The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
IN V E S T M E N T C Y C L IC A L IT Y IN M A N U F A C T U R I N G IN D U S T R IE S B ruce C . Petersen and W illiam A . Strauss W o rkin g Paper Series M acro E co n o m ic Issues R esearch Departm ent Federal R eserve B an k o f C h ic a g o Septem ber, 1989 (W P -89-20 ) In v e s tm e n t c y c lic a lity in m a n u fa c tu r in g in d u s tr ie s Bruce C. Petersen and William A. Strauss* I i well known t a investment f uctuates proportionately by much more than ts ht l t t l output. The evidence on t i i quite dramatic. Consider for example the oa hs s r t o of net investment t GNP over the period 1946 to 1985. The lowest ai o values of t i r t o a l occurred during recession y hs ai l ears; while the mean of the r t o was 5.6 percent, the r t o was 2.9 percent i 1982, 3 percent i 1975, ai ai n .3 n 3.7 percent i 1983, and 4.0 percent i 1976. In c n r s , the r t o tends t be n n otat ai o high i boom periods.1 n In addition, investment closely tracks the business c ycle. This procyclicality of investment i extremely important i accounting for the " h r f l " of GNP s n sotal during downturns i the economy. Robert Barro's calculation of the n difference between actual GNP and a smoothly growing " o e t a " GNP ptnil s r e over the period 1946 t 1985 shows t a i a l categories of investment eis o ht f l are added together, fluctuations i investment account f 88 percent of the n or G NP s o t a l during recessions. hrfl Barro concludes t a "as a f r t ht is approximation, explaining recessions amounts t explaining the sharp o contractions i the investment components."2 n There are many competing views explaining why investment i so procyclical. s Among the most widely known hypotheses are the accelerator model; the neoclassical investment model, emphasizing the cost of ca i a and stock ptl adjustments; and the cash flow model under conditions of imperfect ca i a ptl markets. To d t , there i no widespread agreement on which view of ae s investment i most consistent with the f c s concerning the c c i a i y of s at ylclt investment. In t i a t c e we do not d r c l t s any of the competing theories of h s ril, i e t y et investment. Rather, we explore the c c i a i y of fixed investment a the ylclt t industry lev l within the manufacturing s c o . Very l t l a t ntion has been e etr ite t e given t examining investment a t i l v l The lack of information about o t hs ee. industry behavior i probably due t the f c t a investment studies s o at ht employing firm data typic l y do not have enough data points to produce al estimates of c c i a i y across a wide range of i d s r e . ylclt nutis There are some very basic questions concerning industry investment behavior t a must be addressed. Do a l broadly defined i dustries exhibit roughly the ht l n F R B C H IC A G O W orking P a p e r Septem ber 1989, W P-1989-20 1 same degree of investment c c i a i y over the business cycle? I no , i there ylclt f t s some obvious pattern i the data t a permits a useful organization of n ht industries according t t e r degree of c c i a i y There i no obvious o hi ylclt? s pattern i c c i a i y predicted by investment models t a focus only on the n ylclt ht cost of c p t l On the other hand, i i dustries do exhibit diffe e t investment aia. fn rn patterns over the business cycle, then theories emphasizing e t e firm- or ihr industry-specific determinants of investment may be required. To investigate industry c c i a i y we use a panel of 270 i d s ylclt, n u tries a the t f ur-digit Standard I d s r a Classification (SIC) level for the time period o nutil 1958 to 1986. For most of the issues explored i t i study, we aggregate t i n hs hs panel t the two-digit SIC level of disaggregation. W e find t a most of the o ht 20 two-digit industries do exhibit procyclical investment behavior over the period of our study. There a e however, marked differences across these r, industries both with respect t investment v l t l t and to investment o oaiiy c c i a i y Industries such as food products exhibit l t l or no investment ylclt. ite c c i a i y Our main finding i t a i dustries producing non-durable goods ylclt. s ht n exhibit l s c c i a i y i investment than i d stries producing durable goods. es ylclt n nu Very often the difference i quite s r k n . s tiig The remainder of the a t c e proceeds as follows: The next section br e l ril ify reviews alter a i e views of investment c c i a i y and some of the existing ntv ylclt evidence. The following section describes the panel database employed i the n study and the method used t construct "smoothed" industry investment o s r e . F n lly, we report our r s l s on both the v l t l t and c c i a i y of eis i a eut oaiiy ylclt industry investment. T h e o r ie s o f in v e s tm e n t c y c lic a lit y There are a number of investment theories t a predict t a investment should ht ht be a v l t l component of GNP. Space permits only a cursory overview of oaie three of the leading contenders; we describe here the predictions of the accelerator model, the neoclassical model, and the cash flow model.3 The accelerator model hypothesizes t a the lev l of net investment depends ht e on the change i expected demand for business output. According t t i n o hs theory, a business’ desired stock of c p t l varies dir c l with i s le e of s aia ety t vl output. Thus, when there i an "acceleration" i the economy and expected s n output increases, net investment i p s t v . The opposite occurs when there s oiie i a deceleration and net investment can actually become negative. s F R B C H IC A G O W orking P a p e r Septem ber 1989 , W P -1989-20 2 Depending on the s of the capital-output r t o investment can be several ize ai, times more v l t l and procyclical than output. oaie Neoclassical models have a t e r t c l advantage over the simple accelerator hoeia model i t a they include the cost of c p t l as one of the determinants of the n ht aia desired stock of c p t l and thus the lev l of investment. Some economists aia e explain the v l t l t of investment through the cost- f c p t l channel.4 oaiiy o-aia Their argument i essen i l y t a when the r a r t of i t r s changes, a l s tal ht el ae neet l firms experience a change i t e r desired stock of c p t l Given t a any n hi aia. ht y a ' investment amounts t a small portion of the t t l c p t l s o even a ers o o a a i a t ck, r l t v l small percentage change i the desired stock of c p t l can r s l i eaiey n aia eut n large percentage changes i net investment. Shocks t the r a i t r s r t n o el neet ae can cause firm investment t be very v l t l and industry investment t be o oaie o procyclical. The cash flow model also has a long t a i i n in the investment l t r t r . In rdto ieaue a world of perfect c p t l markets, sources of finance are ir e e a t for the aia rlvn investment decision. However, when there are imperfections i c p t l n aia markets, then i t r a finance generally has a cost advantage over external nenl finance. When t i i t u , then sources of finance do matter. In p r i u a , hs s re atclr the quantity of i t r a finance, or cash flow, should be positively associated nenl with the l v l of investment. Since firm p o i s and cash flows are very ee rft procyclical, the cash flow model of investment also predicts t a investment ht w l be procyclical. Furthermore, i predicts t a investment w l be more il t ht il procyclical f r indu t i s which experience the most procyclicality i p o i s o sre n rft. E v i d e n c e o n th e c y c l i c a l i t y o f in v e s tm e n t There i no widespread agreement on which of these theories i most s s consistent with the f c s concerning the c c i a i y of investment. Over the at ylclt l s three decades, a large number of empirical studies have been undertaken, at many of them with firm d t . An excellent review of the l t r t r before aa ieaue 1970 can be found i Kuh (1971). A review of some of the more recent n l t r t r can be found i Fazzari, Hubbard, and Petersen (1988). ieaue n Many of the e r i r empirical studies such as Kuh (1971), Meyer and Kuh ale (1957), and Meyer and Glauber (1964) focused on accelerator and cash flow models of investment, t pically finding some support f r both explanations. y o In the l s two decades, however, empirical research has s ifted toward at h neoclassical models of investment. The impetus for t i s i t i direction hs hf n F R B C H IC A G O W orking P a p e r Septem ber 1989, W P -1989-20 3 came from the i f u n i l work of Modigliani and Miller (1958) who nleta demonstrated t a under certain conditions, r a investment decisions can be ht el separated from purely f nancial f c o s t a i, t a financial fa t r such as i atr; ht s ht cos cash flow may be irrel v n t investment decisions. Whether t i separation eat o hs of r a investment from f el inancial considerations e i t in practice i silbeing xss s tl debated.5 A review of the empirical l t r t r on the determinants of investment reveals ieaue t a almost no studies systematically consider investment behavior a the ht t industry l v l Studies t pically use e t aggregate investment s r e for the ee. y i her eis whole economy or a sector of the economy or they use firm d t . Firm data aa has many advantages over aggregate data f r examining economic behavior. o However, most studies t a employ firm data do not have enough data points ht t permit estimates of differences i investment behavior across i d s r e . o n nutis This i probably the explanation for the paucity of studies t a compare the s ht investment behavior of a large number of i dustries for a s bstantial time n u period. There a e however, some potentially i t r s i g f c s t a can be learned by r, neetn at ht examining investment behavior a the industry l v l I i well known t a t ee. t s ht i d s r e , even within manufacturing, do not respond a nutis like t the business o c cle. For example, some i d s r e , such as those engaged i the processing y nutis n of food, experience very l t l variation i demand for t e r output over the ite n hi cycle. On the other hand, i dustries t a produce durable goods experience n ht considerable variation i demand and cash flow. n This r i e an i t r s i g t s of models of business investment. Models ass nee t n et which emphasize only the cost of c p t l do not predict systematic differences aia i investment c c i a i y across i d s r e . However, both the cash flow and n ylclt nutis the accelerator models c learly do. In the following sections of t i a t c e we h s ril, seek t s t out some of the f c s about differences i investment behavior a o e at n t the industry l v l ee. T h e d a ta The primary data sources u i i e i t i study are the Census o f Manufactures t l z d n hs and the Annual Survey o f Manufactures (U.S. Bureau of the Census). There are several reasons why these data sources are the best available for examining the c c i a i y of investment a the industry l v l F r t the Census ylclt t ee. is, reports investment data a the f u t o r digit l v l which i very disaggregated. ee, s F R B C H IC A G O W orking P a p e r Septem ber 1989, W P -1989-20 4 Second, Census data assign individual p a t , rather than whole companies, t lns o t e r primary SIC ind s r . Since plants are t pically much more specialized hi uty y than companies, the problem of contamination i ne l g b e F n l y the data s giil. ial, for most Census i dustries are available back t a l a t 1958, allowing for a n o t es panel of s b t u s antial l n t . egh The Census o f Manufactures currently contains approximately 455 fo r d g t u-ii i d s r e , of which 270 are included i our panel. Since, i i e t e nutis n t s ihr impossible or inconvenient t work with the e t r population of Census o nie i d s r e , we excluded in u t i s for any of three reasons. F r t because we nutis dsre is, wished t examine a balanced panel of i d stries covering as many business o nu cycles as possible, we excluded a l i d s l n u tries for which the Census of Manufactures began gathering data l t r than 1958. Second, we excluded a ae number of industries having large gaps i the d t . Fin l y we excluded n aa al, i dustries with inconsistencies i the industry c a s f c t o or d finition over n n lsiiain e time.6 Table 1 provides a summary of the breakdown of our sample of Census industries across the 20 two-digit manufacturing i d s r e . The f r t column nutis is l s s the i e t t of the 20 i dustries t a make up the Census o f Manufactures. it dniy n ht The second column l s s the t t l number of f u - i i indu t i s which made it oa ordgt sre up each of the two-digit Census i dustries i 1986. The t i d column reports n n hr the breakdown of our sample of i dustries across the two-digit i d s r e . The n nutis fourth column indicates the percentage of f u - i i in u t i s contained i ordgt dsre n our database. The f f h and s x h columns s a e what the average r a it it tt el investment (1982 d l a s was for each two-digit industry both for the Census olr) population and our sample of fo r d g t i d s r e . The f n l column u-ii nutis7 ia indicates the percentage of r a investment accounted for by our s t of el e idsre. nutis I can be e t asily ascertained from Table 1 t a our sample contains some 59.3 ht percent of the t t l number of four- i i indus r e currently contained i the oa dgt tis n Census. This percentage varies across two-digit i d s r e , the low being 25.3 nutis percent i SIC 24. Our coverage of t t l manufacturing investment i n oa s considerably higher; over t 1958-1986 period, our sample includes 77.2 he percent of a l investment. Again, t i percentage varies somewhat across the l hs two-digit c t a egories. F R B C H IC A G O W orking P a p e r Septem ber 1989 , W P-1989-20 5 Table 1 Investment breakdown by two-digit SIC code Total number of four-digit industries in 1986 Total manufacturing SIC 20 - Food and kindred products SIC 21 - Tobacco products SIC 22 - Textile mill products SIC 23 - Apparel and related products SIC 24 - Lumber and wood products SIC 25 - Furniture and fixtures SIC 26 - Paper and allied products SIC 27 - Printing and publishing SIC 28 - Chemicals and allied products SIC 29 - Petroleum and coal products SIC 30 - Rubber and plastic products SIC 31 - Leather and leather products SIC 32 - Stone, clay and glass products SIC 33 - Primary metal industries SIC 34 - Fabricated metal products SIC 35 - Machinery, except electrical SIC 36 - Electrical machinery SIC 37 - T ransportation equipment SIC 38 - Instruments and related products SIC 39 - Miscellaneous manufacturing Number of four-digit industries in FRB data base FRB data base, 1958-1986 average investment Percent of total Percent of total 455 270 59.3 57,453.6 44,322.1 77.1 47 38 4 30 4 19 80.9 5,124.1 100.0 63.3 314.9 1,726.6 4,463.2 314.9 1,375.3 100.0 79.7 33 15 17 13 4 7 45.5 602.8 305.0 50.6 23.5 53.8 1,618.5 521.1 984.7 258.1 60.8 49.5 17 17 11 64.7 3,938.3 3,602.9 91.5 33 8 16 47.1 48.5 2,363.2 7,625.1 1,348.8 4,585.7 57.1 60.1 1958-1986 average investment 87.1 6 5 83.3 2,994.3 2,994.3 100.0 6 4 66.7 1,992.1 1,705.4 85.6 11 3 27.3 142.7 47.4 27 23 85.2 2,389.5 2,281.8 26 36 44 18 19 29 69.2 52.8 65.9 5,736.6 3,076.3 5,287.8 5,097.7 1,970.2 4,187.1 33.2 95.5 88.9 64.0 79.2 37 25 67.6 4,522.3 2,848.8 63.0 18 8 44.4 5,607.5 4,919.5 87.7 13 7 53.8 1,256.5 895.2 71.2 20 7 35.0 613.4 136.1 22.2 C o n s t r u c t i n g t h e s m o o t h e d i n v e s t m e n t s e r ie s To examine investment c c i a i y we are going t compare i the next ylclt, o n section each i d s r ' actual investment s r e t a "smoothed" investment nutys eis o s r e , where the smoothed investment s r e i the average of recent eis eis s investment l v l . The logic of our approach i quite straightforward. I an ees s f industry’ actual investment tends t be above i s smoothed investment s r e s o t eis i boom times and below during economic contractions then actual n investment i cl a l procyclical. The degree of c c i a i y i measured by the s ery ylclt s extent t which actual investment deviates from "smoothed" investment o during economic expansions and contractions. F R B C H IC A G O W orking P a p e r Septem ber 1989, W P-1989-20 6 For comparison, we indexed the actual ( e l t d investment for a l two-digit dfae) l i d s r e , s t i g the value i 1958 a 100. To construct the smoothed nutis etn n t investment s r e , we chose the simplest possible technique t a would eis ht accomplish our o jective. The method used, known as a "centered moving b average smoothing" procedure, i given i Equation ( ) below: s n 1 2 2 where It i actual indexed investment i year t I t i the smoothed value of s n ; s indexed investment i year t\ and n i the number of years over which n s investment i averaged.8 W e experimented with alt r a i e values f n, s entv or s t l n on a value of nine as a compromise f r achieving the twin goals of etig o producing a smoothed investment s r e which also responds reasonably eis quickly to changes i the growth r t or trend i industry investment.9 n ae n Graphs of the actual and smoothed investment s r e appear below for a l eis l manufacturing and selected two-digit i d s r e . Figure 1p nutis lots the investment s r e for a l manufacturing over the time period 1958-1986. The actual eis l investment s r e i indicated by the s l d l n while the smoothed s r e i eis s oi ie eis s indicated by the dashed l n . Figures 2-5 report the same information for ie selected two-digit i d s r e . Figures 2-5 a l have the same v r i a scale t nutis l etcl o f c l t t cross-industry comparisons. The i d s aiiae n u tries are as follows: food and kindred products (SIC 20 chemicals and a l e products (SIC 28); i d s r a ); lid nutil machinery and equipment (SIC 3 ); and transportation equipment (SIC 3 ) 5 7. These i dustries have a large share of t t l investment i manufacturing, and n oa n as wi l become apparent, they i l s r t d f e e t types of industry investment l lutae i f r n behavior.10 An inspection of Figures 1 below indicates t a the procedure outlined i -5 ht n Equation ( ) appears t do a s t s a t r job of creating a smoothed 1 o aifcoy investment s r e for each i d stry. To see t i , compare the actual eis nu hs investment s r e f each industry with i s smoothed investment s r e . The e i s or t eis smoothed investment s r e picks up the trend i each i d s r ' investment eis n nutys s r e without being unduly affected by the fluctuations i the actual eis n investment s r e around i st e d eis t rn. F R B C H IC A G O W orking P a p e r Septem ber 1989, W P-1989-20 7 Figure 1 Indexed real investment (Total manufacturing) index, 1958=100 Figure 2 Indexed real investment (SIC 20: Food and kindred products) index, 1958=100 F R B C H IC A G O W orking P a p e r Septem ber 1989, W P -1989-20 8 Figure 3 Indexed real investment (SIC 28: Chemicals and allied products) Figure 4 Indexed real investment (SIC 35: Industrial machinery and equipment) index, 1958=100 F R B C H IC A G O W orking P a p e r Septem ber 1989, W P -1989-20 9 Figure 5 indexed real investment (SIC 37: Transportation equipment) index, 1958=100 In Figures 1-5, the differences between the actual investment series (solid line) and the smoothed investment series (dashed line) illustrate the cyclical behavior of investment. In Figure 1, for total manufacturing, the peaks and valleys in investment over the business cycles between 1958 and 1986 are quite evident. In addition, an inspection of Figures 2-5 indicates a wide range of cyclical investment behavior for SIC 20,28, 35, and 37. V o la t ili t y o f in d u s tr y in v e s tm e n t Before turning to the statistical results on the cyclicality of industry investment, it is of interest to report the differences in the volatility of industry investment. It is quite apparent from Figures 2-5 that some industries exhibit more volatile investment than others.^ To quantify this, we form the ratio of actual to smoothed investment ( I t / I t ) for each year for each industry and compute the coefficient of variation, reported in Table 2.11 Judging by the size of the coefficients, the industry with the most volatile investment series is the transportation industry (SIC 37), closely followed by the petroleum (SIC 29) and tobacco (SIC 21) industries. At the other end of F R B C H IC A G O W orking P a p e r Septem ber 1989 , W P-1989-20 10 the scale, the food industry (SIC 20) has a coefficient of variation about five times smaller than that of the transportation industry. When volatility is measured by output or sales, it is well known that transportation is one of the most volatile industries and that food is one of the least volatile industries. It is apparent from Table 2 that this is also true with respect to their investment. But, high volatility is not necessarily linked to high cyclicality, as we shall see in the next section. While it is linked in the case of the transportation industry, it definitely is not in the petroleum and tobacco industries. T h e c y c l i c a l i t y o f in d u s tr y in v e s tm e n t We turn now to the descriptive statistics on the cyclicality of industry investment. We fit the following relationship to each industry’s investment series: / 2 ) - = I a + bA t-\ + £/ t where I t is actual investment in year r, I t is the smoothed investment series discussed above; A is a measure of the state of the aggregate economy; and e is the error term. The measure of aggregate economic activity is lagged by one period because the peaks and troughs of the aggregate investment cycle typically lag slightly the peaks and troughs of aggregate GNP.12 We considered three alternative measures of A . One measure was the ratio of actual to potential GNP as measured by the Federal Reserve Board.13 A second measure was the ratio of current capacity utilization in manufacturing to average capacity utilization. The final measure was the ratio of the actual rate of unemployment to the natural rate of unemployment. All three measures have potential shortcomings. Fortunately, the results were qualitatively the same for all three measures. Therefore we report results for only the first measure and briefly summarize the results for the other two measures; that is, for each industry, we report results for the following regression: GNP / 3) F R B C H IC A G O W orking P a p e r Septem ber 1989, W P -1989-20 - = a+b /-i POTGNP + 8t t- 11 Table 2 Coefficient of variation of the investment ratio ___________________________________ Coefficient of variation Total manufacturing SIC SIC SIC SIC SIC SIC SIC SIC SIC SIC SIC SIC SIC SIC SIC SIC SIC SIC SIC SIC 20 - Food and kindred products 21 - Tobacco products 22 - Textile mill products 23 - Apparel and related products 24 - Lumber and wood products 25 - Furniture and fixtures 26 - Paper and allied products 27 - Printing and publishing 28 - Chemicals and allied products 29 - Petroleum and coal products 30 - Rubber and plastic products 31 - Leather and leather products 32 - Stone, clay and glass products 33 - Primary metal industries 34 - Fabricated metal products 35 - Machinery, except electrical 36 - Electrical machinery 37 - Transportation equipment 38 - Instruments and related products 39 - Miscellaneous manufacturing 9.9 4.8 22.2 14.2 11.7 18.0 15.7 13.3 11.3 12.9 22.5 18.0 22.2 14.7 18.1 11.8 13.6 12.3 23.2 16.2 12.5 Table 3 shows our findings for the manufacturing sector and its component two-digit industries for the regression given in Equation (2A). To economize on space, we do not report the intercept terms, which were statistically insignificant in all but one regression. For each industry, we report three statistics: the slope coefficient for the state of the economy variable, the standard error of the variable, and the adjusted r-square of the regression. We start with the obvious. For the manufacturing sector as a whole, the estimated coefficient is positive and significant at a very high confidence level. In other words, investment in manufacturing is procyclical. This is not a very surprising result; we would be hard pressed to explain a different finding. What is more interesting is that our regression results indicate that investment in manufacturing is more cyclical than aggregate GNP; our estimated coefficient of 2.23 implies that investment is approximately 2 percent above trend following a period when GNP is 1 percent above potential F R B C H IC A G O W orking P a p e r Septem ber 1989, W P -I989-20 12 GNP. In addition, it is interesting to note that our single regressor is explaining a considerable fraction (40 percent) of the variation of actual investment around trend investment. T a b le 3 Regression results: Investment ratio versus GNP ratio Slope Standard R-square ____________________________________Coefficient_______ error______ (adjusted) Total manufacturing SIC SIC SIC SIC SIC SIC SIC SIC SIC SIC SIC SIC SIC SIC SIC SIC SIC SIC SIC SIC 20 - Food and kindred products 21 - Tobacco products 22 - Textile mill products 23 - Apparel and related products 24 - Lumber and wood products 25 - Furniture and fixtures 26 - Paper and allied products 27 - Printing and publishing 28 - Chemicals and allied products 29 - Petroleum and coal products 30 - Rubber and plastic products 31 - Leather and leather products 32 - Stone, clay and glass products 33 - Primary metal industries 34 - Fabricated metal products 35 - Machinery, except electrical 36 - Electrical machinery 37 - Transportation equipment 38 - Instruments and related products 39 - Miscellaneous manufacturing 2.227 0.502 0.400 0.609 -1.361 1.530 1.825 1.928 1.296 2.129 1.710 1.903 0.976 2.320 1.228 2.299 3.368 2.389 3.022 2.248 3.789 2.528 0.760 0.296 1.450 0.889 0.687 1.138 1.014 0.786 0.683 0.769 1.478 1.105 1.448 0.880 1.008 0.635 0.686 0.691 1.360 0.959 0.813 0.104 -0.004 0.066 0.178 0.063 0.022 0.185 0.158 0.155 -0.021 0.108 -0.010 0.172 0.266 0.320 0.396 0.255 0.195 0.175 -0.005 We turn now to the two-digit industry results. A cursory look at the results indicates a considerable range of point estimates across the 20 industries. The smallest coefficient, -1.36, is for SIC 21 (tobacco products), while the second smallest is for SIC 20 (food products). At the other end of the scale, SIC 37 (transportation) has an estimated coefficient of 3.79, while the next largest coefficient is for SIC 33 (primary metals). For all but SIC 21 (tobacco) the point estimate for the slope coefficient is positive. Of these nineteen industries, all but three (SIC 20, SIC 29, and SIC 39) have estimated slope coefficients of greater than one. F R B C H IC A G O W orking P a p e r Septem ber 1989, W P -1989-20 13 We believe the most interesting finding of our research is the clean separation into two groups, with respect to cyclical investment behavior, of the 20 twodigit industries. The group consisting of SIC 20 through SIC 31 as well as SIC 39 (miscellaneous manufacturing) exhibits slope coefficients of less than the overall manufacturing average of 2.23. The other group, SIC 32 through SIC 38, exhibits slope coefficients greater than the manufacturing average; that is, they exhibit more procyclical investment than average. The first group, SIC 20 through SIC 31, can be characterized approximately as the nondurable-goods sector of manufacturing. With one exception, every one of these industries has an estimated slope coefficient of less than the all manufacturing coefficient of 2.23. For seven of these industries, the estimated standard error is large enough that one cannot reject the hypothesis at a 5 percent confidence level that investment is acyclical. For SIC 23, 26, 27, 28, and 30, the estimated coefficients are large enough to reject the hypothesis of acyclical investment behavior. However, one cannot conclude that their investment is more cyclical than GNP. Finally, it is interesting to note that while the previous section indicated that the petroleum (SIC 29) and tobacco (SIC 21) industries have very volatile investment series, they do not exhibit procyclical investment behavior. The other group, SIC 32 through SIC 38, consists of all durable-goods industries. All of these industries have slope coefficients greater than the manufacturing average, most noticeably for transportation (SIC 37), primary metals (SIC 33), and nonelectrical machinery (SIC 35). These three industries, along with fabricated metal products (SIC 34), have large enough coefficients relative to their standard errors such that one can reject the hypothesis that their slope coefficient is less than one. The transportation industry is particularly noteworthy, given the volatility of its investment series combined with its very high slope coefficient. The durable-goods sector has long been known to have more cyclical output than the nondurable-goods sector. It also appears to be the case that investment across virtually all of the durable-goods two-digit industries is more cyclical than investment in the nondurable-goods industries. This pattern of results was confirmed for all measures of aggregate economic activity that were used as regressors in Equation 2, including capacity utilization and unemployment. Results for capacity utilization appear in Table 4. F R B C H IC A G O W orking P a p e r Septem ber 1989, W P -1989-20 14 Table 4 R eg ressio n resu lts: In ve s tm e n t ratio v e rs u s c a p a c ity u tilizatio n ratio Ratio of capacity utilization divided by average capacity utilization (with one lag) Slope Standard R-square Coefficient_______ error______ (adjusted) Total manufacturing SIC SIC SIC SIC SIC SIC SIC SIC SIC SIC SIC SIC S IC SIC SIC SIC SIC SIC SIC SIC 20 - Food and kindred products 21 - Tobacco products 22 - Textile mill products 23 - Apparel and related products 24 - Lumber and wood products 25 - Furniture and fixtures 26 - Paper and allied products 27 - Printing and publishing 28 - Chemicals and allied products 29 - Petroleum and coal products 30 - Rubber and plastic products 31 - Leather and leather products 32 - Stone, clay and glass products 33 - Primary metal industries 34 - Fabricated metal products 35 - Machinery, except electrical 36 - Electrical machinery 37 - Transportation equipment 38 - Instruments and related products 39 - Miscellaneous manufacturing 1.069 0.227 0.430 0.252 -0.285 0.797 0.700 1.152 0.488 1.059 0.735 0.935 0.240 1.077 0.466 1.253 1.699 1.145 1.388 0.938 1.743 1.050 0.251 0.140 0.682 0.407 0.332 0.510 0.476 0.357 0.322 0.352 0.690 0.513 0.675 0.389 0.450 0.290 0.321 0.333 0.633 0.456 0.381 0.075 -0.030 0.092 0.110 0.128 0.002 0.217 0.130 0.178 -0.032 0.109 -0.019 0.252 0.321 0.343 0.387 0.199 0.191 0.133 -0.021 C o n c lu s io n Studies of investment typically use either aggregate investment numbers or firm level data. We believe, however, that useful knowledge can be obtained by examining the investment behavior at the industry level. Using a panel database of 270 four-digit industries over the period 1958-1986, we have examined the volatility and cyclicality of investment for all 20 of the two-digit Census o f M anufactures industries. We find that there is a great deal of heterogeneity across these industries. Some industries, such as transportation, petroleum, and tobacco, exhibit considerable investment volatility. We show, however, that industries which F R B C H IC A G O W orking P a p e r Septem ber 1989, W P-1989-20 15 have the most volatile investment series do not necessarily exhibit the most cyclical investment series. The major question that our article sought to answer is: Are there important differences in the cyclicality of investment across manufacturing industries? Our findings indicate that there are. With one exception, industries in SIC 20 through SIC 31 have estimated measures of cyclicality that are less than the manufacturing average for our sample. The remaining group of industries, SIC 32 through SIC 38, which consists of durable-goods manufacturers, appears to be more cyclical than the manufacturing average. The transportation industry leads the way followed by the primary metals and nonelectrical machinery. While it has long been known that the durable-goods sector has larger cyclical swings in output and profits than the nondurable-goods sector, it also appears that the durable-goods sector has larger cyclical swings in the accumulation of capital. Thus, our results shed some doubt on the view that our economy’s large swings in aggregate investment are primarily caused by firms' efforts to readjust their capital stocks in response to changes in real rates of interest. Models of investment that focus only on the cost of capital appear to be missing some important determinants of investment behavior. Given the well documented swings in output and profits in the durable-goods sector, the likely missing determinants are accelerator effects and internal finance considerations. F o o tn o te s *Bruce C. Petersen and W illiam A. Strauss are econom ists at the Federal Reserve Bank of Chicago. The authors thank Charles H m im elberg, Ed Nash, and Steve Strongin for com ents. m 1These values are taken from Barro (1987, p. 226), w hich contains a m detailed discussion of ore the facts concerning the cyclicality of aggregate investm ent. 2See Barro (1987, p. 229). ^For a m detailed discussion of these m ore odels of investm ent, see Gordon (1984) or Kopcke (1985). ^See for exam B ple arro (1987, p. 247). ^Recent papers which present evidence supporting the view that fluctuations in cash flow are an im portant source of fluctuation in investm include Fazzari, H ent ubbard, and Petersen (1988), H oshi, Kashap, and Scharfstein (1989), and Kopcke (1985). ^It is well known that the Census periodically changes the definitions of som industries, often by e m erging portions of one industry w pieces of another. This provides the biggest challenge to ith utilizing the Census of Manufactures. Since we did not w our findings to be biased by changes ant F R B C H IC A G O W orking P a p e r Septem ber 1989, W P-1989-20 16 in reported investm arising from industry reclassification, we thought it necessary to exclude ent all industries that underw a significant reclassification. M details on the construction of the ent ore panel can be found in Dom owitz, H ubbard, and Petersen (1986). 7The current dollar investm by two-digit SIG code industries w adjusted for inflation by ent ere dividing each of the series by the producer price index for capital goods. ^The centered m oving average approach that we utilized averages the data for the previous four years, the data for the current year and the data for the next four years. Of course, for the years near our endpoints, fewer years of data were available for com puting this average. See Pindyke and Rubinfeld (1981) for details. 9W e experim ented w different n values for Equation (1). and found that the results reported in ith the article are robust to a wide range of different values for n. ^Charts for the rem aining two-digit industries are available fromthe authors upon request. **The coefficient of variation is the ratio of the stan dard deviation to its m ean. The standard deviation is an absolute m easure of dispersion m easured in units of the original data. By contrast, the coefficient of variation is dim ensionless and m easures relative dispersion. l^We also considered contem poraneous A as well as A lagged by two years. The regression results for total m anufacturing, based on a considerably higher adjusted r-square, prefers A lagged by one period over contem poraneous A. At the two-digit industry level the results of contem poraneous versus one-year lag w roughly the sam H ever, for A w a two-year ere e. ow ith lag, there is no statistically significant relationship betw een investm and the two-year lagged ent state of the econom y. ^Potential GNP is from estim ates m ade by staff m bers of the B em oard of Governors. For the m ethodology underlying these estim ates see Clark (1982). R e fe r e n c e s Barro, Robert J., M a croecon o m ics , New York: John Wiley & Sons, 1987. Clark, Peter K., "Okun’s Law and Potential GNP,M oard o f G overn ors o f the B Federal R eserve System , October 1982. Domowitz, Ian, R. Glenn Hubbard, and Bruce C. Petersen, "Business Cycles and the Relationship Between Concentration and Price-Cost Margins," Rand Journal o f E con om ics, Vol. 17, No. 1, Spring 1986, pp. 1-17. Fazzari, Steven M., R. Glenn Hubbard, and Bruce C. Petersen, "Financing Constraints and Corporate Investment," B rookings P apers on E con om ic A ctivity , Vol. 1,1988, pp. 141-195. Gordon, Robert J., M a croecon o m ics , Boston: Little, Brown and Company, 1984. Hoshi, Takeo, Anil Kashap, and David Scharfstein, "Corporate Structure, Liquidity, and Investment: Evidence from Japanese Panel Data," Fed eral R eserve B oa rd Working P a p er , June 1988. F R B C H IC A G O W orking P a p e r Septem ber 1989, W P -1989-20 17 Kopcke, Richard W., “The Determinants of Investment Spending," Federal Reserve Bank of Boston, N ew England E conom ic R ev iew , July/August 1985, pp. 19-35. Kuh, Edwin, Capital Stock G row th : A M icro-E con om etric A pp roa ch , London: North-Holland Publishing Company, 1971. Meyer, John R., and Robert R. Glauber, Investment D ecision s, E conom ic F orecasting, and Public P o lic y , Boston: Division of Research, Graduate School of Business Administration, Harvard University, 1964. Meyer, John R., and Edwin Kuh, The Investment D ecisio n : An Empirical Study , Boston: Harvard University Press, 1957. Modigliani, Franco, and Merton H. Miller, "The Cost of Capital, Corporation Finance and the Theory of Investment," Am erican E conom ic R ev iew , Vol 48 June 1958, pp. 261-297. Pindyke, Robert S., and Daniel L. Rubinfeld, Econom etric M o d els and E con om ic F oreca sts , New York: McGraw-Hill Book Company, 1981. U.S. Department of Commerce, Census o f M anufactures, selected issues. , Annual Survey o f M anufactures , selected issues. F R B C H IC A G O W orking P a p e r Septem ber 1989, W P-1989-20 18 Federal Reserve Bank of Chicago R E S E A R C H S T A F F M E M O R A N D A , W O R K IN G P A P E R S A N D S T A F F S TU D IE S The following lists papers developed in recent years by the Bank’s research staff. Copies of those materials that are currently available can be obtained by contacting the Public Information Center (312) 322-5111. Working Paper Series—A series of research studies on regional economic issues relating to the Sev enth Federal Reserve District, and on financial and economic topics. Regional Economic Issues Donna Craig Vandenbrink “The Effects of Usury Ceilings: the Economic Evidence,” 1982 David R. Allardice “ Small Issue Industrial Revenue Bond Financing in the Seventh Federal Reserve District,” 1982 WP-83-1 William A. Testa “Natural Gas Policy and the Midwest Region,” 1983 WP-86-1 Diane F. Siegel William A. Testa “Taxation of Public Utilities Sales: State Practices and the Illinois Experience” WP-87-1 Alenka S. Giese William A. Testa “Measuring Regional High Tech Activity with Occupational D ata” WP-87-2 Robert H. Schnorbus Philip R. Israilevich “Alternative Approaches to Analysis of Total Factor Productivity at the Plant Level” WP-87-3 Alenka S. Giese William A. Testa “Industrial R & D An Analysis of the Chicago Area” WP-89-1 William A. Testa “ Metro Area Growth from 1976 to 1985: Theory and Evidence” WP-89-2 William A. Testa Natalie A. Davila “Unemployment Insurance: A State Economic Development Perspective” WP-89-3 Alenka S. Giese “A Window o f Opportunity Opens for Regional Economic Analysis: BEA Release Gross State Product D ata” WP-89-4 Philip R. Israilevich William A. Testa “Determining Manufacturing Output for States and Regions” WP-89-5 Alenka S.Geise “The Opening o f Midwest Manufacturing to Foreign Companies: The Influx of Foreign Direct Investment” WP-89-6 Alenka S. Giese Robert H. Schnorbus “A New Approach to Regional Capital Stock Estimation: Measurement and Performance” •WP-82-1 ••WP-82-2 ^Limited quantity available. ♦♦Out of print. Working Paper Series (cont'd) WP-89-7 William A. Testa “Why has Illinois Manufacturing FaBen Behind the Region?” WP-89-8 Alenka S. Giese William A. Testa “ Regional Specialization and Technology in Manufacturing” WP-89-9 Christopher Erceg Philip R. Israilevich Robert H. Schnorbus “Theory and Evidence of Two Competitive Price Mechanisms for Steel” WP-89-10 David R. Allardice William A. Testa “ Regional Energy Costs and Business Siting Decisions: An Illinois Perspective” Issues in Financial Regulation W P-89-11 Douglas D. Evanoff Philip R. Israilevich Randall C. Merris “Technical Change, Regulation, and Economies o f Scale for Large Commercial Banks: An Application of a Modified Version of Shepard’s Lemma” WP-89-12 Douglas D. Evanoff “ Reserve Account Management Behavior: Impact of the Reserve Accounting Scheme and Carry Forward Provision” WP-89-14 George G. Kaufman “ Are Some Banks too Large to Fail? Myth and Reality” W P-89-16 Ramon P. De Gennaro James T. Moser “ Variability and Stationarity of Term Premia” WP-89-17 Thomas Mondschean “ A Model of Borrowing and Lending with Fixed and Variable Interest Rates” WP-89-18 Charles W. Calomiris “D o Vulnerable" Economies Need Deposit Insurance?: Lessons from the U.S. Agricultural Boom and Bust of the 1920s” Macro Economic Issues W P-89-13 David A. Aschauer “ Back o f the G -7 Pack: Public Investment and Productivity Growth in the Group of Seven” WP-89-15 Kenneth N . Kuttner “Monetary and Non-Monetary Sources o f Inflation: An Error Correction Analysis” WP-89-19 Ellen R. Rissman “Trade Policy and Union Wage Dynamics” WP-89-20 Bruce C. Petersen William A. Strauss “Investment Cyclicality in Manufacturing Industries” ♦Limited quantity available. ♦♦Out of print. 3 StafT Memoranda—A series o f research papers in draft form prepared by members of the Research Department and distributed to the academic community for review and comment. (Series discon tinued in December, 1988. Later works appear in working paper series). ••SM -81-2 George G. Kaufman “ Impact of Deregulation on the Mortgage Market,” 1981 ♦ •SM-81-3 Alan K. Reichert “An Examination of the Conceptual Issues Involved in Developing Credit Scoring Models in the Consumer Lending Field,” 1981 Robert D. Laurent “A Critique of the Federal Reserve’s New Operating Procedure,” 1981 George G. Kaufman “ Banking as a Line of Commerce: The Changing Competitive Environment,” 1981 SM-82-1 Harvey Rosenblum “Deposit Strategies of Minimizing the Interest Rate Risk Exposure of S&Ls,” 1982 •SM-82-2 George Kaufman Larry Mote Harvey Rosenblum “ Implications of Deregulation for Product Lines and Geographical Markets of Financial Instititions,” 1982 •SM-82-3 George G. Kaufman “The Fed’s Post-October 1979 Technical Operating Procedures: Reduced Ability to Control M oney,” 1982 SM-83-1 John J. Di Clemente “The Meeting of Passion and Intellect: A History of the term ‘Bank’ in the Bank Holding Company Act,” 1983 SM-83-2 Robert D. Laurent “Comparing Alternative Replacements for Lagged Reserves: Why Settle for a Poor Third Best?” 1983 ♦ ♦ SM-83-3 G. O. Bierwag George G. Kaufman “A Proposal for Federal Deposit Insurance with Risk Sensitive Premiums,” 1983 •SM-83-4 Henry N. Goldstein Stephen E. Haynes “A Critical Appraisal of McKinnon’s World Money Supply Hypothesis,” 1983 SM-83-5 George Kaufman Larry Mote Harvey Rosenblum “The Future of Commercial Banks in the Financial Services Industry,” 1983 SM-83-6 Vefa Tarhan “ Bank Reserve Adjustment Process and the Use of Reserve Carryover Provision and the Implications of the Proposed Accounting Regime,” 1983 SM-83-7 John J. Di Clemente “The Inclusion of Thrifts in Bank Merger Analysis,” 1983 SM-84-1 Harvey Rosenblum Christine Pavel “ Financial Services in Transition: The Effects of Nonbank Competitors,” 1984 SM-81-4 ♦ ♦ SM-81-5 ’•'Limited quantity available. **Out of print. Staff Memoranda ( co n i'd ) SM-84-2 George G. Kaufman “The Securities Activities of Commercial Banks,” 1984 SM-84-3 George G. Kaufman Larry Mote Harvey Rosenblum “Consequences of Deregulation for Commercial Banking” SM-84-4 George G. Kaufman “The Role of Traditional Mortgage Lenders in Future Mortgage Lending: Problems and Prospects” SM-84-5 Robert D. Laurent “The Problems of Monetary Control Under Quasi-Contemporaneous Reserves” SM-85-1 Harvey Rosenblum M . Kathleen O ’Brien John J. Di Clemente “On Banks, Nonbanks, and Overlapping Markets: A Reassessment of Commercial Banking as a Line of Commerce” SM-85-2 Thomas G. Fischer William H. Gram George G. Kaufman Larry R. Mote “The Securities Activities of Commercial Banks: A Legal and Economic Analysis” SM-85-3 George G. Kaufman “ Implications of Large Bank Problems and Insolvencies for the Banking System and Economic Policy” SM-85-4 Elijah Brewer, III “The Impact of Deregulation on The True Cost of Savings Deposits: Evidence From Illinois and Wisconsin Savings & Loan Association” SM-85-5 Christine Pavel Harvey Rosenblum “ Financial Darwinism: Nonbanks— and Banks—Are Surviving” SM-85-6 G. D. Koppenhaver “Variable-Rate Loan Commitments, Deposit Withdrawal Risk, and Anticipatory Hedging” SM-85-7 G. D. Koppenhaver “A Note on Managing Deposit Flows With Cash and Futures Market Decisions” SM-85-8 G. D. Koppenhaver “ Regulating Financial Intermediary Use o f Futures and Option Contracts: Policies and Issues” SM-85-9 Douglas D. Evanoff “The Impact of Branch Banking on Service Accessibility” SM-86-1 George J. Benston George G. Kaufman “ Risks and Failures in Banking: Overview, History, and Evaluation” SM-86-2 David Alan Aschauer “The Equilibrium Approach to Fiscal Policy” ♦Limited quantity available. ♦♦Out of print. 5 Staff Memoranda (cont'd) SM-86-3 George G. Kaufman “Banking Risk in Historical Perspective” SM-86-4 Elijah Brewer III Cheng Few Lee “The Impact of Market, Industry, and Interest Rate Risks on Bank Stock Returns” SM-87-1 Ellen R. Rissman “Wage Growth and Sectoral Shifts: New Evidence on the Stability of the Phillips Curve” SM-87-2 Randall C. Merris “Testing Stock-Adjustment Specifications and Other Restrictions on Money Demand Equations” SM-87-3 George G. Kaufman “The Truth About Bank Runs” SM-87-4 Gary D. Koppenhaver Roger Stover “On The Relationship Between Standby Letters of Credit and Bank Capital” SM-87-5 Gary D. Koppenhaver Cheng F. Lee “Alternative Instruments for Hedging Inflation Risk in the Banking Industry” SM-87-6 Gary D. Koppenhaver “The Effects of Regulation on Bank Participation in the Market” SM-87-7 Vefa Tarhan “ Bank Stock Valuation: Does Maturity Gap Matter?" SM-87-8 David Alan Aschauer “ Finite Horizons, Intertemporal Substitution and Fiscal Policy” SM-87-9 Douglas D. Evanoff Diana L. Fortier “ Reevaluation of the Structure-ConductPerformance Paradigm in Banking” SM-87-10 David Alan Aschauer “Net Private Investment and Public Expenditure in the United States 1953-1984” SM-88-1 George G. Kaufman “ Risk and Solvency Regulation of Depository Institutions: Past Policies and Current Options” SM-88-2 David Aschauer “Public Spending and the Return to Capital” SM-88-3 David Aschauer “Is Government Spending Stimulative?” SM-88-4 George G. Kaufman Larry R. Mote “ Securities Activities of Commercial Banks: The Current Economic and Legal Environment" SM-88-5 Elijah Brewer, III “A Note on the Relationship Between Bank Holding Company Risks and Nonbank Activity” SM-88-6 G. O. Bierwag George G. Kaufman Cynthia M. Latta “ Duration Models: A Taxonomy” G . O. Bierwag George G. Kaufman “ Durations of Nondefault-Free Securities” •Limited quantity available. ••Out of print. 6 Staff Memoranda (cont'd) SM-88-7 David Aschauer “ Is Public Expenditure Productive?” SM-88-8 Elijah Brewer, III Thomas H. Mondschean “Commercial Bank Capacity to Pay Interest on Demand Deposits: Evidence from Large Weekly Reporting Banks” SM-88-9 Abhijit V. Baneijee Kenneth N. Kuttner “Imperfect Information and the Permanent Income Hypothesis” SM-88-10 David Aschauer “Does Public Capital Crowd out Private Capital?” SM -88-11 Ellen Rissman “Imports, Trade Policy, and Union Wage Dynamics” Staff Studies—A series o f research studies dealing with various economic policy issues on a national level. SS-83-1 **SS-83-2 Harvey Rosenblum Diane Siegel “Competition in Financial Services: the Impact of Nonbank Entry,” 1983 Gillian Garcia “Financial Deregulation: Historical Perspective and Impact o f the Garn-St Germain Depository Institutions Act o f 1982,” 1983 •Limited quantity available. **Out of print.
@FedFRASER