Cash and investment of Brazilian companies: COVID-19 evidence

The COVID-19 outbreak triggered social isolation measures that generated an exogenous shock to both the supply and demand. Mobility restrictions disrupted production and sales, while rising unemployment and income losses reduced consumption (Guerrieri, Lorenzoni, Straub, & Werning, 2022). Consequently, corporate performance deteriorated. In the United States, corporate profits fell by approximately 11.8% in the second half of 2020, the largest decline since 2008 (Tawiah & O'Connor Keefe, 2022), and evidence from more than 100 countries shows that firm performance worsened as COVID-19 cases increased (Hu & Zhang, 2021).

In this environment, firms sought to preserve liquidity to meet financial obligations and sustain investment, a behavior described as a “dash-for-cash.” However, this increased demand for liquidity coincided with a contraction in credit supply. Operating cash flows weakened, and external financing for investment declined by approximately 37% (Tawiah & O'Connor Keefe, 2022). In the United States, the corporate bond market slowed considerably, forcing even highly rated firms to issue debt at elevated costs, while lower-rated firms relied mainly on existing credit lines (Acharya & Steffen, 2020).

Thus, the pandemic simultaneously reduced firms' ability to generate cash and limited their access to external financing. In this setting, firms with larger cash reserves are better able to sustain investment, highlighting the role of liquidity as a buffer against adverse shocks (Ho, Lam, & Wong, 2023; Joseph, Kneer, & van Horen, 2021; Tawiah & O'Connor Keefe, 2022; Zheng, 2022). Nevertheless, the importance of cash holdings varies across firms depending on their degree of financial constraints. The more limited a firm's access to external financing, the more critical internal liquidity is for investment decisions.

Cash reserves are particularly valuable for financially constrained firms because they allow the financing of projects that might otherwise be delayed or abandoned (Denis & Sibilkov, 2010). These firms accumulate precautionary cash to hedge against illiquidity and cash flow volatility (Almeida, Campello, & Weisbach, 2004; Han & Qiu, 2007). Consistent with this view, Duchin, Ozbas, and Sensoy (2010) show that firms with low cash reserves and high dependence on external finance reduced investment by approximately 6.4% during the 2007–2008 financial crisis.

Accordingly, the relationship between pre-crisis cash holdings and investment is expected to be stronger for financially constrained firms than for unconstrained firms. Prior evidence supports this prediction, showing that firms facing tighter financing conditions systematically maintain higher cash reserves as a precaution (Bates, Kahle, & Stulz, 2009; Opler, Pinkowitz, Stulz, & Williamson, 1999).

Thus, this study examines whether cash reserves accumulated before the pandemic affected the investment behavior of publicly traded Brazilian firms during the COVID-19 crisis and whether this effect was more pronounced among financially constrained firms. Brazil provides a particularly relevant setting for examining the role of financial constraints in shaping corporate investment during the COVID-19 shock. First, the Brazilian financial system is characterized by a relatively high dependence on bank-based financing and limited access to capital markets for a large share of firms, especially those that are financially constrained (Kirch, Procianoy, & Terra, 2014). This reliance on bank credit may increase firms' exposure to credit supply contractions during periods of stress, thereby amplifying financial constraints and increasing internal liquidity.

Second, Brazil has historically experienced recurrent macroeconomic instability, including high inflation, sovereign risk, and credit tightening, which increases firms' exposure to liquidity shocks. These conditions reinforce the precautionary motives for holding cash, particularly among financially constrained firms (Almeida et al., 2004; Opler et al., 1999). Moreover, Brazil's relatively high interest rates raise the cost of external finance and intensify informational friction, increasing firms' reliance on internal funds (Fazzari, Hubbard, Petersen, Blinder, & Poterba, 1988; Stiglitz & Weiss, 1981).

Using a difference-in-differences approach that treats the pandemic as an exogenous shock in a fixed-effects panel model, the results indicate that higher pre-pandemic cash reserves positively affected corporate investment during the crisis, particularly for financially constrained firms.

This analysis is especially relevant given the pandemic's heterogeneous effects across countries. Çolak and Öztekin (2020) argue that reductions in credit supply increase the importance of cash for sustaining investments. In Brazil, the pandemic generated severe health and economic consequences, including a 4.1% contraction in GDP in 2020 (Valor Econômico, 2023). Although GDP grew by 4.6% in 2021 and returned to pre-pandemic levels, cumulative growth between 2020 and 2022 remained below that of 96 out of 191 countries, placing Brazil near the global average (Poder360, 2022).

Given this context, cash reserves played a critical role in sustaining corporate investment. This study relates closely to, Duchin et al. (2010), Tawiah and O'Connor Keefe (2022), and Zheng (2022), which examine how internal liquidity mitigates investment declines during periods of financial stress. In particular, Zheng (2022) investigates the effect of pre-crisis cash holdings on investment during the COVID-19 pandemic using a similar empirical framework.

This study contributes to the literature in two ways. First, it extends this evidence to an emerging market, where financial constraints tend to be more severe and capital markets are less developed. Second, it explicitly examines the differences between financially constrained and unconstrained firms, providing additional evidence on the mechanisms through which cash holdings affect investment.

These findings have important practical implications. By showing that pre-crisis cash reserves support investment, especially among financially constrained firms, the results highlight the importance of prudent liquidity management. For firms, this underscores the need for robust liquidity strategies to withstand future shocks. For investors, it provides evidence for assessing financial resilience and improving capital allocation decisions.

To motivate the hypotheses and empirical strategy, understanding why firms accumulate cash is important. The agency view argues that excess cash may destroy value, whereas the precautionary view emphasizes its importance for financially constrained firms facing costly external financing and liquidity risk.

From an agency perspective, Jensen (1986) excess cash can encourage empire building and delay payouts, creating conflicts between managers and shareholders. Similarly, Stulz (1990) shows that abundant cash may lead to overinvestment, whereas insufficient internal funds can result in underinvestment. Therefore, in normal times, large cash holdings may exacerbate agency costs.

However, during periods of economic distress, the role of cash shifts toward financial flexibility. Financing frictions make internal funds central to investment decisions. Fazzari and Athey (1987) argue that prices and interest rates do not fully adjust under asymmetric information, while Fazzari et al. (1988) show that financially constrained firms exhibit greater investment-cash flow sensitivity because of the wedge between internal and external financing costs. In the Brazilian context, Terra (2003) finds that credit constraints significantly affect investment decisions, with cash flow playing a particularly important role among firms with limited access to external financing.

These frictions may lead to the credit rationing. Stiglitz and Weiss (1981) show that higher interest rates can worsen adverse selection, inducing lenders to restrict credit, even for viable projects. Similarly, Holmström and Tirole (1998) argue that liquidity is underprovided by private markets, leaving firms vulnerable to aggregate liquidity shocks. These frictions are particularly relevant in Brazil. Aldrighi and Bisinha (2010) find that investment remains sensitive to internal funds, even among large publicly traded firms, suggesting that financial constraints extend beyond small firms and reflect broader institutional and informational frictions.

In this environment of informational friction, cash retention emerges as insurance. Almeida et al. (2004) show that financially constrained firms systematically save cash to hedge against liquidity shocks and financing friction. Asset characteristics also matter: Almeida and Campello (2007) find that higher asset tangibility increases investment-cash flow sensitivity among constrained firms by easing contracting frictions. Denis and Sibilkov (2010) further documents that cash is more valuable for constrained firms, as it allows them to finance profitable projects that would otherwise be delayed or abandoned. Complementing this perspective, Kirch et al. (2014) provide evidence consistent with a credit multiplier mechanism, showing that constrained firms' investments respond more strongly to internal liquidity when asset tangibility relaxes financing frictions, which is consistent with Almeida and Campello (2007).

Thus, cash provides financial flexibility to sustain investment when external finance is costly or unavailable. However, in emerging markets, this mechanism is also shaped by macroeconomic conditions and institutional features. Oliveira (2019) shows that financial constraints and unexpected monetary shocks significantly affect investment dynamics in Brazil, while access to long-term credit from development banks mitigates these effects, reinforcing the importance of liquidity in constrained environments.

Crisis episodes reinforce these patterns. During the COVID-19 crisis, Acharya and Steffen (2020) document a widespread “dash for cash,” with BBB and speculative-grade firms drawing on credit lines as bond markets froze. Evidence from the 2007–2008 crisis shows that low-cash firms reduced investment by about 6.4% (Duchin et al., 2010). More recent studies confirm that firms with higher pre-pandemic cash holdings were more resilient during COVID-19 and sustained higher investment levels (Tawiah & O'Connor Keefe, 2022; Zheng, 2022). The evidence from Brazil is consistent with these findings. Franzotti and Valle (2020) show that financial crises have stronger negative effects on investment among financially constrained firms and that internal liquidity becomes more important for sustaining investment during such periods.

Taken together, these results reveal a clear tradeoff. In tranquil periods, excess cash may increase agency costs (Jensen, 1986; Stulz, 1990). In contrast, under asymmetric information, credit rationing, and systemic liquidity shocks, precautionary cash, especially for financially constrained firms, supports investment and creates value (Almeida et al., 2004; Almeida & Campello, 2007; Denis & Sibilkov, 2010). Evidence from Brazil reinforces that financial constraints are pervasive, heterogeneous across firms, and strongly influenced by institutional and macroeconomic conditions (Aldrighi & Bisinha, 2010; Franzotti & Valle, 2020; Kirch et al., 2014; Oliveira, 2019; Terra, 2003). Crisis episodes further highlight that pre-existing liquidity mitigates investment cuts when external financing becomes scarce.

Despite this evidence, an important question remains: Did firms' pre-pandemic cash positions play a decisive role in sustaining investment during the COVID-19 crisis? Prior studies show that firms accumulate cash when they are financially constrained and face the risk of future underinvestment (Almeida et al., 2004) and that low-cash firms reduce investment more sharply during crises (Duchin et al., 2010). Conversely, firms with higher pre-crisis cash holdings exhibit greater investment resilience (Tawiah & O'Connor Keefe, 2022; Zheng, 2022). Moreover, Acharya and Steffen (2020) argue that lockdown measures worldwide created the conditions for a widespread liquidity crisis. Consequently, the COVID-19 pandemic can be viewed as an exogenous shock that increased firms' default risk and adversely affected expected cash flows and refinancing conditions (He & Xiong, 2012).

Given this, this study investigates whether the cash position prior to the pandemic contributed to the maintenance of investments of non-financial companies listed on B3 during a credit supply shock, in this case, the COVID-19 crisis. Although the COVID-19 pandemic affected both supply and demand conditions, there is substantial evidence that it also led to a tightening of the credit supply, particularly for firms with weaker financial positions. In this sense, the pandemic can be interpreted as a shock that increased the relevance of internal liquidity by restricting access to external funds. Importantly, this heterogeneity is central to our identification strategy. If the results were driven purely by differences in investment demand, one would not expect the effect of cash holdings to vary systematically with firms' degree of financial constraints. Therefore, examining how this effect differs across constrained and unconstrained firms allows us to determine whether the underlying mechanism is related to financial frictions rather than demand-side factors.

In this sense, the following hypotheses were elaborated:

This study examines the impact of pre-pandemic cash reserves on corporate investment among publicly traded Brazilian firms during the pandemic. The sample is constructed from the consolidated quarterly financial statements of firms listed on B3, obtained from the Refinitiv® Eikon database. The analysis covers the last quarter of 2018 and all quarters of 2019 and 2020, with 2019 representing the pre-pandemic period and 2020 the pandemic period.

The initial sample includes 719 firms in total. The exclusions comprise: (1) non-Brazilian firms listed on B3; (2) firms without GICS industry classification; and (3) firms classified under GICS Code 40 (Financials), resulting in 298 firms. Firms lacking the data required to compute the variables or with fiscal year-ends other than December 31 were also excluded, which resulted in 287 companies for the final sample.

To test Hypotheses 2a and 2b, firms are classified according to their degree of financial constraints using two standard proxies: the SA index and credit ratings.

The SA index follows Hadlock and Pierce (2010), based on Kaplan and Zingales (1997), and is computed as:

where $Sizei,t$ is the natural logarithm of total assets, and $Agei,t$ is the firm's age. Firms in the highest (lowest) three deciles of the SA index distribution are classified as financially constrained (unconstrained), consistent with Almeida et al. (2004).

The SA index, proposed by Hadlock and Pierce (2010), is used as a proxy for the financial constraints. This measure is based on firm size and age, which are widely recognized as key determinants of financing friction. Although originally developed using U.S. data, the SA index has been widely applied in different institutional contexts, including emerging markets (e.g. Machokoto, 2021; Mansilla-Fernández & Milgram-Baleix, 2023). In the Brazilian context, Balan and Norden (2024) provide evidence supporting its validity. Therefore, consistent with prior literature, this study adopts the SA index to classify firms as financially constrained or unconstrained.

Credit ratings provide an alternative measure: companies rated as investment grade (AAA to BBB−) are classified as unconstrained, while speculative-grade firms (BB + to CCC−) are considered to be financially constrained.

The empirical specification is given by:

Where $Capexi,t$ is the ratio between capital expenditure and total assets; $αi$ and $γt$ represent firm and quarter fixed effects, respectively; $Cashi,2018Q4$ measures cash and cash equivalents over total assets in the fourth quarter of 2018; $Covid$ is a dummy variable equal to one for all quarters of 2020 and zero otherwise; $Controli,t$ represents all control variables, namely: i) Size (log of total assets); ii) Leverage (total debt over total assets); iii) Cash Flow (net profit plus depreciation, amortization, and depletion over total assets); and iv) Market-to-Book (ratio between market value and accounting value); Finally, $εi,t$ represents the error term of the model.

Following Zheng (2022), the model links firms' pre-crisis financial positions to investment during the crisis. The panel spans from 2019Q1 to 2020Q4, while the cash variable is fixed at its 2018Q4 value. Although using a cash measure closer to the pandemic onset might seem preferable, fixing cash in 2018Q4 allows for balanced comparisons across the pre-crisis and crisis periods, yielding eight quarters of observation. Subsequent analysis shows that cash positions are persistent over time, supporting this choice.

A difference-in-differences approach treats the COVID-19 pandemic as a fully exogenous shock. All estimations were obtained from a fixed-effects panel data model, in which the Huber-White robust variance estimator was employed to mitigate heteroskedasticity, with firm and quarter fixed effects included. According to the hypothesis prepared in the previous section, the $β1Cashi,2018Q4*Covid$ is positive and statistically significant, thus showing that the cash level maintained by companies before the pandemic was a determining factor for maintaining their investments during the pandemic period.

The descriptive statistics presented in the next section indicate that firms' average financial positions remained stable between 2018Q4 and 2019Q4, which gives greater validity to the methodological procedure adopted in this study. Table 1 summarizes the variables used in the analysis.

It should be noted that all continuous variables were winsorized in percentiles 2.5 and 97.5 to reduce the influence of outliers.

Table 2 reports descriptive statistics for the main variables.

The Capex variable is right-skewed, with a mean (0.008) above the median (0.005), indicating that some firms invest substantially more than average. Compared with Franzotti and Valle (2020), the lower mean investment in this study may reflect the downward trend identified by the authors in the earlier periods. Cash2018 exhibits a similar pattern, with a mean of 0.12 and median of 0.09, suggesting substantial heterogeneity in liquidity positions. Cash Flow is highly dispersed, with both negative and positive values, a mean of 0.012, and a standard deviation of 0.077. Finally, Size ranges from 17.53 to 25.34, with a median of 21.89 and a relatively low standard deviation (1.86), indicating a concentrated distribution around the mean.

Table 3 compares the descriptive statistics between the pre-pandemic (2019) and pandemic (2020) periods.

Consistent with prior crisis literature, average Capex, Size and Cash Flow. This pattern is consistent with Almeida, Campello, Laranjeira, and Weisbenner (2011), Duchin et al. (2010) and Zheng (2022), who document investment contractions during crisis periods. Leverage increased slightly in 2020.

To provide an initial assessment of the relationship between pre-pandemic cash and investment, firms were classified as cash-strong or cash-weak according to whether their cash holdings were above or below the sample average, respectively. Figures A1 and A2 (listed in Supplementary_Material) present the evolution of the average investment for these groups.

Figure A1 compares firms classified as cash-strong and cash-weak based on their cash holdings in 2018Q4. Figure A2 extends this classification to include all quarters of 2019. Based on these groups, the average Capex was calculated for each period, allowing a comparison of investment dynamics between firms with persistently high and low cash reserves.

According to Figure A1, which shows the investment levels throughout the analyzed period, both groups of companies showed reductions in their investment levels between the last quarter of 2019 and the first quarter of 2020. However, the decline was less abrupt for companies with above-average cash levels at the end of the last quarter of 2018.

Figure A2 shows the investment levels of companies with above- and below-average cash reserves from 2018Q4 to all quarters of 2019. As shown in Figure A1, both groups of companies showed reductions in their investment levels in the first quarter of 2020 compared to the previous quarter, but the most affected companies were those with the lowest cash reserves.

Overall, Figures A1 and A2 show that the decline in corporate investment during the pandemic was less pronounced for firms with higher cash reserves. Consistent with this evidence, Acharya and Steffen (2020) found that although leverage increased across firms during the pandemic, BBB- and speculative-grade firms substantially increased their cash holdings due to the heightened risk of downgrades and financial constraints.

To demonstrate the persistence of cash positions over time, t-tests and Mann-Whitney tests were performed. The results indicate that the average cash position in 2018Q4 was statistically equal to that in 2019Q4, as the null hypothesis of equal means cannot be rejected (p-values of 0.5130 and 0.3928, respectively). In addition, the correlation between cash holdings in 2018Q4 and 2019Q4 was 0.7556, indicating strong persistence in firms' cash positions across periods.

Table 4 presents the baseline regression results with quarter fixed effects.

The variable $Cashi,2018Q4$ is omitted from the regression because it is time-invariant, which results in a zero coefficient. The Covid dummy is positive and significant at the 5% level and should be interpreted jointly with the interaction term. The variable of interest, Cash_Covid, which captures the interaction between Covid and $Cashi,2018Q4$⁠, is also positive and significant at the 5% level. This finding supports the literature and evidence from the graphical and descriptive analyses, indicating that cash reserves positively affect corporate investment during crises. Specifically, a stronger pre-crisis financial position mitigated the pandemic's adverse effects on investment, with an estimated effect of approximately 0.5%. Among the control variables, only Size and Cash Flow are significant at the 1% level.

The results indicate that, for the overall sample of non-financial firms listed on B3, cash reserves play an important role in sustaining investments during crises. Prior studies document a “dash-for-cash,” particularly among firms facing financial constraints. Acharya and Steffen (2020) show that BBB- and speculative-grade firms increased their leverage in early 2020 to build cash buffers, while Almeida (2021) argues that these firms accumulated liquidity beyond their immediate needs as a precaution against future funding shortages.

Therefore, the results in Table 4 support Hypothesis 1, indicating a positive and statistically significant relationship between pre-pandemic cash reserves and corporate investment during the COVID-19 crisis.

As an additional test, we created a dummy variable, HighCash, which is equal to 1 for firms that maintained above-average cash reserves in the last quarter of 2018 and throughout 2019. These correspond to the cash-strong firms illustrated in Figure A2.

We also constructed the interaction variable HighCash_Covid, defined as the interaction between HighCash and Covid. This specification tests whether firms that consistently held high cash reserves before the pandemic were better able to sustain investment during the crisis, thereby assessing the importance of cash-holding policies in periods of heightened uncertainty. The regression model is presented in Equation 3, and the results are reported in Table A1 (listed in the Supplementary_Material).

The variable of interest, HighCash_Covid, has a positive and statistically significant coefficient at the 1% level. The results indicate that firms that maintained above-average cash reserves in the last quarter of 2018 and throughout 2019 invested more during the COVID-19 pandemic than firms with below-average cash reserves. This finding confirms the results in Table 4, showing that cash reserves positively affect corporate investment during periods of liquidity uncertainty and limited access to external financing. As in the baseline model, only Size and Cash Flow were statistically significant, both at the 1% level.

This study chose to verify the impact of cash reserves on investment during the pandemic using a sample of companies that are theoretically financially constrained and those that are not. This study aims to determine the importance of cash reserves when companies are under threat or facing financial constraints. To this end, the SA index and the companies' ratings were used. The aim is to test hypothesis H2a, which states that for constrained companies, there is a positive relationship between maintaining high levels of cash and investment during the pandemic. Furthermore, we sought to test hypothesis H2b, which states that for unconstrained companies, no relationship between cash and investment is expected during the pandemic. The regression results are presented in Table 5.

For financially constrained firms, the interaction between pre-pandemic cash and Covid remains positive and statistically significant, with larger coefficients than those in the full sample (0.924% using the SA index and 0.959% using ratings). In contrast, the interaction term is not significant for unconstrained firms.

These findings suggest that maintaining high cash reserves is particularly important for financially constrained firms. According toAlmeida (2021), some firms engage in precautionary borrowing, accumulating liquidity beyond their immediate needs to protect themselves against future funding shortages and potential credit downgrades. However, for unconstrained firms, the financial position does not appear to significantly affect investment during the pandemic, likely because concerns about future liquidity are less severe.

Therefore, the results support Hypotheses H2a, indicating a positive relationship between cash and investment for financially constrained firms during the pandemic, and H2b, which predicts no significant relationship for unconstrained firms.

As a robustness check, a placebo test was conducted in which the proposed model was replicated for a non-crisis period, specifically the quarters of 2018 and 2019, with cash holdings fixed at 2017Q4. In practice, this involved replacing the variable $Cashi,2018Q4$ with $Cashi,2017Q4$ and substituting the analysis period of 2019–2020 with 2018–2019 in the regression model specified in Equation 2. The results indicate no statistical significance at the 5% level for the interaction term of interest, namely, the interaction between the cash variable and the dummy variable that equals 1 for the crisis period (in the placebo test, the quarters of 2019) and 0, otherwise.

When replicating the same specification for the regressions in Table 5, that is, applying the placebo test to financially constrained and unconstrained firms based on the SA Index and credit ratings, no statistical significance was found for the interaction variable of interest. Overall, this strengthens the hypotheses H1, H2a, and H2b.

As an additional check, we include industry and time fixed effects to control for industry-specific characteristics and common time trends. The results are presented in Table A2 (listed in Supplementary_Material).

The coefficient of interest remains positive but loses its statistical significance in this more restrictive specification. This outcome is expected because the inclusion of industry-time fixed effects absorbs much of the variation in investment associated with industry-specific dynamics, particularly during the COVID-19 pandemic when shocks differed substantially across industries.

These findings highlight the importance of industry characteristics in shaping investment behavior during the pandemic. Firms within the same industry appear to follow similar investment patterns, reflecting common exposure to industry-specific shocks and opportunities.

This study examines whether pre-pandemic cash reserves mitigated the negative impact of the COVID-19 crisis on corporate investment, particularly among financially constrained firms. Using a difference-in-differences approach that treats the pandemic as an exogenous shock, the analysis focused on the interaction between firms' cash positions in 2018Q4 and the COVID-19 period.

The results show that cash reserves positively affected corporate investment during the crisis. Firms with stronger pre-pandemic financial positions experienced smaller investment contractions, with an estimated effect of approximately 0.5% in the full sample, consistent with Tawiah and O'Connor Keefe (2022) and Zheng (2022). The effect is substantially larger for financially constrained firms, reaching 0.924% using the SA index and 0.959% using credit ratings. These findings are consistent with Acharya and Steffen (2020) and Almeida (2021), who argue that firms facing greater funding risks accumulate liquidity as a precaution against future credit shortages.

Additional robustness tests reinforce these results. Placebo tests for the 2018–2019 period, using cash holdings from 2017Q4, yielded no significant interaction effects, suggesting that the documented relationship is specific to the COVID-19 shock. Moreover, when industry-time fixed effects were included, the coefficient of interest remained positive but became statistically insignificant, indicating that heterogeneous industry conditions influenced investment dynamics during the pandemic.

By providing evidence from Brazil, this study contributes to the literature on cash holdings and investment during crises, particularly in emerging markets, which remain underrepresented in the empirical research. The findings highlight the importance of internal liquidity as a buffer against investment disruptions when external financing is scarce.

Future research could examine the sources of liquidity used by Brazilian firms during the pandemic, including government-subsidized credit and special lending programs. Comparing firms that accessed these programs with those that did not may provide additional insights and allow for closer comparisons with firms that participated in Federal Reserve programs during the COVID-19 crisis.

The supplementary material for this article can be found online.