Investment is the foundation for future economic growth under capitalism, but public company investment isn’t what it used to be. This is measurable. US public companies have, in aggregate, reduced their level of investment since 1985 by about 10% , while increasing cash disbursements to shareholders. That’s roughly USD $100 billion per year.
Where did 10% of the US jetpack future fund money go? Was the shareholder revolution simply an excuse to give up innovation and use public companies as cash machines for shareholders? Was it linked, as JW Mason and others argue, to institutional and ideological changes promoting a more short-term outlook, such as the theory that a firm’s purpose is to maximize shareholder value? Or did the money get reinvested in private firms, in world-changing tech startups and failing legacy companies rebuilt with private equity tough love?
On the specific question of whether the money taken out of public companies went into private investment, a decent estimate, explained below, is that at most half went to venture capital, and the rest went on the economic investment equivalent of beer and hookers. (Or perhaps painkillers and football tickets. But not cancer research and bridges.)
The Theory of the Rentier Firm
Economic growth creates wealth and moderates inequality . Investment – building things – is also where change in society and industry manifests through the economy. Investment is a necessity for innovation and bringing new inventions into use.
Mason  finds that since 1985, there is a statistically significant drop of 10.7% in investment by US public companies (0.6% of total corporate assets) and a slightly higher increase in payouts to shareholders. This net decrease in corporate assets has been accompanied by greater use of corporate debt to ensure payouts, rather than fund investment. Mason names this corporate financing management approach the “rentier firm”.
For this note we take the drop of investment shown by Mason as given, and investigate one particular explanation involving private equity. Mason, and earlier Lazonick , have argued that this drop in investment was a consequence of the “shareholder revolution” of the 1980s and accompanying institutional change. One regulatory change cited by both is the 1982 adoption by the Securities and Exchange Commission (SEC) of Rule 10b-18, legalizing large scale share repurchases. This was one change as part of a broader deregulatory trend in financial markets, with the academic success of mathematical finance theory such as the Capital Asset Pricing Model (CAPM)  also influential. The connection between the shareholder revolution and mathematical finance theory is pretty well established, and a good history of it is Mackenzie’s An Engine, Not A Camera .
The assertion that the shareholder revolution resulted in a drop in investment is more controversial. Similarly the drop in public company investment that Mason observes is quantitative and hard to dispute, but due to the data available, the explanation currently depends on coincidence in time. Policy seems a fairly reasonable place to put the blame, but there could be alternative causes. An aging rich world population might be drawing down US investment for spending on the costs of retirement. Investment might have focused on firms outside the US, for instance in the developing world during a period of free trade and globalization. The oil shocks of the 1970s may have altered investment preferences to favour cash over more capital-intense firms. The explanations can even be combined: eg, an aging baby boomer cohort coming through middle age may have been more inclined to policies which allowed them to draw down investment for shorter term priorities.
Michael Jensen  also drew on Sharpe’s CAPM to build his model of firm management, describing agency problems that could arise between the management of a firm and their shareholders. Jensen  would go on to argue in 1989 that privately held corporations were a superior organizational form to public companies listed on the stockmarket. A particular focus is the reduction of wasteful investment, and returning cash to shareholders (whether private or public):
The public corporation is not suitable in industries where long-term growth is slow, where internally generated funds outstrip the opportunities to invest them profitably, or where downsizing is the most productive long-term strategy.
[Management of a private corporation] is built around highly leveraged financial structures, pay-for-performance compensation systems, substantial equity ownership by managers and directors, and contracts with owners and creditors that limit both cross-subsidization among business units and the waste of free cash flow. Consistent with modern finance theory, these organizations are not managed to maximize earnings per share but to maximize value, with a strong emphasis on cash flow.
— Jensen, Eclipse of the Public Corporation
Jensen’s second paragraph can be re-used as the definition of Mason’s rentier firm, whether public or private. They describe the same structure, Jensen favorably, Mason unfavorably. Both have a degree of theoretical and empirical support: Bonen  provides a more recent model of the firm where shareholder value maximization does not ensure firm profit maximization. Rent-seeking by shareholders can be considered a risk, as well as the rent-seeking by managers described by Jensen.
Proponents of this approach to firm management, and the principle of maximizing shareholder value behind it, may sometimes argue that it derives from a universal moral of property rights, but more common is an argument from utility, specifically the productive use of capital:
A central weakness and source of waste in the large public corporation is the conflict between shareholders and managers over the payout of free cash flow—that is, cash flow in excess of that required to fund all investment projects with positive net present values when discounted at the relevant cost of capital. For a company to operate efficiently and maximize value, free cash flow must be distributed to shareholders rather than retained.
— Jensen (ibid)
So cash machines were installed in public companies, with the understanding that cash rightfully belonged to the shareholders, and they would find the best ways to invest the money. Though it’s possible to argue the coincident drop in investment had other causes, this would be either arguing against the actual justification provided for the policy changes by proponents at the time, or that some third force was at work, which also fit the same time period, and had a greater suppressive effect on reinvestment of the cash in public companies. For now, let’s take the proposal that policy changes intended to make cash more readily disgorged from public firms actually had that effect as a reasonable working hypothesis.
Private Equity Investment
The justification for drawing down cash in the rentier firm is that it will be more productively invested elsewhere. However, based on the history since 1985, and netted out, that destination isn’t other public companies. Could we find it outside the stock market? Given the theoretical link between the rentier firm and private equity, it is seems particularly worth looking there.
The early proponents of the shareholder revolution hoped that financial managers at investment banks and private equity funds could entirely replace the top layers of corporate management, but these dreams were not fulfilled. […] Contrary to Michael Jensen’s hopeful forecasts that fund managers would replace corporate headquarters entirely, shareholders do not make the day-to-day decisions in American corporations. They do, however, sharply limit the decisions made by the managers, with the result that corporate finance in a rentier-dominated firm looks quite different than in a managerial one.
Let’s reframe this implication as a hypothesis:
H1 Investment behaviour across public and privately owned firms is consistent.
It’s useful to break private equity into two categories for this analysis: Leveraged Buy-Out (LBO) and venture capital.
H1-1 Investment behaviour across public and privately owned firms created by leveraged buy out (LBO) is consistent.
H1-2 Investment behaviour across public and privately owned firms is consistent for venture capital.
Some existing research addresses this, and public accounting data is a second source of information. Ideally, we would crunch statistical correlations using extensive time series, but it’s awkward data, and a bit beyond the scope of this note. The proportions are also huge enough that confidence intervals are kind of overkill.
Early research, in 1993, by Long and Ravenscraft , on the then fairly new phenomenon of the private firm and leveraged buyout (LBO), showed that levels of Research and Development investment were 40% lower after buyouts. They also found that firms targeted for buyouts are significantly less R&D intensive before buyouts occur, and that those R&D intensive firms that are bought out improve performance afterwards. So while there is evidence investment decreases in accounting terms, there is also an argument the capital may have been used wastefully before the buyout.
There are two recent studies looking at invention in privately owned firms created from leveraged buyouts, using patent quantity and quality. Lerner et al  looks at the US while Amess et al  look at European firms. By contrast with Long and Ravenscraft, these find “no evidence for a decrease” in investment after a buyout, some indication for increase in patent quality, and a slight increase in volume. The test for patent quality is particularly important given the five-fold explosion in patent applications over the last few decades (see below), which is usually treated as inflationary rather than representing a revolutionary age of invention. Both studies note that any improvement shown may be due to selection bias introduced by the private equity process itself – i.e. the selection of underperforming firms.It is limited data, but none of it suggests private equity invests more than public firms. For now, let’s say private equity invests no more than public firms, at least in its leveraged buyout form. This supports H1-1: Investment behaviour across public and privately owned firms created by leveraged buy out (LBO) is consistent.
Feeding The Young Hatchlings of Venture Capital
Cash withdrawn from public companies may be spent elsewhere or invested elsewhere. In particular, the cash disgorged from public companies may have been regurgitated into the hungry innovative mouths of young startups by the nurturing mother birds of private equity. Matt Levine has suggested this , and we can take it as a second hypothesis:
H2 Cash taken from public company investment went into venture capital.
If true, this would contradict consistency of behaviour across public and private firms, ie H1.
A related literature exists on private equity investment as a whole, particularly for insight into “risk capital” or “venture capital”. These forms of private equity should by definition be more focused on investment and innovation. Kortum and Lerner  and Popov et al  look at the US and Europe respectively. The venture capital industry in the US dates back to the 1940s, but its existence at scale emerged in the late 1970s as part of the same deregulatory trends sketched above, particularly a 1979 US Department of Labor reinterpretation of the “prudent man” rule allowing pension funds to invest in risk capital.
For US firms, an elasticity of 0.09 in successful patent applications to venture capital disbursements was found. For European firms, “a 1% increase in private equity invested increases the number of USPTO patents by between 0.04% and 0.05% … PE accounts for as much as 12% of industrial innovation”. Though there is further work to be done on causality, filter effects and the political structure of venture capital, it does seem a genuine source of investment comparable to that in the Flow of Funds, used by Mason. So our H1-2 hypothesis is gone: Investment behaviour across public and privately owned firms is not consistent for venture capital.
That venture capital firms invest capital is perhaps news from the obvious, but it’s also good to have some empirical confirmation. Is it enough to account for the whole investment, gap, though? Some basic accounting arithmetic provides a test.
Since LBO private equity investment seems at best consistent with public companies, the gap found by Mason would need to be taken up entirely by venture capital. Venture capital would also have to cover the investment gap for the entire private equity sector of the economy.
In Mason Table 3, he summarises mean investment in the pre- and post-1985 periods as 5.6% and 5.0% of total corporate assets respectively. Below we use the simple difference, 0.6%.
The graph in Mason Figure 5 actually appears to show higher levels of investment than the figures quoted in Table 3. For example the visual minima of investment in the figure is above the reported mean. The graph appears to directly report percentages from the Flow of Funds  table F.103. I was unable to reconstruct the figures in Table 3. I emailed Mason, but it was probably the victim of a spam filter or busy schedule. As Table 3 seems more conservative, this effect may be greater than described below.
Using the most recent Flow of Funds table L.103, we can project the absolute notionals in reduced investment vs a hypothetical continuation of pre-1985 investment behaviour using the Mason estimator. This would be USD $108 billion for 2015 (hence my $100 billion a year in the intro), with a brief historical summary below.
Secondly, for the size of private equity, we use the Prequin 2016 report  which reports a total size of private equity of $2.4 trillion in 2015. This is in the same order of magnitude ballpark as the market capitalisation of public firms. The World Bank reports US stockmarket capitalisation in 2015 as $25 trillion , making private equity 9.6% of this size. While this is real money, it is a small proportion, and the difficulty of getting a clear LBO versus venture capital measurement is consistently reported in the literature. Rather than introduce another crude estimator, we ignore it in the calculation below for a more conservative result.
Thirdly, for the size of venture capital, we use the National Venture Capital Association (NVCA) 2016 report . Figure 3.12 from that report provides annual levels of investment with some historical data and $59 billion for 2015.
|Investment Gap (Mason estimator)||84.1||87.1||90.5||97.6||102.9||108.2|
|Venture Capital investment||23.5||29.9||27.6||30.3||50.8||59|
|VC as percent of gap||27.9%||34.3%||30.5%||31.1%||49.4%||54.5%|
Table 1. Mason investment gap and venture capital investment (USD billions) and comparison, 2015 real terms.
I haven’t extended the table back further in time, deferring the grunt work of calculating real terms beyond 2010, because the more extensive VC data set shows the 2010-15 years to be historical maxima.
Venture capital investment over recent years covers at best half of the investment gap found by Mason. The only year which even approaches covering the entire annual gap would be the outlier year of 2000, the height of the dot-com boom, twice the next largest year (2015). Recent years also show relatively high VC investment by historical standards. This means both H1 and H2 are wrong: investment isn’t consistent across public and private US firms, but venture capital doesn’t make up the difference.
It’s true that tests of statistical significance on an expanded data set would be good, and would also yield measures of correlation and variance. The investment levels from Compustat / Flow of Funds and the NVCA series should make such an extension possible. Compustat doesn’t appear to be available outside an institutional context, though. An inflation adjusted time series based on open data would be technically possible from the flow of funds, that work is not done here.
Though venture capital investment accounts for only a fraction of the cash disgorged from public companies since 1985, it is a sizable fraction. Shareholders may have been using public companies as an ATM, but they did throw some of the cash they took out to engineers starting new companies in their garage. Some of it went to Google, some of it went to Juicero. The rest of the jetpack fund, they spent.
10% of investment loss a year is a funny middling number, enough to be billions of dollars of compounding effect, but hardly a collapse. It’s also US-specific. Beyond that, it’s worth noting that tech giants which began as startups have much higher rates of R&D intensity and other capital investment than their public company peers (eg Amazon 11.7% and Facebook 26.9% vs more typical values of 5%) . So an argument could be made that a long term structural transformation is going on which will ultimately bring more investment, finally returning to 1980 levels in 2045, after half a century. Even if true, it looks less like an unleashing of investment than recovery from a long debilitating illness.
 Mason, J. W. “Disgorge the cash: the disconnect between corporate borrowing and investment. The Roosevelt Institute. Link (2015).
 Piketty, Thomas. “Capital in the 21st Century.” Cambridge: Harvard Uni (2014).
 Mason, ibid.
 Mason, J. W. “Understanding Short-Termism.” (2015). The Roosevelt Institute. http://rooseveltinstitute.org/wp-content/uploads/2015/11/Understanding-Short-Termism.pdf
 Lazonick, William. “Profits without prosperity.” Harvard Business Review 92.9 (2014): 46-55.
 Sharpe, W. F. (1964). Capital asset prices: A theory of market equilibrium under conditions of risk. The journal of finance, 19(3), 425-442.
 Modigliani, F., & Miller, M. H. (1958). The cost of capital, corporation finance and the theory of investment. The American economic review, 48(3), 261-297.
 Black, F. (1976). The dividend puzzle. The Journal of Portfolio Management, 2(2), 5-8.
 Jensen, Michael C., and William H. Meckling. “Theory of the firm: Managerial behavior, agency costs and ownership structure.” Journal of financial economics 3.4 (1976): 305-360.
 MacKenzie, Donald. An engine, not a camera: How financial models shape markets. Mit Press, 2008.
 Jensen, Michael C. “Eclipse of the public corporation.” Harvard Business Review (Sept.-Oct. 1989), revised (1997).
 Bonen, Anthony. “Shareholders as Rent-Seekers: Institutional Realities of Corporate Governance and the Implications for Economic Theory.” (2015).
 Long, William F., and David J. Ravenscraft. “LBOs, debt and R&D intensity.” Strategic Management Journal 14.S1 (1993): 119-135.
 Lerner, J., Sorensen, M., & Strömberg, P. (2011). Private equity and long‐run investment: The case of innovation. The Journal of Finance, 66(2), 445-477.
 Amess, K., Stiebale, J., & Wright, M. (2015). The Impact of Private Equity on Firm’s Innovation Activity. Düsseldorf Institute for Competition Economics (DICE).
 US Patent Statistics Chart Calendar Years 1963-2015. Retrieved from http://www.uspto.gov/web/offices/ac/ido/oeip/taf/us_stat.htm
 Kortum, S., & Lerner, J. (2000). Assessing the contribution of venture capital to innovation. RAND journal of Economics, 674-692.
 Popov, A. A., & Roosenboom, P. (2009). Does private equity investment spur innovation? Evidence from Europe.
 Levine, M. (2015). Private Companies Will Take Money Public Companies Don’t Want. Bloomberg View . https://www.bloomberg.com/view/articles/2015-03-24/private-companies-will-take-money-public-companies-don-t-want
 Board of Governors of the Federal Reserve System, Z.1 Financial Accounts of the United States, First Quarter 2016.
 2016 Preqin Global Private Equity & Venture Capital Report, Sample, https://www.preqin.com/docs/reports/2015-Preqin-Global-Private-Equity-and-Venture-Capital-Report-Sample-Pages.pdf
 World Bank, World Federation of Exchanges database, http://data.worldbank.org/indicator/CM.MKT.LCAP.CD?end=2015&start=1975&view=chart
 National Venture Capital Association, Yearbook 2016, http://nvca.org/?ddownload=2963
 Price Waterhouse Coopers, 2016 Global Innovation 1000 Study, https://www.strategyand.pwc.com/innovation1000