Financing R&D Spending: The Role of Corporate Cash Holdings

Research and development (R&D) investment and productivity growth have long been recognized as central drivers of economic growth. Yet innovation is inherently difficult to finance externally: It is intangible, risky and subject to severe asymmetric information. As a result, firms with higher R&D intensity—R&D expenditures divided by total assets—tend to hold a larger share of their assets in cash.Researchers Thomas W. Bates, Kathleen M. Kahle and René M. Stulz documented that the secular increase in cash holdings from 1980 to 2006 was connected to an increase in R&D intensity. Researchers James R. Brown and Bruce C. Petersen found that firms most likely to face financing frictions rely extensively on cash holdings to smooth R&D. Another important source of funding is venture capital. Economists Jeremy Greenwood, Pengfei Han and Juan M. Sánchez documented that venture capital financing also plays an important role in economic growth, especially in the early stages of development. They highlighted empirically that venture capital financing is positively associated with firm innovation and growth in the United States. Their model analysis indicates that the increased efficiency offered by venture capital for financing inventive startups is important for long-run growth and welfare (measured in terms of consumption).

In recent years, with the need to invest in artificial intelligence (AI), it has become more important to understand the role of cash. For example, “hyperscalers” have announced their investment plans on their AI-related projects at unprecedented levels.For example, Meta Platforms said in November 2025 that it will invest $600 billion in U.S. infrastructure and jobs over the next three years, including AI data centers. In December 2025, Microsoft unveiled $23 billion in new AI investments, with the bulk earmarked for India. In this blog post, we will explore how cash holdings and R&D intensity of U.S. firms have changed over time. We will also discuss possible implications of these trends for the future given the increase in AI-related investment.

Using Data on U.S. Incorporated Firms

For our analysis, we studied firms that are U.S. incorporated and publicly traded between 1979 and 2024 using S&P Compustat data. We focused on nonfinancial and nonutility firms because financial firms are likely to hold cash for different reasons, such as liquidity requirements, and utility firms are likely to be regulated. Variables were inflation-adjusted to 2017 U.S. dollars and were converted into U.S. dollars if reported in other currencies. Since the fiscal year varies by firm, we defined a year based on the last date of the company’s fiscal year. In addition, we removed observations if the following information was negative or missing: the firm’s total assets, sales, cash and short-term investments; R&D expenditures; and net property, plant and equipment. Observations with missing information on the firm’s age were also excluded. Observations whose cash ratio—defined as cash and short-term investments divided by total assets—is above 1 were also excluded. Lastly, we focused on firms whose inflation-adjusted total assets were above $10 million.

How Corporate Cash Holdings Have Changed Over Time

The first figure displays the distribution of the cash ratio among firms from 1979 to 2024, showing the time series of the 25th, 50th, 75th and 90th percentiles as well as the mean, all weighted by firm assets.Asset-weighted percentiles are calculated by sorting firms by their cash ratio and identifying the percentile thresholds based on cumulative asset shares. The cash ratio is defined as cash and short-term investments divided by total assets. Weighting by firm assets gives greater weight to larger firms, so the statistics reflect where most corporate assets were held rather than treating all firms as equal in size. In simple terms, the figure shows how much cash firms held relative to their asset size, and how this differed across the distribution—from firms holding relatively little cash (25th percentile) to those holding substantially more (90th percentile), as well as the typical firm (the 50th percentile, or median) and the average firm (mean).

The figure illustrates a pronounced upward trend in cash holdings, particularly between 1990 and 2010.Researchers Juan M. Sánchez and Emircan Yurdagül studied determinants of the rising cash holdings of U.S. firms during this period. Over that period, the mean and median increased by 9.1 percentage points and 8.2 percentage points, respectively. More recently, however, the pattern has partially reversed. From 2020 to 2024, the mean and median declined by 3.4 percentage points and 1.3 percentage points, respectively. The decline is even sharper at the top of the distribution, with the 90th percentile falling by 14.9 percentage points.

How R&D Intensity Has Changed over Time

The next figure displays the same statistics but for R&D intensity, defined as R&D expenditures divided by total assets. In other words, it tracks how aggressively companies invest in developing new products and technologies relative to their size.

R&D intensity has risen steadily over the past four decades, with the increase again concentrated in the upper tail of the distribution. The figure shows a particularly notable acceleration between 2013 and 2024: The mean and median rose by 2.0 percentage points and 0.9 percentage points, respectively, while the 75th and 90th percentiles increased by 3.4 percentage points and 6.9 percentage points, respectively.

In recent years, as firms have ramped up investment in AI, the figures show a sharp increase in R&D intensity alongside a decline in cash ratios. These trends are particularly pronounced at the top of the distribution. The evidence is consistent with the view that the cash accumulated between 1990 and 2010 was, at least in part, intended to finance future innovation.

Possible Implications for Future Financing of R&D

Looking ahead, if AI-related investment continues to expand, firms may enter a period with lower internal liquidity and greater reliance on external financing. Because innovation is typically intangible, risky and difficult to pledge as collateral, such a shift could create financing challenges in the years to come.

Notes

1. Researchers Thomas W. Bates, Kathleen M. Kahle and René M. Stulz documented that the secular increase in cash holdings from 1980 to 2006 was connected to an increase in R&D intensity. Researchers James R. Brown and Bruce C. Petersen found that firms most likely to face financing frictions rely extensively on cash holdings to smooth R&D. Another important source of funding is venture capital. Economists Jeremy Greenwood, Pengfei Han and Juan M. Sánchez documented that venture capital financing also plays an important role in economic growth, especially in the early stages of development. They highlighted empirically that venture capital financing is positively associated with firm innovation and growth in the United States. Their model analysis indicates that the increased efficiency offered by venture capital for financing inventive startups is important for long-run growth and welfare (measured in terms of consumption).
2. For example, Meta Platforms said in November 2025 that it will invest $600 billion in U.S. infrastructure and jobs over the next three years, including AI data centers. In December 2025, Microsoft unveiled $23 billion in new AI investments, with the bulk earmarked for India.
3. Asset-weighted percentiles are calculated by sorting firms by their cash ratio and identifying the percentile thresholds based on cumulative asset shares.
4. Researchers Juan M. Sánchez and Emircan Yurdagül studied determinants of the rising cash holdings of U.S. firms during this period.