Do Corporations Retain Too Much Cash? Evidence from a Natural Experiment

Many corporations have been criticized for accumulating large amounts of cash on their balance sheets. Does holding on to all this cash represent a good use of corporate resources? On the one hand, consistent with the label “cash hoarding” often used in this context, firms’ cash holdings could be excessive and harmful to their valuations. That would be the case, for example, if having large cash balances increases the likelihood that this money will be spent unwisely (Jensen, 1986) or if more productive alternative uses exist for the cash, such as investing in new projects or paying it out to investors. On the other hand, firms also hold cash for many legitimate reasons, in which case large cash holdings could be optimal even if they appear excessive. For instance, having an ample cash buffer can help a firm to better ride out recessions or industry downturns. Because of this tension, it is difficult to tell if firms are holding on to too much cash or just the right amount.

To evaluate whether firms retain too much cash, we ideally need an exogenous shock to firms’ cash retention policies and a way to measure the effect on valuations. If a firm retains the right amount of cash, any exogenous change away from this level will reduce the firm’s value. But if a firm retains too much cash, then a decrease in retention can raise the firm’s value.

In our paper, Do Corporations Retain Too Much Cash: Evidence from a Natural Experiment that was recently published in the Review of Financial Studies, we seek to shed light on this question through the lens of a natural experiment. For our empirical analysis, we study the effects of a 2014 tax reform in Korea that acted as a shock to firms’ incentives to accumulate cash. This reform was introduced with the specific intent of reducing firms’ cash retention by imposing a new tax on firms if they retained too much of their earnings.

An important feature that we exploit in our empirical analysis is that the new tax applied only to firms that met at least one of two conditions: 1) those with book equity greater than 50 billion won (roughly US$50 million), or 2) those belonging to large business groups known as chaebol. This allows us to compare the reactions of firms and investors among the firms that were subject to the tax (“treated” firms) and those that were not (“control” firms).

For the treated firms, the reform imposed a ten-percent tax on “excess retention,” which was defined as earnings above a given threshold after deductions for payouts, investments, and wage increases. Specifically, firms that did not spend at least 80% of their earnings on such payouts, investments, or wage increases during a year would have to pay the 10% tax on the residual. (An alternative rule that firms could choose did not allow for any deduction for investments but instead required firms to spend at least 30% of earnings on either payout or wage increases to avoid the tax.) While some firms already were spending enough of their earnings to avoid paying the tax, most firms did not, which means that the new tax pushed a majority of firms to change their behavior.

In the paper, we conduct three different types of tests to map out the effects of this natural experiment. First, we study the effects on firms’ cash policies using a difference-in-differences framework by comparing changes in cash retention from two years before the reform (2013-14) to two years after (2015-16) between the treated and control firms. This analysis shows that the treated firms responded to the reform by significantly cutting their annual flow of cash retention by around half. This corresponds to an average annual reduction of about 4 billion won (US$4 million) per treated firm, or 16 trillion won (US$16 billion) annually in aggregate. Achieving such a change was, of course, the stated goal of the reform.

We next study the valuation effects of the reform, employing an event study framework. We measure how the treated and control firms’ valuations differentially changed around key dates when the reform was proposed and passed. Our findings show that the treated firms’ cumulative abnormal returns (CAR) around the key dates were positive, around 2%, which raised the aggregate market value for those treated firms that were publicly listed by around 20 trillion won (US$20 billion). This positive reaction from investors is consistent with these firms’ cash savings having been excessive before the reform. That is, if a firm’s cash policies were previously optimal, then the new tax would be distortive and would result in lower values (negative CARs), either from distorted decisions if the firm retains less cash to avoid the tax, from higher taxes paid, or both. However, if a firm had been retaining too much cash and the reform pushed it to retain less, its valuation could increase even despite higher taxes, which is what we observe in the data on average.

Earnings not retained as cash must be spent elsewhere, so we use the difference-in-differences methodology to study how this reform affected alternative uses of cash. Specifically, we examined changes to firms’ spending on investments, wage increases, and payouts, as the reform explicitly targeted those outcomes. Our findings show that the treated firms, on average, increased payouts to investors by around 30%, or around 6 trillion won (US$6 billion) per year. Capital expenditures increased by 20%-30%, and the treated firms raised wages more.

Since the previous tests suggested excessive cash retention before the tax reform, we next investigated two underlying mechanisms for why firms could have acted in that way. We focused on two mechanisms: (1) behavioral biases stemming from memories from the 1997 Asian financial crisis, and (2) agency problems due to poor corporate governance. As a first step in this analysis, we establish that firms with adverse memories from the 1997 crisis and those firms with weak governance tended to hold more cash than their industry peers in the years leading up to the reform.

We then examine how these specific groups of firms and their investors reacted to the reform. Here, we find that firms with past experiences from the Asian financial crisis responded more strongly to the reform by making larger cuts to their cash retention and larger increases to payouts (more so than the other treated firms without such past experiences). These firms also experienced more positive valuation effects. These results suggest that the tax reform had a positive ‘nudge’ effect for firms holding excessive cash for behavioral reasons. And consequently, the reform resulted in a particularly positive impact on their valuations.

Conversely, even though treated firms with poor governance also reduced their cash retention, such firms instead primarily started spending more on new investments and did not increase payouts. No increase in firm value was observed for these treated firms with poor governance. These findings imply that the reform might have had the unintended consequence of spurring firms that previously had saved cash excessively due to agency problems to start investing this money in potentially negative-NPV projects instead.

Overall, our findings provide supportive evidence for the frequent critique that corporations’ cash retention has been excessive—not only for a few anecdotal extremes but across a wide sample of firms in an economy. Generalizing our results to settings beyond Korea is always challenging. Still, our findings indicate that corporations across many other developed nations exhibit cash and investment policies that are relatively similar to those in Korea. Other countries, including Japan and China, have also considered instituting similar policies that target cash retention. However, an important takeaway for economic policy is that utilizing policy levers to discourage excessive cash retention is not a silver bullet and can have unintended consequences. In particular, for poorly governed firms, our findings suggest that “forced” dissaving could further exacerbate value destruction, as firms might be incentivized to spend more money on inefficient empire-building.