ASC 842 Study Finds More Reported Non-Lease Leverage, Not Proof That Total Debt Fell

A year-end balance sheetbalance sheetA financial statement showing a company's assets, liabilities and owners' equity at a particular reporting date. is a photograph, not a video. Imagine a company drawing on a short-term credit line during the year, repaying it before the reporting date and borrowing again afterward. The closing photograph can show no loan even though the income statement still records interest. Kallen calls this intra-period borrowingintra-period borrowingDebt issued and repaid within a reporting period, potentially leaving interest expense without a loan balance on the period-end balance sheet. [Federal Reserve paper](https://www.federalreserve.gov/econres/ifdp/files/ifdp1441.pdf#page=7), but the study does not observe the individual loans. It infers them from interest expense that appears high relative to reported debt. [Measurement discussion](https://www.federalreserve.gov/econres/ifdp/files/ifdp1441.pdf#page=9)

ASC 842ASC 842A US accounting standard that generally requires a company leasing an asset to recognize a lease asset and liability when the lease lasts more than 12 months. [FASB](https://storage.fasb.org/FIF%20ASU%202016-02%20Leases%20%28Topic%20842%29%20%28Rev%206-3-20%29.pdf) changed what appeared in a different part of that photograph. Under the previous model, operating leases such as a long office lease could remain outside the balance sheet while payment commitments appeared in the notes. ASC 842 generally requires lessees to recognize an asset and liability for leases lasting more than 12 months. For calendar-year public companies, the standard became effective on January 1, 2019. [FASB summary](https://storage.fasb.org/FIF%20ASU%202016-02%20Leases%20%28Topic%20842%29%20%28Rev%206-3-20%29.pdf)

The study uses Compustat financial-statement data. It constructs non-lease leveragenon-lease leverageThe paper's measure of debt after removing recognized lease liabilities, divided by total assets. [Table 10 definition](https://www.federalreserve.gov/econres/ifdp/files/ifdp1441.pdf#page=32) by taking reported debt, removing capital leases and recognized operating-lease liabilities, and dividing the result by total assets. It separately constructs an implied interest rateimplied interest rateFor the annual analysis, interest expense divided by average reported debt in the current and previous fiscal years, multiplied by 100. [Measurement definition](https://www.federalreserve.gov/econres/ifdp/files/ifdp1441.pdf#page=32) from annual interest expense divided by average reported debt in the current and previous fiscal years. [Variable construction](https://www.federalreserve.gov/econres/ifdp/files/ifdp1441.pdf#page=9) [Table 10 definitions](https://www.federalreserve.gov/econres/ifdp/files/ifdp1441.pdf#page=32)

Neither measure is a complete tally of economic debt. The non-lease measure omits the lease liabilities by design and cannot include an intra-period loan that was gone by the reporting date. The implied rate can signal that unreported borrowing may have occurred, but it cannot identify a lender, principal balance, borrowing date or repayment date. [Federal Reserve paper](https://www.federalreserve.gov/econres/ifdp/files/ifdp1441.pdf#page=9)

The study observes balance-sheet outcomes and interest-based clues. It does not observe debt moving dollar-for-dollar from a hidden bucket into a reported one.

The paper compares companies classified as having a large balance-sheet impact from lease recognition with other public nonfinancial companies before and after adoption. In the specification controlling for stable company characteristics and common fiscal-year shocks, the exposed companies' non-lease debt-to-assets ratio rose by an estimated 0.030, or 3.0 percentage points of assets, relative to the comparison groupcomparison groupCompanies used as a benchmark because they did not receive the study's main exposed-company classification.. The regression contains 39,957 firm-year observations. [Table 10](https://www.federalreserve.gov/econres/ifdp/files/ifdp1441.pdf#page=32)

Cumulant's calculated 95% confidence intervalconfidence intervalA range showing the uncertainty around an estimate under the statistical model. runs from 0.452 to 5.548 percentage points. That calculation uses the reported 3.0 percentage-point coefficient and 1.3 percentage-point standard error. The interval describes statistical uncertainty under the model; it does not correct for an invalid research design or measurement error. [Table 10](https://www.federalreserve.gov/econres/ifdp/files/ifdp1441.pdf#page=32)

The positive estimate is not a mechanical addition of lease liabilities to the numeratornumeratorThe number above the line in a fraction, such as debt in a debt-to-assets ratio. because the paper removes recognized lease liabilities from its non-lease debt measure. But it remains a ratio. Debt is the numerator and total assets are the denominator, and ASC 842 also recognizes a right-of-use assetright-of-use assetThe balance-sheet asset representing a lessee's contractual right to use leased property or equipment.. The regression therefore does not tell readers how many dollars of conventional debt were added or how much the asset denominator changed. [Paper's variable definition](https://www.federalreserve.gov/econres/ifdp/files/ifdp1441.pdf#page=32) [FASB recognition rule](https://storage.fasb.org/FIF%20ASU%202016-02%20Leases%20%28Topic%20842%29%20%28Rev%206-3-20%29.pdf)

The paper's second result moves in the opposite direction. Interest expense divided by average reported debt fell by an estimated 2.119 percentage points for exposed companies relative to comparison companies. The firm fixed-effects regression contains 34,472 firm-year observations. Kallen interprets the decline as evidence that exposed companies curtailed intra-period borrowing. [Table 10](https://www.federalreserve.gov/econres/ifdp/files/ifdp1441.pdf#page=32)

The difficulty is in the denominatordenominatorThe number below the line in a fraction, such as assets in a debt-to-assets ratio.. ASC 842 increased reported debt when operating-lease liabilities entered the balance sheet. Even if interest expense did not change, dividing it by a larger reported-debt figure would push the implied rate lower. The paper explicitly identifies this mechanical effect as a concern. [Appendix discussion](https://www.federalreserve.gov/econres/ifdp/files/ifdp1441.pdf#page=47)

That does not make the result useless. It changes what the result can establish on its own. The decline is consistent with reduced intra-period borrowing, but the headline proxyproxyAn indirect measure used when the thing researchers want to study cannot be observed directly. cannot separate a behavioral change from the accounting-driven denominator change without additional evidence.

The paper's appendix tries to escape the denominator problem in several ways. A log excess-interest measure adjusts for observed interest rates, credit spreads and modeled default risk; its post-adoption interaction is negative 0.37 log points. A short-term implied-rate measure divides estimated short-term interest expense by average notes payablenotes payableShort-term written borrowing obligations that include certain loans, overdrafts and commercial paper in the dataset used by the study. [Table A6 note](https://www.federalreserve.gov/econres/ifdp/files/ifdp1441.pdf#page=47) rather than total reported debt; its estimate is negative 6.54 percentage points. [Table A6](https://www.federalreserve.gov/econres/ifdp/files/ifdp1441.pdf#page=48)

The short-term estimate is not precise. Limited reporting of long-term interest expense reduces that regression to 3,036 observations, compared with 35,018 for the appendix's annual implied-rate regression. Its reported standard errorstandard errorA measure of how much a statistical estimate would be expected to vary across repeated samples. is 6.70 percentage points, slightly larger than the estimated effect itself. The paper describes the retained sample as about 10% of the original. [Appendix discussion and Table A6](https://www.federalreserve.gov/econres/ifdp/files/ifdp1441.pdf#page=47)

Two further tests examine notes payable, a short-term debt category that the paper says is not mechanically increased by operating-lease liabilities. After adoption, exposed companies became less likely to report zero notes payable and reported more notes payable relative to assets. Those results fit a story in which companies stopped clearing as much short-term debt before the reporting date. They still do not reveal the proposed within-period loans themselves. [Table A6 and variable note](https://www.federalreserve.gov/econres/ifdp/files/ifdp1441.pdf#page=48)

Before ASC 842, 29.2% of firms were flagged for substantial operating leases over 1995-2018. In the same window, 11.5% were flagged for suspiciously high interest expense, and 8.217% had both flags. Multiplying the first two shares produces an independence benchmark of 3.358%, well below the observed overlap. [Table 2](https://www.federalreserve.gov/econres/ifdp/files/ifdp1441.pdf#page=17)

The clustering makes the paper's explanation plausible: some companies may have combined off-balance-sheet leases with borrowing that disappeared before the reporting-date photograph. But the overlap is descriptive. Company size, industry, financing needs or other shared characteristics could contribute to both flags, and Table 2 does not show that one practice caused the other. [Paper's caution on noncausal analysis](https://www.federalreserve.gov/econres/ifdp/files/ifdp1441.pdf#page=17)

The paper distinguishes companies with a history of substantial operating leases from companies showing a large balance-sheet impact when ASC 842 took effect. In the final regression sample, 47.5% had substantial lease history during 2011-2018, while 17.2% received the large-impact classification. Only 12.3% had both classifications. [Table 7](https://www.federalreserve.gov/econres/ifdp/files/ifdp1441.pdf#page=27)

The main exposure flag uses unusually large adoption-period changes in leased property, plant and equipmentproperty, plant and equipmentLong-lived physical assets such as buildings, machinery and equipment, abbreviated as PPE. or, for companies with high operating leases, unusually large changes in reported leverage. Because that classification is measured from balance-sheet changes occurring at adoption rather than assigned randomly in advance, the design relies on the assumption that comparison companies provide a credible picture of what would otherwise have happened. [Research design](https://www.federalreserve.gov/econres/ifdp/files/ifdp1441.pdf#page=26)

An alternative regression enters lease history and large balance-sheet impact separately. For non-lease leverage, the post-adoption coefficient is 0.03 for the lease-history flag and 0.01 for the large-impact flag; only the lease-history coefficient carries the paper's conventional significance marker. For the implied interest rate, both interactions are negative. This supports the broad direction of the main story while showing that the label exposed combines related but nonidentical concepts. [Table A2](https://www.federalreserve.gov/econres/ifdp/files/ifdp1441.pdf#page=43)

The causal analysis covers 2011-2024, with adoption occurring in 2019 or 2020 depending on a company's fiscal-year timing. The main regressions compare changes for exposed and comparison companies while controlling for fiscal-year effects; the preferred debt-ratio and implied-rate specifications also include firm fixed effectsfirm fixed effectsStatistical controls that compare each company with itself over time, removing company characteristics that remain constant.. [Research design and Table 10](https://www.federalreserve.gov/econres/ifdp/files/ifdp1441.pdf#page=32)

The appendix's full pre-trend tests do not reject equal pre-adoption paths for non-lease leverage or the implied interest rate. However, the more recent pre-trend test for non-lease leverage is marked as significant at the paper's 10% level, a warning that the groups may not have moved perfectly in parallel immediately before adoption. The implied-rate measure passes both reported pre-trend tests. [Table A1](https://www.federalreserve.gov/econres/ifdp/files/ifdp1441.pdf#page=41)

ASC 842 was announced in 2016, before its 2019 or 2020 implementation. The paper's transition-period analysis says most effects began after recognition, but the shift in intra-period borrowing may have started earlier. That timing makes implementation a less perfectly isolated shock than a rule announced and imposed on the same day. [Robustness discussion](https://www.federalreserve.gov/econres/ifdp/files/ifdp1441.pdf#page=33)

The evidence supports a visibility shift more strongly than it supports deleveraging. Exposed companies reported a higher non-lease debt-to-assets ratio after adoption. Their interest-based borrowing proxies declined, and their reporting of short-term notes moved in the direction expected if less debt were being repaid just before the balance-sheet photograph. [Table 10](https://www.federalreserve.gov/econres/ifdp/files/ifdp1441.pdf#page=32) [Table A6](https://www.federalreserve.gov/econres/ifdp/files/ifdp1441.pdf#page=48)

What the evidence does not establish is a fall in total economic debttotal economic debtIn this article, all borrowing obligations that economically burden a company, including borrowing that may not be present on the reporting-date balance sheet.. The study has no complete series combining reported debt, recognized leases and directly observed average intra-period loan balances. It therefore cannot determine whether one dollar of previously hidden borrowing was replaced by one dollar of reported debt, whether companies borrowed more overall, or whether they reduced borrowing in both categories.

For investors, the practical lesson is narrower and more useful than a claim of deleveraging. A cleaner year-end balance sheet does not necessarily mean a company carried less debt through the year. Interest expense, short-term notes, lease obligations and the timing of reporting all matter when evaluating leverage. The new paper provides evidence that ASC 842 changed the photograph. It does not supply the missing video.

ASC 842 made important obligations harder to leave outside the balance sheet. This study suggests other borrowing also became more visible, but it does not prove that the total debt burden declined.