Abstract
In this paper, we show that the churning generated by listing and delisting is a key mechanism for disseminating new financial norms regarding cash and debt among listed nonfinancial firms in the U.S. after 1980. Over this period, cash holdings have grown markedly across the distribution of listed U.S. nonfinancial corporations, while indebtedness has fallen in less indebted firms. Surprisingly, these trends are particularly dramatic among the most financially fragile listed firms, which generate insufficient cash flows to service their financial obligations (what Minsky terms ‘Ponzi’ firms). In this paper, we use quantile decompositions to show that these long-term trends in cash and debt are driven by churning among listed firms during the 1980s and 1990s, when listing and delisting rates are high. The reason is that firms list with more cash and less debt, and delist with less cash and more debt, than continuing firms. This mechanism is particularly strong among listed Ponzi firms. Our results highlight the importance of institutional changes surrounding access to equity finance and Initial Public Offerings, rather than changing behavior within continuing firms, for trends in listed firms’ financing behavior since 1980.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Data Availability
The dataset generated during the current study is not publicly available as it contains proprietary information. Information on how to obtain it and reproduce the analysis is available from the corresponding author on request.
Notes
These calculations use Compustat data. We discuss the data and our sample in Section 2.
See Minsky (1986, Ch.9) and Wray (2016) for introductions to the Financial Instability Hypothesis; Taylor and O’Connell (1985), Lavoie (1986), Gatti and Gallegati (1990), Keen (1995), Skott (1995), Ryoo (2010, 2016) and Pedrosa and Lang (2021) for examples of papers from the large theoretical literature in the Minskian tradition; Davis et al. (2019), Nishi (2018), Pedrosa (2019), and Laborda et al. (2021) for recent empirical work in a Minskian framework; and Nikolaidi and Stockhammer (2017) for a survey of this literature.
In contrast, additive decompositions (e.g. shift share decompositions) rely on the linearity of the weighted mean and, therefore, cannot be extended to decompositions of the median or other quantiles.
See, for example, Palley (1994) for a discussion of Minskian dynamics at business cycle frequency, and Ryoo (2010), Bhattacharya et al. (2015), and Davis et al. (2019) for analyses of Minskian cycles as long waves. Wray (2009), similarly, analyzes the 2008 financial crisis as resulting from a long-term expansion of financial fragility. In Minsky’s own work, see Minsky (1957) for an analysis of the business cycle and Minsk (1964) and Minsky (1995) for work on long waves of institutional change.
See Schlingemann and Stulz (2022) for a discussion of their data adjustments to improve comparability between firm-level and national accounting data.
Nikolaidi and Stockhammer (2017) survey theoretical models of macroeconomic cycles from a Minskian perspective, including those that analyze the interaction between stock prices and aggregate demand.
This cash flow-based definition of financing regimes underlies the definitions of hedge, speculative, and Ponzi finance that we use in this paper. The recent empirical literature studying Minsky’s financial instability hypothesis generally also starts from cash flow-based regime classifications (Nishi 2018; Davis et al. 2019; Pedrosa 2019). Similarly, theoretical work on Minsky has used cash flow-based regime classifications, particularly in models that analyze cash flow-based financing regimes and aggregate leverage (see, for example, Foley 2003; Lima and Meirelles 2007; Di Guilmi and Carvalho 2017). See also Nikolaidi and Stockhammer (2017) for a review of the theoretical literature on Minsky models. Most importantly for our purposes, the stock-flow consistency of this cash flow-based application of Minskian financing regimes ensures that our measures of financing regimes are linked to expected changes in assets and liabilities. However, it is important to note that, if our focus instead lay in analyzing the contribution of each regime to systemic fragility and the incidence or severity of crises, this cash flow-based definition must be supplemented with stock-based measures of liquidity and margins of safety, as well as with information on the size distribution of firms (to capture, for example, that Ponzi firms with large cash reserves pose lower systemic risk than Ponzi firms that are forced to sell illiquid assets, and that small Ponzi firms pose lower systemic risk than large Ponzi firms). For an example of an empirical paper that supplements a cash flow-based approach with margins of safety, see Nishi (2018).
To address the fact that Compustat assigns funds from operations a missing value when any subcomponents are unreported, we follow Davis et al. (2019) and construct this variable as the sum of its components. We take the major components of funds from operations ‘as is’ (income before extraordinary items, depreciation and amortization, and extraordinary items and discontinued operations) and impute zeros for missing values of smaller items (e.g. deferred taxes and equity in net loss). See Appendix Table 4 for details. Like Davis et al. (2019), we also include three categories of cash inflows for which we cannot fully distinguish operational decisions from responses to financial distress (sale of property, plant and equipment, sale of investments, and extraordinary items and discontinued operations). These categories are ambiguous because a company may, for example, sell a subsidiary either as part of its standard operational decisions, or because it is strapped for cash and needs to meet financial obligations. By including these categories, we define an upper bound on cash inflows (and, thus, a lower bound on the share of Ponzi firms). However, Davis et al. (2019) show that, because these cash inflows are small relative to funds from operations, their inclusion does not substantively change the incidence of Ponzi finance among listed U.S. firms.
About 80% of these firms leave Compustat within three years of ceasing to report enough information for a regime assignment. Of these cases, over 97% begin when they stop reporting operational income.
We show in Appendix C that our results are robust to using listing firms as the reference group. In this case, we compute a counterfactual distribution for period t (rather than \(t+1\)) that replaces the observations for delisting firms with those of listing firms, and assess the contribution of continuing firms against the distribution of listing firms.
This path dependence is a well-documented property of decompositions based on counterfactual samples (Fortin et al. 2011). To see this path dependency in our context, consider the following decomposition of the effect of firms that either list as Ponzi or transition into Ponzi from a more robust regime:
$$\begin{aligned} q_{0.5,t+1} - \tilde{q}^C_{0.5,t+1} = (q_{0.5,t+1} - \tilde{q}^{C,T}_{0.5,t+1})+ (\tilde{q}^{C,T}_{0.5,t+1} - \tilde{q}^C_{0.5,t+1}) \end{aligned}$$(3)where \(\tilde{q}^{C,T}_{0.5,t+1}\) is the median of a counterfactual sample that includes continuing (C) firms and firms that transitioned into Ponzi from other regimes (T). The first term on the right-hand side is the contribution of new lists; the second term is the contribution of firms transitioning from other financing regimes. Path dependency arises because each term is a different comparison: the first contribution is assessed against both continuing and transitioning firms, while the second is assessed only against continuing firms. However, a group may contribute positively to the median when assessed against continuing firms, but negatively when assessed against continuing firms and another group. Furthermore, while one could in principle choose a common reference group, the resulting decomposition would not add up to the observed change it is meant to decompose.
For any given observation, the recentered influence function at the median is \(RIF(Y, 0.5)= q_{0.5} + \frac{0.5 - \theta \{ Y \le q_{0.5}\}}{f_y(q_{0.5})}\), where \(f_Y(q_{0.5})\) is the probability density function of Y evaluated at the median, and \(\theta \) is the indicator function. The second term on the right-hand side is the influence function, which yields the effect of an individual observation on the median of Y. We obtain the recentered influence function through the addition of \(q_{0.5}\), such that its expectation is equal to the median. The regressions in Eq. 3 use a non-parametric (kernel) method to estimate \(f_Y(q_{0.5})\) to compute the RIF. For further discussion, see Firpo et al. (2009).
Consider the residual \(e_I=\tilde{q}^C_{0.5,t+1} - \hat{\alpha }_I\). The function \(RIF(Y, 0.5)_{t+1}\), which we compute over the full distribution of Y (including both continuing and incoming firms), has a sample mean equal to the unconditional sample median of Y. However, its sample mean conditional on \(I^L_{t+1}=0\) and \(I^T_{t+1}=0\), which is equal to the fitted intercept \(\hat{\alpha }_I\), differs from the sample median of Y conditional on \(I^L_{t+1}=0\) and \(I^T_{t+1}=0\). Thus, \(\hat{\alpha }_I \ne \tilde{q}^C_{0.5, t+1}\) and \(e_I \ne 0\). Analogous reasoning applies to \(e_O\).
These subperiods roughly correspond to decades. The discussion is not sensitive to the specific time cutoffs that we use to present the results.
Appendix A shows decompositions for hedge and speculative firms, which indicate that our primary conclusion about the relative importance of compositional versus within-firm effects also holds for less fragile firms. Appendix B, which extends the analysis to a large set of quantiles, shows that the decomposition results also hold away from the median. Appendix C uses listing firms as the reference group. Finally, Appendix D shows the results are robust to the restricted definitions of listing and delisting discussed in Section 2.1.2.
Since we compute the average annual change in the ratio for each pair of adjacent years, it generally differs from the total change between the first and the last year divided by the number of years in the period.
Table 3 also shows that decomposition residuals are negative, but generally small relative to the independent effects of both listing and transition and, therefore, do not dominate/change these qualitative results.
The average duration of Ponzi finance for a listed firm that then transitions into a speculative regime is 1.9 years, and the average duration of Ponzi finance for a listed firm that then delists is 2.8 years.
Table 5 in the Appendix decomposes changes in median interest payments as a share of sales for Ponzi firms, and corroborates these patterns: as continuing Ponzi firms become more indebted over time, they also face growing interest burdens, whereas churning within the group of Ponzi firms reduces the median interest burden.
Davis et al. (2019) document that a high share of Ponzi firms are Ponzi due to negative sources of cash net of operational expenses – i.e. before even considering their financial obligations. In our sample, 84.1% of listed Ponzi firms have negative sources of cash. While a large share of Ponzi firms with negative sources of cash do subsequently transition to a more robust regime, this share is more than ten percentage points lower (71.8%) than in the full group of Ponzi firms.
There is a positive association between being in a more fragile financing regime and the likelihood of delisting. Davis et al. (2019) show, for instance, that being in a Ponzi regime enhances a firm’s likelihood of delisting in the subsequent year, whereas being in a more robust regime reduces the likelihood of delisting.
The fact that each cohort of newly listed firms over this period has persistently higher idiosyncratic risk (Brown and Kapadia 2007) and lower profitability (Davis and de Souza 2022a) also highlights that young firms in recent decades look substantively different than the young firms of the early 1980s, and further distinguishes the processes discussed here from standard life cycle effects, wherein firms list with fragile structures and become more robust over time as their income grows.
Because of a small number of Ponzi firms in 1970-1974, delisting rates in that group are noisy prior to 1975. We do not observe delisting in 2017 (the last year of our sample).
As noted in Section 2.1, voluntary delists, in which listed firms are taken private, are – in contrast – rare. Doidge et al. (2017) distinguish three reasons for delisting: due to mergers and acquisitions, for cause (due to bankruptcy or a failure to meet the listing requirements of an exchange), and delists that are voluntary. They show that an average of 8.22% of listed firms delist each year between 1975 and 2012, where 56.4% of these delists are due to mergers and acquisitions, 40.8% are for cause, and only 2.7% are voluntary.
Doidge et al. (2017) show that 73% of the increased delist rate after 1997 reflects mergers and acquisitions, whereas for cause and voluntary delists account for 17% and 18% of this increase respectively.
While not reported, decompositions over the full set of periods analyzed above indicate that the qualitative results regarding the relative importance of compositional versus within-firm effects again hold for a wide range of quantiles, and especially in the lower half of the distributions.
For example, from 1993-2003 the median debt-to-asset ratio averages 23.9% while the 20\(^{\text {th}}\) quantile averages only 2.2%. This difference magnifies the relative impact of arguably similar absolute changes in these quantiles, which average \(-\)0.13 percentage points for the median and \(-\)0.25 percentage points for the 20\(^{\text {th}}\) quantile.
References
Alti A (2006) How persistent is the impact of market timing on capital structure? The Journal of Finance 61(4):1681–1710
Arikan AM, Stulz RM (2016) Corporate acquisitions, diversification, and the firm’s life cycle. The Journal of Finance 71(1):139–194
Asker J, Farre-Mensa J, Ljungqvist A (2015) Corporate investment and stock market listing: A puzzle? The Review of Financial Studies 28(2):342–390
Bates TW, Kahle, KM, Stulz RM (2009). Why do U.S. firms hold so much more cash than they used to? The Journal of Finance 64(5):1985–2021
Bhattacharya S, Goodhart C, Tsomocos D, Vardoulakis A (2015) A reconsideration of Minsky’s Financial Instability Hypothesis. Journal of Money, Credit and Banking 47(5):931–973
Brown G, Kapadia N (2007) Firm-specific risk and equity market development. Journal of Financial Economics 84(2):358–388
Campbell JY, Lettau M, Malkiel BG, Xu Y (2001) Have individual stocks become more volatile? An empirical exploration of idiosyncratic risk. The Journal of Finance 56(1):1–43
Comin D, Philippon T (2005) The rise in firm-level volatility: Causes and consequences. NBER Macroeconomics Annual 20:167–201
Davis L, de Souza J (2022a). Stylized facts on the evolution of profit rates in the U.S.: Evidence from firm-level data. University of Massachusetts Boston Working Paper 2022-01
Davis L, de Souza JPA (2022b). Churning and profitability in the U.S. corporate sector. Metroeconomica 73(3):924–957
Davis LE (2016). Identifying the “financialization” of the nonfinancial corporation in the U.S. economy: A decomposition of firm-level balance sheets. Journal of Post Keynesian Economics 39(1):115–141
Davis LE (2018) Financialization and the non-financial corporation: An investigation of firm-level investment behavior in the united states. Metroeconomica 69(1):270–307
Davis LE, de Souza JPA, Hernandez G (2019). An empirical analysis of Minsky regimes in the U.S. economy. Cambridge Journal of Economics 43(3):541–583
Decker RA, Haltiwanger J, Jarmin RS, Miranda J (2016a) Declining business dynamism: What we know and the way forward. American Economic Review 106(5):203–07
Decker RA, Haltiwanger J, Jarmin RS, Miranda J (2016b) Where has all the skewness gone? The decline in high-growth (young) firms in the U.S. European Economic Review 86:4–23
Di Guilmi C, Carvalho L (2017) The dynamics of leverage in a demand-driven model with heterogeneous firms. Journal of Economic Behavior & Organization 140:70–90
DiNardo J, Fortin NM, Lemieux T (1996) Labor market institutions and the distribution of wages, 1973–1992: A semiparametric approach. Econometrica 64(5):1001–1044
Doidge C, Karolyi GA, Stulz RM (2017). The U.S. listing gap. Journal of Financial Economics 123(3):464–487
Dudley E, James CM (2013). Capital-structure changes around IPOs. SSRN Working Paper
Fama EF, French KR (2004) New lists: Fundamentals and survival rates. Journal of Financial Economics 73(2):229–269
Firpo S, Fortin NM, Lemieux T (2009) Unconditional quantile regressions. Econometrica 77(3):953–973
Foley CF, Hartzell JC, Titman S, Twite G (2007) Why do firms hold so much cash? A tax-based explanation. Journal of Financial Economics 86(3):579–607
Foley DK (2003) Financial fragility in developing economies. In: Dutt A, Ros J (eds) Development Economics and Structuralist Macroeconomics. Edward Elgar
Fortin N, Lemieux T, Firpo S (2011) Decomposition methods in economics. In: Ashenfelter O, Card D (eds) Handbook of Labor Economics, volume 4 of Handbook of Labor Economics chapter 1. Elsevier, pp 1–102
Gatti DD, Gallegati M (1990) Financial instability, income distribution, and the stock market. Journal of Post Keynesian Economics 12(3):356–374
Gutierrez G, Philippon T (2016). Investment-less growth: An empirical investigation. Working Paper 22897, National Bureau of Economic Research
Hirtenstein A (2021). Companies are hoarding record cash amid delta fears. Wall Street Journal
Kahle KM, Stulz RM (2017). Is the U.S. public corporation in trouble? Journal of Economic Perspectives 31(3):67–88
Keen S (1995) Finance and economic breakdown: Modeling Minsky’s “financial instability hypothesis’’. Journal of Post Keynesian Economics 17(4):607–635
Laborda J, Salas V, Suarez C (2021) Financial constraints on R &D projects and Minsky moments: Containing the credit cycle. Journal of Evolutionary Economics 31:1089–1111
Lavoie M (1986) Systemic financial fragility: A simplified view. Journal of Post Keynesian Economics 9(2):258–266
Lima GT, Meirelles A (2007) Macrodynamics of debt regimes, financial instability and growth. Cambridge Journal of Economics 31(3):563–580
Minsky HP (1957) Monetary systems and accelerator models. The American Economic Review 47(6):860–883
Minsk HP (1964) Longer waves in financial relations: Financial factors in the more severe depressions. The American Economic Review 54(3):324–335
Minsky HP (1986) Stabilizing an Unstable Economy. McGraw Hill
Minsky HP (1992). The financial instability hypothesis. The Jerome Levy Economics Institute Working Paper, (74)
Minsky HP (1995) Longer waves in financial relations: Financial factors in the more severe depressions II. Journal of Economic Issues 29(1):83–96
Nikolaidi M, Stockhammer E (2017) Minsky models: A structured survey. Journal of Economic Surveys 31(5):1304–1331
Nishi H (2018) An empirical contribution to Minsky’s financial fragility: Evidence from non-financial sectors in Japan. Cambridge Journal of Economics 43(3):585–622
Opler T, Pinkowitz L, Stulz R, Williamson R (1999) The determinants and implications of corporate cash holdings. Journal of Financial Economics 52(1):3–46
Orhangazi Ö (2008) Financialisation and capital accumulation in the non-financial corporate sector: A theoretical and empirical investigation on the US economy: 1973–2003. Cambridge Journal of Economics 32(6):863–886
Palley TI (1994) Debt, aggregate demand, and the business cycle: An analysis in the spirit of Kaldor and Minsky. Journal of Post Keynesian Economics 16(3):371–390
Pedrosa Í (2019) Firms’ leverage ratio and the Financial Instability Hypothesis: An empirical investigation for the US economy (1970–2014). Cambridge Journal of Economics 43(6):1499–1523
Pedrosa I, Lang D (2021). To what extent does aggregate leverage determine financial fragility? New insights from an agent-based stock-flow consistent model. Journal of Evolutionary Economics 31
Ryoo S (2010) Long waves and short cycles in a model of endogenous financial fragility. Journal of Economic Behavior & Organization 74(3):163–186
Ryoo S (2016). Household debt and housing bubbles: A Minskian approach to boom-bust cycles. Journal of Evolutionary Economics 1–36
Schlingemann FP, Stulz RM (2022) Have exchange-listed firms become less important for the economy? Journal of Financial Economics 143(2):927–958
Skott P (1995) Financial innovation, deregulation and Minsky cycles. Studies in Investment, Saving and Finance, Macroeconomic Policy After the Conservative Era, p 255
Stockhammer E (2004) Financialisation and the slowdown of accumulation. Cambridge Journal of Economics 28(5):719–741
Taylor L, O’Connell SA (1985) A Minsky crisis. The Quarterly Journal of Economics 100:871–885
Wray LR (2009) The rise and fall of money manager capitalism: A Minskian approach. Cambridge Journal of Economics 33(4):807–828
Wray R (2016). Why Minsky matters: An introduction to the work of a maverick economist. Princeton University Press
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflicts of interest
The authors report that there are no competing interests to declare that are relevant to the content of this article, and that no funding was received to assist with the preparation of this manuscript.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendices
Appendix A: Additional tables
Appendix B: Extending to a large set of quantiles
The results in Sections 4.1 and 4.2 show that, for the median firm, changes in the composition of firms produce rising cash-to-sales and declining debt-to-asset ratios, despite offsetting within-firm changes over time. To show the robustness of this main result across the distributions of these variables, Figs. 7 and 8 generalize the analysis in Sections 4.1 and 4.2 to a wide set of quantiles, by plotting average annual changes in the 20\(^{\text {th}}\) quantile to the 90\(^{\text {th}}\) quantile in increments of ten. To keep the analysis manageable, we focus on the aggregate decomposition that divides changes over time into within-firm and composition effects, and the 1980s and 1990s (specifically, the 1982-1992 and 1993-2003 periods used above).Footnote 26 To facilitate comparisons across quantiles, we divide the average annual changes by the average levels of the corresponding quantiles, and report the results in percentages. For example, the 40\(^{\text {th}}\) quantile of the cash-to-sales ratio for all firms was 2% during 1982-1992, implying that the average annual increase in the cash-to-sales ratio at the 40\(^{\text {th}}\) quantile was 2% of its average level from 1982-1992.
All firms: Decompositions across quantiles, as % of average levels per period
Ponzi Firms: Decompositions across quantiles, as % of average levels in each period
First, Fig. 7 shows that – consistent with the trends in Section 2 – the cash-to-sales ratio increases almost uniformly across the distribution, while the debt-to-asset ratio declines until the 60\(^{\text {th}}\) percentile. In addition, for the debt-to-asset ratio, the lowest quantiles display the steepest relative declines. This pattern reflects that the distribution of debt is skewed, and, therefore, magnifies the effects of any absolute changes in these quantiles when assessed relative to their corresponding levels.Footnote 27 In turn, these decompositions analysis confirms our main findings: during 1982-1993, the within-firm effect works to lower all quantiles of the cash-to-sales ratio and raise all quantiles of the debt-to-asset ratio, while the composition effect works in the opposite direction. The same pattern holds during 1993-2003, with the exception that the within-firm effect also makes a slight positive contribution to raising the cash-to-sales ratio until the 40\(^{\text {th}}\) quantile.
Second, Fig. 8 shows that the patterns among Ponzi firms are even more pronounced: the relative declines in the lowest quantiles of the debt-to-asset ratio are between two and four times larger than those in the full sample, as are the contributions of the within-firm and composition effects. Likewise, the negative within-firm contribution to the change in the cash-to-sales ratio is about twice as large as in the full sample, without any sign reversals in either period.
Appendix C: Changing the reference group to incoming firms
Appendix D: Decomposition results based on alternative sample definitions
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Davis, L.E., de Souza, J.P.A. & Hernandez, G. Listing, delisting, and financial norms: a quantile decomposition of firm balance sheets. J Evol Econ 33, 1259–1302 (2023). https://doi.org/10.1007/s00191-023-00815-9
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00191-023-00815-9