<<

c COPYRIGHT

by

John S. Schuler

2017

ALL RIGHTS RESERVED ii

A STUDY OF THE CANTILLON EFFECT by John S. Schuler

ABSTRACT

There is economic evidence for monetary neutrality in the long run but not the short run. Rather, in the short run, monetary shocks tend to increase the variance of changes in prices. This is a study of this “Cantillon Effect” using methods of multiple time series and analysis of variance. iii

ACKNOWLEDGEMENTS

I would like to acknowledge the generous support of the National Science Founda- tion. Further, I would like to acknowledge the help of my committee: Michael Baron, Jun

Lu, and Elizabeth Malloy. iv

TABLE OF CONTENTS

ABSTRACT ...... ii

ACKNOWLEDGEMENTS ...... iii

LIST OF TABLES ...... vi

LIST OF FIGURES ...... vii

CHAPTER

1. Introduction ...... 1

1.1 Literature on the Cantillon Effect ...... 3

1.2 Economic Equilibrium, the Quantity Theory of , and the Can- tillon Effect ...... 4

1.3 Equilibrium ...... 4

1.4 The Quantity Theory of Money ...... 5

1.5 The Statistical Problem ...... 7

2. Data and Analysis ...... 8

2.1 Description of the Data ...... 8

2.2 Description of the Statistical Model ...... 9

2.2.1 Construction of the Autoregressive System ...... 12

2.3 Multivariate Time Series Model ...... 12

2.4 Impulse Response Function ...... 17

2.4.1 Choleski Decomposition: Theory ...... 18

3. Results ...... 20

3.1 Tests of Equality of Variances ...... 21 v

3.2 Probability Distribution of the Impulse Response of Prices ...... 21

4. Conclusion ...... 25

REFERENCES ...... 62 vi

LIST OF TABLES

Table Page

2.1 Unit Root Tests ...... 14

2.2 Coefficient Estimates ...... 17

2.3. Correlation Matrix ...... 18

3.1. Standard Deviations of % Price Changes (1 Standard Deviation Shock) . . 22

3.2. Levene’s Test for Difference in Variance of % Price Changes (1 Standard Deviation Shock) ...... 22

3.3. Standard Deviations of % Price Changes (1.5 Standard Deviation Shock) . 22

3.4. Levene’s Test for Difference in Variance of % Price Changes (1.5 Standard Deviation Shock) ...... 23

3.5. Standard Deviations of % Price Changes (2 Standard Deviation Shock) . . 23

3.6. Levene’s Test for Difference in Variance of % Price Changes (2 Standard Deviation Shock) ...... 23

3.7. Standard Deviations of % Price Changes (2 Standard Deviation Shock) . . 23

3.8. Levene’s Test for Difference in Variance of % Price Changes (2 Standard Deviation Shock in M2) ...... 23

3.9. Standard Deviations of % Price Changes (2 Standard Deviation Shock in Industrial Production) ...... 24

3.10. Levene’s Test for Difference in Variance of % Price Changes (2 Standard Deviation Shock in Industrial Production) ...... 24 vii

LIST OF FIGURES

Figure Page

1.1. Supply and Demand ...... 4

2.1. Macroeconomic Indicator Series ...... 10

2.2. Distribution of Month over Month % Change in Price ...... 11

3.1. Distribution of Impulse Responses ...... 22 1

CHAPTER 1

INTRODUCTION

The question of monetary neutrality has occupied macroeconomists for a long time.

The concept goes back to the writings of (Patinkin, 1989; Lucas Jr, 1996).

The term “monetary neutrality” itself was introduced by FA Hayek (Hayek, 1932). To claim that money is neutral is to claim that money has no direct effect on the productive capacity of the economy. In other words, while a monetary system is required to sustain highly specialized trade, its actual details of the monetary system are independent of the patterns of specialization that emerge in an economy and the of that economy. Economists have long debated the extent to which the hypothesis of monetary neutrality holds.

Any number of theories of the boom and bust cycle are theories of short-run mon- etary non-neutrality (Hayek, 1932; Keynes, 2016; Shah, 1997). Most discussion among economists focuses on the question of whether money is neutral in the long run. The evidence for this proposition is mixed. There is evidence against long run non-neutrality in Indonesia, Taiwan, and Thailand (Puah et al., 2008). Proponents of monetary non- neutrality could cite evidence from New Zealand (Robertson and Orden, 1990). Practi- cally all of the work on this topic leverages methods of multiple time series. This raises questions about the robustness of various specifications and dependence on unit root assumptions (Poitras, 1997). These questions will become important later. 2

There is another perennial question in relevant to our present pur- pose. Most macroeconomic models involve relationships among macroeconomic variables; that is, among statistical summaries which are assumed to be representative of the entire economy and thus have no direct relationship to any particular individual, firm, or orga- nization within the economy. Thus, a natural question involves whether theories hold up at lower levels of aggregation; perhaps industry by industry as opposed to the whole econ- omy. Gauger (1988) reviews the evidence on industry by industry monetary neutrality.

The present project is a contribution to the literature in two ways. Firstly, it provides evidence for a particular model of short-run price dynamics and therefore short-run mon- etary non-neutrality. It also bears on questions related to the suitability of methods of multiple time series for this sort of analysis. Finally, it suggests the direction for future work on industry by industry non-.

The Cantillon effect, named for is an anomaly occurring in eco- nomics. Classical price theory predicts that given a fixed quantity of goods being bought and sold, an increase in the quantity of circulating money will raise prices. This does hold in the long run. Of course in reality, prices are not fixed but rather move somewhat from month to month. Our main empirical finding is that in the aftermath of a monetary shock, there is an increase in the variance of the month over month percentage change in prices. Our method has a few main parts. Firstly, we will build a multiple time series model of prices to test the appropriateness and feasibility of this way of modeling price dynamics. This model may be re-expressed so as to allow us to define monetary shocks in terms of the probability distribution of month over month changes in the Federal Funds

Rate. A monetary shock of size k in our sense refers to a month over month change in the Federal Funds rate occupying the tails of this distribution defined by that magnitude k.

The present paper is partially a replication of Balke and Wynne (2007). In that 3 paper, they give a method for the detection of monetary shocks and give evidence that the probability distribution of month over month price changes is altered substantially in the aftermath of these shocks. We have adopted their method for the detection of these shocks but made a few changes to it. Their time series model contains seasonal effects without any indications that multiple specifications were tested. The result is a highly complex model with no justification of its complexity. We have retained their 12 lags to fit each monthly period on a year of data. Secondly, Balke and Wynne (2007) build Gaussian models of the distribution of price changes as a way to detect a change in variance. Since the distribution of price changes is not even remotely Gaussian, we have tested the equality of these variances directly. The analysis of Balke and Wynn only extends through about 2003. We have an additional decade of data on which to corroborate their results.

1.1 Literature on the Cantillon Effect

The literature most relevant to our present purposes involves that which studies heterogeneity in the effects of money across sectors. For instance, there is evidence that the responses of prices to monetary shocks is not uniform across sectors even in the long run (Lastrapes (2004)). This is calls the very concept of a into question. There is also evidence of widely varying mechanisms by which price changes propagate through the economy (Baumeister et al., 2013). Further, the choice of tools appears to affect relative price variability (Berument et al., 2009). Additionally, Balke

(2012) and Karim et al. (2011) give further evidence of price change heterogeneity in the

United States and Malaysia respectively.

It is an open question as to whether it is better when forecasting inflation to use aggregate measures or individual prices. This question revolves around the statistical properties of the aggregates and the extent to which they capture the variation of . 4

In any case, heterogeneity of prices complicates these measurements (Demers et al., 2005).

In any case, there is also microeconomic evidence for asymmetries in price adjustment by individual firms in the economy (Rather et al., 2015). This confirms a theoretical hypothesis laid out in (Ball and Mankiw, 1994).

1.2 Economic Equilibrium, the Quantity Theory of Money, and the Cantillon Effect

The purpose of this section is to familiarize the non-economist reader with the basics of the Economic theory. There are three main topics as stated in the title. We will consider these each in turn.

1.3 Equilibrium

Figure 1.1. Supply and Demand

All who have taken an introductory course are familiar with supply and demand curves as displayed in Figure 1.1. The basic intuition is that for good α, we 5

may express the demand for good α as a function of the price of α, Dα = gα (pα) where

0 consumers demand fewer units of this good as price increases, mathematically, Dα (pα) ≤ 0. We may express upward sloping supply curves analogously. These curves are defined over the entire range of prices but are not necessarily smooth. The most basic requirement is the demand curves do not jump up and supply curves do not fall down. The price at which these two curves intersect is known as the market or equilibrium price. If we are simultaneously considering all goods in the economy, we find there is a price vector

~p for which all markets clear. This is known as a general equilibrium (Kreps, 2012). An important aspect of the price vector which a general equilibrium yields is that it is not unique. If ~p is market clearing, so is β~p for β > 0. Since general equilibrium is believed to be a reasonable model of the economy in the long run– that is, at any given time, prices are adjusting in an attempt to approximate it, it may be used as a justification for long-run monetary neutrality- the idea that money has no long run effect on the (Caplin and Spulber, 1987).

1.4 The Quantity Theory of Money

At base the quantity theory of money refers to the tautologous equation:

P y = MV where P is the price level, V is the “velocity” of money, y is the inflation adjusted output of the economy, that is, all produced goods and services multiplies by their price level

(Friedman, 2010). It is this quantity that GDP attempts to measure. M is the money

P y supply, roughly, the number of dollars in circulation. Velocity is defined as V = M 1 but it can be conceived of as the frequency with which money changes hands. P is the purchasing power of the dollar. If prices are quoted in dollars, then the more money required to buy a bundle of goods, the lower the purchasing power and the greater the price level. The important aspect here is that, if we assume that velocity is constant, 6 an increase in the will raise the price level and thus lower the purchasing power of the dollar. This does happen in the long run. It is our contention that in the short run, this equation does not hold. Rather, when there has not been sufficient time to reach equilibrium, the very concept of a price level does not apply. To understand this, recall that prices are relative. Suppose an apple is worth two oranges in equilibrium. If a monetary shock sends the price of apples up, this relative price will change temporarily.

In the long run, provided economic fundamentals have not changed, the price of apples will return to twice the price of oranges.

There is a as for other goods. Indeed, if the money market is in equilibrium, then money demanded equals money supplied; that is Md = Ms. Economic theory tends to assume that the money supply is decided by the and therefore

Ms is fixed at any given time. Money demand on the other hand can be written as

P y Md = V . That is, the quantity of money demanded is a function of , the price level, and real output; that is, the total amount of buyable things in the economy.

Then, in equilibrium, if Ms = Md = M, the quantity theory of money is an equilibrium condition. Thus, if we increase M assuming velocity stays constant, we increase P in equilibrium. Equilibrium is not restored immediately. Obviously restaurants don’t reprint their menus with higher prices the minute the orders an increase in the money supply. Our purpose is to study price dynamics in the disequilibrium stage. The

Cantillon Effect, then, more precisely refers to two observations anomalous with respect to a monetary equilibrium model. Firstly, some prices move in the “wrong” direction; that is, prices fall in the aftermath of a monetary increase. Secondly, the variance of month over month price changes increases. It is primarily the second of these we wish to test. 7

1.5 The Statistical Problem

Therefore, our problem is to find evidence that sudden increases in the money supply induce these price dynamics. Thus, we must solve several problems.

1. We must detect monetary shocks statistically.

2. We must make a case that this statistical incidence of a monetary shock is exogenous.

In other words, we need to make the case that it was due to the actions of the Federal

Reserve and not due to a change in the underlying macroeconomy. An example of such a change would be a shock in the velocity of money.

3. We must test the equality of variances in the aftermath of these shocks versus other- wise. 8

CHAPTER 2

DATA AND ANALYSIS

2.1 Description of the Data

Following Balke and Wynne (2007) we identify monetary shocks using a five-variable system of macroeconomic indicators consisting of an industrial production index, a per- sonal expenditure consumption deflator, a general commodity price index, the nominal

Federal Funds rate, and the measure M2 of the money stock. All of these indicators come from the St Louis branch of the Federal Reserve, excluding the general commodity price index which is from the Bureau of Labor Statistics. All data are reported monthly. The data may be accessed through their system FRED (Federal Reserve Economic Data).

The industrial production index is a measurement of the level of industrial production and helps us to control for the monetary effects of business cycles (in slow economic times, there is a tendency to save case thus decreasing the volume of circulating money). The personal expenditure consumption deflator is a measurement of the aggregate amount consumers spend on goods and services. If prices increase, consumers hold more money and thus the amount of money in circulation decreases. The nominal Federal Funds rate is the rate at which depository institutions lend overnight to other depository institutions.

Here “nominal” refers to the fact that this quantity is not inflation-adjusted. Finally, the

M2 money measure is the Federal Reserve’s measure of the stock of money circulating in the economy. The Federal Funds rate is a percentage. We take the natural logarithm 9 of the other four variables so we can build models relating percentage changes. Assume a toy model log (p(t)) = β log (p(t)). Let p(˙t) refer to the time derivative of p(t). Then,

p(˙t) p(t−˙ 1) differentiating both sides: p(t) = β p(t−1) and β refers to a percentage change. Our other main source of data is a set of price indices from the Bureau of Labor

Statistics. I have used only the indices with 8-digit codes which are the most product- specific available indices. There are many gaps in these data as BLS rules prohibit re- porting a price index with fewer than 3 usable price quotations (Balke and Wynne, 2007).

These indices are for very specific classes of products. Wine grapes have a separate index from food grapes. We have used 50 price series chosen for their data coverage. A complete description of the available data is given in the appendix.

The sole purpose of the macroeconomic variables is the detection of monetary shocks. We are interested only in the behavior of Federal Funds rate as we use it as our measurement of Federal Reserve monetary increases. The remaining macroeconomic variables are used to control for other sources of variation in money supply so we can isolate the impact of the Federal Reserve. This will be clarified below.

The Federal Funds rate, plotted in Figure 2.1, changes frequently but there is a long term trend.

The distribution of month over month percentage price changes, plotted in Figure

2.2, has a distribution so irregular that a histogram representation is not helpful. There is an extremely high central peak at 0 and very long tails. This can be seen from the plot below where the percentage of data within a given Z-score of the center is plotted.

2.2 Description of the Statistical Model

Since we are working with a vector of 50 prices and five macroeconomic variables, we cannot naively apply multivariate time series as the degrees of freedom of the system 10

Figure 2.1. Macroeconomic Indicator Series would overwhelm the data. Thus, we will make an assumption that, having conditioned on the system of five macroeconomic variables, the prices are orthogonal. The justification for this approach is in Balke and Wynne (2007). In summary, as mentioned earlier, it is common practice in macroeconomic models to assume that macroeconomic variables are functions of other macroeconomic variables. Thus, the primary justification for this method is its commonality. The assumption more concretely is that the price changes influence each other only through broader macroeconomic processes that can be detected in statistical aggregates. In order to evaluate this assumption, we would need to actually test this orthogonality assumption which would overwhelm the data. Secondly, the usage 11

Figure 2.2. Distribution of Month over Month % Change in Price of statistical aggregation implicitly makes independence assumptions that are likely untrue if there exist complex relationships between individual prices. A true evaluation of this assumption would involve the construction of alternative models with explicit network features which is well beyond our present scope.

The mathematics of this approach is original to Lastrapes (2005) though it will be reviewed here for logical completeness. The upshot of this method is we may use univariate time series methods on each of the prices and then assemble them into a multivariate system. 12

2.2.1 Construction of the Autoregressive System

Following Lastrapes (2005), we consider the system:

          m A 0 m A 0 m  t   1   t−1   12   t−12    =     + ... +     + t pt B1 C1 pt−1 B12 C12 pt−12

where the ~mt is estimated using a multivariate time series and each element of ~pt is esti- mated separately using a univariate time series. Our assumption of orthogonality of prices implies this is equivalent to estimating a 55 variable system at once. Further, we have   0 0 E [ ] = Ω. mt = log INDPRO log M2 log PCEPI log PRC INDX FEDFUNDS   ~ and pt is the vector of prices. Each row of matrix B : bi bi1 bi2 bi3 bi4 b15 is es- timated from the univariate time series model p(t) = b1m1,t−1 + b2m2,t−1 + b1,3m3,t−1 + b1,4m4,t−1 +b1,5m5,t−1 ++c1pt−1 +...+b12,1m1,t−12 +b12,2m2,t−12 +b12,3m3,t−12 +b12,4m4,t−12 + b12,5m5,t−12+c12pt−12 . The remaining autoregressive coefficients {cij} for i ∈ {1, 2,..., 12} and j ∈ {1, 2, . . . , n} are entered into the matrix:

  c 0 0 ... 0  i1     0 c 0 ... 0   i2     ..  C =  0 0 . 0 0  .    . .   . . 0 c 0   i,n−1    0 0 ... 0 cin

We use 12 lags to capture a year of variation. Since the covariances between any two prices over time are set to zero, the C matrix is constrained to be diagonal.

2.3 Multivariate Time Series Model

We now proceed to fit the model described above. Table 2.2 displays the results of the Dicky-Fuller Unit Root Test to evaluate stationarity (Dickey and Fuller, 1979). 13

Our orthogonality assumptions will ensure that individual stationarity implies joint sta- tionarity. The Dicky Fuller test fails to reject the null hypothesis of non-stationarity for industrial production, personal consumption expenditure deflator, Federal Funds Rate and

M2. Since individual stationarity is necessary but not sufficient for joint stationarity, joint stationarity also fails (L¨utkepohl, 2005). We could address this with further differencing but this would add considerable complexity to the model and make economic interpreta- tion difficult. These macroeconomic variables primarily serve two purposes: they allow us to detect monetary shocks and serve as the basis for the orthogonality argument which reduces the degrees of freedom. Further, in truth, money propagates through an economy from firm to firm based on network relationships among firms. Time series models implic- itly assume that, at the resolution of the data in question, the system is “well-mixed”, or rather more formally, network effects do not matter. If, in fact, these networks are strong determinants of behavior, then it would be expected that a propagating money shock will change the behavior of a firm when the new money reaches that firm. As stationarity amounts to an assumption of symmetry between inter-temporal covariances, we believe it is unlikely to well-approximate this phenomenon. Further research is needed on this topic.

Dickey-Fuller Unit Root Tests

Variable Type Rho Pr

L INDPRO Zero Mean 0.15 0.7173 5.12 0.9999

Single Mean 0.12 0.9627 0.22 0.9736

Trend −5.17 0.8065 −1.82 0.6956

L PCEPI Zero Mean 0.15 0.7181 8.20 0.9999

Single Mean −1.21 0.8653 −5.70 <.0001

Trend −2.84 0.9430 −2.57 0.2937

L PRC INDX Zero Mean −262.87 0.0001 −11.53 <.0001 14

Dickey-Fuller Unit Root Tests

Variable Type Rho Pr

Single Mean −298.31 0.0001 −12.23 <.0001

Trend −299.63 0.0001 −12.24 <.0001

FEDFUNDS Zero Mean −2.89 0.2428 −1.90 0.0548

Single Mean −6.54 0.3041 −2.10 0.2464

Trend −15.54 0.1605 −2.87 0.1728

L M2 Zero Mean 0.17 0.7220 8.94 0.9999

Single Mean −0.64 0.9161 −2.31 0.1687

Trend −3.85 0.8940 −2.53 0.3123

Table 2.1: Unit Root Tests

The parameter estimates are as follow:

AR Coefficient Estimates

Lag Variable L INDPRO L PCEPI L PRC INDX FEDFUNDS L M2

1 L INDPRO 0.85630 0.55179 −0.05628 0.00391 0.08519

L PCEPI 0.03195 1.06221 0.04505 0.00107 −0.02711

L PRC INDX 0.20576 0.84098 −0.14620 0.00345 −0.06563

FEDFUNDS 2.64265 18.40210 −2.50783 1.33421 −3.71258

L M2 0.11167 −0.52926 0.03169 −0.00214 1.00646

2 L INDPRO 0.25673 −0.68189 −0.06981 −0.00480 0.04236

L PCEPI −0.02617 −0.32880 0.02774 −0.00162 0.00323

L PRC INDX −0.11298 −1.13996 −0.14807 −0.00351 −0.16312

FEDFUNDS 4.73093 −25.67120 1.94187 −0.45409 18.48232

L M2 −0.11965 0.20131 0.08998 0.00022 0.05768 15

AR Coefficient Estimates

Lag Variable L INDPRO L PCEPI L PRC INDX FEDFUNDS L M2

3 L INDPRO 0.06842 0.12474 −0.09002 −0.00032 −0.09183

L PCEPI 0.00199 0.18129 0.01221 0.00020 −0.03372

L PRC INDX −0.03590 0.42889 0.01686 0.00039 −0.04551

FEDFUNDS −2.67028 28.82644 −0.99953 0.17331 −20.09572

L M2 −0.05292 0.02372 0.05395 −0.00040 0.03217

4 L INDPRO −0.04889 0.01334 0.00727 0.00270 0.01646

L PCEPI −0.00042 −0.06289 0.02353 0.00109 0.04670

L PRC INDX −0.03584 −0.60690 0.12233 −0.00066 0.19742

FEDFUNDS −2.28996 −4.41356 −3.10016 −0.16331 12.05025

L M2 0.09201 0.04981 0.07814 0.00192 −0.11080

5 L INDPRO −0.07707 −0.38291 0.02362 −0.00346 0.02138

L PCEPI 0.00684 0.06527 0.02379 −0.00137 0.02460

L PRC INDX 0.00900 0.26223 0.08754 −0.00304 0.09872

FEDFUNDS −1.59808 2.66344 −1.84860 0.04364 −2.08774

L M2 0.01151 0.06500 0.02391 0.00112 0.06007

6 L INDPRO −0.00290 0.21631 0.03357 0.00145 −0.10198

L PCEPI −0.01454 0.11375 0.01059 0.00084 −0.03932

L PRC INDX 0.05197 −0.21102 −0.06534 0.00299 −0.20583

FEDFUNDS 2.52295 19.99126 −5.55722 −0.05522 −6.87743

L M2 0.08471 0.12768 0.05126 −0.00204 0.02353

7 L INDPRO −0.12362 0.11526 0.04635 0.00101 0.31238

L PCEPI 0.01443 −0.12076 0.01248 −0.00001 0.03964

L PRC INDX −0.03146 −0.03353 0.17022 −0.00044 0.28886 16

AR Coefficient Estimates

Lag Variable L INDPRO L PCEPI L PRC INDX FEDFUNDS L M2

FEDFUNDS 9.39869 −45.32886 2.23749 −0.03802 7.43908

L M2 −0.04308 0.49940 −0.00354 0.00130 −0.17002

8 L INDPRO −0.03979 0.22502 0.08711 −0.00251 −0.29423

L PCEPI 0.00044 0.04080 0.01753 0.00014 −0.06028

L PRC INDX 0.04456 0.46120 0.08176 0.00249 −0.30763

FEDFUNDS −17.76764 11.35680 −2.43441 0.30372 −14.98917

L M2 −0.03316 −0.38449 −0.04877 −0.00103 0.11286

9 L INDPRO 0.09994 −0.23838 0.00334 0.00299 0.06032

L PCEPI −0.01941 0.10355 0.00831 −0.00082 0.04520

L PRC INDX 0.04993 0.71352 −0.11228 −0.00327 0.26102

FEDFUNDS 7.11715 9.35153 0.46584 −0.06833 20.33772

L M2 −0.07198 0.33401 −0.02948 0.00211 0.23674

10 L INDPRO 0.05795 0.26235 −0.00338 −0.00076 0.08167

L PCEPI −0.01083 −0.06777 0.02326 0.00038 0.00221

L PRC INDX −0.21241 −0.04165 0.00532 0.00018 −0.03592

FEDFUNDS −0.66673 −8.83665 −2.15916 −0.19262 −14.47692

L M2 −0.03982 −0.37530 −0.03268 −0.00087 −0.22811

11 L INDPRO −0.04767 0.19757 −0.06470 −0.00001 0.02091

L PCEPI −0.00624 0.08924 0.00262 0.00028 −0.00422

L PRC INDX −0.00004 −0.09449 −0.09834 0.00135 0.01444

FEDFUNDS 2.04286 29.84953 −3.88585 0.10793 3.88746

L M2 0.08123 −0.12717 −0.04323 0.00252 −0.19819

12 L INDPRO −0.02588 −0.35745 −0.02932 −0.00080 −0.16263 17

AR Coefficient Estimates

Lag Variable L INDPRO L PCEPI L PRC INDX FEDFUNDS L M2

L PCEPI 0.02062 −0.08980 −0.00894 0.00021 0.01120

L PRC INDX 0.09567 −0.63378 −0.00854 −0.00004 −0.02297

FEDFUNDS −3.04834 −35.75774 −0.99566 −0.04873 −0.42389

L M2 −0.00783 0.09222 −0.00355 −0.00228 0.18319

Table 2.2: Coefficient Estimates

We now have a model of month over month price changes. We can now build an impulse response function.

2.4 Impulse Response Function

We now have the current vector of macroeconomics variables and prices expressed as a function of the lagged vectors. It is more convenient for the present purpose to re-express this autoregressive model as a moving average. This allows us to model the response of these variables to a “sudden” movement in a previous period. In other words, it allows us to model the current vector as a function of the error term of the autoregressive process. Since monetary shocks will manifest themselves in this error term, this is the most convenient model for their detection. More generally, this allows us to interpret a change the month over month percent change in prices as a linear function of how far off previous predictions were. That is, the slope parameters may be interpreted as the partial derivative of change in price with respect to a given element in the error term. Since our hypothesis does not predict a change in a single direction, we expect the distribution of impulse responses to include both positive and negative changes but not necessarily

Gaussian. 18

L INDPRO L PCEPI L PRC INDX FEDFUNDS L M2 L INDPRO 1.00000 0.96631 0.05679 −0.71603 0.96392 L PCEPI 0.96631 1.00000 0.01965 −0.81630 0.98102 L PRC INDX 0.05679 0.01965 1.00000 0.04777 0.02606 FEDFUNDS −0.71603 −0.81630 0.04777 1.00000 −0.80312 L M2 0.96392 0.98102 0.02606 −0.80312 1.00000 Table 2.3. Correlation Matrix

For the purpose of detecting monetary shocks, we are interested in sudden changes in the Federal Funds rate or rather large values in the 4th element in the error vector.

There is an additional issue here. We are concerned with changes in the Federal Funds rate due to Federal Reserve action. Other macroeconomic phenomena can also shift the Federal Funds Rate. Thus, we will not use the error term directly in our analysis.

Rather, we will orthogonalize it to isolate those changes in the Federal Funds rate which may reasonably be interpreted as monetary shocks. The macroeconomic variables are sometimes correlated. Consider the correlation matrix in Table 2.3. The Federal Funds rate has large correlations with everything but the price index. Industrial Production and

Personal Consumption Expenditure Deflator are also highly correlated. Only the overall price index seems nearly orthogonal. Clearly there is a lot of redundant information in these macroeconomic variables.

We thus will run Gramm-Schmitt orthogonalization on the error term of the moving average. Our means to do this is the Choleski Decomposition.

2.4.1 Choleski Decomposition: Theory

From linear algebra, we have the Choleski decomposition as a technique for orthog- onalization. Given a matrix Ω, we may decompose it into Ω = SS0. Now, given the standard theory of multivariate autoregressive processes:

n X ~yt = Ai ~yt−i + t i=1 19

We may rewrite this in terms of lag polynomials L(i)yt = yt−i.

" n # X I − AL(i) ~yt = t i=1 We may rewrite this expression in terms of an orthogonal error term:

" n # X −1 I − AL(i) ~yt = S ηt i=1 Then, rewriting as a moving average, we may calculate the impulse response function of the orthogonal error term:

" n #−1 X −1 ~yt = I − AL(i) S ηt i=1

Then, the (i, j) component of this matrix is ∂yt+1 . ∂ηj For now, let us consider the orthogonal error term η. We orthogonalize in the order Industrial Production, Personal Consumption Expenditure Deflator, the overall commodity price index, the federal funds rate, and the M2 measure of the money supply.

In this situation, we the correlated components of each variable get attributed to the first variable in order. Thus, a shock to the orthogonal federal funds rate may reasonably represent a monetary shock in so far as any overlap with previous variables is attributed to those variables and not the Federal Funds rate. An interesting point of comparison is whether a shock to M2 also works the same way. M2 is a measure of the money supply but newly injected money will not show up immediately. Since any component of M2 correlated with the Federal Funds rate will be attributed to the Federal Funds rate, we would not expect shocks in M2 to represent monetary shocks and thus Cantillon Effects should not be observed. 20

CHAPTER 3

RESULTS

We define a monetary shock of size k as a k standard deviation move in the or- thogonalized Federal Funds rate. Thus, we are comparing the periods of time where the orthogonalized Federal Funds rate is in the tails of the overall distribution as defined by k. We check the variance of the price distribution for three months after the shock. We hypothesize that for sufficiently large shocks in the Federal Funds Rate, the variance of percent changes in price will increase. More precisely, if the orthogonalized Federal Funds rate F has mean µF and standard deviation σF and Π represents the probability distribu- tion of month over month changes in price for the three months following the observation of the Federal Funds rate, we are testing the null hypothesis:

    |F − µF | |F − µF | H0 : Var Π ≥ k = Var Π < k . σF σF

We also suggest that this should not be true for a shock in the M2 supply as its movement is either due to the Federal Reserve and thus attributed to the Federal Funds rate by our orthogonalization or else due to a change in economic fundamentals and thus not properly understood as a shock. The standard deviation of both month over month percentage changes in price is calculated in the three months following a shock of a given size and for all other periods. 21

3.1 Tests of Equality of Variances

We find from a Levene test that the variances are indeed unequal for sufficiently large shocks. We run the test for 1, 1.5 and 2 standard deviation shocks to the Federal

Funds rate. We find that the second and the third are significant. The results for 1 standard deviation are listed in Tables 3.1 and 3.2.

The same results for a 1.5 standard deviation shock are displayed in Tables 3.3 and

3.4

Finally, Tables 3.5 and 3.6 contain the results for 2 standard deviations.

As Tables 3.7 and 3.8 show, we do not get similar results for a 2 standard deviation shock in M2.

Finally, as a robustness check, we will run the same test for a 2 standard deviation shock in the industrial production index. The results are in Tables 3.9 and 3.10

3.2 Probability Distribution of the Impulse Response of Prices

Figure 3.1 is a plot of the absolute value of the z-scores of the impulse response of the prices. The distribution has a high peak at zero and very long tails but is roughly symmetrical. The table is large and adds nothing to the presentation. Its coefficients are often not statistically significant. The fact that these impulse responses have different signs and hover around 0 is consistent with the Cantillon Effect in the sense that we do not reliably observe an immediate trend in prices due to monetary shocks. A positive monetary shock should result in long run inflation and thus increases in prices. A nega- tive shock should result in the reverse. 22

Figure 3.1. Distribution of Impulse Responses

Analysis Variable : DELTA Variance of % Price Change Post Monetary Shock N Obs N Mean Std Dev Minimum Maximum No 265593 264429 0.000272261 0.0978748 −4.5726710 4.5861411 Yes 79675 79155 −0.000893942 0.0981973 −1.8758426 1.5672051 Table 3.1. Standard Deviations of % Price Changes (1 Standard Deviation Shock)

Levene’s Test for Homogeneity of DELTA Variance Source DF Sum of Squares Mean Square F Value Pr >F SHCK 1 0.000243 0.000243 0.01 0.9114 Error 343582 6734.2 0.0196 Table 3.2. Levene’s Test for Difference in Variance of % Price Changes (1 Standard Deviation Shock)

Analysis Variable : DELTA Variance of % Price Change Post Monetary Shock N Obs N Mean Std Dev Minimum Maximum No 313579 312185 0.000197961 0.0971091 −4.5726710 4.5861411 Yes 31689 31399 −0.0019289 0.1059335 −1.8212353 1.4731928 Table 3.3. Standard Deviations of % Price Changes (1.5 Standard Deviation Shock) 23

Levene’s Test for Homogeneity of DELTA Variance Source DF Sum of Squares Mean Square F Value Pr >F SHCK 1 0.0916 0.0916 4.67 0.0307 Error 343582 6733.8 0.0196 Table 3.4. Levene’s Test for Difference in Variance of % Price Changes (1.5 Standard Deviation Shock)

Analysis Variable : DELTA Variance of % Price Change Post Monetary Shock N Obs N Mean Std Dev Minimum Maximum No 329108 327605 0.000169000 0.0973147 −4.5726710 4.5861411 Yes 16160 15979 −0.0033877 0.1101235 −1.7791011 1.4134964 Table 3.5. Standard Deviations of % Price Changes (2 Standard Deviation Shock)

Levene’s Test for Homogeneity of DELTA Variance Source DF Sum of Squares Mean Square F Value Pr >F SHCK 1 0.1075 0.1075 5.49 0.0192 Error 343582 6733.5 0.0196 Table 3.6. Levene’s Test for Difference in Variance of % Price Changes (2 Standard Deviation Shock)

Analysis Variable : DELTA Variance of % Price Change Post Monetary Shock N Obs N Mean Std Dev Minimum Maximum No 325674 324282 0.000088073 0.0983731 −4.5726710 4.5861411 Yes 19594 19302 −0.0014158 0.0905414 −1.7357652 1.2375771 Table 3.7. Standard Deviations of % Price Changes (2 Standard Deviation Shock)

Levene’s Test for Homogeneity of DELTA Variance Source DF Sum of Squares Mean Square F Value Pr >F SHCK 1 0.0399 0.0399 2.04 0.1536 Error 343582 6734.3 0.0196 Table 3.8. Levene’s Test for Difference in Variance of % Price Changes (2 Standard Deviation Shock in M2) 24

Analysis Variable : DELTA Variance of % Price Change Post Monetary Shock N Obs N Mean Std Dev Minimum Maximum No 322252 320750 0.000096046 0.0980333 −4.5726710 4.5769140 Yes 23016 22834 −0.0012951 0.0967697 −4.1584570 4.5861411 Table 3.9. Standard Deviations of % Price Changes (2 Standard Deviation Shock in Industrial Production)

Levene’s Test for Homogeneity of DELTA Variance Source DF Sum of Squares Mean Square F Value Pr >F SHCK 1 0.00130 0.00130 0.07 0.7971 Error 343582 6735.0 0.0196 Table 3.10. Levene’s Test for Difference in Variance of % Price Changes (2 Standard Deviation Shock in Industrial Production) 25

CHAPTER 4

CONCLUSION

We conclude that we have weak evidence for the Cantillon effect in this data. There are many margins on which this technique could be improved. The presented specifica- tion is a baseline without seasonal effects and other controls. Further, while these are standard techniques in macroeconometrics, linear models such as these ignore network feedback effects. These methods have this in common with most macroeconomic models but nonetheless, the assumption we made of no feedback from the price vector to the vector of macro variables in the strictest sense cannot be literally true. It is increasingly common for macroeconometrics to make use of explicit network structures in macroeco- nomic modeling and some of these may be statistically estimated. This seems a promising future route for macroeconometrics and a logical followup to the present project. 26

APPENDIX: DETAILED DESCRIPTION OF DATA

Macroeconomic Variables

Variable Description

Federal Funds Rate The at which banks lend

funds held at the Federal Reserve to each

other over night. This is the rate the

Federal Reserve Board targets in setting

monetary policy. Industrial Production A statistical summary of industrial

production. Full description available at:

https://www.federalreserve.gov/releases

/g17/About.htm Personal Consumption Expenditure Defla- A statistical summary of how much is

tor spent on consumption versus savings.

Full description available in

http://www.bea.gov/national

/pdf/nipaguid.pdf. 27

Variable Description

M2 “M2 includes a broader set of financial

assets held principally by households. M2

consists of M1 plus: (1) savings deposits

(which include money market

accounts, or MMDAs); (2)

small-denomination time deposits (time

deposits in amounts of less than

$100,000); and (3) balances in retail

money market mutual funds (MMMFs).

Seasonally adjusted M2 is computed by

summing savings deposits,

small-denomination time deposits, and

retail MMMFs, each seasonally adjusted

separately, and adding this result to

seasonally adjusted M1.” ? Commodity Price Index An overall commodity price index

produced by the Bureau of Labor

Statistics. Described under WP00000000

at https://download.bls.gov/

pub/time.series/wp/wp.txt 28

Price Indexes

Code Description

WP01110104 Lemons

WP01110211 Golden delicious apples

WP01110222 Strawberries

WP01130110 Dry pinto beans

WP01130111 Dry great northern beans

WP01130112 Dry pink beans

WP01130113 Dry pea beans

WP01130120 Dry peas

WP01130121 Dry lentils

WP01130211 Cabbage

WP01130212 Carrots

WP01130213 Celery

WP01130214 Sweet corn

WP01130215 Lettuce

WP01130216 Dry onions

WP01130217 Tomatoes

WP01130218 Snap beans

WP01130222 Broccoli

WP01130223 Cauliflower

WP01130228 Green peppers

WP01130231 Squash

WP01130301 Sweet potatoes

WP01130602 Round white potatoes 29

Code Description

WP01130603 Russet potatoes

WP01130604 Round red potatoes

WP01210101 Hard red winter wheat

WP01210102 Hard red spring wheat

WP01210103 Soft white wheat

WP01210104 Soft red winter wheat

WP01220101 Barley

WP01220205 Corn

WP01220205 Corn

WP01220311 Oats

WP01220501 Sorghum

WP01310199 Slaughter steers and heifers

WP01310299 Slaughter cows and bulls

WP01310399 Slaughter vealers

WP01320199 Slaughter barrows and gilts

WP01320299 Slaughter sows

WP01330199 Slaughter lambs

WP01410299 Slaughter chickens

WP01420199 Slaughter turkeys

WP01430199 Slaughter ducks

WP01510101 Raw cotton

WP01610102 Raw milk

WP01710701 Eggs, jumbo

WP01710702 Eggs, extra large 30

Code Description

WP01710703 Eggs, large

WP01710704 Eggs, medium

WP01710705 Eggs, small

WP01710801 Breaker stock

WP01710802 Checks and undergrades

WP01810101 Alfalfa hay

WP01830131 Soybeans

WP01830131 Soybeans

WP02110703 Other sweet goods, excluding frozen

WP02110803 Soft cakes, excluding frozen

WP02110903 Pies (fruit, cream, and custard),

excluding frozen WP02112103 Cookies, wafers, and ice cream cones and

cups (excluding frozen) WP02112104 Crackers, biscuits, and related products

WP02120301 Wheat flour

WP02120401 Flour base mixes and doughs

WP02130201 Milled rice (incl second heads, screenings,

brewers, bran, sharps, rice flour, and

byproducts) WP02140907 Manufactured starch

WP02140908 Wheat mill products, corn mill products,

and other grain mill products except flour WP02210133 Beef, fresh/frozen, primal and subprimal

cut, slaughtering 31

Code Description

WP02210579 Canned meats, excluding dog, cat, and

baby food WP02220333 Young chickens, including bulk, chilled,

frozen, whole, and in parts WP02220611 Turkeys, including frozen, whole, and

parts WP02220811 Canned, cooked, smoked or prepared

poultry WP02220911 Other poultry/small game (duck, goose,

rabbit) WP02230101 Haddock

WP02230131 Flounder

WP02230132 Cod

WP02230133 Pollock

WP02230135 Rockfish

WP02230199 Other finfish

WP02230429 Canned seafood (including soups, stews,

and chowders) WP02230501 Shrimp

WP02310301 Fluid whole milk

WP02310302 Reduced fat and lowfat milk (1/2-2)

WP02310303 Fat free or skim milk

WP02310304 Other fluid milk related products,

packaged, incl. cartons, bottles, cans, and

dispenser cans WP02310401 Cottage cheese 32

Code Description

WP02310501 Other milk products

WP02310601 Bulk fluid milk and cream

WP02320114 Butter

WP02330212 Natural cheese, except cottage cheese

WP02330312 Process cheese and related products

WP02340201 Ice Cream and frozen desserts

WP02350201 Dry milk products, including feed grade

WP02350301 Consumer-type canned milk products

WP02350303 Bulk liquid milk products, including feed

grade WP02410105 Canned fruits, excluding baby foods

WP02410239 Canned and fresh fruit juices, nectars,

and concentrates WP02420209 Frozen fruits

WP02420301 Frozen concentrated orange juice,

consumer and institutional WP02440102 Canned vegetables

WP02440127 Canned catsup and other tomato based

sauces WP02440139 Canned vegetable juices

WP02450202 Frozen potato products (French-fried,

patties, puffs, etc.) WP02450509 Frozen vegetables, other than potato

products WP02540104 Chocolate coatings, made from cacao

beans 33

Code Description

WP02540105 Other chocolate and cocoa products,

made from cacao beans WP02540107 Corn sweeteners

WP02550201 Chewing gum, bubble gum, and chewing

gum base WP02550301 Chocolate and chocolate-type

confectionery products WP02550302 Nonchocolate-type confectionery products

WP02550304 Nuts and seeds (salted, roasted, cooked

or blanched) WP02610101 Bottled beer and ale

WP02610103 Canned beer and ale

WP02610105 Beer and ale in barrels and kegs

WP02620609 Soft drinks, non-carbonated

WP02630103 Coffee, concentrated, including coffee

substitutes WP02630104 Roasted coffee

WP02630313 Tea in consumer packages, exc. canned

ice tea WP02640101 Malt and malt by-products

WP02640111 Liquid beverage bases, for sale by soft

drink bottlers WP02780109 Margarine, butter blends, and butter

substitutes WP02810102 Canned jams, jellies, and preserves

WP02820106 Pickles and other pickled products,

including horseradish 34

Code Description

WP02830111 Processed eggs, liquid, dried, or frozen

WP02840102 Canned dry beans

WP02840104 Canned soups and stews, excluding

frozen and seafood WP02850109 Frozen bakery products

WP02850111 Frozen dinners and nationality foods

WP02850113 Other frozen specialties, excluding

seafood WP02860103 Prepared sauces, except tomato

WP02890102 Mayonnaise, salad dressings, and

sandwich spreads WP02890148 Dry mix food preparations

WP02890149 Perishable non-frozen prepared food,

including tortillas WP02890151 Flavoring extracts, emulsions and other

liquid flavors WP02890156 Table (evaporated), pepper (white

and black), and other spices WP02890158 Peanut butter

WP02890161 Ice

WP02890162 Dairy product substitutes

WP02890172 Chips (potato, corn, etc.)

WP02890175 Other food preparations

WP02930102 Chicken and turkey feed, supplements,

concentrates, and premixes WP02930118 Other poultry and livestock feeds 35

Code Description

WP02940202 Dog and cat food

WP02940203 Other pet and specialty feeds

WP02940301 Meat meal and meat and bone meal

WP02940325 Other animal feeds, incl. fertilizer

byproducts and feather WP03150231 Glass fiber, textile-type

WP03260304 Spun synthetic yarns

WP03260402 Thrown filament yarns, except textured

WP03260403 Textured, crimped, or bulked filament

yarns WP03370104 Greige cotton broadwoven fabrics

WP03370304 Greige manmade fiber broadwoven fabrics

WP03380301 Greige weft (circular) knit fabrics

(excluding hosiery) WP03390101 Tire cord and tire fabric

WP03420104 Finished cotton broadwoven fabrics

WP03420304 Finished manmade fiber, silk and other

natural fiber (excl. cotton and wool)

broadwoven fabrics WP03440204 Woven narrow fabrics

WP03450321 Nonwoven fabrics

WP03460102 Vinyl coated fabrics, including expanded

vinyl coated WP03460103 Other coated or laminated fabrics and

coated yarns, incl. impregnated WP03470105 Schiffli machine embroideries 36

Code Description

WP03470108 Screen printing on garments, apparel

accessories, and other fabric articles,

except labels WP03810407 Fur products

WP03810441 Apparel and accessories, n.e.c.

WP03810623 Women’s and girls’ skirts

WP03810624 Women’s and girls’ tailored jackets and

vests, except fur and leather WP03810645 Women’s and girls’ robes and dressing

gowns WP03810646 Women’s and girls’ brassieres, corsets

(exc. surgical), girdles, and combinations WP03830351 Fabricated textile products, n.e.c.

WP03910201 Recovered fibers, processed mill waste,

and related products WP04310501 Men’s nonathletic footwear

WP04320501 Women’s nonathletic footwear

WP04410112 Luggage

WP04410128 Women’s and children’s handbags and

purses WP04410132 Other personal leather goods

WP04450101 All other leather goods

WP04450111 Leather belts

WP05110117 Prepared anthracite shipped

WP05310105 Natural gas 37

Code Description

WP05320104 Propane

WP05320105 Butane

WP05412101 Residential electric power

WP05422101 Commercial electric power

WP05432101 Industrial electric power

WP05512101 Residential natural gas

WP05522101 Commercial natural gas

WP05532101 Industrial natural gas

WP05542101 Natural gas to electric power

WP05610102 Crude petroleum (domestic production)

WP05720201 Kerosene

WP05720201 Kerosene

WP05720301 Jet fuel

WP05730201 Home heating oil and distillates

WP05730201 Home heating oil and distillates

WP05730302 No. 2 diesel fuel

WP05760303 Lubricating grease not from petroleum

refineries WP05760401 Lubricating and similar oils not from

petroleum refineries WP05810319 Other petroleum and coal products,

including coke oven products, n.e.c. WP06130201 No Description

WP06130209 Aluminum compounds

WP06130213 Lime 38

Code Description

WP06130252 Barite

WP06130271 Rock salt

WP06130301 Natural sodium carbonate and sulfate

WP06140399 Other basic organics, n.e.c.

WP06210201 OEM finishes excluding marine coatings

WP06210301 Special purpose coatings, incl. marine,

industrial and construction coatings WP06220206 Iron oxide pigments

WP06220209 Titanium pigments

WP06220298 Synthetic organic pigments

WP06220299 Other inorganic pigments

WP06220407 Kaolin and ball clay

WP06371801 In-vitro diagnostics

WP06380202 Analgesics

WP06380202 Analgesics

WP06510501 Mixed fertilizers

WP06520136 No Description

WP06630601 Thermosetting resins and plastics

materials WP06710401 Soaps and detergents, commercial,

industrial, and institutional WP06710402 Household detergents

WP06710403 Soaps, excluding specialty cleaners,

household 39

Code Description

WP06720103 Polishing preparations and related

products WP06750206 Shaving preparations

WP06750306 Perfumes, toilet waters, and colognes

WP06750407 Hair preparations (including shampoos)

WP06751401 Creams, lotions and oils, excluding

shaving, hair, and deodorant WP06751501 Other cosmetics and toilet preparations

WP06790301 Acetylene

WP06790302 Carbon dioxide

WP06790401 Natural base glues and adhesives

WP06790402 Synthetic resin and rubber adhesives

WP06790501 Textile and leather assistants and finishes

WP06790502 Surfactants (bulk surface active agents)

WP06790904 Salt, evaporated and solar

WP06790918 Carbon black

WP06790919 Printing ink

WP06790961 Water-treating compounds

WP06790999 Other chemical preparations, n.e.c.

WP07120105 Truck and bus (including

off-the-highway) pneumatic tires WP07130604 All other industrial rubber products

WP07130606 All other miscellaneous rubber goods

WP07130607 Rubber sponge, expanded and foam

rubber products 40

Code Description

WP07130608 Rubber floor and wall coverings

WP07130612 Rubber compounds or mixtures for sale

or interplant transfer WP07210606 Other plastic construction products

WP07260305 Foam components for furniture

WP07290196 Custom compounding of purchased

plastic resins (incl color concentrates and

resin pellets) WP07290197 All other reinforced and fiberglass

plastics products WP07290198 Products made of foam other than

polystyrene or polyurethane WP08110503 Softwood cut stock and dimension

WP08120311 Hardwood cut stock and dimension

WP08120401 Oak and maple hardwood flooring

WP08210112 Wood window units

WP08210122 Wood sash

WP08210132 Wood window and door frames

WP08210142 Wood doors, flush and panel, interior and

exterior WP08210152 Other wood doors, incl. garage, screen,

storm, etc. WP08210162 Wood moldings

WP08210183 Other wood millwork products

WP08420101 Nailed and lock-corner wood boxes

WP08490901 All other miscellaneous wood products 41

Code Description

WP08610101 Prefabricated stationary wood buildings,

components WP08610102 Prefabricated stationary wood buildings,

precut packages WP08610103 Prefabricated stationary wood buildings,

shipped in panel form WP08610104 Prefabricated stationary wood buildings,

shipped in three-dimensional assemblies WP08710101 Wood poles, piles, and posts owned and

treated by the same establishment WP08710102 Other wood products owned and treated

by the same establishment WP09130291 Newsprint

WP09130321 Coated and laminated single and

multi-web paper WP09130322 Coated and laminated single and

multi-web film WP09140551 Corrugated paperboard in sheets and

rolls, lined and unlined WP09141104 Semichemical paperboard

WP09141105 Recycled paperboard

WP09141107 Unbleached kraft packaging and

industrial converting paperboard WP09150123 Sanitary paper products, including stock

WP09150214 Uncoated single-web paper grocers’ bags

and sacks and variety and shopping bags WP09150216 Specialty bags, pouches and liners 42

Code Description

WP09150218 Shipping sacks and multiwall bags, all

materials, excpet textiles WP09150301 Corrugated and solid fiber boxes

WP09150322 Setup (rigid) paperboard boxes

WP09150337 Paperboard fiber drums with ends of any

material WP09150441 Other corrugated and solid fiber

products, incl. containers, etc. WP09150636 Envelopes

WP09150655 Paper supplies for business machines and

other misc. unprinted office supplies WP09150814 Molded pulp goods, including egg

cartons, florists pots, food trays, etc. WP09150901 Pasted, lined, laminated, or

surface-coated paperboard WP09150999 Other sanitary paper/paperboard food

containers/boards/trays WP09220123 Particleboard

WP09220124 Waferboard and oriented strandboard

(OSB) WP09220131 Medium density fiberboard (MDF)

WP09230102 Cellulosic insulating fiberboard

WP09450101 Unit set business forms, loose or bound,

including label/form type WP09450102 Manifold books (incl. sales) and

pegboard accounting systems 43

Code Description

WP09450103 Custom continuous business forms

WP09450104 Stock continuous business forms

WP09470102 Magazine and periodical printing

(lithographic) (offset) WP09470201 Label and wrapper printing (letterpress)

WP09470202 Label and wrapper printing

(lithographic) (offset) WP09470203 Label and wrapper printing (gravure)

WP09470302 Catalog and directory printing

(lithographic) (offset) WP09470402 Financial and legal printing

(lithographic) (offset) WP09470502 Advertising printing (lithographic)

(offset) WP09470601 Other commercial and general job

printing (letterpress) WP09470603 Other commercial and general job

printing (gravure) WP09470609 Other commercial and general job

printing WP09471101 Screen printed materials

WP09471102 Engraving (printing)

WP09471103 Digital commercial printing

WP09480201 Edition, library and other hardcover book

binding WP10110555 Iron ores 44

Code Description

WP10121191 Heavy melting scrap

WP10121192 Carbon steel scrap bundles

WP10121193 Shredded carbon steel scrap

WP10121194 Cut plate and structural scrap

WP10121195 Other carbon steel scrap

WP10150211 Pressure pipe and fittings, ductile iron

WP10150501 Standard and pearlitic malleable iron

castings WP10151329 Seamless rolled ring forgings, ferrous

WP10151359 Closed die forgings, ferrous

WP10210201 Copper ores

WP10220131 Copper cathode and refined copper

WP10230101 No. 1 copper scrap, including wire

WP10230102 No. 2 copper scrap, including wire

WP10230104 Other copper and brass scrap

WP10230201 Solids and clippings, new aluminum base

scrap WP10230205 Used beverage can scrap

WP10230206 Other old aluminum base scrap

WP10240229 Secondary aluminum

WP10240669 Secondary precious metals

WP10250129 Aluminum plate

WP10260301 Electronic wire and cable

WP10260314 Copper wire and cable

WP10260333 Fiber optic cable 45

Code Description

WP10270111 Nonferrous forge shop products

WP10280204 Sand castings, aluminum and

aluminum-base alloy WP10280205 Permanent and semi-permanent mold

castings, aluminum and aluminum-base

alloy WP10280207 Aluminum die-castings

WP10410311 Motor vehicle hardware

WP10410341 Other transportation equipment

hardware WP10540211 Bath and shower fittings

WP10540218 Lavatory and sink fittings

WP10540223 Other plumbing fixture fittings and trim

WP10610106 Cast iron heating boilers, radiators and

convectors WP10610112 Steel heating boilers (15 psi or less) and

all hot water heating boilers (except

parts) WP10630161 Other heating equipment, non-electric,

including parts WP10640141 Domestic heating stoves

WP10660101 Household water heaters, electric, for

permanent installation WP10710313 Metal windows (except storm sash)

WP10710515 Metal combination screen, storm sash,

and storm doors 46

Code Description

WP10720104 Storage and other non-pressure tanks

WP10720122 Gas cylinders

WP10720152 Metal tanks and vessels, custom

fabricated and field erected WP10740803 Iron, steel, and aluminum stairs,

staircases and fire escapes WP10740811 Metal grilles, registers and air diffusers

WP10740813 Steel and aluminum fences, gates (not

wire), and railings and window guards WP10740814 Open metal flooring, grating and studs

WP10750103 Heat exchangers and steam condensers

(except for nuclear applications) WP10790101 Prefabricated metal building systems

(excl. farm service bldgs, residential

bldgs, and parts) WP10790201 Farm service buildings and other

prefabricated and portable buildings,

steel and aluminum WP10790354 Panels, parts, and sections for

prefabricated buildings, steel and

aluminum WP10810206 Hex bolts, including heavy, tap-and-joint

WP10810231 Cap, set, machine, lag, flange, and

self-locking screws, except aircraft types WP10810236 Tapping screws and wood screws, except

aircraft types 47

Code Description

WP10810262 Other metal bolts, incl. square, round,

plow, high-strength structural, and bent WP10810263 Other externally threaded metal

fasteners, including studs, except aircraft

types WP10810312 Internally threaded fasteners, except

aircraft types WP10810424 Nonthreaded metal fasteners, except

aircraft types WP10810648 Formed products, except fasteners, made

by cold-, warm-, or hot-heading processes WP10830225 Residential electric lighting fixtures,

except portable WP10830522 Outdoor lighting equipment (including

parts and accessories) WP10830524 All other miscellaneous electric and

nonelectric lighting equipment, incl parts

and accessories WP10850101 Ammunition, except for small arms

WP10860101 Ordnance and ordnance accessories, n.e.c.

WP10880101 Ferrous wire rope, cable, forms and

strand WP10880701 Ferrous wire cloth, other woven wire

products WP10890424 Precision mechanical wire springs

WP10890425 Other light gauge wire springs

WP10890507 Automotive job stampings 48

Code Description

WP10890521 Precision turned products, automotive

WP10890522 Precision turned products, except

automotive WP10890564 Metal powders, paste, and flake

WP10890571 Powder metallurgy parts, except

bearings, gears, etc. WP10890589 Other fabricated metal products

WP10890701 Metal job stampings, except automotive

WP10890811 Flexible packaging foil

WP11140611 Harvesting machinery (except hay and

straw) and attachments WP11140711 Haying machinery and attachments

WP11330175 Arc welding machines, components, and

accessories, excluding electrodes WP11330276 Resistance welders, components,

accessories, and electrodes WP11330378 Arc welding electrodes, metal

WP11350243 Precision measuring tools

WP11350501 Other machine tool attachments and

accessories WP11360105 Nonmetallic sized grains, powders, and

flour abrasives WP11360311 Nonmetallic abrasive products (including

diamond abrasives) WP11360505 Nonmetallic coated abrasive products and

buffing wheels, polishing wheels, and laps 49

Code Description

WP11382611 Metal punching and shearing (power and

manual), and bending and forming

machines (power only) WP11382631 Metalworking presses (except forging and

die-stamping presses) WP11382651 Other metal-forming machine tools

(except presses) WP11410212 Industrial pumps, except hydraulic fluid

power pumps WP11410701 Parts and attachments for air and gas

compressors and vacuum pumps WP11410801 Industrial spraying equipment

WP11411211 Parts and attachments for pumps

WP11411215 All other pumps, including domestic

sump, oil well, and oil field pumps WP11420221 Parts and attachments for elevators and

moving stairs (sold separately) WP11430312 Non-aerospace hydraulic and all

aerospace-type fluid power cylinders WP11430313 Non-aerospace pneumatic fluid power

cylinders WP11430315 Parts for hydraulic and pneumatic

cylinders and actuators WP11430406 Fluid power hose and tube fittings

WP11430501 Parts for fluid power valves 50

Code Description

WP11440212 Unit handling conveyors and conveying

systems WP11440214 Parts and accessories for unit handling

conveyors and conveying systems, sold

separately WP11440216 Bulk material handling conveyors and

conveying systems WP11440218 Parts and accessories for bulk material

handling conveyors and conveying

systems WP11440378 Parts and attachments for industrial

trucks and tractors WP11440481 Hoists

WP11440485 Overhead traveling cranes and monorail

systems WP11450108 Loose gear, pinions and racks

WP11450201 Plain bearings and bushings

WP11470143 Dust collection and other air purification

equipment for cleaning incoming air WP11470144 Dust collection and other air purification

equipment for industrial gas cleaning

systems WP11480222 Unitary air-conditioners, except air

source heat pumps WP11480331 Commercial refrigeration equipment 51

Code Description

WP11480631 All other miscellaneous refrigeration and

air-conditioning equipment WP11480901 Parts and accessories for air conditioning

and heat transfer equipment WP11490201 Gates, globes, angles, and checks

WP11490202 Industrial ball valves, incl. manual and

power operated WP11490203 Industrial butterfly valves, incl. manual

and power operated WP11490204 Industrial plug valves

WP11490205 Plumbing and heating valves (low

pressure) WP11490208 Solenoid valves

WP11490209 Other industrial valves, including nuclear

WP11490211 Automatic regulating and control valves

WP11490534 Mounted bearings, except plain

WP11490535 Parts and components for ball and roller

bearings, incl. balls and rollers WP11490537 Tapered roller bearings (including cups

and cones), unmounted WP11490538 Other roller bearings, unmounted

WP11490701 Industrial patterns

WP11510114 Personal computers and workstations

(excluding portable computers) WP11510115 Portable computers, laptops, tablets and

other single user computers 52

Code Description

WP11520101 Computer storage devices (except parts,

attachments and accessories) WP11540501 Parts, attachments and accessories for

POS terminals and computer peripherals WP11621201 Textile machinery (except parts,

attachments, and accessories) WP11627701 Textile machinery parts and attachments

WP11660632 Foundry machinery and parts

WP11660638 Semiconductor manufacturing machinery

and parts WP11670502 Parts for packing, packaging and bottling

machinery WP11680111 Commercial and industrial vacuum

cleaners WP11710143 Current-carrying switches for electrical

circuitry (including vehicular switches) WP11710144 Current-carrying wire connectors for

electrical circuitry WP11720501 Test equipment for electrical, radio, and

communication circuits and motors WP11730312 Fractional horsepower motors and

generators WP11730405 Integral horsepower motors and

generators (excluding land transportation

types) WP11730901 Parts and supplies for motors and

generators 53

Code Description

WP11740999 Power and distribution transformers,

except parts WP11741101 Commercial, institutional and industrial

general-purpose transformers, all voltages WP11741103 Fluorescent lamp ballasts

WP11750799 Industrial controls and related parts and

accessories WP11760121 Switching equipment

WP11760141 Wireline voice and data network

equipment WP11760301 Broadcast, studio and related equipment

WP11760302 Radio station and wireless

communication equipment WP11760303 Intercommunications, alarm and traffic

control systems WP11782116 Relays for electronic circuitry, industrial

control overload, and switchgear type WP11782599 Magnetic and optical recording media

WP11782890 Filters, crystals, and transducers

WP11784901 Bare printed circuit boards

WP11785399 Electronic coils, transformers, and other

inductors WP11785603 All other miscellaneous electronic

components WP11790103 Storage batteries, lead acid type, BCI

dimensional group 8D or smaller 54

Code Description

WP11790104 Storage batteries, lead acid type, larger

than BCI dimensional group 8D WP11790512 Irradiation equipment

WP11790551 Electronic hearing aids

WP11792902 Rectifying apparatus

WP11840102 Integrating and totalizing meters for

gases or liquids WP11840103 Counting devices

WP11840104 Motor vehicle indicating instruments

WP11850111 Aeronautical, nautical, and navigational

instruments WP11860111 Sighting, tracking, and fire-control

equipment, optical type WP11860211 All other miscellaneous optical

instruments and lenses WP11860311 Laboratory analytical instruments

WP11930656 -operated mechanisms and other

parts for automatic merchandising

machines WP11930753 No Description

WP11930757 No Description

WP11940811 Diesel, semidiesel, and dual-fuel engines,

automotive WP11941301 Parts and accessories for internal

combustion engines WP11950501 Other machine shop products 55

Code Description

WP12110102 Metal household dining room and kitchen

furniture WP12110104 Other metal household furniture

WP12140114 Mattresses, other types, including crib,

foam, waterbed mattresses and mattress

inserts WP12150112 Metal porch, lawn, outdoor and casual

furniture WP12160101 Nonupholstered household furniture,

except wood and metal WP12210112 Wood office seating

WP12210113 Wood office desks and extensions

WP12210114 Wood office storage units, files and tables

WP12210116 Other wood office furniture

WP12210161 Fixtures for stores, banks and offices

WP12220202 Nonwood commercial storage units, files,

and tables WP12220321 Nonwood commercial desks and

extensions WP12220325 Nonwood office seating

WP12220326 Nonwood office furniture panel systems,

and other nonwood office furniture WP12220405 Shelving and lockers, except wood

WP12220407 Storage racks and accessories, except

wood WP12220409 Fixtures (bank, office, and store) except

wood 56

Code Description

WP12230101 School furniture (except stone and

concrete), excluding library furniture WP12230102 Public building and related furniture,

excluding school and restaurant furniture WP12310101 Carpets and rugs

WP12410220 Household laundry equipment and parts

WP12440178 Small electric household appliances,

except fans WP12440228 Electric fans, except industrial-type

WP12450135 Portable residential lighting fixtures,

including parts and accessories WP12670141 Cutlery, scissors, shears, trimmers, and

snips WP12680102 Stamped and spun kitchen utensils,

except aluminum WP12690101 Window shades and window shade

accessories and rollers WP12690102 Venetian blinds

WP12690103 Other shades and blinds, including

curtain and drapery fixtures, poles, and

rods WP13130111 Machine-made pressed and blown

lighting, automotive, and electronic

glassware WP13210121 Crushed and broken stone

WP13220161 Cement, hydraulic 57

Code Description

WP13311135 Structural concrete block

WP13312101 Decorative concrete block (such as screen,

split, slump block, etc.) WP13313101 Concrete brick

WP13314101 Concrete pavers (including grid,

interlocking, etc.) WP13320108 Concrete pipe

WP13330101 Ready-mix concrete

WP13340106 Precast concrete products

WP13350107 Prestressed concrete products

WP13420101 Face brick and building brick

WP13440131 Clay floor and wall tile, glazed and

unglazed WP13450101 Vitrified clay sewer pipe and fittings

WP13450199 All other structural clay products, except

clay refractories, n.e.c. WP13620121 Roofing asphalts, pitches, coatings, and

cement WP13710102 Gypsum building plasters, boards and

laths WP13810104 Glass containers

WP13920201 Mineral wool for industrial, equipment,

and appliance insulation WP13950111 Dressed dimension granite (including

gneiss, syenite, diorite, and cut granite) WP13980111 Nonmetallic gaskets and gasketing 58

Code Description

WP13980112 Metallic gaskets and machined seals

WP13980211 Compression packings

WP13980212 Molded packing and sealing devices

WP13990101 Industrial glass sand

WP13990111 Industrial molding sand

WP13990121 Hydraulic fracturing sand and all other

industrial sand WP13990209 Minerals and earths ground or treated

WP13990211 Dimension stone mining and quarrying

WP13990214 Clay, excluding kaolin and ball clay

WP13990299 Talc, gypsum, and other powdered clay

mining and quarrying WP14110131 Passenger cars

WP14110401 Motorcycles, including three-wheel

motorbikes, motorscooters, mopeds, and

parts WP14120513 Carburetors, pistons, piston rings, and

valves WP14120514 Motor vehicle electrical and electronic

equipment WP14140602 Truck trailers and chassis, axle rating

10,000 lb or more WP14150101 Motor homes, built on purchased chassis

WP14160101 Travel trailers

WP14160201 Camping trailers, campers, pick-up

covers, and parts 59

Code Description

WP14230101 Aircraft engine and engine parts

WP14310301 Nonmilitary self-propelled ships, new

construction WP14310401 Nonpropelled ships, new construction

WP14911101 Self-propelled golf carts and industrial

in-plant personnel carriers and parts WP14911104 Automobile and light truck trailers

WP14911105 All other miscellaneous transportation

equipment, including all-terrain vehicles WP15110152 Nonelectronic games and puzzles (excl.

toys), and electronic toys (incl. parts) WP15110154 Other nonelectronic toys, including parts

WP15110155 Models, science and craft kits/supply,

and collectors’ miniatures WP15110156 Dolls, toy animals, action figures and

stuffed toys, incl parts WP15120103 Fishing tackle and equipment

WP15120127 Golf equipment, excluding apparel and

shoes WP15120182 Bicycles, adult tricycles, unicycles and

parts WP15120185 Playground equipment

WP15120192 Gymnasium and exercise equipment

WP15120193 Other sporting and athletic goods

WP15130301 Small arms ammunition primers

WP15240208 Stemmed and redried tobacco 60

Code Description

WP15320102 Zippers and slide fasteners

WP15320103 Buckles, non-slide fasteners, needles, and

pins WP15410701 Photographic and photocopying

equipment WP15420601 Photographic supplies

WP15530301 Manufactured homes (mobile homes), all

sizes (incl. multisection) WP15909011 No Description

WP15930117 Other musical instruments and parts

WP15950210 Pens, markers, mechanical pencils, and

associated parts WP15950307 Lead pencils, art goods, office supplies

and small office equipment, excluding

paper WP15970501 Brooms, mops, and dusters

WP15970502 Paint and varnish brushes and rollers

WP15970503 Other brushes, excl. paint and varnish

brushes WP33110101 Sales of textbooks

WP33110201 Sales of technical, scientific, and

professional books WP33110301 Sales of religious books

WP33110402 Sales of general adult and juvenile books

WP33110501 Sales of general reference books

WP33310101 Greeting card publishing sales 61

Code Description

WP33410101 Calendars, yearbooks, and other

miscellaneous publishing sales WP36110101 Advertising space sales in specialized,

business, and professional periodicals WP36110102 Advertising space sales in general and

consumer periodicals WP59110111 Metal plating and polishing

WP59110222 Metal coating and allied services

WP60110101 Metal mining services

WP60110201 Support activities for coal mining

WP60110301 Drilling oil and gas wells services

WP60110401 Support activities for oil and gas

operations WP60110501 Nonmetallic minerals mining services

WP61110101 Commission throwing, texturing, or

winding filament yarns WP61110201 Commission finishing of broadwoven

fabrics WP61110204 Commission knit or knit and finishing of

circular and warp fabric WP61210101 Contract and commission receipts for

tanning or finishing leather owned by

others 62

REFERENCES

Abel, Andrew B and Frederic S Mishkin (1983). “An Integrated View of Tests of Ratio- nality, Market Efficiency and the Short-Run Neutrality of Monetary Policy.” Journal of 11(1), 3–24.

Balke, Nathan S (2012). “Sectoral Effects of Aggregate Shocks.” Advances in Economet- rics 30, 299–357.

Balke, Nathan S and Mark A Wynne (2007). “The Relative Price Effects of Monetary Shocks.” Journal of Macroeconomics 29(1), 19–36.

Ball, Laurence and N Gregory Mankiw (1994). “A Sticky-price Manifesto.” In Carnegie- Rochester Conference Series on Public Policy, Volume 41, pp. 127–151. Elsevier.

Baumeister, Christiane, Philip Liu, and Haroon Mumtaz (2013). “Changes in the Ef- fects of Monetary Policy on Disaggregate Price Dynamics.” Journal of Economic Dynamics and Control 37(3), 543–560.

Berument, MH, A Sahin, and B Saracoglu (2009). “The Choice of Monetary Policy Tool(s) and Relative Price Variability: Evidence from Turkey.” Journal of Applied Sciences 9(12), 2238–2246.

Caplin, Andrew S and Daniel F Spulber (1987). “Menu Costs and the Neutrality of Money.” The Quarterly Journal of Economics 102(4), 703–725.

Demers, Fr´ed´erick, Annie De Champlain, et al. (2005). Forecasting Core Inflation in Canada: Should We Forecast the Aggregate or the Components? Bank of Canada.

Dickey, David A and Wayne A Fuller (1979). “Distribution of the Estimators for Au- toregressive Time Series with a Unit Root.” Journal of the American statistical association 74(366a), 427–431.

Friedman, Milton (2010). “Quantity Theory of Money.” In Monetary Economics, pp. 299–338. Springer.

Gauger, Jean (1988). “Disaggregate Level Evidence on Monetary Neutrality.” The Review of Economics and Statistics, 676–680.

Hayek, Friedrich August (1932). Prices and Production, Volume 107. Institute. 63

Karim, Zulkefly Abdul, Mohd Azlan Shah Zaidi, and WNW Azman-Saini (2011). “Rela- tive Price Effects of Monetary Policy Shock in Malaysia: a SVAR Study.” National Univeristy of Malaysia 15.

Keynes, John Maynard (2016). General Theory of Employment, Interest and Money. Atlantic Publishers & Dist.

Kreps, David M (2012). Microeconomic Foundations I: Choice and Competitive Markets, Volume 1. Princeton University Press.

Lastrapes, William D (2004). “Inflation and the Distribution of Relative Prices: the Role of Productivity and Money Supply Shocks.

Lastrapes, William D (2005). “Estimating and Identifying Vector Autoregressions under Diagonality and Block Exogeneity Restrictions.” Economics letters 87(1), 75–81.

Lucas Jr, Robert E (1996). “Nobel Lecture: Monetary Neutrality.” Journal of political economy 104(4), 661–682.

L¨utkepohl, Helmut (2005). New Introduction to Multiple Time Series Analysis. Springer Science & Business Media.

Patinkin, Don (1989). “Neutrality of Money.” In Money, pp. 273–287. Springer.

Poitras, Marc (1997). “Expectations and Monetary Neutrality: an Empirical Reexami- nation.” Southern Economic Journal, 920–928.

Puah, Chin-Hong, Muzafar Shah Habibullah, and Shazali Abu Mansor (2008). “On the Long-Run Monetary Neutrality: Evidence from the SEACEN Countries.

Rather, Sartaj Rasool, S Raja Sethu Durai, and M Ramachandran (2015). “Asymmetric Price Adjustment–Evidence for India.” The Journal of Economic Asymmetries 12(2), 73–79.

Robertson, John C and David Orden (1990). “Monetary Impacts on Prices in the Short and Long Run: Some Evidence from New Zealand.” American Journal of Agricultural Economics 72(1), 160–171.

Shah, Parth J (1997). “The Theory of Business Fluctuations: New Keynesians, Old Monetarists, and Austrians.” In Advances in Austrian economics, pp. 33–62. Emerald Group Publishing Limited.