Political Uncertainty and Finance: A Survey

2.1 Equity Markets

Prior studies analyze the asset pricing implications of political uncertainty in a theoretical setting (e.g., Pástor and Veronesi, 2012, 2013). Pástor and Veronesi (2012) develop a general equilibrium model and prove analytically that the expected value of the stock return at the announcement of a policy change is negative. In Pástor and Veronesi (2012)'s equilibrium model, a firm's profitability follows a stochastic process and can be impacted by the prevailing government policy. Meanwhile, the government acts as a quasi-benevolent decision maker who considers both investors’ welfare and relevant political costs when making its policy decision. At the announcement of a policy decision, the change in stock prices largely depends on two factors: whether a policy change happens and the extent to which this result is unexpected. Solving for the optimal government policy decision, a policy change is most likely to occur if the old policy has led to unexpected low realized profitability. Therefore, policy change can increase firms’ expected profitability, thereby pushing stock prices up, that is, the cash flow effect. However, due to the new policy's higher uncertainty, policy change can also increase discount rates, thereby pushing stock prices down, that is, the discount rate effect. The authors show analytically that, on average, the discount rate effect will outweigh the cash flow effect and stock prices will fall at the announcement of a government policy change. The reason is that positive announcement returns tend to be small as policy changes that have the potential to push stock prices up are mostly expected by investors. Moreover, as new policies introduce more uncertainty and raise the volatility of the stochastic discount factor, government policy changes will increase stock return volatilities and correlations.

Closely related to Pástor and Veronesi (2012, 2013) analyze the risk premium induced by political uncertainty. Pástor and Veronesi (2013) extend the general equilibrium model developed by Pástor and Veronesi (2012) in two key aspects. First, unlike Pástor and Veronesi (2012) who assume political cost is unknown to investors, Pástor and Veronesi (2013) allow investors to learn about the political costs of the new policies through the stream of political news. Second, Pástor and Veronesi (2012) assume that all government policies are identical a priori, whereas Pástor and Veronesi (2013) allow the government to choose from a set of heterogeneous policies. The impacts of both, the learning about political costs and policy heterogeneity, are larger in weak economic conditions when a policy change is more likely to happen.

Accordingly, in Pástor and Veronesi's (2013) analytic model, stock returns are driven by capital shocks, impact shocks, and political shocks, and correspondingly the equity risk premium can be decomposed into three components based on the three types of shocks. As the optimal choice for the government is to change its policy when the old policy is perceived as sufficiently unfavorable, the authors show analytically that the composition of the risk premium varies depending on the economic state. In weak economic conditions when the probability of a policy change is high, the equity risk premium is largely driven by the political risk premium. However, in strong economic conditions when a policy change is unlikely to happen, the risk premium is mostly influenced by impact shock-related risk premiums. In intermediate economic conditions when it is uncertain whether the prevailing policy will be replaced, the equity risk premium reaches a peak due to the concurrence of the three types of risk premium components. Moreover, the authors also find that stock returns are more volatile and more correlated in bad economic conditions and in situations where there is a larger heterogeneity among potential new policies.

Using the policy uncertainty index constructed by Baker et al. (2016) as a proxy for political uncertainty, several studies empirically investigate the impacts of political uncertainty on stock prices (e.g., Pástor and Veronesi, 2013; Brogaard and Detzel, 2015; Baker et al., 2016). Baker et al.'s (2016) policy uncertainty index is a composite measure based on three components: (i) the percentage of news articles related to policy-related uncertainty in large newspapers; (ii) the magnitude of federal tax code provisions set to expire; and (iii) dispersion of economic forecasts of the consumer price index and purchases of goods and services by governments.

Using the stock return data from the US market, Pástor and Veronesi (2013) conduct empirical analyses to test the predictions derived from their analytic model. They find that the empirical evidence is consistent with their predictions. First, results show that the uncertainty index is negatively associated with economic conditions, as measured by the Chicago Fed National Activity Index, industrial production growth, the Shiller price–earnings ratio, and the default spread. Second, the uncertainty index is positively associated with return volatilities and return correlations, and these positive associations are more pronounced in bad economic times when political uncertainty is higher. Finally, by regressing future excess stock returns on the interactions between the policy uncertain index and the measures of economic conditions, they find that political uncertainty commands a higher risk premium in bad economic times.

Pástor and Veronesi (2013) emphasize the role of political uncertainty in different economic conditions. Brogaard and Detzel (2015) focus on the unconditional association between policy uncertainty and expected returns and conclude that policy uncertainty is an economically important risk factor in the pricing of assets. In particular, based on the framework of Merton's (1973) intertemporal capital asset-pricing model (ICAPM), Brogaard and Detzel's (2015) theoretical work suggests that if innovations in policy uncertainty deteriorate investment opportunity, then economic policy uncertainty (EPU) should carry a negative risk premium. The authors first use time series to investigate whether EPU helps to forecast future stock returns. Results show that an increase of one standard deviation of the uncertainty index leads to an increase of 1.5% in three-month expected excess return in the US equity market. Next, the authors investigate whether EPU is priced in the cross-section of stock returns. Using Fama-French 25 portfolios formed on size and momentum returns as test assets, results show that EPU earns a significant negative risk premium, after controlling for the Carhart four factors, the Pástor-Stambaugh traded liquidity factor, as well as other measures of economic uncertainty. Baker et al. (2016) also find that their policy uncertainty index is positively associated with stock price volatility, and Berkman et al. (2011) show that political crises lead to an increase in volatility of world stock market returns.

In another stream of literature on political uncertainty, researchers employ the empirical strategy based on various political-associated events. By comparing the stock returns of firms in industries with different exposure to government spending, Belo et al. (2013) investigate the impact of the presidential partisan cycle on stock returns through the channel of government spending. They find that in the United States, firms exposed more to government spending experience higher (lower) stock returns during Democratic (Republican) presidencies. Next, focusing on an investment strategy that features the observed pattern in returns of government exposure portfolios across presidential terms, the authors document non-zero abnormal returns of the investment strategy, ranging from 3.3% to 6.9% per annum using different asset-pricing specifications. A further investigation into the distribution of abnormal returns within presidencies reveals that abnormal returns concentrate mainly in the middle of the presidential term, suggesting that the market underreacts to changes in government spending policies.

Exploiting the unexpected 2012 Bo Xilai political scandal that occurred in China as an exogenous shock to political stability, Liu et al. (2017) examine the impact of political uncertainty on stock prices. Using a three-day window cumulative abnormal returns centred on Bo's event to measure the market reactions to the shock, Liu et al. (2017) find that stock prices declined as a response to the scandal, especially for politically sensitive firms. They further investigate the reason for the decrease in stock prices. On the one hand, they find that the decrease in both expected cash flows and realized cash flows following the event tended to be more substantial for less policy-sensitive firms, which is inconsistent with the cash flow effect explanation. On the other hand, they document a large increase in stock volatility right after the event for more policy-sensitive firms. These combined results show that this political effect is mainly driven by a change in the discount rate rather than being the result of a cash flow news effect.

In a recent international study, Brogaard et al. (2019) utilize the US election cycle as a proxy for global political uncertainty and find that in 50 non-US countries, stock prices fall and volatilities rise during the periods prior to US elections. This global political effect increases with the level of uncertainty in US election outcomes and a country's equity market exposure to foreign investors, but does not vary with the country's international trade exposure, again suggesting a discount rate news effect rather than a cash flow news effect.

Two recent empirical papers focus on the economic consequences of political uncertainty on other properties of stock prices. Using a large sample of firm-year observations from 38 countries from 1982 to 2012, Li et al. (2018a) investigate how national elections impact individual companies’ stock risk tail around the world. A major advantage of using national elections to proxy for political uncertainty is that elections are pre-scheduled and therefore can be treated as exogenous events. Results show that stock price crash risk is lower (higher) during (after) national elections worldwide. They attribute this inter-temporal pattern to the hoarding of bad news before elections and the subsequent release of bad news after elections. They also document cross-country variations in the impact of political uncertainty on tail risk. In particular, the association between tail risk and political uncertainty is more pronounced in countries with less monitoring and more uncertainty and in firms with a higher level of information asymmetry.

Also employing Baker et al.'s (2016) policy uncertainty index to measure political uncertainty, a recent US-based study by Duong et al. (2018a) focuses on the impact of political uncertainty on stock liquidity. Using alternative measures of stock liquidity, Duong et al. (2018a) consistently document a negative impact of policy uncertainty on stock liquidity. In addition, this negative impact was more severe during the global financial crisis period when there was a higher level of policy uncertainty, and for firms whose returns were more sensitive to government policy changes. The authors further explore reasons that help to explain the negative impact of political uncertainty on firm-level stock liquidity. They find that information asymmetry, cash flow and return volatility, and funding constraint are the three approaches through which policy uncertainty reduces stock liquidity.

2.2 Other Financial Markets

In addition to studies that document the effect of political uncertainty on equity markets, an emerging line of research also examines this effect in other financial markets.

For example, guided by the theoretical model of Pástor and Veronesi (2013), Kelly et al. (2016) document that political uncertainty is also priced in the equity options market. To investigate the political uncertainty pricing implications in the options market, Kelly et al. (2016) exploit the variation in options market variables around national elections and global summations. As these political events can be viewed as exogenous events, a major advantage of this method is that it at least partially isolates political uncertainty from the broader uncertainty in economic fundamentals. In examining the value of options protection against price risk, tail risk, and variance risk associated with political events, the authors focus on three options market variables, namely, implied volatility, implied volatility slope, and variance risk premium. Using the options data of 20 countries, they show that options live across political-related events (e.g., national elections and global summits) tend to be more expensive because they provide protection against risks related to political events. Moreover, the protection provided by options tends to be more valuable in a weaker economy and in a higher uncertainty environment.

Also using elections to proxy for political uncertainty, Gao and Qi (2013) study how political uncertainty arising from US gubernatorial elections influences the municipal bond markets. Using a sample of 121 503 municipal bonds issued between 1990 to 2000, the authors find that the yields of municipal bonds increased by about 6–8 basis points in periods leading up to elections and reversed in post-election periods. Further, by differentiating elections concurrent with economy contractions and those concurrent with economy expansions, they find that the offering yield for bonds issued during the former is significantly higher than that for bonds issued during the latter. Their findings are consistent with the predictions based on Pástor and Veronesi's (2013) theoretical work. That is, political uncertainty commands a risk premium and this risk premium is higher when political uncertainty increases. The authors employ three alternative measures of the degree of political uncertainty to corroborate their finding. Results show that the impacts of political uncertainty on public financing are more pronounced when election outcomes are less predictable, when states have more outstanding debt, and when there are few restrictive institutional constraints in place.

Relying on the measure of political uncertainty constructed by Baker et al. (2016), Kaviani et al. (2018) investigate whether and how policy uncertainty affects credit risk. Using US bond data covering the period 2002–2015, they find that policy uncertainty is positively associated with corporate bond spreads, controlling for bond-issue and firm characteristics, firm and credit-rating fixed effects, as well as macroeconomic conditions and economic uncertainty. The authors also adopt several approaches to address endogeneity concerns that might affect causality, including computing policy uncertainty residuals, identifying instrumental variables, and using elections to proxy for policy uncertainty. In examining how policy uncertainty impacts credit risk, the authors identify both direct and indirect channels through which policy uncertainty exerts influence on credit spreads. Regarding direct channels, they find that the positive association between policy uncertainty and corporate bond spread is stronger for firms in regulation-intensive industries, for firms with higher tax rates, and for firms that depend on government spending. In addition, Kaviani et al. (2018) also find that firms’ dependence on external financing could be an indirect channel through which policy uncertainty heightens corporate credit spreads.

Liu and Zhong (2017) and Wang et al. (2019) also focus on the link between political uncertainty and credit risk. Unlike Kaviani et al.'s (2018) study, which uses bond yield spread to measure credit risk, these two studies measure an individual firm's credit risk by using credit default swap (CDS) spreads. The main reason, as argued by the authors, is that CDS yields perform better than corporate bond spread yields in reflecting individual firms’ default risk. That is, corporate bond yield spreads are affected by various factors which are not directly related to a firm's default risk, such as liquidity risk, debt–equity agent risk, and bond contractual provisions.

Using national elections as a proxy for political uncertainty and using a sample of firms with single-name CDS across 30 countries, Liu and Zhong (2017) find the positive impact of national elections on CDS spreads in national election years, but the positive impact starts to vanish after an election year. Moreover, they also find that the impact of political uncertainty on individual firms’ credit risk is more severe for short-run credit risk, for firms with no political connections or poor international diversification, and for firms in countries with poor investor protection. Wang et al. (2019) employ the political uncertainty index of Baker et al. (2016), and document similar evidence that increases in the uncertainty index lead to increases in the CDS spreads, and the impacts can persist for up to eight quarters. They also investigate the effect of policy uncertainty on liquidity provision and find that a 10% increase in the uncertainty index is associated with a 4% decrease in the number of market makers in the CDS market. The impact of political uncertainty on CDS spreads and liquidity provision in the CDS market is more pronounced during recession years and for policy-sensitive firms. The results of the two studies suggest that when political uncertainty increases, credit protection becomes costlier for investors.

The abovementioned studies show that political uncertainty has pricing implications for financial markets, such as equities and bonds. Hou et al. (2018) concentrate on the effects of political uncertainty on the pricing of commodities, assets which are situated between the real economy and financial markets. As political uncertainty exerts influence on both the demand and supply of commodities, the direction of the effect is unclear a priori and is ultimately determined by the equilibrium between demand and supply. Hou et al. (2018) find that in a large sample of 87 commodities traded in 12 countries over the period 1960–2017, global commodity prices decreased in the period leading up to a US presidential election. The decrease is deemed to be driven by the shrink in demand for commodities caused by the high level of political uncertainty surrounding presidential elections. This effect remains consistent across countries and commodity categories, and it is stronger for elections with smaller victory margins, and for elections during weaker economic periods. However, in a sub-sample of countries which dominate the production of commodities but contribute very little to the consumption of commodities, the authors find that national elections in those countries push commodity prices up. This test isolates the impact of political uncertainty on the supply side of commodities from the demand side.

Brogaard et al. (2019) also examine non-equity financial markets worldwide, and show that prior to US elections, non-US currencies depreciate, and non-US sovereign bond returns increase, suggesting an increase in investors’ aggregate risk aversion, a flight-to-safety effect.

In sum, the empirical evidence on asset pricing documented in the existing literature generally shows a negative effect of political uncertainty on stock prices and liquidity, and a positive effect on stock return volatilities and correlations, as well as the spread of municipal and corporate bonds.

3.1 Financing Activities

Some studies explore the political uncertainty effect on firms’ financing decisions in the equity market (e.g., Chan et al., 2017; Çolak et al., 2017). Çolak et al. (2017) are among the first to document the association between political uncertainty and initial public offering (IPO) activities. Instead of using national elections to proxy for political uncertainty (e.g., Liu and Zhong, 2017; Hou et al., 2018; Li et al., 2018a; Brogaard et al., 2019), they use US gubernatorial elections to study the impact of political uncertainty on firms’ IPO decisions. Compared with national elections, gubernatorial elections provide several advantages for the authors to examine their research question. First, in the United States, a firm's activities, including its financing activities, can be hugely impacted by its state-level governor's policies regarding taxes, subsidies, budgets, or purchases, and therefore the uncertainty arising from gubernatorial elections can be very important to a firm's IPO decision. Second, the frequency of gubernatorial elections is much higher than the frequency of national elections, which helps to enlarge the sample size. Third, as gubernatorial elections in different states happen in different years, the authors can compare the IPO activities of firms in neighboring states to control for the impact of unobservable economic conditions on their results. Using a sample of 5727 IPOs that occurred during the period 1988–2011, Çolak et al. (2017) find fewer IPOs in a state during the period when the state is scheduled to have an election than during the off-election period. This election effect does not apply to neighboring states that do not hold elections at the same time but it is stronger for firms with more businesses in the focal state, for firms that are more dependent on state contracts, and for hard-to-value firms. Moreover, they also find that political uncertainty leads to higher cost of capital. That is, IPO offer prices are found to be lower during an election period.

Following Çolak et al. (2017), Chan et al. (2017) explore the effects of political uncertainty on equity financing in the context of seasoned equity offerings (SEOs). As there is no limit on the times of equity issuance through SEOs, the pricing and timing of SEOs are more sensitive to the continuously changing environment than those of IPOs. Therefore, to better address their research question, Chan et al. (2017) employ the economic policy uncertainty index to measure the change in policy uncertainty over time. Using SEO data from the period between 1985 and 2014, Chan et al. (2017) find that the economic policy uncertainty index is positively associated with SEO discount, which is the price change from the pre-offer day closing price to the offer price multiplied by minus one. Similar to the findings documented in Çolak et al. (2017), they also find that the policy uncertainty effect is stronger when the economy is weaker, when firms’ stock prices are less informative, and when firms are more dependent on government contracts. Regarding the timing of SEOs, they find that the number of SEOs decreases with a higher degree of political uncertainty, and the length of the period between the filing date and the offering date is positively related to the uncertainty index.

Other studies, such as Waisman et al. (2015), Tran and Phan (2017), and Ben-Nasr et al. (2017), investigate the impact of political uncertainty on firms’ financing activities in debt markets. Waisman et al. (2015) examine the effect of political uncertainty on the cost of corporate debt in the US setting. Using a sample of over 20 000 corporate bond issues over the period 1980–2012, they show that increased political uncertainty leads to higher public debt costs proxied by corporate bond spreads, especially in the years closer to a campaign. This effect holds for the uncertainty measured using US presidential elections as well as the economic policy uncertainty index. However, when measuring political uncertainty using US gubernatorial elections, they fail to document significant results that firms in states with gubernatorial elections face higher financing costs compared with firms in states without gubernatorial elections. The main reason, as argued by the author, is that a large number of firms operate in the national market and therefore the impact of gubernatorial elections is very limited compared with that of national elections. Another explanation is that the bond market is not active enough to compound all the political risk-related information into the pricing of debt.

Tran and Phan (2017) examine the debt contracting effects of political uncertainty in the United States, with a primary focus on the impact of policy uncertainty on debt maturity. They present two competing views regarding the debt maturity structure when policy uncertainty is high. From the supply-side's viewpoint, creditors are more likely to issue short-term debt as it is easier for them to monitor the borrowers when lending in the short term. However, from the demand side's viewpoint, firms are more likely to prefer long-term debt to reduce their refinancing risks. Using a policy uncertainty measure based on Baker et al. (2016) and a sample of 6433 firm-years over the period 1985–2015, Tran and Phan (2017) document results that favor the supply-side explanation. That is, policy uncertainty is negatively associated with the maturity of new bond issues and positively related to the number of debt covenants and yield spreads. These results are stronger for higher default risk firms, as proxied by firms with poor credit ratings or no ratings. The results suggest that bondholders are concerned about the ability of firms to make payments and therefore prefer to lend short term and charge higher risk premiums.

Lastly, Ben-Nasr et al. (2017) study the impact of national elections on the choice of debt financing between public and private debts. Compared with public debtholders, banks are less information asymmetry-sensitive for two reasons. First, banks are more able to access private information, thereby monitoring the actions of borrowers on a timely basis. Second, banks are more efficient in renegotiating or restructuring debt contracts when adverse situations occur. Therefore, it is expected that during election periods when policy uncertainty and information asymmetry is high, firms are more likely to rely on bank debt instead of public debt. In a sample of firms from 35 countries in the 1990–2012 period, Ben-Nasr et al. (2017) show that in periods of national elections, there is a substitution of bank debt for public debt, especially for firms with higher information opacity and for firms facing more financial constraints. Further, the positive association between national elections and the degree of reliance on private debt is more pronounced in firms with strong shareholder rights, but less pronounced in firms with strong labor protection and firms with strong creditor rights. Lastly, this positive relationship is also stronger when policy uncertainty is high, that is, when elections are closely contested, and governments are constrained.

The abovementioned studies show that policy uncertainty adversely affects firms’ access to external financing markets, including both the equity market (e.g., Çolak et al., 2017) and the debt market (Waisman et al., 2015). From a different perspective, two inter-related studies, Duong et al. (2017, 2018b), investigate how firms manage their internal funds when faced with higher political uncertainty. The premise underlying these studies is that when political uncertainty is high, the frictions faced by firms in raising external funds increases, which then motivates firms to seek funds through internal channels to maintain their business operation.

For example, using a sample of US firms from 1987 to 2015, Duong et al. (2017) investigate firms’ tax avoidance behavior during periods of heightened policy uncertainty. As an important source of internal financing, tax saving has its advantages as it does not require the firm to cut off operation-related expenditure. Using the cash effective tax rate to measure tax avoidance, they show that policy uncertainty is positively associated with the firm's level of tax avoidance. The effect of policy uncertainty on corporate tax avoidance initially increases and then begins to decline after three years when policy uncertainty starts to resolve. They then conduct further analyses to provide evidence supporting the idea that firms engage in tax avoidance activities to mitigate the financial constraints caused by policy uncertainty. First, they show that the market credit conditions tighten when policy uncertainty heightens. Second, they find that firms with more cash reserves are less likely to increase tax avoidance. Finally, they show that the tax savings induced by policy uncertainty are used for capital expenditure reinvestments rather than dividend pay-outs. These combined findings suggest that policy uncertainty heightens corporate tax avoidance through its impact on firm financial constraints.

Duong et al. (2018b) directly focus on managers’ cash holding decisions in response to the negative effects of increased policy uncertainty on external financing. Using a sample of US firms over the period 1985–2014, Duong et al. (2018b) find that firms’ cash holdings are positively associated with the policy uncertainty index, controlling for the effects of investment opportunities and the effects of macro-economic uncertainties. They also document that the positive impact of policy uncertainty on a firm's cash holdings is stronger for policy-sensitive firms. In exploring the channels through which this positive association takes effect, they find strong supporting evidence for the “financial constraints” channel but little supporting evidence for the “investment irreversibility” channel. That is, the increase in cash reserves when policy uncertainty surges are due to firms’ incentives to save more cash when facing financial constraints driven by policy uncertainty rather than a lack of investment in the presence of policy uncertainty. Finally, regarding the usefulness of increasing cash holdings, the authors find that reserving more cash in response to policy uncertainty surges enhances shareholder value and helps to mitigate the dampening effects of policy uncertainty on investment and firm innovation. The results of Duong et al. (2018b) suggest that managers strategically hold more cash in response to increased policy uncertainty.

3.2 Corporate Investment Decisions

Julio and Yook (2012) use a sample of national elections in 48 countries in the period 1980–2005 to examine the impact of political uncertainty on firms’ investment behavior. It is argued that in a high uncertainty period, firms are concerned about potential bad outcomes on investment payoff arising from uncertainty and therefore tend to delay investment until the uncertainty is resolved. Using national elections to proxy for political uncertainty, Julio and Yook (2012) investigate the effect of political uncertainty on corporate investment by comparing corporate investment behavior in the period leading up to a national election with that in a non-election period. They find that there is a significant decline in corporate investment in the years leading up to elections compared to non-election years. They also document variations in magnitude of the election effect. In particular, at the country level, the decline is larger in countries with weaker investor protection, with a parliamentary system, with a less stable government, and with a larger central government. Within countries, the election effect is more pronounced for politically sensitive firms, in marginal-win elections, and in situations where the incumbent leader is deemed to be “market-friendly.” Further analyses rule out other possible channels through which firms’ investment behavior can be affected by national elections. That is, the decline in investment during election years is not driven by the political business cycle hypothesis (Nordhaus, 1975) and the political connection hypothesis (Bertrand et al., 2006).

Similar findings are documented by Gulen and Ion (2016) and Jens (2017), who focus on US firms and use the policy uncertainty index and gubernatorial elections as proxies for uncertainty, respectively. Specifically, Gulen and Ion (2016) use the policy uncertainty index to measure the continuously changing overall level of policy uncertainty in the economy and find that a doubling in the level of the uncertainty index is associated with an 8.7% decrease in the investment ratio in the next quarter, and this uncertainty effect can last for up to eight quarters into the future. Moreover, they find that the negative effect of policy uncertainty on capital investment is stronger for firms with more irreversible investments and for firms relying more on the government to maintain their profits.

Jens (2017) uses 328 gubernatorial elections in the United States from 1984 to 2008 to examine the impact of political uncertainty on investment. Examining the research question in the context of gubernatorial elections allows Jens (2017) to use a difference-in-differences model, with firms headquartered in states with (without) an upcoming election as a treatment (control) group. Accordingly, the difference-in-differences model nets out any pre-existing differences in investment levels between firms in the control group and those in the treatment group, which allows the author to identify the effects of political uncertainty on firm investment in a cleaner manner. Similar to the finding documented in Julio and Yook (2012), Jens (2017) also finds that investment declines before elections, especially for firms that are more susceptible to policy changes. However, unlike Julio and Yook (2012) who document a bounce back in investment in the year immediately after an election, Jens (2017) finds that the level of investment in the post-election period is dependent on whether an incumbent leader is re-elected. As most of the firms in Julio and Yook's (2012) sample operate in the parliamentary system, the inconsistency in the results suggests that political uncertainty can be resolved in different patterns in different political systems.

Other studies that focus on investment efficiency include, for example, Durnev (2012) and Drobetz et al. (2018). Durnev (2012) concentrates on the change in corporate investment-to-price sensitivity during election years compared to non-election years. Durnev (2012) considers both the “information view” of investment and the “political view” of investment to explain why managers would pay less attention to stock prices in making their investment decisions during election years when political uncertainty is high. From the information perspective, policy uncertainty arising from elections makes stock prices noisier than non-election years and lowers the amount of private firm-specific information contained in stock prices. From the political perspective, during election years, firms’ investment behavior can be more politically motivated, which also results in a decrease in information reflected in stock prices. Using a binary election variable to capture the effects of political uncertainty during election years, the results suggest that stock-to-price sensitivity decreases during election years in a panel of 79 countries from 1980 to 2006, and the degree of decrease varies with the level of corruption, state ownership, and disclosure by politicians across countries. Further, when viewing stock-to-price sensitivity as a proxy for investment efficiency, Durnev (2012) finds that political uncertainty is associated with a decline in investment efficiency, which further lowers firms’ performance in the years following elections.

Based on the real option theory, prior studies, such as Julio and Yook (2012), Gulen and Ion (2016), and Jens (2017), find that political uncertainty negatively impacts the level of investment. Bloom (2014) argues that uncertainty also affects the sensitivity of the factors that impact a firm's investment behavior by decreasing investors’ sensitivity to changes in economic conditions. Following Bloom's (2014) argument, Drobetz et al. (2018) study the moderating role of political uncertainty in the inverse relation between cost of capital and investment as shown in the theoretical work by Abel and Blanchard (1986) and the empirical work by Frank and Shen (2016). Results demonstrate that the sensitivity of investment to the cost of capital also decreases with the economic policy uncertainty index across 21 countries over the period 1989–2012. After decomposing the measure of the cost of capital into debt component and equity component, they find that the moderating effect of policy uncertainty on the link between the cost of debt and investment is comparable in magnitude to that on the link between the cost of equity and investment. The moderating effect is stronger for high government-dependent firms and for less transparent firms.

From a different perspective, Nguyen et al. (2018) address the impact of political uncertainty on firms’ investment behavior by focusing on the investment decisions that firms make to alleviate the risks resulting from political uncertainty. Specifically, using a sample of 881 non-financial firms from eight East Asian countries from 2003 to 2013, Nguyen et al. (2018) examine the link between policy uncertainty and firms’ foreign direct investment (FDI), and the link between policy uncertainty and firms’ use of derivatives. Regarding FDI, they find that firms are more likely to make FDI in countries with a lower level of economic policy uncertainty. Regarding derivatives use, they find that firms’ derivative use intensity is positively associated with the polity uncertainty index, which is consistent with the explanation that policy uncertainty increases firms’ risk exposure and therefore increases firms’ incentives to use derivatives to hedge risk and manage uncertainty. Turning to the effectiveness of derivatives use, employing a market model based on Adler and Dumas (1984) and Jorion (1990) to estimate firms’ uncertainty exposure, they find that the use of derivatives efficiently reduces firms’ exposure to policy uncertainty. Moreover, the use of derivatives significantly mitigates the negative impact of policy uncertainty on firm performance.

The abovementioned studies focus primarily on the effect of uncertainty on corporate tangible asset investment expenditure. The existing literature also explores the uncertainty effect on corporate investment in intangible assets, such as corporate innovation investment. For example, Bhattacharya et al. (2017) compare the relative importance of policy and policy uncertainty for technological innovation activities. Using industry-level information of US patents to measure the different dimensions of innovation activities (i.e., quantity, quality, originality, riskiness, and exploratory versus exploitative patents), Bhattacharya et al. (2017) document significant decreases in the quantity, quality, and originality of a country's innovation activities a year after elections for a sample of 43 countries between 1976 and 2010. However, they fail to find evidence supporting the impact of policy on most dimensions of innovation activities. To strengthen the causal effects of policy uncertainty on innovation, the authors employ two approaches to address endogenous concerns. They first repeat their empirical test by using a more exogenous measure of political uncertainty, that is, close presidential elections, and find that the results remain unchanged. Next, they partition the sample based on ethnic fractionalization and find that ethnic heterogeneity amplifies the negative effect of political uncertainty on innovation. Taken together, the results imply that the prevailing policy does not significantly affect innovation activities as businesses adapt themselves to policies, whereas political uncertainty has real adverse economic consequences.

Bhattacharya et al. (2017) find that political uncertainty adversely affects innovative investment in the same way as regular investment. Another recent working paper, Ni (2018), revisits this documented negative relationship in a different setting and finds that uncertainty can encourage innovation if firms have an understanding of future policy changes. Using turnovers of local city heads to proxy for policy uncertainty, Ni (2018) examines firms in China and shows that state-owned enterprises (SOEs) may exploit their information advantages, which help them better assess future policy changes, and thus invest more in innovative activities in cities experiencing governor turnovers. The results remain constant after addressing endogeneity concerns. The author also shows that the patents filed during uncertainty periods bring about real economic benefits for SOEs. Further tests rule out alternative explanations of the results. That is, the results are not driven by the new governor's incentives to increase economic growth or by the alleviation in firms’ financial constraints after the new governor assumes office. However, the author concludes that these findings do not imply that political turnover is always beneficial – the positive impact of policy uncertainty on innovation applies only to SOEs that are more willing and more able to adapt to changes, that is, the SOEs facing higher uncertainty, facing more competition, and possessing more growth opportunities.

Another line of literature focuses on mergers and acquisitions (M&As), another important form of corporate investment. Compared with other forms of investment, M&As are usually larger in magnitude and more difficult to reverse. Therefore, in line with the real option theory, it is predicted that policy uncertainty is likely to adversely affect firm acquisitiveness. Consistent with this prediction, two recent studies, Nguyen and Phan (2017) and Bonaime et al. (2018), both document a negative association between the policy uncertainty index and M&A activities in the United States. In particular, they show that policy uncertainty reduces the likelihood of a firm making an M&A announcement. The two studies further examine different aspects of the impacts of policy uncertainty on M&As.

Using a sample of 88 768 firm-year observations over the period 1986–2014, Nguyen and Phan (2017) further consider the effect of policy uncertainty on other characteristics of M&A deals. Specifically, they find that policy uncertainty increases the time for the acquirers to complete M&A deals. Moreover, considering the potential adverse effects resulting from policy uncertainty, the authors find that acquirers tend to be prudent with M&As. That is, acquirers are more likely to use stock rather than cash for M&A payments to maintain their liquidity, and they tend to lower the bid premium. Nguyen and Phan (2017) also document that policy uncertainty is positively associated with both short-term and long-term acquirers’ shareholder value, which can be attributed to two aspects. First, high policy uncertainty motivates acquirers to screen potential targets carefully and choose targets with better expected returns. Second, when policy uncertainty is high, there is value transfer from financially constrained targets to acquirers through acquisition.

Also using US M&A deals between 1985 and 2014, Bonaime et al. (2018) focus on the channels through which policy uncertainty affects M&A activities and document overwhelming evidence confirming the real option theory. That is, policy uncertainty affects M&A decisions primarily through increasing the value of the option of a wait. Specifically, they find that the negative effect of policy uncertainty on M&As is stronger for more irreversible investments, for deals in industries with higher competition and therefore more costly to delay, and for targets whose firm value is more sensitive to policy shocks. These findings consistently lend support to the real option channel through which policy uncertainty affects acquisition decisions. The authors also explore other channels, namely, the interim risk channel, the empire-building channel, and the risk management channel, but only document some evidence supporting the idea that policy uncertainty operates through the risk management channel. Specifically, they find that policy uncertainty increases the likelihood of engaging in M&As that have the potential to reduce the acquirer's risk exposure, such as acquiring a foreign firm. However, the authors note that risk management cannot be the dominant approach through which policy uncertainty affects M&A decisions.

Consistent with the risk management channel through which policy uncertainty affects M&A decisions documented by Bonaime et al. (2018), in an international setting with 47 countries in the period 2001 to 2013, Cao et al. (2017) show that political uncertainty proxied by national elections deters foreign acquirers from purchasing local targets but encourages local acquirers to conduct outbound cross-border acquisitions. Specifically, they find that in the year leading up to national elections when political uncertainty is high, the volume of inbound cross-border acquisitions decreases by 6.8% compared with non-election years, whereas the volume of outbound cross-border acquisitions increases by 7.5% relative to other years. They also show that these changes in the volume of cross-border acquisitions are not driven by changes in macro-economic conditions. Moreover, in selecting a target country, acquirers prefer to conduct outbound acquisitions in countries that can offset the political uncertainty they are facing in their home countries, such as countries with free-trade agreements or military allies, and countries with better shareholder protection.

In a similar vein, Chen et al. (2018) use the gubernatorial election cycle in the United States to study the impact of political uncertainty on M&A deals, as well as the channels through which political uncertainty affects M&A decisions. Consistent with the finding revealed in prior literature that political uncertainty adversely affects investment (e.g., Julio and Yook, 2012; Jens, 2017), Chen et al. (2018) find that acquirers themselves are less likely to engage in acquisition activities if there is an upcoming election in their headquartered state, suggesting that political uncertainty increases acquirers’ risk premia around elections, thereby leading them to pull back M&A investment. The authors also show that acquirers tend to select fewer targets from a state with an upcoming election. This result implies that political uncertainty can also affect M&A investment by changing project characteristics that matter to investment decisions. Further analyses show that the negative impact of political uncertainty on M&As is stronger for smaller deals, for more financially constrained acquirers, and for projects that require higher integration costs.

3.3 Other Corporate Real Decisions

In this sub-section, we will review the political uncertainty literature with regard to other real decisions made by corporate managers in relation to dividend, risk-taking, hedging, tax avoidance, and corporate lobbying activities.

To the best of our knowledge, Huang et al. (2015) is the only study that examines the association between political uncertainty and payout policy. They use a large sample of 35 countries for the period between 1990 and 2008. They show that global political crises increase the market's perceived uncertainty. As a result, this leads to a higher likelihood of dividend termination for existing payers, and a lower likelihood of dividend initiation for non-payers. This finding is consistent with our previous review for the literature on firms’ financing and investment policies, in that the increased difficulty of raising funding from the capital markets makes managers more cautious about cash outflows through investment and payout channels.

An area related to corporate investment is firms’ risk-taking behavior in response to high levels of political uncertainty. Using a sample of US firm-year observations spanning the period 1986–2016, Ion and Yin (2017) explore whether managers take actions to reduce firm-specific risks when faced with policy uncertainty. Their research rests on the premise that CEOs would be motivated to take actions to reduce their firm's risk exposure arising from policy uncertainty if their wealth portfolios were closely related to the riskiness faced by their firm. To begin addressing the research question, they first document a negative relation between the policy uncertainty index and future stock return volatility, suggesting that firms react to policy uncertainty by taking risk-reducing actions. To more directly address the incentives of CEOs in reducing firm-specific risks, the authors examine the mediating effect of CEOs’ financial capital and human capital on this negative relation separately. They find that the negative relation is stronger for CEOs with a higher delta (i.e., the CEO's higher exposure to changes in firm value), and for CEOs with less transferable skills (i.e., the CEO is closely tied to the firm). Ion and Yin (2017) provide further supportive evidence that when policy uncertainty is high, CEOs take actions to reduce risks at both the firm level and the personal level, such as engaging in hedging activities and cross-border mergers, selling own-firm stocks, and exercising fewer options.

Using Brexit as an exogenous political event that can cause political uncertainty, Hill et al. (2019) examine the variation in UK firms’ exposure to political uncertainty. The authors first construct two alternative measures of firm-level uncertainty exposure in the context of Brexit, namely, the sensitivity of the firm's stock returns to changes in the probability of Brexit (i.e., the Brexit beta) and the firm's stock price reaction to the referendum result. Based on the two measures of uncertainty exposure and a sample of firms listed on the London Stock Exchange (LSE), Hill et al. (2019) show that there is variation in individual firms’ uncertainty exposure relating to Brexit. Moreover, they find that more internationalized firms are less affected by the uncertainty arising from Brexit, suggesting that internationalization can be used as a diversification mechanism to help reduce firms’ domestic risk exposure. In addition to internationalization, the authors find that other firm-level characteristics, such as firm growth, firm size, and firm profitability also affect Brexit uncertainty exposure. At the industry level, firms in the financial and consumer sectors experience the highest level of risk exposure, whereas firms in the basic materials and healthcare sectors have the lowest exposure to Brexit risk. This industry-level variation can partially be attributed to the difference in firms’ dependence on domestic economic conditions.

Two political-related corporate decisions are examined by Shang et al. (2018) and Duong et al. (2017). Using a large sample of US firms from 1999 to 2015, Shang et al. (2018) explore how political uncertainty impacts non-lobbying firms’ lobbying initiation decisions. Under high policy uncertainty conditions, non-lobbying firms face two competing forces that can affect their decision regarding whether to initiate lobbying or not. On the one hand, high policy uncertainty can motivate lobbying initiation. Prior studies have shown that policy uncertainty adversely affects firms’ real economic activities. As a result, firms have the incentive to engage in lobbying to influence policy outcomes in their favor and to manage potential risks arising from high policy uncertainty. On the other hand, non-lobbying firms can also be discouraged from initiating lobbying as they are likely to face higher entry barriers when policy uncertainty is high, such as higher upfront costs arising from the asymmetric supply and demand of lobbying services and lower returns arising from a lack of experience in lobbying. Empirical results show that non-lobbying firms are less likely to initiate lobbying when policy uncertainty is higher, which is consistent with the explanation that lobbying entry barriers will increase in periods characterized by higher political uncertainty. The results suggest that although lobbying is most beneficial during high policy uncertainty periods, the high upfront costs can price non-lobbying firms out of the lobbying market.

3.4 Corporate Disclosure Decisions

An emerging stream of literature examines the impact of political uncertainty on firms’ financial reporting policies. Nagar et al. (2018) employ a sample of nearly 7000 US public firms from 2003 to 2016 and find a positive effect of economic policy uncertainty on voluntary management disclosures with regard to management forecasts and 8-K filings, which partially mitigate the increased information asymmetry between investors and managers in higher uncertainty periods.

Similarly, Bird et al. (2017) study the corporate disclosure response to information asymmetry driven by the uncertainty of government policies around the US gubernatorial elections in the years 1995–2014. They find that the frequency and content of voluntary 8-K filings and managerial forecast increase in periods leading up to a close US gubernatorial election. This election effect is more pronounced for elections when the incumbent governor terms out and when the recent party flips. Moreover, they show that the effect of uncertainty on voluntary disclosure reverses in the post-election period.

Boone et al. (2018) also examine US-based firms in a sample period between 1997 and 2013 and find that firms in states experiencing gubernatorial elections provide more frequent and informative 8-K filings, mainly stemming from the Regulation Fair Disclosure filings that contain information such as product development, customers, and key employees. Their cross-sectional analyses show that these increase disclosures are concentrated in firms with more investment, higher information demand, and lower proprietary disclosure costs.

Jiang et al. (2017) study the policy uncertainty effect on the textual disclosures in 10-K and 10-Q filings over the period 1995 to 2015. In a sample of 89 846 firm-quarter observations with 8047 unique firms, they demonstrate that uncertainty increases the length of disclosure but reduces readability. In addition, they find that the uncertainty effect on length is stronger for firms with higher institutional ownership in the post-SOX period, while the effect on readability is weakened in firms with higher analyst coverage and audited by Big 4 CPA firms. These results imply that during high uncertainty periods managers work hard to disclose more information to the market, but they fail to improve the readability and informativeness of the disclosures.

Dai and Ngo (2019) examine the effect of political uncertainty on accounting conservatism, using a sample of 199 110 firm-year observations between 1963 and 2016. They find that the asymmetric timeliness of news recognition in earnings increases in periods leading up to US gubernatorial elections because higher political uncertainty causes an increased contracting demand for accounting conservatism. Moreover, they show that the effect on accounting conservatism increases with firms’ industry exposure to contracting needs, leverage, and the strength of internal governance mechanisms, but decreases with managerial ownership.

Using conference call scripts, Hassan et al. (2017) document that managers and analysts devote more time to discussing topics on political risks prior to or during presidential and congressional election quarters. In a related note, both Bird et al. (2017), and Baloria and Mamo (2018) show that information production by one important information intermediary – financial analysts – declines in high-uncertainty periods, that is, fewer analysts following, larger forecast errors, and higher forecast dispersions, possibly due to the increased complexity of forecasting tasks.

Taken together, the results from the studies on corporate policies suggest that political uncertainty imposes a negative impact on firms’ financing activities, while as a response, managers increase cash holdings to alleviate such a negative impact. Furthermore, political uncertainty not only dampens corporate general investment activities proxied by investment expenditures but this effect also extends to specific riskier investment domains, that is, innovation activities and M&A markets, while also influencing other real corporate activities and financial reporting decisions.