Over the weekend, both Professors Barro and Mankiw wrote on investment in the New York Times. As Modeled Behavior observed, the focus on business fixed investment (BFI) or nonresidential investment was somewhat odd because BFI behavior had not been particularly anomalous in the recovery. Here, I extend upon that analysis, and draw some policy implications.
Investment Behavior in Context
Figure 1 below highlights the depth of the last recession, and the dropoff in nonresidential fixed investment.
Figure 1: Four quarter growth rate in GDP (dark blue) and in nonresidential fixed investment (red). Growth rates calculated as log differences. NBER defined recession dates shaded gray. Source: BEA, 2011Q2 2nd release, NBER and author’s calculations.
Consider the relationship between q/q growth rate of investment on current and three lags of q/q output growth over periods of expansion:
Δ bfi t = -0.001 + 1.253 Δ y t + 0.188 Δ y t-1 + 0.332 Δ y t-2 + 0.417 Δ y t-3 + u t (1)
Adj. R2 = 0.22, SER = 0.015, Breusch-Godfrey LM(2) = 3.18 [p-val.=0.204]. Bold face denotes significance at 10% level.
Whatever instability exists in this relationship, it is not evident in the most recent expansion:
Figure 2: One step ahead recursive residuals test for equation (1). Source: Author’s calculations.
The instability seems more marked earlier in the sample, in the 1980’s. A CUSUM test also fails to reject stability overall.
In fact, it is interesting that nonresidential investment has performed better in this recovery from recession as opposed to the preceding recovery from the 2001Q1-2001Q4 recession.
Figure 3: Log difference of nonresidential investment and GDP, versus 2009Q2 (blue) and versus 2001Q4 (dark red). Source: BEA, 2011Q2 2nd release and author’s calculations.
One might want to take into account the fact that the recession has been particularly deep. I have normalized on the beginning of the downturn (2008Q2 and 2000Q4) instead of the trough.
Figure 4: Log difference of nonresidential investment and GDP, versus 2008Q2 (blue) and versus 2000Q4 (dark red). Source: BEA, 2011Q2 2nd release and author’s calculations.
Hence, taking into account the depth of the recession, one finds that BFI is doing better than in the recovery from the previous recession (as well as the previous recession+recovery). It makes one wonder about this argument that great regulatory uncertainty is dampening business enthusiasm for capacity increases. After all, this bivariate logic would imply that regulatory uncertainty was greater in the years after the first Bush recession, relative to now.
As noted in several previous posts, investment in the previous recovery from the first Bush recession (2001Q1-01Q4) was itself lackluster   . Hence, one needs to take seriously the arguments regarding how to encourage more rapid investment.
Here is where I found the two op-eds this weekend of interest. Professor Mankiw notes the demand side motivation for investment, while stressing the user cost of capital:
The relationship between investment and the overall economy is what an engineer would call a positive feedback loop. Greater business investment would increase hiring, both by those who produce the investment goods and those who buy them. Greater employment would mean more workers taking home paychecks, which in turn would increase the overall demand for goods and services. When businesses saw more customers coming through their doors, they would then increase investment spending yet again.
WHAT can policy makers do to stoke animal spirits and encourage businesses to invest?
One obvious step would be a cut in the taxation of income from corporate capital. According to a 2008 study by the Organization for Economic Cooperation and Development, “Corporate taxes are found to be most harmful for growth.” Tax reform that reduced the burden on capital income and shifted it toward consumption would improve prospects for long-run growth and, in so doing, encourage greater investment today.
I had a dream that Mr. Obama and Congress enacted this fiscal reform package — triggering a surge in the stock market and a boom in investment and G.D.P. — and that he was re-elected.
His fiscal reform package includes abolishing the corporate and estate taxes, implementing a VAT, and cutting government spending. He also argues for reducing regulatory uncertainty
I think all of these factors in principle could be determinants of investment. The question is really the magnitude, empirically, of these effects. One way of assessing the relative importance of each variable is to conduct a horse race of the various models; I cited a Kopcke study from 2001 in this post.
we run horse races involving a number of forecasting models of US real fixed private
non-residential investment spending growth over the 1995:1–2004:2 out-of-sample period, a volatile
period marked by an investment ‘boom and bust’ cycle. The forecasting models we consider are
based on the familiar Accelerator, Neoclassical, Average Q, and Cash-Flow models of investment
spending, and we also consider two forecasting models suggested by the more recent work of Barro
(1990) and Lettau and Ludvigson (2002). The Average Q model produces the most accurate forecasts
at the 1-quarter horizon, while the Barro (1990) Stock Price model generates the most accurate
forecasts at horizons beyond 1 quarter. At longer forecast horizons, forecast encompassing tests
indicate that the Stock Price model contains most of the information useful for forecasting US business
fixed investment spending growth for the 1995:1–2004:2 out-of-sample period relative to the
other models. These results point to an important predictive role for the stock market with respect
to the recent course of US business fixed investment spending growth. While stock prices appear
important in forecasting the recent behavior of US business investment spending growth, robustness
checks show that other variables are often more important over the 1975:1–1984:4 and
1985:1–1994:4 out-of-sample periods. Furthermore, we obtain extensive in-sample evidence of multiple structural breaks in all of the forecasting models. These results question whether the Stock Price model will continue to produce the most accurate forecasts outside the 1995:1–2004:2 period, and they suggest that structural instability will make it difficult in general to forecast US business fixed investment spending growth using conventional and several recently proposed models.
Barro’s assertion that a stock price boom would induce an investment increase seems at least somewhat validated by this analysis. However, the outperformance is not uniform, as the authors note. And it is important to highlight that while Barro asserts fiscal reform would cause the boom in the stock market, that particular step is more conjecture (hope?) than analysis.
Recall in the risk neutral version of stock price valuation model, the stock price is the present discounted value of dividends and the discount rate used for equities . At some level, the amount of dividends paid out depends on profits, and these in turn will depend on the level of economic activity. Hence, if fiscal reform incorporates contractionary fiscal policy, one may very well get a stock market decline (so I am arguing against expansionary fiscal contraction)    .
Now, it could be that fiscal reform that induces lower interest rates and hence a lower discount rate for equities that results in a stock market boom. But interest rates, both nominal and real, are at quite low rates. Hence, the scope for fiscal consolidation to effect a stock market boom via real rates over the relevant period seems limited.
Professor Barro also stresses eliminating regulatory uncertainty as an impediment to investment. It could be that regulatory uncertainty is depressing stock prices (and hence investment). It would be interesting to see this effect quantified, in a manner that controlled for concurrent macro uncertainty.
By the way, the fact that investment responds to real stock prices most, rather than the user cost of capital key in the neoclassical model, suggests that further tax measures are unlikely to have a big impact. That’s not an exact formulation, since some tax provisions that affect the user cost of capital can affect after tax cash flow and hence stock prices. But the fact that much movement in stock prices has occurred with high corporate-to-GDP ratios suggests that this link, if it exists, is not very operative at the moment. At the very least, fantastical effects used in the Heritage Foundation’s Center for Data Analysis “analysis” of the Ryan Plan     are unlikely to materialize. (Where is William Beach these days?)