The Two Visibility Gaps
For nonfinancial corporations, hidden leverage is not one accounting trick. This chapter traces the two mechanisms studied by the Federal Reserve paper: pre-2019 operating leases, whose payment obligations were kept outside recognized debt, and intra-period borrowing, whose liability could disappear before the reporting date. The goal is to understand what crosses the boundary between a corporation's economic financing and the number shown on its balance sheet—and why the two mechanisms must be analyzed separately before their connection can be judged.
Operating leases: an obligation separated from the debt line
A lease gives a corporation the right to use an asset in exchange for payments over time. Economically, that arrangement can support the same operating capacity that a purchased asset financed with borrowing would support. But under the historical accounting distinction, not all leases entered the balance sheet in the same way. Capital leases appeared on the asset and liability sides. Operating leases were excluded from the balance sheet and disclosed in accompanying notes.
That difference changed the starting point for a leverage calculation. A capital lease increased recognized assets and liabilities, so a reader looking at reported debt could see it. An operating lease still required future payments, but those payments were not part of the debt line used in the headline ratio. The corporation had not escaped the obligation; the obligation had been placed in a different layer of the financial statements.
The distinction was not arbitrary in its economic consequences. Leasing can be a sensible alternative to buying an asset. It may provide flexibility when a corporation does not want to own equipment for its full useful life. It can have tax advantages, share some asset risk with the lessor, or function as a highly collateralized form of financing because the lessor can recover the asset if the lessee defaults. A corporation choosing an operating lease is therefore not automatically trying to disguise leverage. The instrument can be useful even when the accounting presentation is completely understood.
The visibility problem arises because the economic usefulness of the lease and the salience of its liability are different questions. The notes disclosed the schedule of operating-lease payments, allowing a determined analyst to estimate what the commitment was worth today. A lender or credit-rating agency with the data, expertise, and incentive to make that adjustment might treat the lease much like debt. Another reader might compare recognized debt across companies and never convert the note schedule into a comparable liability. A board or compensation formula could also rely on the reported ratio rather than on an adjusted measure.
The note disclosure therefore acts as a second interface between the corporation and its audiences. It contains information, but it does not automatically put that information into every decision. A reader must notice that the commitment is material, determine how its payments are distributed over time, and decide how to compare those payments with other debt. Each step requires attention that a headline ratio does not demand. The older accounting presentation did not make reconstruction impossible; it made reconstruction optional for the person using the statements.
This creates an uneven information environment around the same corporation. One audience sees a company with a manageable debt-to-assets ratio. Another sees the ratio plus a stream of contractual payments that limits future cash available for other uses. Neither audience is necessarily looking at fabricated data; they are answering different questions with different parts of the statement package. The first question is “what liabilities are recognized today?” The second is closer to “what financing commitments will compete for cash over the life of the asset?”
The older classification also created room for firms to structure contracts so that they remained operating leases rather than capital leases. That possibility matters because an accounting boundary can affect reported leverage even when two arrangements have similar economic functions. It still does not establish why any particular corporation selected its classification. The paper combines this institutional possibility with firm-level patterns and later changes in behavior; it does not infer intent simply from the presence of an operating lease.
The eventual move to on-balance-sheet treatment under ASC 842 changed the common view of these commitments. The important point for the mechanism is not yet the later result, but the location of the gap: before recognition, the obligation could be found in notes without appearing in recognized debt. The balance sheet was therefore not a complete map of lease-related financing, even though the corporation's future payments remained real and the information was not necessarily unavailable.
Intra-period borrowing: a liability that leaves before the photograph
Intra-period borrowing creates a different kind of gap. The corporation borrows on a short-term basis during the reporting interval and repays the debt before the balance sheet is prepared. When the reporting date arrives, the liability is no longer outstanding. The period-end balance sheet can consequently show little or no trace of the principal even though borrowed cash helped the corporation operate during the period and interest was paid for using it.
The sequence matters. At the beginning of a quarter, a company may have a financing need. It draws on commercial paper, a credit line, a short-term note, or another temporary source. The cash supports payroll, inventory, a lumpy payment, or some other timing mismatch. The company then repays the borrowing before quarter-end. A reader who sees only the endpoint observes the repayment, not the financing episode. A reader who also sees the income statement may notice interest expense that seems high relative to the debt reported at the preceding or ending date.
This is not the same as a company carrying a high-cost loan continuously. A continuously outstanding loan raises the debt balance that appears in the snapshot, even if its interest rate is high. With intra-period borrowing, the principal can be missing precisely when the balance sheet is measured, so the comparison between interest expense and reported debt can become unusually large. The paper's diagnostic begins from that mismatch: interest expense is a record of financing cost over the interval, while reported debt is a record of principal at a boundary.
The paper uses a broad market pattern to make the timing issue visible. In its Federal Reserve commercial-paper series, nonfinancial commercial paper outstanding at quarter-end has understated the amount outstanding within the quarter by approximately 10 percent since 2005. That is a meaningful difference between a stock measured at one date and activity observed over the interval. It does not prove that every company in the series was deliberately manipulating its reporting profile. The same pattern could be examined alongside other evidence about supply, demand, and disclosure rules, and the study argues that several alternative supply-side explanations do not fit all of the observed timing and cross-sectional facts.
The alternatives are important because quarter-end behavior can have more than one cause. The paper notes that the pattern emerged before the Great Financial Crisis and related Dodd–Frank effects that might have constrained dealer balance sheets. It also reports that the pattern does not appear in asset-backed commercial paper, cannot explain the firm-level variation in intra-period borrowing, and does not line up with the shifts around disclosure-rule changes. These comparisons do not turn the aggregate chart into a direct observation of hidden borrowing, but they make a purely supply-side explanation less sufficient for the broader set of facts.
There are also legitimate reasons for a corporation to borrow and repay within a quarter. Short-term finance can match cash availability to expenses that arrive unevenly. Commercial paper or a credit line can bridge a temporary need without creating a year-round liability. If the borrowing is fully understood by lenders and other decision-makers, the timing may be ordinary treasury management rather than an attempt to fool anyone.
The issue becomes more pointed when the reporting date is treated as a summary of the whole period. A corporation that repeatedly borrows and repays can show a low end-of-quarter debt balance while paying interest that reflects more activity than that balance suggests. The paper's question is whether unusually high interest relative to reported debt identifies a pattern consistent with this behavior. That is an inference, not a direct record of every borrowing episode. High interest can also reflect genuinely expensive financing, financial distress, or other forms of implicit financing such as overdue accounts payable. Those possibilities have to be addressed rather than silently folded into the definition of hidden borrowing.
The two gaps therefore differ in both state and evidence. An operating lease, under the earlier presentation, was a continuing commitment whose payment schedule could remain in the notes for a careful reader. Intra-period borrowing was a temporary liability that could be gone before the reporting date, leaving interest expense as a more indirect trace. The first gap concerns recognition: which debt-like obligation is placed on the balance sheet? The second concerns timing: when is the liability observed? Treating them as identical would obscure the reason each requires a different diagnostic.
They also create different kinds of persistence. The operating-lease commitment can continue after the reporting date, so its future payments remain relevant even if no new borrowing occurs. The intra-period loan may be gone, but the cost of using it remains in the period's results. One trail points forward through contractual obligations; the other points backward through financing expense. A serious assessment of leverage may need both directions: future cash claims and evidence of what happened during the period.
Yet the mechanisms can be connected at the corporation level. Both affect the leverage profile that a reader sees at a particular moment, and both can be more consequential when the corporation relies on short-term funding or when the audience does not reconstruct the missing exposure. The paper finds that substantial operating-lease exposure and suspicious interest expense overlap more than independence would predict. This does not show that every firm planned a coordinated two-part strategy. It does show that the financing and reporting environment can support more than one visibility gap in the same firm, so identifying one form of exposure does not guarantee that the full picture has been recovered.
That connection also explains why the paper keeps the mechanisms separate while studying them together. If operating leases and intra-period borrowing were simply interchangeable labels for the same liability, a change in lease presentation would tell us little about temporary borrowing. If they were completely unrelated, their overlap and joint response to later disclosure would need a different explanation. The study instead asks whether a corporation's broader information position matters: when one form of hidden leverage becomes visible, does the value of maintaining another form change as well?
A final boundary is essential. “Off-balance-sheet” does not mean “outside the financial statements,” and “hidden” does not mean “illegal.” Operating-lease commitments were disclosed in notes. Intra-period borrowing can leave an interest expense footprint. What is hidden is the exposure from the headline balance-sheet view at the moment many decisions are made. The practical consequence is an information differential, not a conclusion about misconduct.
The next question is how to turn those different traces into comparable evidence. A lease-payment schedule must be translated into a debt-like amount. Interest expense must be compared with a risk-adjusted expectation rather than treated as suspicious merely because it is high. Only then can the paper ask how common the patterns are and which conclusions the data can support.