence would predict.
Those figures need their denominators and definitions attached. They are not estimates of all corporations, all debt-like commitments, or all forms of off-balance-sheet finance. The operating-lease flag identifies unusually substantial exposure, not every corporation that signs a lease. The interest flag identifies an unusual relationship between expense and reported debt, not every loan that moved in and out within a quarter. Different sample periods and reporting frequencies also mean the percentages should not be compared as if they measured the same population.
At the same time, “minority” should not be confused with “unimportant.” Operating leases were common in the data, but the exposure was concentrated. Between 1999 and 2018, the 75th percentile of firms' off-balance-sheet operating-lease share remained above 50 percent. In other words, for more than a quarter of the firms in that analysis, the estimated operating-lease liabilities exceeded total reported debt. A corporation in that group could look lightly indebted under the recognized balance sheet while carrying a much larger debt-like payment obligation once the lease schedule was considered.
The accounting change later made that concentration visible in aggregate figures. The paper reports that leased property, plant, and equipment on balance sheets rose by roughly $500 billion in 2019, from less than 2 percent of gross property, plant, and equipment to more than 5 percent. Using its own Moody's Baa discounting approach, it estimates $823 billion of operating-lease liabilities in 2018. Those are not identical measures—the asset-side accounting increase and the present-value estimate use different constructions—but together they show why recognition could change the apparent scale of a firm's financing exposure without creating a new lease overnight.
The overlap matters for the same reason. If one corporation is flagged for substantial leases and suspicious interest expense, an analyst who checks only recognized debt may miss two different ways of reaching a low reported ratio. Conversely, a combined flag still does not prove that management designed a coordinated concealment strategy. It tells us that the corporation's reporting profile deserves a closer look across both continuing commitments and within-period activity. The research question is about the information pattern first; motive must be argued separately.
That is why the subject matters beyond the arithmetic of one ratio. Reported leverage can influence a lender's assessment of capacity, an investor's view of risk, a board's response to management, or a covenant and compensation threshold. If those judgments depend on a number that captures only obligations outstanding on one date, two corporations with similar reported leverage may not have similar financing exposure across the interval. Conversely, a higher adjusted number does not automatically show bad faith; it may reveal a reasonable financing choice whose significance was distributed unevenly across audiences.
The useful question, then, is not whether the balance sheet should be discarded. It is what additional evidence is needed before treating its leverage ratio as decision-sufficient. That requires separating the two visibility gaps, tracing what each leaves behind, and asking whether the patterns survive explanations such as ordinary working-capital timing, expensive but observable distress financing, or legitimate lease economics. Once those distinctions are in place, the evidence can address a harder question: when exposure becomes visible to more people, does the corporation merely report differently, or does it change how it invests and borrows?
The next step is to follow each mechanism on its own terms—one through lease classification and continuing payments, the other through borrowing that disappears before the reporting date—before judging how they interact.