Resilience and asset pricing in COVID-19 disaster

Daadmehr, Elham Article Resilience and asset pricing in COVID-19 disaster Economies Provided in Cooperation with: MDPI – Multidisciplinary Digital Publishing Institute, Basel Suggested Citation: Daadmehr, Elham (2025) : Resilience and asset pricing in COVID-19 disaster, Economies, ISSN 2227-7099, MDPI, Basel, Vol. 13, Iss. 5, pp. 1-35, https://doi.org/10.3390/economies13050123 This Version is available at: https://hdl.handle.net/10419/329403 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Academic Editor: Gheorghe H. Popescu Received: 21 March 2025 Revised: 22 April 2025 Accepted: 23 April 2025 Published: 1 May 2025 Citation: Daadmehr, E. (2025). Resilience and Asset Pricing in COVID-19 Disaster. Economies,13(5), 123. https://doi.org/10.3390/ economies13050123 Copyright: © 2025 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/ licenses/by/4.0/). Article Resilience and Asset Pricing in COVID-19 Disaster Elham Daadmehr Department of Economics and Management “Marco Fanno”, University of Padua, 35123 Padova, Italy; [email protected] Abstract: The COVID-19 pandemic potentially affected stock prices in two non-mutually exclusive ways: discount rates and cash flows. This paper focuses on the latter and analyzes it through the lens of an asset-pricing model. It shows how workplace resilience and financial resilience interacted and significantly affected asset prices. The model-based equity premium increases with the probability of a disaster. The results suggest the significant amplification of workplace resilience by financial resilience. Specifically, the dividend growth of low-resilience firms is significantly more responsive to workplace flexibility and suffers more severely than that of high-resilience firms. Keywords: financial resilience; workplace resilience; dynamic functional principal components; Markov switching; COVID-19 disaster; equity premium 1. Introduction COVID-19 has profoundly affected the economy and induced tremendous uncertainty in the financial markets. Governments adopted different types of social distancing policies to control the spread, especially in the first wave and the fever period of COVID (February to April 2020). These social distancing rules and lockdowns effectively influenced the working environment and firms’ performance (Koren & Pet˝o,2020, among others). The fastgrowing literature asserts that firms with fewer labor constraints in the lockdown-restricted situation featured better performance (Bretscher et al.,2020), as firms with more flexibility in their workforce are expected to be less financially vulnerable in such situations, since they are less likely to face additional costs due to lockdowns and social distancing rules. Koren and Pet˝o (2020) propose different dimensions of the firms’ workplace flexibility that played an important role in their cost of production, as well as the fluctuations in asset prices in response to the COVID-19 shock (Pagano et al.,2023). In theory, COVID-19 can affect stock prices through two non-mutually independent channels: discount rates and cash flows. Pagano et al. (2023) focus mainly on the impact of the increase in perceived risk on expected excess returns (first channel). Back to the story of COVID, industries saw massive business disruption due to social distancing and lockdowns as a consequence of the pandemic, which affected the cost of production and especially the output of firms with less flexibility in their workplace. In such a turbulent market, conservative investors primarily focus on the price of an equity claim to the output of such firms as risky assets—that is, the expected future cash flows. Daadmehr (2024) shows that risks related to the working environment of a company, including the impact of the communication mode, teamwork, and physical presence, to which a company is exposed, can create heterogeneity in expected cash flows. According to Stulz (2025), managing such risks as consequences of the COVID pandemic crisis can potentially affect the value of firms. This paper concentrates on the second channel and, in a novel work, quantifies expected cash flows not just to fill the gap in the risk management literature, but to show Economies 2025,13, 123 https://doi.org/10.3390/economies13050123 Economies 2025,13, 123 2 of 35 how the impact of COVID-19 and corporate resilience can carry over from cash flows to expected returns. The characterization of resilience heterogeneity in expected cash flow sheds light on how this paper bridges a gap and links the real part of the economy, where the exogenous COVID-19 shock originated, to the financial market. This paper analyzes the asset-pricing implications of the COVID-19 crisis, including its impact on the economy and firms’ production costs, in the context of a model with (i) a fictitious representative investor with Epstein–Zin–Weil preferences, who may prefer an early resolution of uncertainty in disasters 1 ; and (ii) an exogenous dividend stream sensitive to the consequences of the COVID-19 disaster and to its contractionary effects on the real part of the economy. This study considers the cross-sectional time-varying impact of COVID-19 on the dividend stream as the interaction of two components: the crosssectional firm-level impact of workplace resilience and the time-varying impact of aggregate economic contraction, as a control for the macro time effect of COVID-19. Specifically, it shows that the dividend growth of low-workplace-resilience firms is significantly more sensitive to workplace resilience and suffers more severely than that of high-workplaceresilience firms. However, the impact of corporate financials was quite complex during the COVID-19 outbreak. Firms started raising capital just because of cash flow dry-up fears or to strengthen their ability to overcome difficulties during the first wave. Meanwhile, the provided credit, especially in small firms, could affect the capital structure and increase leverage. Consequently, the financial characteristics of firms, such as capital structure and liquidity, also played an important role in the performance of firms and were considerably influenced by a wide range of policies adopted in response to COVID-19, from corporate policies to public policies, including bank-loan guarantees and additional mitigation packages, for example, the Paycheck Protection Program, PPP loans (Fahlenbrach et al.,2020; Pagano & Zechner,2022). This suggests that these characteristics may have also contributed to amplifying the degree of corporate resilience and, as a result, the response of asset prices (Daadmehr,2024). Some recent articles have shed light on the impact of corporate financial amplification on asset prices during COVID-19 period (Daadmehr,2024;Ramelli & Wagner,2020) and have found significant heterogeneity in resilience in expected returns by introducing a new “composite financial resilience” index that contains both workplace flexibility and “financial-based resilience” (Daadmehr,2024). The novelty of this paper lies in the fact that it provides a simple example as evidence of the amplification effect and shows that crosssectional and time-varying corporate financials can potentially amplify the overall effect of exogenous consequences of COVID-19, consistent with the evidence that better-financed firms before and during COVID-19 can better overcome the side effects of lockdown and the associated mitigation of COVID restrictions (Daadmehr,2024;Fahlenbrach et al.,2020). Then, it quantifies “financial resilience” to capture the footprints of the impact of a wide range of different policies on the financial status of firms. This paper proposes a new asset-pricing model with COVID-19 disaster embedding workplace resilience and financial resilience to investigate and track the impact of firms’ characteristics on asset-pricing implications. The novelty of this paper is directly related to how it quantifies the consequences of the exogenous COVID-19 crisis on the dividend stream that affects asset prices depending on the resilience of these two types of firms. In other words, it clarifies how resilience practically contributes to asset pricing and characterizes the resilience-heterogeneous equity premium by providing tractable formulas, showing that it increases with the probability of a disaster. In line with Daadmehr (2024), this article shows that the effect of financial resilience of firms significantly amplifies the impact of workplace resilience and aggregate economic Economies 2025,13, 123 3 of 35 contraction due to COVID-19. The novel proposed an exogenous dividend stream, and its estimation provides an opportunity to compare the impact of firms’ financial resilience and workplace resilience. Meanwhile, estimated dividend growth highlights that the heterogeneous effect of workplace resilience is dominant, although the impact of firms’ financial resilience is statistically significant, demonstrating the need for characterization of “financial resilience”. The novel application of Dynamic Functional Principal Component Analysis (DFPCA) enables us to distinguish not only the main time-varying elements of financial resilience but also those that create significant cross-sectional variation. Finally, this paper empirically proves that valuation, liquidity, and solvency ratios play key roles in firms’ financial resilience and the corresponding amplification of workplace resilience. The results of this part shed light on possible corporate policies. The paper is structured as follows. Section 2clarifies how this paper contributes to the literature and postpones the introduction of resilience intuitions to Section 3. The model of the economy and the assumptions about the exogenous dividend stream at the firm level are presented in Section 4. The solution of the model appears in Section 5, where the closed form of the resilience-heterogeneous equity premium is presented. The results on both the effect of the macroeconomic contraction of COVID-19 and the estimated dividend stream are included in Section 6. The paper proposes the main components of financial resilience in Section 7and then concludes. 2. Contribution to Literature As the main novelty of this paper is to investigate resilience heterogeneity in expected cash flows and its role in asset pricing, this paper contributes to two main strands of literature: (i) corporate resilience and the impact of COVID-19 on the cross-section of stock returns, and (ii) asset pricing models with rare events. On the one hand, this article attempts to discuss the impact of workplace resilience and financial resilience as an overall indicator of firms’ financial status. Since the emergence of COVID-19, many studies have started to explain the flexibility of firms or industries in such a pandemic crisis, specifically in response to health mitigation policies, social distancing rules, and lockdowns (among all Dingel & Neiman,2020;Hensvik et al.,2020;Koren & Pet˝o,2020). This paper contributes to these studies more from the perspective of asset pricing than from corporate resilience and considers workplace resilience in the spirit of Koren and Pet˝o (2020). As a key feature, this paper uses the data on workplace resilience and explains how and to what extent it significantly contributes to the asset-pricing model during the COVID pandemic. Meanwhile, this paper takes into account financial resilience, as the fast-growing literature has already emphasized the importance of corporate financials on asset prices in response to a wide range of public and corporate policies (Ding et al.,2021;Fahlenbrach et al.,2020;Pagano & Zechner,2022;Ramelli & Wagner,2020). Daadmehr (2024) compares these two intuitions of resilience and provides a measure of corporate resilience, called the composite-financial resilience index, applicable in times of pandemics. This cross-sectional measure allows for the categorization of assets into more risky and less risky groups and enables investors to manage their resources. She provides different evidence of resilience heterogeneity in expected returns and expected future cash flows that can potentially originate from workplace resilience and financial resilience as an additional source of variation. This paper deviates from Daadmehr (2024) and quantifies the cross-sectional “time-varying” financial resilience, although the aim is not to propose an index. Contrary to the major part of these studies in corporate finance, this paper proposes a mechanism to estimate a mixed specification for exogenous dividend stream as an expected future cash flow. Since there is no empirical study on the link between these two types of resilience, Economies 2025,13, 123 4 of 35 nor theoretical work or background, this paper (Section 3.2) starts with simple empirical evidence to motivate the characterization of the exogenous dividend stream as a bridge between these two types of resilience. According to the literature, it is noteworthy to mention and clarify the impact of the possibility of disaster on the macro time effect of the exogenous COVID-19 pandemic as the third source of variation. Among all, Gourio (2012), Gabaix (2012), and Wachter (2013) declare the time-varying probability of disaster that generates covariation in the equity premium. Ghaderi et al. (2022) develop the literature and consider the gradual unfolding disasters. They explain that investors are not aware of the true state of the economy and introduce a Bayesian learning framework showing that updating investors’ beliefs captures the effect of slowly unfolding disasters, as prices truly react to the consumption decline. They show that updating the agent’s belief accords with the true state of the economy. This paper deviates from Ghaderi et al. (2022) by considering disaster states as bad times of the economy. It empirically proves that COVID-19 has a tremendous impact on the economy and captures the impact of disaster and controls the time variation of expected cash flows for macroeconomic sensitivity to COVID-19 using the Markov-Switching approach. As opposed to Wachter and Zhu (2024) who use the jump Poisson process to capture the low and high intensity of disaster that defines disaster states based on two high and low amounts of disaster intensity, and apply a Markov-Switching simulation to investigate the impact of learning in asset pricing of rare disasters, this study considers disaster states as the bad times of the economy and “estimates” the probability of disaster, states, and the duration of regimes based on monthly GDP. From an asset pricing perspective, this paper uses the general framework proposed by Barro (2006) but with a special case of EZ preferences. Contrary to Barro (2006), who proposes economic contraction due to rare events as a “random variable” and calibrates it, this article considers it as a “stochastic process” and provides an estimate for each month, using the Markov-Switching approach. The empirical analysis on this part shows that dividend growth was significantly sensitive to the overall economic contraction due to the COVID-19 phenomenon, with a conditional probability of 2 percent, in line with Barro (2006), who calibrates the disaster probability parameter. Another main difference is directly related to the idea of resilience. “Resilience” in Asset Pricing: In the asset pricing literature, many studies introduce resilience in the rare-disaster framework. Gabaix (2012) considers a deterministic aggregate consumption growth in the absence of disaster; however, consumption growth is magnified by a positive macroeconomic recovery rate when a disaster occurs. He presents an asset-specific dividend process magnified by a positive rate of surviving in a disaster period. The definition of timevarying “resilience” in his paper is an increasing function of the asset-specific survival rate. In the framework he proposed, resilience is a linearity-generating process that sees shock, uncorrelated with disaster occurrence. Since the definition of resilience highly depends on the type of disaster, this article, in contrast to Gabaix (2012), considers cross-sectional workplace resilience due to the natural feature of the COVID-19 pandemic and its effect on the workforce and firm costs through social distancing rules and lockdowns (similar to Daadmehr,2024;Pagano et al.,2023). This paper deviates from Pagano et al. (2023) in two aspects: First, it uses data on workplace resilience (in the spirit of Koren & Pet˝o,2020) rather than considering it as a parameter in the model. Moreover, this study considers workplace resilience, affecting the dividend stream cross-sectionally, and quantifies the cross-sectional timevarying impact of corporate financials as an additional part called financial resilience. Second, it provides monthly estimates for disaster probability rather than theoretical Economies 2025,13, 123 5 of 35 intuition for the probability of disaster as a parameter in the model. It should also be noted that this paper provides an empirical test as a prerequisite to having a particular definition of disaster state and the corresponding probabilities based on the intrinsic characteristic of COVID-19 and the Poisson distribution, as theoretically proposed by Daadmehr (2025). The striking difference is directly related to the role of the estimated Markov switching process as a control for the macro time effect of COVID-19, a necessary part as Barro (2006) declares. This paper demonstrates the dominant heterogeneous effect of workplace resilience, showing that the dividend growth for low-resilience firms is more responsive to workplace resilience than that of high-resilience firms. Meanwhile, the impact of cross-sectional timevarying financial resilience is not negligible and significantly amplifies the impact of COVID19 on firms’ production growth and dividend stream. Section 3briefly introduces these two intuitions of resilience and provides some pieces of initial evidence of amplification, which sheds light on the characterization of the dividend stream as expected cash flows. 3. Resilience This paper investigates how resilience affects asset-price fluctuations in the COVID-19 pandemic. At first glance, it seems a little bit tricky to clarify “resilience”. Due to the pathological features of the COVID-19 pandemic and its severe impact on the labor force and the workplace, which leads to huge business disruptions, this article considers the resilience of the workplace as the capacity to absorb the disturbance in the COVID-19 outbreak. 3.1. Workplace Resilience After the emergence of COVID-19, many studies started to interpret to what extent firms’ performance depends on communication restrictions and social distancing rules (among all Dingel & Neiman,2020;Hensvik et al.,2020;Koren & Pet˝o,2020). Many of them tried to propose a measure of workplace flexibility. Koren and Pet˝o (2020) provide a theorybased measure for the dependency of US businesses on human interaction, based on three dimensions of occupation: teamwork intensive, customer facing, and physical presence. Their model of communication reveals the sensitivity of production costs to an increase in face-to-face interaction and determines firms with less efficient performance from home. They explain the impact of face-to-face communication on costs of production, introduce the average ‘affected share’, and interpret that a higher firm’s affected share implies less flexibility towards social distancing restrictions during the COVID-19 pandemic. The important feature of workplace resilience is that the resilience of firms depends on their own workplace characteristics and flexibility towards the new social distancing rules and lockdown policies, which is not implied by the workplace resilience of other firms. Despite all the prominent features of this resilience measure, Daadmehr (2024) shows the shortcoming of this type of resilience to exhibit “significant” resilience heterogeneity in the firm’s implied discount rate as the proxy for expected return. 3.2. Is Workplace Resilience Adequate Enough? Although the necessity and adequacy of workplace resilience are fully investigated by Daadmehr (2024) using several pieces of empirical evidence, Figure 1shows the evolution of analysts’ expectation of future cash flows for highand low-resilience 2 firms in the spirit of Koren and Pet˝o (2020) in the first panel. They propose a proxy called “affected share” to show to what extent businesses rely on human interaction. From the analysts’ point of view, low-resilience firms experienced lower expected cash flows 3 compared to high-resilience firms. The first panel reveals that aggregate expectations better reflect the earnings expectation of low-resilience firms, especially before and during the fever period Economies 2025,13, 123 6 of 35 of COVID-19, and suggests that workplace resilience can potentially be an important source of heterogeneity in firms’ expected cash flows. Figure 1. The evolution of expected future cash flows in the fever period of COVID-19 (the impact of workplace resilience and leverage): The first panel shows the standardized earnings expectation (ExtEPSi,2020 −EPSi,2019)/EPSi,2019 of highand low-resilience firms, in the sense of workplace flexibility, for current fiscal year of 2020. ExtEPSi,2020 stands for an earnings expectation of firm i at time t (similar to Daadmehr,2024;Koren & Pet˝o,2020;Landier & Thesmar,2020). Firms with an ‘affected share’ less than 40 are assigned to the high-resilience group, and ones with greater than 65 are assigned to the low-resilience one. The second panel shows the standardized earnings expectations of firms with different levels of leverage. Firms with higher leverage than the 80th percentile are assumed high-levered (Q5), and firms with lower than the 20th percentile are the low-levered ones (Q1). Data source: Compustat/CRSP merged, WRDS for fundamentals, and Refinitiv-Eikon (Thomson Reuters) I/B/E/S forecasts for daily consensus analysts’ earnings. Although this type of resilience is in accordance with the type of pandemic crisis and shows to what extent firms can survive when their productivity is affected by human loss (Koren & Pet˝o,2020), the financial status of firms can provide a type of flexibility for firms to handle additional production costs. In addition to all the evidence provided by Daadmehr (2024), the second panel of Figure 1shows the analysts’ expectations of future cash flows separately for highand low-levered firms and provides evidence of the importance of capital structure and firms’ leverage on the evolution of expected earnings. This panel exhibits that not only did earnings expectations decline more for high-levered firms, but also this decline for high-levered firms was persistent and associated with higher oscillations in the following fiscal years. This is consistent with much of the previous evidence that firms with less strong balance sheets experienced greater difficulties during Economies 2025,13, 123 7 of 35 and after the fever period of COVID-19, such as Pettenuzzo et al. (2023), which show how leverage and cash holdings are related to firm performance, especially those with less profitability and lower revenue growth. Each panel of this figure emphasizes that workplace resilience and firms’ corporate financials can separately explain heterogeneity in expected cash flows. Meanwhile, firms with low workplace resilience would be more capable of handling and managing production costs if they already have a suitable financial position. They see less reduction in the average earnings expectations (Daadmehr,2024); However, analysts were quite pessimistic about the rebound in earnings of these firms with lower financial status, as Daadmehr (2024) explains. So, it is crucial to find a mechanism in which these two intuitions jointly affect the expected cash flows and the price fluctuations of an equity claim to the output of such firms. Figure 2motivates and suggests the existence of such interaction. It shows the average earnings expectations for four categories of firms: “high workplace-resilience and high-levered” firms, “high workplace-resilience and low-levered” firms, “low workplaceresilience and high-levered” firms, and “low workplace-resilience and low-levered” firms. The evidence emphasizes two prominent impacts of firms’ financial status: (i) Among lowlevered firms, those with a more flexible workforce not only have less reduction in average earnings expectations but also see less severe fluctuations in the following months after the onset (first panel). (ii) A higher leverage appears to weaken the benefit of high workplace resilience. The second panel, compared to the first one, suggests that high leverage reduces the earnings expectation surplus of high workplace resilience and makes fluctuations more severe. These two pieces of evidence highlight that firms’ financial characteristics can potentially magnify the impact of their workplace resilience on expected cash flows. 3.3. Financial Resilience Moreover, back to the story of the impact of a wide range of policies, any financial ratios from any category, including, e.g., profitability and solvency ratios, can be effective on all these associations. This re-motivates to quantify the overall impact of all corporate financials. This paper suggests a machine learning definition for financial resilience based on Dynamic Functional Principal Component Analysis (DFPCA) as a solution for such quantification (Section 4.2.2). The following sections show how cross-sectional time-varying dynamic functional PCs contribute to the asset-pricing model and to what extent financial resilience elements interact at the dividend level (Section 4.2) and possibly not only amplify the impact of workplace resilience, but also create significant heterogeneity in dividend growth (Section 6). Figure 2. Cont. Economies 2025,13, 123 8 of 35 Figure 2. The evolution of expected future cash flows in the fever period of COVID-19 (amplification effect): Both panels show the standardized earnings expectation, (ExtEPSi,2020 −EPSi,2019)/EPSi,2019 , for four groups of firms during the fever period. ExtEPSi,2020 stands for the earnings expectation of firm i at time t for the current fiscal year of 2020. Firms with an ‘affected share’ less than 40 are assigned to the high-resilience group, and ones with greater than 65 are assigned to the low-resilience one. Firms with higher leverage than the 80th percentile are assumed to be high-levered, and firms with lower leverage than the 20th percentile are the low-levered ones. The categorization of firms into highand low-resilience firms in the sense of workplace flexibility is based on Koren and Pet˝o (2020) and follows Daadmehr (2024). Data source: Compustat/CRSP merged, WRDS for fundamentals, and Refinitiv-Eikon (Thomson Reuters) I/B/E/S forecasts for daily consensus analysts’ earnings. 4. Model and Data This section introduces the standard asset-pricing framework with the exogenous dividend stream, embedding the impact of COVID-19 on firms’ productivity. It clarifies how the paper quantifies resilience as well as how it controls the macro time effect of economic contraction. The data description is presented at the end of the corresponding subsections. 4.1. The Economy The COVID-19 pandemic caused massive business disruption due to social distancing rules and lockdowns that almost all governments imposed. Specifically, firms in some industries were affected even more because they were not really flexible in their workforce, or they could not run tasks in the hybrid mode, simply because such tasks needed more human interaction or face-to-face communication with a higher physical presence. All of these increased the cost of production and affected the output of such low workplaceresilient firms. This situation created a sort of additional uncertainty in the market, which is quite important from an asset-pricing perspective. For a representative consumer (investor), it is important to know what happened to the price of an equity claim to the output of these firms. Following Mehra and Prescott (1985) and Barro (2006), this paper considers recursive preferences of Epstein and Zin (1989) and Weil (1989) for representative-consumer Lucas’ fruit-tree model of asset pricing with an exogenous stochastic dividend stream 4 . Based on Campbell (1993) with total wealth at the beginning of t+1, Wt+1=Wt−Ct as an intertemporal budget constraint and Mt+1=β∗θ(Ct+1 Ct)−θ ψ as the stochastic discount factor with time discount β∗; in partial equilibrium, the standard Euler equation is5: 1=Et[β∗θ(Ct+1 Ct )−θ ψRi,t+1]. (1) Economies 2025,13, 123 15 of 35 very high probability of a disaster state, there is no interest in the risky asset, so the equity premium increases as compensation to cover the additional risk. The third panel shows that the model-based equity premium is an increasing function of the probability of disaster. Moreover, in line with Barro (2006), the equity premium is a decreasing function of risk aversion γ . In what follows, the paper presents the estimation of exogenous dividend growth and its parameters used in the calibration exercise. Figure 3. Cont. Economies 2025,13, 123 16 of 35 Figure 3. Price-to-dividend ratio, risk-free rate, and equity premium: This figure plots the log P/D ratio, log risk-free rate logRf it , and equity premium logEtRit −logRf it as a function of probability of disaster. The log P/D ratio, log risk-free rate, and equity premium are computed based on estimated parameters α , δ , β1,..., βM (Section 6) and calibration exercise for γ , ψ , and σε . Data source: Compustat/CRSP merged and financial ratios, WRDS. 6. Results This section presents an estimated exogenous dividend stream for model-based assetpricing implications. The first part (Section 6.1) clarifies that COVID-19 is a disaster and provides the estimation of the macro time effect of COVID-19 ( ηs t ) from 2013 to 2022, the period over which the exogenous dividend stream is estimated. The second step (Section 6.2) quantifies the impact of financial resilience components. It also estimates dividend growth (Equation (8)) as well as the fixed-effects α and βm s, which are coefficients of macroeconomic contraction, ˆ ηs t , and financial resilience components, PCit,m , respectively, and δ as the heterogeneous effect of workplace resilience, ln(φi) , using the Restricted Maximum Likelihood method, REML, over 2013–2022. 6.1. Macroeconomic Sensitivity to COVID-19 Disaster In the proposed approach, the first step is to estimate the macroeconomic contraction due to COVID-19 to control for the aggregate time effect, as explained in Section 4.2.3. Figure 4shows the fitted two-regime Markov-switching (MS) model for the monthly GDP of the United States from 1960 to 2022 and specifies the disaster regimes. This figure provides an opportunity to empirically prove that this pandemic was a disaster with significant macroeconomic consequences and exhibits the COVID-19 pandemic period as a disaster regime. Furthermore, according to Table 1, the estimation for controlling the macro time effect of COVID-19, ˆ ηs t, can be obtained from: ˆ ηt=   100.69 +0.97ˆ ηt−1 100.69 +1.03ˆ ηt−1 1−pt:Non −Disaster.state pt:Disaster.state with the estimated transition probabilities in Table 2. The significant switching AR(1) coefficients in Table are a sign of severe economic contraction in disaster states, specifically, Economies 2025,13, 123 17 of 35 the estimated coefficient in disaster states (1.03) shows that such a macro time effect is not mean-reverting in disasters. The LRT statistic provided in Table empirically proves the significance of nonlinear two-regime MS-AR(1). The evidence on optimal choice of the number of regimes is presented in Tables A1 and A2 in Appendix A. Figure 4. Monthly GDP, fitted the two-regime MS-AR(1) and the one-step prediction from 1960 to 2022: The blue columns show the disaster regimes (states and the duration). Data source: normalized seasonally adjusted GDP, Federal Reserve Bank of St. Louis, Economic Research Division. Table 1. Estimated parameters of MS-AR(1): This table presents the maximum likelihood estimation of the two-regime Markov-switching AR(1) and the conditional probability of disaster states. It provides the Likelihood Ratio Test to examine linear vs. nonlinear two-regime MS-AR(1). Significant codes: 0 ‘***’, 0.001 ‘**’. Nonlinear Markov Switching Coefficients (StDev) t-Value Intercept: ρ0100.69 (0.013) 592.00 *** AR-1: ρ(Disaster state) 1.03 (0.01) 65.40 *** AR-1: ρ(Non-Disaster state) 0.97 (0.003) 172.00 *** p(Disaster|Disaster) 0.91 (0.02) 63.20 *** p(Disaster|Non-Disaster) 0.02 (0.005) 4.51 *** log-likelihood statistics 431.80 LRT statistics 1102.7 ** Table 2. Estimated transition matrix: This table shows the conditional probability of disaster states estimated by two-regime MS-AR(1). Transition Probability Disaster State at Time tNon-Disaster State at Time t Disaster state at time t+ 1 0.91 0.02 Non-Disaster state at time t+ 1 0.08 0.97 Economies 2025,13, 123 18 of 35 Figure 5shows the evolution of the probability of a disaster state, pt . As can be clearly seen and in line with Figure 1, the probability of a disaster state increased to around 0.9 in the fever period of COVID-19, followed by a reduction due to the impact of good news about vaccines. Although the probability of a disaster state at time t being conditional on the non-disaster state for the previous month is 2 percent, which is in line with the calibrated static disaster probability of 1.7 percent proposed by Barro (2006), the economy will remain in the disaster regime due to the low transition probability of 8 percent. Table 2 shows that switching from disaster states to non-disaster ones happens with a probability of 0.08. Figure 5. Evolution of estimated probability of disaster state based on MS-AR(1), from 1960 to 2022:The blue columns show the disaster regimes (states and the duration). In addition to Figure 4, which graphically shows the goodness of fit and the appropriateness of the estimated economic contraction, ˆ ηs t , Table A3 (in the Appendix A) verifies these results and contains not only the estimated disaster states st and the duration of the regimes, but also the evidence from the Fed reports. It compares the estimated disaster regimes with the corresponding actual events. The estimation of disaster regimes accords with the historical information in (Burger,1969;Hoxworth et al.,1983;Supel,1978). Moreover, Figure A1 (in the Appendix A) shows the estimated distribution for macroeconomic sensitivity, ηs t , and compares the bimodal distribution with the corresponding normal distribution. It provides another form of verification of the number of regime switches. 6.2. Justification for Dividend Stream and Asset-Pricing Moments To interpret the impact of corporate financials and to investigate whether and to what extent the financial status of firms amplifies the consequences of COVID-19 on asset prices, this paper starts with around 70 financial ratios of 5833 US firms over 2013–2022 at monthly frequency and employs Dynamic Functional Principal Component Analysis (DFPCA) to capture the impact of firm’s financial status, as explained in Section 4.2.2. By computing the filter sequences and dynamic functional principal components, it is possible to provide the scree plot and decide on the number of components required to include most of the variation originating from all corporate financials that possibly affected dividend growth. Figure 6shows the portion of variance explained by each dynamic functional PC for all firms, separately, in one diagram. It suggests that the first five components explain the most variation (over 90 percent) induced by financial ratios for almost all firms. Based on Equation (8) and the first five dynamic principal components (PCs) 16 , the results of the estimated dividend growth are summarized in Table 3. This table presents the quantified effect of workplace resilience and the impact of the firm’s financial resilience as the elasticity of dividend growth to these two intuitions of resilience. Economies 2025,13, 123 19 of 35 Figure 6. A scree plot of the Dynamic Functional Principal Component Analysis (DFPCA) of financial ratios: This figure shows the portion of variance explained by each component (eigenvalues). Each colored line is the scree plot of one firm, separate from others. The sample contains around 5833 US firms. Data source: firm-level financial ratios, WRDS. Table 3. Dividend growth estimation: This table provides estimation for coefficients of financial resilience components and the impact of COVID-19, including the macro time effect of COVID (ln ˆ ηs t) and workplace resilience (Equation (8)). It presents the fixed-effects (β1 ,..., β5) of the first five cross-sectional time-varying dynamic functional PCs (PC1 ,..., PC5) and the heterogeneous effect (δ) of workplace resilience ln(φi) , by REML estimating method. The industry sector codes from “2” to “6” belong to “Mining, Utility and Construction”, “Manufacturing”, “Trade, Transportation and Warehousing”, “Information, Finance, Management, and Remediation Services”, “Educational, Health Care and Social Assistance”, respectively. Each column shows the estimated result for each industry separately. The numbers in parentheses are standard deviations of corresponding estimated coefficients. Significant codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05. Dependent Variable: Dividend Growth All Industries Industry Sector (NAICS Code) 2 3 4 5 6 PC10.0007 ** (0.0002) 0.0027 (0.0014) 0.0041 *** (0.0003) −0.0026 ** (0.0008) −0.0029 *** (0.0005) 0.0113 *** (0.0025) PC2−0.0029 * (0.0004) 0.0066 ** (0.0021) −0.0045 *** (0.0005) 0.0052 *** (0.0011) −0.0022 *** (0.0006) −0.0082 * (0.0032) PC3−0.0025 * (0.0005) 0.0010 (0.0028) −0.0013 (0.0007) 0.0025 (0.0015) −0.0050 *** (0.0008) 0.0132 ** (0.0042) PC4−0.000007 *** (0.0022) −0.0092 * (0.0037) −0.0014 (0.0008) −0.0029 (0.0020) 0.0030 ** (0.0011) 0.0026 (0.0058) PC50.00001 (0.0008) −0.0060 (0.0045) 0.0002 (0.0011) 0.0055 * (0.0025) −0.00007 (0.0013) 0.0310 *** (0.0069) ln ˆ ηs t −2.3149 *** (0.0889) −0.7308 (0.4882) −1.8041 *** (0.1191) −2.4705 *** (0.2832) −3.2298 *** (0.1399) −3.2044 ** (0.0085) Average of workplace resilience (heterogeneous effect) 10.3690 *** (0.4622) 2.6394 *** (0.5561) 8.0623 *** (0.4327) 11.2225 *** (0.5193) 14.5646 *** (0.3182) 14.0970 *** (0.6553) F-statistics 127.48 *** 4.013 * 79.816 *** 19.303 *** 105.565 *** 11.4737 *** Economies 2025,13, 123 20 of 35 6.2.1. Interpretation of Workplace Resilience Impact It can be clearly seen that the workplace resilience has a significant positive average heterogeneous effect of 10.36 on dividend growth for a sample of all industries. For individual industries, the corresponding coefficient of workplace resilience in the specification of dividend growth varies on average from 2 to 14, respectively, in “Mining, Utility and Construction” and “Information, Finance, Management, and Remediation Services”. The key result on the average heterogeneous effect of workplace resilience can be seen in Figure 7 . Based on workplace resilience, firms are categorized into two, three, and four groups. In each case, the first and last groups are considered lowand high-resilience firms, respectively. This figure indicates that the average heterogeneous effect of workplace resilience for low-resilience firms is greater than the one for highresilience firms (the red line is below the blue line in Figure 7), meaning that the elasticity of dividend growth with respect to workplace resilience ˆ δ for firms with a very low degree of workplace resilience is much higher than the one for very high-resilience firms. On the other hand, it can be clearly seen in Figure 7that the greater the difference in the workplace resilience of firms (an increase in the number of groups, equivalently), the greater the difference in the averaged heterogeneous effect or the corresponding elasticity (an increase in vertical distance between the red point and the blue one); as a result, for the same amount of increase in workplace resilience, there is a greater change in dividend growth of low-resilience firms based on Equation (8). Daadmehr (2025) theoretically proves a similar statement for expected returns and shows that an increase in COVID intensity increases the expected return of low-resilience firms much more than that of high-resilience firms. To sum up, Figure 7suggests that in low-resilience firms, a one-percent improvement in workforce flexibility increases dividend growth much more than in the case of highresilience firms, since the average estimated heterogeneous coefficient ˆ δ for low-resilience firms is much higher. Summary statistics and empirical results on the heterogeneous effect of workplace resilience are provided in Table 4. This table provides statistical tests to reveal these differences in heterogeneous effect, ˆ δ , for these two groups of firms. The results in this table implicitly examine the significant differences in the elasticity of dividend growth to workplace resilience between highand low-workplace-flexible firms. This table empirically proves that for any number of groups (K), the heterogeneous effect of the workplace resilience of high workplace-resilient firms is “significantly” different from that of the low workplace-resilient ones. Consequently, there are significant discrepancies in dividend growth of highand low-resilience firms created by the heterogeneous effect of workplace resilience. In other words, this indicates that dividend growth for low-resilience firms is significantly more sensitive to workplace resilience than that of high-resilience firms, technically proving the existence of significant resilience heterogeneity in expected cash flows. Economies 2025,13, 123 21 of 35 Figure 7. The average heterogeneous effect of workplace resilience: This figure exhibits how the difference in workplace resilience of highand low-resilience firms changes the average heterogeneous effect of workplace resilience, by sorting and equally splitting firms into K groups based on their workplace resilience. The first group and the last one are considered firms with low and high workplace resilience, respectively. Table 4. The summary statistics of heterogeneous effect of workplace resilience: This table provides the summary statistics on the heterogeneous effect of workplace resilience for high-resilience and low-resilience firms, including the results of group comparisons. Based on workplace resilience, firms are sorted and split into K groups. The first group and the last one are considered firms with low and high workplace resilience, respectively. It compares the heterogeneous effect of two groups of firms using the nonparametric Wilcoxon test. The heterogeneous effect of high-resilience firms is significantly different from the low-resilience ones (at the level of 0.001, indicated by ***). K Workplace Resilience Minimum 1st Qu. Median Mean 3rd Qu. Maximum Group Comparison Test p-Values 2High 9.88 10.21 10.29 10.33 10.52 10.52 0.00 *** Low 9.35 9.16 10.41 10.38 10.69 10.51 3High 10.05 10.05 10.22 10.19 10.22 10.33 0.00 *** Low 9.35 10.15 10.41 10.36 10.72 11.41 4High 10.05 10.05 10.21 10.17 10.22 10.29 0.00 *** Low 9.35 10.16 10.44 10.34 10.72 11.41 6.2.2. Interpretation of Macro Time Effect of COVID-19 Table 3emphasizes the importance of macroeconomic COVID sensitivity, which implies a significant reduction in dividend growth not only at the level of “All industries”, but also within each industry, except “Mining, utilities and construction”. The estimated coefficient of ln(ˆ ηs t) is statistically significant, showing that all sectors are significantly sensitive to the recession caused by COVID-19, except “Mining, Utility and Construction”. Economies 2025,13, 123 22 of 35 Does the low amount of estimated workplace resilience, ˆ δ , imply such an insignificant effect of macroeconomic contraction due to COVID-19? In this sector, the reason for the lack of statistical significance of ˆ α is related to the low amount of estimated average heterogeneous effect of workplace resilience of 2.6. Firms in this sector have a much lower average heterogeneous effect of workplace resilience compared to its average amount plotted in Figure 7(the red line is below the blue line in Figure 7and the average amount of 2.6 for this industry is much smaller than the average in the case of “all industries”, in Figure 7). Consequently, this suggests that firms in this industry are more workplaceresilient, on average. Hence, in this industry, social distancing restrictions are not as intrusive as they are in other sectors, so it is not surprising to see such an insignificant impact of the macro time effect of COVID-19. 6.2.3. Interpretation of Financial Resilience Table 3also shows the results of elements of financial resilience. The significance of dynamic functional principal components of financial ratios not only suggests the significant effect of firms’ financial status on dividend growth but also proves the significant amplification of workplace resilience by corporate financials 17 , Git =f(FRit)git , in Section 4. This table reveals that financial resilience, especially the first two principal components, which contain most variations originating from financial ratios, directly affects dividend growth and makes its resilience more heterogeneous. These small estimated effects, β1,..., βM , is not catastrophic in such a pandemic crisis but is significant enough. Then, overall resilience heterogeneity is not just from the workforce resilience perspective but also based on what firms financially experienced before and during the COVID-19 outbreak. The next section introduces the major elements with more contribution to firms’ financial resilience. On top of all this, since the averaged heterogeneous effect of workplace resilience is greater than the estimated coefficients of financial resilience elements (PCs), dividend growth is more elastic and responsive to workplace resilience. Equivalently, the role of workplace flexibility is more prominent in explaining the resilience heterogeneity in dividend growth. Moreover, the empirical results in this section show “to what extent” cash flows can be resilience-heterogeneous and the solution of the proposed model in Section 5, sheds light on “how” such significant resilience-heterogeneity, specifically the heterogeneous effect of workplace resilience and the amplification effect by financial resilience, can be transferred to expected returns as well as all asset-pricing implications. The calibrated exercise compares model-based asset-pricing moments with the corresponding values from historical data. The model-based equity premium (5.269) is close to the average equity premium from the data (5.147). The result holds for the risk-free rate (1.137 vs. 1.006 from historical data). The model-based standard deviation of the log risk-free rate (2.531) is in line with the corresponding amount presented by Ghaderi et al. (2022) using historical data from 1950 to 201918. 7. Major Elements of Financial Resilience Section 6explained the significant role of financial resilience of assets in amplification of the dominant heterogeneous effect of workplace resilience on exogenous dividend growth, as well as asset-pricing implications (Section 5). The key application of Dynamic Functional Principal Component Analysis (DFPCA) determined the first five major PCs as components of financial resilience FRit , at the firm level over time, including the COVID-19 era. This section clarifies which financial ratios mostly drive fluctuations in these firms’ financial resilience components. Economies 2025,13, 123 23 of 35 Figure 8shows the weights of the financial ratios for each of PC1 ,..., PC5 . It provides an opportunity to compare the relative importance of financial ratios in determining the firm’s resilience. This figure reveals major ratios with over 90 percent average weight (above the black dash line) in the specification of at least one of the first five principal components, PCit,m=∑k∈Zϕ′ i,mkXi,t−k , for m= 1,..., 5. It determines Shiller’s Cyclically Adjusted P/E Ratio, Price/Operating Earnings (Basic), Price/Operating Earnings (Diluted), P/E (Diluted, Excl. EI), P/E (Diluted, Incl. EI) and Price/Sales (valuation ratios); Cash Conversion Cycle (liquidity ratio); and Interest Coverage Ratio (solvency ratio) as main elements of dynamic functional PCs and financial resilience as well. Figure 8. Overall weights of financial ratios (average of filter sequences, ϕ , for the first five PCs): This figure shows the standardized weights of all financial ratios obtained by DFPCA. The sample contains 5833 firms in all industries. The black dash line is the threshold of 90 percent for weights. Data source: firm-level financial ratios, WRDS. Daadmehr (2024) explains to what extent the valuation and liquidity ratios are significantly correlated with the proposed financial-based resilience index and emphasizes the necessity of workplace flexibility to define a novel “Composite-Financial Resilience Index”. The machine-based (DFPCA) choice of valuation ratios is in line with Glossner et al. (2024), who emphasize the important amplification role of institutional investors in valuation and the severe price decline in COVID-19. Furthermore, having a liquidity ratio as one of the important ratios determined by DFPCA is consistent with Pagano and Zechner (2022), who mention the significant change in liquidity levels of listed US firms from before the emergence of the pandemic to after the onset. The choice of Interest Coverage Ratio is in line with Palomino et al. (2019), who interpret countercyclicality and its negative relationship with economic activity. In what follows, there is an interpretation of the relation between these ratios, workplace resilience, and firms’ vulnerability and riskiness. Economies 2025,13, 123 24 of 35 7.1. Valuation Ratios By definition, valuation ratios are appropriate to measure the relationship between market value and some stream of fundamentals. Figure 9shows the time variation in different types of valuation ratios with higher than 90 percent weight (averaged ϕ in Equation (4)) in elements of firms’ financial resilience ( PC1 ,..., PC5 ) after the onset of the COVID pandemic, diagnosed by Dynamic Functional PCA (DFPCA). The first panel shows that the time trend for almost all of these ratios is the same, especially different types of price-to-earnings ratios that are commonly used as good financial metrics to obtain a better understanding of the overall picture, and are accessible to a wide range of investors. In this paper, DFPCA technically proved its significant role in resilience-heterogeneous dividend growth through the firm’s financial resilience (Section 6.2). This figure compares the descriptive behavior of these valuation ratios, also separately for highand low-resilience firms (the second and third panels), in the sense of workplace resilience. It can be clearly seen from the second panel that these ratios have a more homogeneous trend for high-resilience firms. This homogeneity is less clear in the case of low-resilience firms. To make a better comparison between the valuation ratios of highand low-resilience firms, one of these ratios is selected as a representative. The Dynamic Functional PCA determines “Shillers Cyclically Adjusted P/E Ratio” as an effective main element of either PC1 or PC5 with a weight of more than 90 percent (Figure 8). In particular, DFPCA implicitly mentions the inflated P/E due to low or even negative earnings during economic downturns like COVID-19 and refers to the cyclicality of earnings during these periods. It highlights the importance of cyclically adjusting P/E and selects “Shillers Cyclically Adjusted P/E Ratio” as the most promising valuation ratio among different definitions of P/E ratio. The first panel of Figure 10 shows that the adjusted P/E ratio for low-resilience firms is higher than that of high-resilience firms during the COVID-19 outbreak, except for a short time at almost the end of the fever period in the first wave of this pandemic. The flip point in the fever period is consistent with Pagano et al. (2023). Generally speaking, when a firm has a high P/E ratio, it implies that investors are willing to pay a premium for its stock relative to its current earnings. Although high P/E ratios signal growth expectations, they also introduce risk. Investors should carefully take these risks into account in their investment decisions. Simply, a firm with a high P/E ratio can be seen as risky for several reasons: (i) Uncertainty: The stock price may suffer if the company fails to meet those expectations. (ii) Market Sentiment: Any negative news can lead to a sharp decline in the price of stock with higher expectations. Then, investor sentiment plays an important role. (iii) Volatility: The price of stocks with high P/E ratios reacts more strongly to market events. (iv) Missed Expectations: It is disappointing for investors if the company loses its growth targets, leading to a potential sell-off. Economies 2025,13, 123 31 of 35 Table A5. Financial ratios and categorization:Data source: Financial ratios, WRDS database. Financial Ratio Variable Name Category Capitalization Ratio capital_ratio Capitalization Common Equity/Invested Capital equity_invcap Capitalization Long-term Debt/Invested Capital debt_invcap Capitalization Total Debt/Invested Capital totdebt_invcap Capitalization Asset Turnover at_turn Efficiency Inventory Turnover inv_turn Efficiency Payables Turnover pay_turn Efficiency Receivables Turnover rect_turn Efficiency Sales/Stockholders Equity sale_equity Efficiency Sales/Invested Capital sale_invcap Efficiency Sales/Working Capital sale_nwc Efficiency Inventory/Current Assets invt_act Financial Soundness Receivables/Current Assets rect_act Financial Soundness Free Cash Flow/Operating Cash Flow fcf_ocf Financial Soundness Operating CF/Current Liabilities ocf_lct Financial Soundness Cash Flow/Total Debt cash_debt Financial Soundness Cash Balance/Total Liabilities cash_lt Financial Soundness Cash-Flow Margin cfm Financial Soundness Short-Term Debt/Total Debt short_debt Financial Soundness Profit Before Depreciation/Current Liabilities profit_lct Financial Soundness Current Liabilities/Total Liabilities curr_debt Financial Soundness Total Debt/EBITDA debt_ebitda Financial Soundness Long-term Debt/Book Equity dltt_be Financial Soundness Interest/Average Long-term Debt int_debt Financial Soundness Interest/Average Total Debt int_totdebt Financial Soundness Long-term Debt/Total Liabilities lt_debt Financial Soundness Total Liabilities/Total Tangible Assets lt_ppent Financial Soundness Cash Conversion Cycle (Days) cash_conversion Liquidity Cash Ratio cash_ratio Liquidity Current Ratio curr_ratio Liquidity Quick Ratio (Acid Test) quick_ratio Liquidity Accruals/Average Assets Accrual Other Research and Development/Sales RD_SALE Other Avertising Expenses/Sales adv_sale Other Labor Expenses/Sales staff_sale Other Effective Tax Rate efftax Profitability Gross Profit/Total Assets GProf Profitability After-tax Return on Average Common Equity aftret_eq Profitability After-tax Return on Total Stockholders’ Equity aftret_equity Profitability After-tax Return on Invested Capital aftret_invcapx Profitability Gross Profit Margin gpm Profitability Net Profit Margin npm Profitability Operating Profit Margin After Depreciation opmad Profitability Operating Profit Margin Before Depreciation opmbd Profitability Economies 2025,13, 123 32 of 35 Table A5. Cont. Financial Ratio Variable Name Category Pre-tax Return on Total Earning Assets pretret_earnat Profitability Pre-tax return on Net Operating Assets pretret_noa Profitability Pre-tax Profit Margin ptpm Profitability Return on Assets roa Profitability Return on Capital Employed roce Profitability Return on Equity roe Profitability Total Debt/Equity de_ratio Solvency Total Debt/Total Assets debt_assets Solvency Total Debt/Total Assets debt_at Solvency Total Debt/Capital debt_capital Solvency After-tax Interest Coverage intcov Solvency Interest Coverage Ratio intcov_ratio Solvency Dividend Payout Ratio dpr Valuation Forward P/E to 1-year Growth (PEG) ratio PEG_1yrforward Valuation Forward P/E to Long-term Growth (PEG) ratio PEG_ltgforward Valuation Trailing P/E to Growth (PEG) ratio PEG_trailing Valuation Book/Market bm Valuation Shillers Cyclically Adjusted P/E Ratio capei Valuation Dividend Yield divyield Valuation Enterprise Value Multiple evm Valuation Price/Cash flow pcf Valuation P/E (Diluted, Excl. EI) pe_exi Valuation P/E (Diluted, Incl. EI) pe_inc Valuation Price/Operating Earnings (Basic, Excl. EI) pe_op_basic Valuation Price/Operating Earnings (Diluted, Excl. EI) pe_op_dil Valuation Price/Sales ps Valuation Price/Book ptb Valuation Notes 1 This type of preference could better capture the investors’ preference in an uncertain situation like COVID-19. The results of the calibration exercise support this, although the Euler equation is a general form of power-utility function with a specific version of EZ preferences. 2Term “resilience” without mentioning its type, refers to workplace intuition of resilience. 3 This paper considers analysts’ earnings expectation as a proxy for future cash flows, following Daadmehr (2024) and Landier and Thesmar (2020). 4 Implicitly, it assumes Ct is equal to production (all output is consumed at each time) and the risky asset pays Dt=Ct , which is a claim to aggregate consumption in each period t. 5 Equation (1) is the simplified version of Equation (13) in Campbell (1993), with constant gross simple return on wealth invested from period t to period t + 1, as an additional assumption to be able to solve the model analytically (Appendix A). This can be considered realistic due to two separate pieces of evidence: (1) Ghaderi et al. (2022) show the wealth-to-consumption ratio varies almost not significantly by time-varying beliefs. (2) The wealth-to-consumption ratio varies with interest rates (Lustig et al.,2013), and interest rates did not affect either the market crash or the market rebound in COVID time (Cox et al.,2020). Moreover, the model and calibration exercise are in line with Barro (2006), who fully explains that the EZW framework ends up as simple as the power-utility setting, and it is in accordance with a broader set of asset-pricing facts. 6To compute empirical spectral density, this paper considers Bartlett kernel (e.g., Brockwell and Davis (1991)). 7 The letter, s is used to emphasize that the overall contraction depends on the state of the economy. It is eliminated for ease in the rest of this subsection. Economies 2025,13, 123 33 of 35 8 Statistical tests provided in the Appendix A(Table A1) guarantee the existence of significant regime switching in the COVID-19 outbreak. Table A1 summarizes the results of the Likelihood Ratio Test (LRT) of the linearity of the model. The null hypothesis of linearity is rejected in favor of a nonlinear Markov-switching model with regime shifts. 9To clarify, the estimated economic contraction is the fitted values of MS-AR process, η. 10 Federal Reserve Bank of St. Louis, Economic Research Division. 11 It is important to mention the impact of the generated regressor on asymptotic variance. 12 Table A4 in the appendix guides statistical model selection, especially containing the results of the Hausman test to verify the existence of the heterogeneous effect. 13 The statistical model is Equation (8). 14 Implicitly, this paper assumes asymptotic expected return where the arbitrary period length tends to zero, similar to Barro (2006). 15 In standard literature, EZ parameters, γ and ψ , are interpreted as risk aversion and elasticity of intertemporal substitution, respectively. However, this interpretation may not be strictly satisfied when γ differs from the reciprocal ψ (Garcia et al. (2006) and Hansen et al. (2007)). The Euler equation and the consequent calibration exercise, as expected, highlight that the model is based on a special case of power utility with a bit of parameter relaxation (consistent with Barro (2009)). 16 In cases of interest, the results based on the first ten principal components can be provided. 17 In line with the intuition of financial-based resilience and its role in composite-financial resilience index in Daadmehr (2024). 18 All values are reported in percentage terms. 19 The P/E ratio can present insights into investors’ expectations for a firm’s future growth prospects. A high P/E ratio implies that investors anticipate strong earnings growth in the future, which increases the risk of possible missed expectations. 20 The standard definition of Cash Conversion Cycle = DIO + DSO − DPO. Increasing DPO, decreasing DSO, or decreasing DIO results in quicker conversion. 21 The ideal target ratio may vary by industry. References Baayen, R., Davidson, D., & Bates, D. (2008). Mixed-effects modeling with crossed random effects for subjects and items. Journal of Memory and Language,59(4), 390–412. (Special Issue: Emerging Data Analysis). [CrossRef] Bansal, R., Kiku, D., & Yaron, A. (2012). An Empirical Evaluation of the Long-Run Risks Model for Asset Prices. Critical Finance Review, 1(1), 183–221. [CrossRef] Barro, R. J. (2006). Rare Disasters and Asset Markets in the Twentieth Century. The Quarterly Journal of Economics,121(3), 823–866. [CrossRef] Barro, R. J. (2009). Rare Disasters, Asset Prices, and Welfare Costs. American Economic Review,99(1), 243–264. [CrossRef] Bretscher, L., Hsu, A., Simasek, P., Tamoni, A., & Roussanov, N. (2020). COVID-19 and the Cross-Section of Equity Returns: Impact and Transmission. The Review of Asset Pricing Studies,10(4), 705–741. [CrossRef] Brillinger, D. R. (2001). Time series. Society for Industrial and Applied Mathematics. [CrossRef] Brockwell, P. J., & Davis, R. A. (1991). Time series: Theory and methods. Springer. Burger, A. E. (1969). A historical analysis of the credit crunch of 1966. Review,51, 13–30. Campbell, J. Y. (1993). Intertemporal Asset Pricing without Consumption Data. American Economic Review,83(3), 487–512. Chib, S. (1998). Estimation and comparison of multiple change-point models. Journal of Econometrics,86(2), 221–241. [CrossRef] Cox, J., Greenwald, D. L., & Ludvigson, S. C. (2020). What explains the COVID-19 stock market? (NBER Working Papers No. 27784). National Bureau of Economic Research, Inc. [CrossRef] Daadmehr, E. (2024). Workplace Sustainability or Financial Resilience? Composite-Financial Resilience Index. Risk Management,26, 7. [CrossRef] Daadmehr, E. (2025). COVID-19 Intensity, Resilience, and Expected Returns. Risks,13(3), 60. [CrossRef] Dempster, A. P., Laird, N. M., & Rubin, D. B. (2018). Maximum Likelihood from Incomplete Data Via the EM Algorithm. Journal of the Royal Statistical Society: Series B (Methodological),39(1), 1–22. [CrossRef] Ding, W., Levine, R., Lin, C., & Xie, W. (2021). Corporate immunity to the COVID-19 pandemic. Journal of Financial Economics,141(2), 802–830. [CrossRef] [PubMed] Dingel, J., & Neiman, B. (2020). How many jobs can be done at home? Journal of Public Economics,189(C), 104–235. [CrossRef] Epstein, L. G., & Zin, S. E. (1989). Substitution, Risk Aversion, and the Temporal Behavior of Consumption and Asset Returns: A Theoretical Framework. Econometrica,57(4), 937–969. [CrossRef] Fahlenbrach, R., Rageth, K., & Stulz, R. M. (2020). How Valuable Is Financial Flexibility when Revenue Stops? Evidence from the COVID-19 Crisis. The Review of Financial Studies,34(11), 5474–5521. [CrossRef] Gabaix, X. (2012). Variable Rare Disasters: An Exactly Solved Framework for Ten Puzzles in Macro-Finance. The Quarterly Journal of Economics,127(2), 645–700. [CrossRef] Economies 2025,13, 123 34 of 35 Garcia, R., Renault, É., & Semenov, A. (2006). Disentangling risk aversion and intertemporal substitution through a reference level. Finance Research Letters,3(3), 181–193. [CrossRef] Gelman, A. (2005). Analysis of Variance: Why It Is More Important than Ever. The Annals of Statistics,33(1), 1–31. [CrossRef] Ghaderi, M., Kilic, M., & Seo, S. B. (2022). Learning, slowly unfolding disasters, and asset prices. Journal of Financial Economics,143(1), 527–549. [CrossRef] Glossner, S., Matos, P. P., Ramelli, S., & Wagner, A. F. (2024). Do institutional investors stabilize equity markets in crisis periods? Evidence from COVID-19. (Swiss Finance Institute Research Paper No. 20-56. Forthcoming at Management Science.) European Corporate Governance Institute. Gourio, F. (2012). Disaster Risk and Business Cycles. American Economic Review,102(6), 2734–2766. [CrossRef] Hamilton, J. D. (1990). Analysis of time series subject to changes in regime. Journal of Econometrics,45(1), 39–70. [CrossRef] Hansen, L. P., Heaton, J., Lee, J., & Roussanov, N. (2007). Intertemporal Substitution and Risk Aversion. In J. Heckman & E. Leamer (Eds.), Handbook of Econometrics (Vol. 6, chap. 61). Elsevier. [CrossRef] Henderson, C. R. (1982). Analysis of covariance in the mixed model: Higher-level, nonhomogeneous, and random regressions. Biometrics,38(3), 623–640. [CrossRef] Hensvik, L., Barbanchon, T. L., & Rathelot, R. (2020). Which jobs are done from home? Evidence from the american time use survey. Available online: https://ssrn.com/abstract=3574551 (accessed on 13 April 2020). Houmes, R., & Chira, I. (2015). The effect of ownership structure on the price earnings ratio — returns anomaly. International Review of Financial Analysis,37, 140–147. [CrossRef] Hoxworth, D. H., Miller, G. H., & Mitchell, K. (1983). The U.S. economy and monetary policy in 1983. Economic Review,68, 3–21. Available online: https://ideas.repec.org/a/fip/fedker/y1983idecp3-21nv.68no.10.html (accessed on 11 May 2022). Hörmann, S., Kidzi´nski, Ł. & Hallin, M. (2015). Dynamic functional principal components. Journal of the Royal Statistical Society Series B, 77(2), 319–348. [CrossRef] Jitmaneeroj, B. (2017). The impact of dividend policy on price-earnings ratio. Review of Accounting and Finance,16(1), 125–140. [CrossRef] Jordà, Ò., Singh, S. R., & Taylor, A. M. (2022). Longer-Run Economic Consequences of Pandemics. The Review of Economics and Statistics, 104(1), 166–175. [CrossRef] Koren, M., & Pet˝o, R. (2020). Business disruptions from social distancing. PLoS ONE,15(9), e0239113. [CrossRef] Krolzig, H. (1997). Markov switching vector autoregressions modelling: Statistical inference and application to business cycle analysis. Springer. Landier, A., & Thesmar, D. (2020). Earnings Expectations during the COVID-19 Crisis. The Review of Asset Pricing Studies,10, 598–617. [CrossRef] Lopes, I. A., & Narciso, A. (2020, July 29–31). Does Earnings Management Influence Dividend Policies? An Approach with Unlisted Firms. Xx usp international conference in accounting. Available online: https://congressousp.fipecafi.org/anais/ 20UspInternational/ArtigosDownload/2352.pdf (accessed on 26 April 2025). Lustig, H., Nieuwerburgh, S. V., & Verdelhan, A. (2013). The Wealth-Consumption Ratio. The Review of Asset Pricing Studies,3(1), 38–94. [CrossRef] Mehra, R., & Prescott, E. C. (1985). The equity premium: A puzzle. Journal of Monetary Economics,15(2), 145–161. [CrossRef] Pagano, M., Wagner, C., & Zechner, J. (2023). Disaster resilience and asset prices. Journal of Financial Economics,150, 103712. [CrossRef] Pagano, M., & Zechner, J. (2022). COVID-19 and Corporate Finance. The Review of Corporate Finance Studies,11(4), 849–879. [CrossRef] Palomino, F., Paolillo, S., Perez-Orive, A., & Sanz-Maldonado, G. (2019). The information in interest coverage ratios of the us nonfinancial corporate sector (Economic Research). FEDS Note. Board of Governors of the Federal Reserve System. [CrossRef] Pettenuzzo, D., Sabbatucci, R., & Timmermann, A. (2023). Payout suspensions during the COVID-19 pandemic. Economics Letters,224, 111024. [CrossRef] [PubMed] Ramelli, S., & Wagner, A. F. (2020). Feverish Stock Price Reactions to COVID-19. The Review of Corporate Finance Studies,9(3), 622–655. [CrossRef] Rogerson, W. P. (2008). Intertemporal Cost Allocation and Investment Decisions. Journal of Political Economy,116(5), 931–950. [CrossRef] Staehle, M., & Lampenius, N. (2013). What is driving the price-to-earnings ratio: The effect of conservative accounting and growth. Available online: https://ssrn.com/abstract=2239208 (accessed on 26 April 2025). Stulz, R. M. (2025). Risk, the Limits of Financial Risk Management, and Corporate Resilience. Annual Review of Financial Economics,17. [CrossRef] Supel, T. M. (1978). The U.S. Economy in 1977 and 1978. Quarterly Review,2(1), 1–6. [CrossRef] Wachter, J. A. (2013). Can Time-Varying Risk of Rare Disasters Explain Aggregate Stock Market Volatility? The Journal of Finance,68(3), 987–1035. [CrossRef] Wachter, J. A., & Zhu, Y. (2024). Learning with rare disasters. Available online: https://ssrn.com/abstract=3407397 (accessed on 12 November 2024). Economies 2025,13, 123 35 of 35 Weil, P. (1989). The equity premium puzzle and the risk-free rate puzzle. Journal of Monetary Economics,24(3), 401–421. [CrossRef] Weitzman, M. L. (2005). A unified Bayesian theory of equity ‘puzzles’. Available online: https://conference.nber.org/confer/2005/mes05/ weitzman.pdf (accessed on 26 April 2025). Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.