One key component of the supply side secular stagnation is the notion that in recent years, corporate investment in the U.S. and other advanced economies have declined on a secular trend (or structurally). With low investment, there is low productivity growth and weak wages growth. The end result is not only lower economic growth, but also declining long term potential growth.
Since the thesis of supply side secular stagnation started making rounds in the economic policy literature, quite a few economists jumped into the debate proposing various explanations to the phenomena. To-date, however, there have not been an empirical study that looked at all reasonably plausible explanations on offer to assess which can account for the decline in capital investment.
German Gutierrez Gallardo and Thomas Philippon, in there paper “Investment-Less Growth: An Empirical Investigation” published this month by NBER do exactly that. The authors “analyze private fixed investment in the U.S. over the past 30 years.”
First, the authors establish that indeed, “investment is weak relative to measures of [firm] profitability and valuation – particularly Tobin’s Q, and that this weakness starts in the early 2000’s.” In other words, whilst firms remain profitable, they simply do not reinvest their profits at the same rate today as in the 1990s.
Per authors, there are “two broad categories of explanations: theories that predict low investment because of low Q, and theories that predict low investment despite high Q.”
As a reminder, Tobin’s Q is a ratio of total market value of the firm to total asset value of asset held by the firm. In simple terms, higher Q means that market value of the firm is higher relative to the cost of replacing the capital and other assets owned by the firm. Thus, a Q between 0 and 1 means that the cost to replace a firm's assets is greater than the value of its stock, so the stock is considered to be undervalued. A Q greater than 1 in contrast implies that a firm's stock is more expensive than the replacement cost of its assets, so the stock is overvalued.
So under the fist argument, if we observe low Q, firms are undervalued by the market and have no incentive to invest as they cannot raise capital for such investment from the markets that perceive the firm’s asset value to be already high (or above the firm value established in the market).
Under the second argument, something other than market valuations drives firm decision to invest or not. What that ‘something other’ is is a matter of various theories.
- Some theories postulate that in the presence of financial market imperfections (high costs, low liquidity supply, high risk premiums etc), low investments prevail even when Q is high (market value of the firm >> total assets value).
- Other theories, including the one that is currently most favoured as an explanation for dramatic decline in productivity growth in recent years (over the alternative explanation of the ‘secular stagnation’ thesis), the problem is that even with high Q, there might be low investment because there is mis-measurement in the markets as to the value of total assets of the firm. This can happen when there are intangible (hard to value) assets held by the firm, or when assets are dispersed across different currencies, markets and geographic, making them hard to value. It is worth noting that the argument of intangibles is commonly used today to argue that there is no real secular stagnation or decline in productivity growth because “things are simply not measured properly anymore”.
- Another view is that decreased competition (either due to technology - e.g. mega aggregators platforms such as google and apple, or due to regulation, or due to trade wars raging on, or broader trend of regionalisation of trade, etc) can reduce investment even in the times of higher Q (high market valuations).
- Finally, there is always a view that firms might under-invest because of short-termism in management strategies or due to restrictive investment climate induced by tighter risk governance (the latter point may overlap with regulatory constraints).
The authors find no support for the first argument. In other words, they find that low Q is not causing low investment. No surprise here, as markets are hardly in the mood of attaching low value to firms. In fact, we have been going through a massive uplift in M&As and equity valuations.
Which means that low investment is happening despite high market valuations - we are in the second set of arguments.
The authors “do not find support for theories based on risk premia, financial constraints, or safe asset scarcity”. They also find “only weak support for regulatory constraints.”
“Globalization and intangibles explain some of the trends at the industry level, but their explanatory power is quantitatively limited,” and does not provide support for aggregate - across economy - explanation of low investment.
So here comes the kicker: “we find fairly strong support for the competition and short-termism/governance hypotheses. Industries with less entry and more concentration invest less, even after controlling for current market conditions. Within each industry-year, the investment gap is driven by firms that are owned by quasi-indexers and located in industries with less entry/more concentration. These firms spend a disproportionate amount of free cash flows buying back their shares.”
Let’s sum this up: short-termism is a problem that holds firms from investing more, and it is more pronounced in industries with less competition. Firms which are owned by investors or funds that focus on indexing (pursue investment returns in line with broader indices, e.g. benchmarking to S&P500) invest less. The investment part of secular stagnation thesis, therefore, is linked at least indirectly to financialization of the economies: the greater is the weight of broad markets in investor decision-making, the lower the investment and the shorter is the time horizon, it appears.
Full paper: Gutierrez Gallardo, German and Philippon, Thomas, Investment-Less Growth: An Empirical Investigation (December 2016). NBER Working Paper No. w22897. https://ssrn.com/abstract=2880335