Independent Work

(1)

Independent Work

Nationalekonomi Economics

Title

Business Cycle and Barter Trading in Modern Economics: Cointegration approach

Tommaso Rotolo

(2)

Institutionen för samhällsvetenskap

Business Cycle and Barter Trading in Modern Economics:

Cointegration approach

Författare: Tommaso Rotolo Handledare: A. Khalik Salman Kurs: NA006G Nationalekonomi GR (C) Kandidatuppsats I nationalekonomi 15 hp

Termin: 2013

(3)

Abstract

This thesis analyses the causes of the resurgence of barter, its organization and its relationship with macroeconomic variables during the business cycles. After a first part mostly descriptive of the functioning of nowadays barter systems, the focus moves to the analysis of many studies about the macroeconomic influence on this market. Data from barter exchange organization in Switzerland and macroeconomic variables are used to test the significance of their reciprocal relation. The Ordinary Least Squares and Cointegration tests are employed to analyse the data. This study tested two main macroeconomic factors as a proxy for the business cycle; they are Gross Domestic Product growth rate and unemployment rate.

The empirical results show a significant negative relationship between unemployment rate and Wir turnover percentage change and an insignificant positive relationship between Gross Domestic Product growth rate and Wir turnover percentage change.

Purpose

The purpose of this paper is to analyse whether an economy could get better involving in modern barter activities. The work is divided into two different parts.

The focus of the first part is on previous studies about the relationships between barter values and economic aggregates in some case studies.

The second part concerns statistical estimations about barter spread, Gross Domestic Product and Unemployment with the aim of finding relationships between barter and macroeconomic variables during business cycles.

The methodology of this study used two alternative methods of estimation. The first is the Ordinary Least Squares regression; the second is the cointegration analysis to test the long-run relationship between the dependent variable and the independent variable. The Augmented Dickey-Fuller method is used to test the variables an choose the better estimation method accordingly. We used as variables the percentage changes of Wir turnover, the annual Gross Domestic Product Growth and the annual Unemployment Rate.

(4)

1 Introduction

Barter trading was the first form of commerce in the human history, it was a simple exchange system where people could swap goods and services freely. With the civilisation people needed more and more economic advances to meet their social needs. Hence barter was not enough to meet the requirements of a raising civilization and society. Indeed among the several issues submitted by barter there is the problem of the double coincidence of wants; in fact in the bilateral barter system it is very rare to find commercial partners having what you need and who need the goods or services you can offer. There is the problem of finding agreements about the values of the goods; in fact there is not a common measure of value for products, thus often people can not find agreement about the value of a good in terms of another good. There is the problem of the lack of information about the characteristics and attributes of goods available for exchange. Finally, the gap time problem, that is exchanges occur only when both partners have got the goods in the same time, it is not possible to have exchange deferred in time.

All these issues led the human civilization to the discovery of money. Money simply solved these difficulties becoming the unitary medium of exchange for most economic transactions. Money allows unilateral transactions, it covers the role of unit of account that is it measures all goods’ values, it is a store of value, it allows to save worth over time and to make deferred transactions. In short it allowed much faster and easier circulation of goods.

A rediscovery of bartering happened during the Great Depression. In the 1930s all the economies of the western world were affected by a heavy crisis leading to an overall lack of liquidity and unemployment. Hence people had not money and jobs anymore, but they still had goods and skills. Consequently they began bartering again; they exchanged work for food, goods for other goods and so on. They organized themselves into some kind of corporations to help each other to survive the hard conditions. The biggest associations in the U.S. were the Unemployed Citizens League of Seattle, the Unemployed Citizens League of Denver, the Minneapolis Organized Unemployed Inc. and the Natural Development Association of Salt lake City.

(6)

At the same time also in Europe there was a resurgence of barter. It happened in Switzerland in 1934, when a group of business men affected by the crisis decided to organize themselves into an economic cooperative based on barter exchanges.

They named it WIR, the German word for “we” (It is also called Wirtschaftsring that is “economic circle”). It is still working and it is considered the most successful barter system in the world. These “Modern Barter” trading systems have worked well for lot of years.

This barter association overcome the classic concept of barter. The creators knew that the original barter system could not meet the needs of an advanced economy.

So they used an innovation to solve problems because of barter had been put aside. They moved to a multilateral barter network based on a credit system laying the basis of modern barter trading. They issued an alternative currency (usually named Scrip), parallel to the official one. So the members of the network could sell or purchase goods using this credit currency as medium of exchange. In most cases barter organizations chose and choose virtual credit currency which is worth as the official currency (e.g. one credit is worth one dollar).

In the 1980s a downturn occurred because of a oil crisis giving a further push to the spread of barter. Indeed the barter networks grew more and more thank to many advances; in fact technologies began to cover a main role in barter transactions, just think about electronic communications, such as internet, which made easier and faster each transaction.

Modern barter organizations are profit-seeking corporations which control all transactions occurring in the network and keep track of the credits and debits of members. They earn because of fees required to join the network and for commissions they impose on each transaction, which are always paid in cash.

They perform a coordination function of everything happening within the organization, such as a sort of central bank of the system; it means that they ensure the recognition of all the credits and debits and they keep credit account for each member of the network. Furthermore in most cases they cover a brokerage function, helping customers finding trade partners to sell or buy goods.

(7)

Finally, nowadays barter organizations also perform functions of commercial bank issuing trade credit loans to the customers to stimulate trade activity.

Partners can sell their own goods earning just credit, which are often represented by some kind of alternative currency, virtual money worth only inside the barter circuit. Then, it is possible to purchase products or services using the credits earned before. This kind of transactions are not supposed to be completely made with these Scrips, but it is up to the parts decide whether use for part of the value the official currency. These multilateral compensation systems allow the economies to carry on even in crisis time, when liquidity is scarce and people do not have money enough to purchase goods, from the first households’ needs to the factors of production of companies.

(8)

2 Previous studies

The aim of this section is to show the previous studies about the relationships between Barter spread and macroeconomics factors such as GDP and unemployment rate.

Marvasti and Smyth (1998, 2011) wrote many works about barter trading, mostly examinations of the determinants of barter and its relation with business cycles in U.S. economy.

Data from the International Retail Trade Association say that last year in the United States about 470.000 firms joined actively the barter systems, for $12 billion in annual transactions.

The figures 1 and 2 show the strong raise of barter between 1974 and 1994 in the United States. Figure 1 contains the trend of barter volume differentiating large barter exchanges, small barter exchanges and total barter exchanges. Figure 2 show the increase of the number of clients joining the barter systems in the same period. These graphs display a quite constant rate of growth for each variable considered.

Fig. 1. Trade volume (in millions of barter dollars): (─□─) large barter exchanges, (─∆─) small barter exchanges, (─○─) total barter exchanges.

Source: International Reciprocal Trade Association.

(9)

Fig. 2. Number of clients: (─■─)

Source: International Reciprocal Trade Association.

Marvasti and Smyth (1998, 2011) distinguish and examine barter with the division between commercial barter and corporate barter.

In fact it is possible to distinguish modern barter in three large categories:

commercial barter, corporate barter and countertrade.

Commercial barter, or retail barter, deals with households and small firms; it takes place sometimes on haphazard and personal contacts and it is managed by organizations working mostly by web networks.

Corporate barter deals with large firms. Usually this kind of network has also a sort of brokerage function, helping members to find partners to sell or to purchase goods.

Countertrade is the exchange between different countries, where whole or part of the goods exchanged are paid using a credit compensative or barter system. It is convenient for multinationals or governments that have difficulties with currency conversion and that have few financial resources but enough commodities to trade.

Marvasti and Smyth (1998, 2011) point out that for large firms barter is a continuous process, used as a supplement to money transactions to extend business and to entry into new markets. Conversely, for small firms it is a tool to deal with economic downturns. Small firms face the fluctuating demand reducing excess inventories in the barter markets instead of selling products at less than wholesale prices.

Many reasons support that firms’ size affects interest in barter. Big companies are clearly less likely to fail.

Big companies have an higher safety to get the trust of banks, they have easier access to credit. Conversely, small firms have a worse treatment in the credit

(10)

market, so in crisis time they feel an harder lack of liquidity and they are more likely to start bartering to sell inventories and to keep on the activity.

Therefore, big companies are less likely to maintain the same level of unemployment and capacity utilisation in crisis time; in fact empirical evidences show that during economic booms large corporations create more job and during recession they Fire more workers. Instead, small firms are less likely to change their level of employment with business cycles.

Furthermore market power is a factor for the strength of firms. Indeed high market power allows large companies to cut prices in order to keep stable sales and to avoid bartering in recession.

Marvasti and Smyth (1998, 2011) found that the number of barter exchanges and the number of clients is slightly related to business inventories and strong related to inflationary trend. Yet, contrary to what most of the economists think, empirical evidences show that the values of the products inside the barter network are higher than in the cash market. The reason is easy to understand, even if the worth of goods is estimate in the official currency, for instance dollars, one credit-dollar has a lower utility than one cash dollar. The sellers in the network earn credits, but often they have to wait for some time to find fine goods to buy. In that waiting credit dollars are immobilized, they do not earn any interest and they cannot be used out of the network. Furthermore the others two “defects” which lower the utility of the credit dollars are the limitation of choices and the existence of fees and commissions charged by barter organizations..

But the main results of Marvasti and Smyth studies (1998, 2011) show that corporate barter is pro-cyclical and commercial barter is counter-cyclical.

Corporate barter spread is slightly pro-cyclical, it means that it is positive related to the economic trend and its diffusion over time gave a contribution to the economic growth. Conversely, their study find that commercial barter is countercyclical, that is when the economy suffers a crisis, inflation and unemployment rise, GDP drops, commercial barter spreads more keeping on the economic activities for small firms.

(11)

Noguera (2004) analyses the trend of barter in Russia during the 1990s. The particularity of the Russia case is that empirical data show a negative correlation between barter exchanges and inflation, contrary to the empirical evidence of most studies.

Yet the most impressive data is shown in figure 3, that is the incredible rise of barter in industrial sales from 5-10% in 1992 to 54% peak in 1998, in the same month of the Ruble collapse. Since this moment onward barter transactions drop constantly until reaching the initial value.

The case of bartering in Russia is different from the American IRTA or the Swiss Wir, in fact in Russia several goods were usually used as medium of exchange, such as cigarettes, vodka, cars, etc. Hence basically Russia lived a resurgence of old barter, with no credit, no unit of account and no multilateral trade.

The explanation of this phenomenon could be finding in the tightening of credit market. Noguera (2004) explains this trend with two reasons: the switch from seignorage to borrowing as a source of government financing and lenders withdrawal from the market as the economy started dropping.

Therefore, firms looked for alternative transactions systems to survive the lack of liquidity, rediscovering the utility of bartering. The main object of the research is finding whether bartering is an optimal decision in credit market crisis or it is an Hobson’s choice, that is the unique possibility without alternatives.

Noguera (2004) identifies three different sceneries.

First, if barter technologies are expensive and hard to meet, firms remain in cash market until credit market becomes too tight to carry on, and then credit rationing makes them moving to barter as the only possible solution.

Second, if barter is easily achievable, as soon as credit market starts tightening less risky firms move to barter avoiding the credit rationing, and this is an optimal choice.

Third, when market becomes tighter and credit rationing appears, barter is an optimal choice for less risky firms and a Hobson’s choice for the riskier ones.

(12)

Fig. 3: Barter and depreciation rate in Russia (1992-2003)

Source: Real exchange rate, depreciation rate. Russian economic Barometer.

Stodder (2007) scans the functioning of the WIR system in Switzerland, the world largest barter network.

Table 1 shows the incredible spread of the Wir currency in the last fifty years.

Since 1948, when barter had an insignificant weight in the Swiss economy, bartering began being more and more spread in the economy. The peak was in 1993, when barter trading volume was about 2300 % of the volume in 1948. Since 1993 the barter variables remain in a costant situation, together with a period of good economic conditions.

Stodder (2007) focus on the macroeconomic effects of this virtual currency using several econometric tools to test the data, such as the Error Correction Model and the Cointegration model. This study uses the log linear model to analyse the relationships between the Wir turnover (Wir virtual money in circulation) and the main economic variables: M2, GDP, numbers of unemployed and exchange rate.

(13)

Table 1. Participants, Total Turnover, Credit, and Credit/Turnover, WIR-Bank, 1948-2003 (Total Turnover and Credit Denominated in Millions of Current Swiss Franks)

Stodder’s analysis (2007) points out different results with short or lung run. In the long run it finds a positive relationship between M2 and Wir. It shows that in the long run the quantity of Wir in circulation is positive related to the Gross Domestic Product and with the number of unemployed (as does also the official Swiss Franc). Conversely, in the short run the Wir turnover is negative related to M2 and Gross Domestic Product.

(14)

The significant meaning of this achievements is that Wir brings benefits for the economy in the long run, it is positive related to the economic growth (GDP). Yet its biggest benefit is the countercylicity. It means that when the economy is in crisis, and then the GDP decreases, the Wir turnover increases. In other words, Wir gives its best in economic downturns, when hard times deny economic activity.

Indeed in financial crisis, when the lack of liquidity is a huge problem for firms, the Wir spreads more helping the economy carrying on.

(15)

3 Empirical Analysis

The previous studies use the analysis of time-series data with the aim of estimate relationships between barter and macroeconomic variables.

This empirical analysis is parallel to the studies of Stodder about Swiss Wir.

Contrary to his work, this estimation considers the percentage growth rate of Switzerland using only three different variables between 1980 and 2003: Wir turnover percentage change, Annual Gross Domestic Product growth rate and Annual Unemployment Rate. Furthermore, this work use the Ordinary Least Square model, the Augmented Dickey-Fuller test and the Cointegration methods to test the long – run relationships of the data.

The following figures 4 and 5 contain the trends of the single variables one by one.

The graphs show that since the end of the 1980’s barter trading had lived a strong raise for about ten years. After the peak in 1993, when barter volume reached the level of 2521 millions of Swiss Franc, it started dropping to the 2003. The same is shown by the fig. 6, which represents the percentage change in barter volume in the same period. We can compare this trend with business cycles. In fact in the 1980’s a deep economic recession troubled the United States, leading many banks to the failure and affecting the whole world. Yet in 2000’s economy was in good conditions, exactly in the same time Wir volume decreased. This relationship seems to confirm the conclusions of the previous works, which is barter trading, is countercyclical in the short run.

Fig. 4. Wir turnover 1980-2003

Barter trade turnover in million of current Swiss Franks

0 500 1000 1500 2000 2500 3000

1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003

Year

Wir turnover

(16)

Fig. 5. Participants in Wir network 1980-2003

Fig. 6. Percentage change in Wir turnover and in number of participants 1980-2003

The data used for this analysis are Wir turnover percentage change, Gross Domestic Product percentage growth and Unemployment percentage rate of Switzerland between 1980 and 2003 (table 2).

Year Wir turnover % change Annual GDP % Annual Unemp. %

1980 3,151 5.110 0,197

1981 7,794 1.601 0,185

1982 19,912 -1.309 0,435

1983 31,000 0.639 0,869

1984 20,980 3.008 1,153

1985 28,680 3.674 0,983

1986 22,734 1.859 0,883

1987 28,934 1.585 0,796

1988 24,788 3.278 0,718

1989 16,854 4.331 0,559

1990 15,132 3.675 0,503

Partecipants in Wir network

0 10000 20000 30000 40000 50000 60000 70000 80000 90000

1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003

Year

Partecipants

Percentage change in Wir turnover and partecipants

-10 -5 0 5 10 15 20 25 30 35

1980 1982

1984 1986

1988 1990

1992 1994

1996 1998

2000 2002 Year

Percentage change in Wir turnover Percentage change in number of partecipants

(17)

1992 17,440 0.100 2,465

1993 4,866 -0.185 4,391

1994 -0,476 1.191 4,645

1995 -6,137 0.350 4,186

1996 -3,949 0.628 4,638

1997 -7,824 2.076 5,204

1998 -5,227 2.639 3,887

1999 -7,236 1.311 2,732

2000 -3,218 3.582 1,962

2001 -3,720 1.152 1,805

2002 -0,995 0.443 2,668

2003 -2,424 -0.198 3,834

Table 2. Annual Wir Turnover percentage change, Annual Gdp growth rate, Annual Unemployment rate.

3.1 The Model

The model considered by this work is the following:

Y = α + βX + γZ + μ

Where:

Y = Wir turnover

X = Annual GDP growth rate

Z = Switzerland Unemployment rate

α , β , γ = parameters associated with the variables e = error term

3.2 Regression Results

In these sections we present the most important results of the study. First, model has been created using the Ordinary Least Square (OLS) method with all included variables at levels. This equation represents the short term relationship between Wir turnover percentage change, Gross Domestic Product growth rate and Unemployment rate. The results from the OLS are shown below.

(18)

Model 1: OLS, observations 1980-2003 (T = 24) Dep. variable: Wir_turnover__

Coefficient Std. Error t-ratio p-value

const 20.6219 4.12136 5.0037 0.00006 ***

Annual_GDP__ 0.000130899 0.00132792 0.0986 0.92241 Annual_Unemp___ -5.58757 1.25437 -4.4545 0.00022 ***

Table 3. Ordinary Least Squares estimation.

The estimated p-value for the GDP is 0,9224, that is non-significant because 0,9224 > 0,05, which is the level of significance chosen alpha (α). Furthermore the GDP percentage change coefficient is zero, and these facts do not allow us to reject the null hypothesis. Conversely, the results show a significant p-value of 0, 0002 for unemployment rate. Hence we can reject the null hypothesis for unemployment rate.

The coefficient of unemployment rate is -5,5876; it means that Wir turnover percentage change and unemployment rate are linked by a negative relationship, and that a 5,5876% increase in Wir turnover corresponds to a 1% drop in Unemployment rate.

The regression also shows a R-squared coefficient of determination equal to 0, 5059, that is a good approximation of the real data considering the cross-sectional regression. The Durbin-Watson statistic find a value of 0,5575. To reject the hypothesis of autocorrelation the D.W. test result should be around 2. Hence our estimations say that the variables have autocorrelation.

3.3 Augmented Dickey-Fuller Test

Hence we use the Augmented Dickey Fuller test to seek out whether there is or not a unit root in the time-series. As shown in table 4, we test the variable series for levels at first and then for the first differences. We consider the critical values for α

= 5%; the results show that is possible to reject the null hypothesis only for the first Mean dependent var 8.981066 S.D. dependent var 13.18748 Sum squared resid 1975.976 S.E. of regression 9.700212

R-squared 0.505996 Adjusted R-squared 0.458948

F(2, 21) 10.75489 P-value(F) 0.000608

Log-likelihood -86.98369 Akaike criterion 179.9674

Schwarz criterion 183.5016 Hannan-Quinn 180.9050

rho 0.653633 Durbin-Watson 0.557522

(19)

is stationary for the first differences. Conversely, the test on the level data indicates that the variable Wir turnovers percentage change and Gross Domestic Product have a unit root.

Test for stationarity - calculated ADF statistic.

Variables Levels First Differences

WIR turn. % ^-2,4147 ^-5,07625

GDP % ^-3,0191 ^-5,40713

U. Rate % ^-3,43871 ^-3,30469

Table 4. Augmented Dickey-Fuller test.

The test has been conducted with a trend and a constant. The numbers of lags is one because of the few observations and the annual nature of the data.

Since the estimations for the levels can not reject the unit root hypothesis for all variables, we prefer to consider the cointegration method in order to avoid spurious regression. Furthermore the Cointegration test reports to long run elasticity, instead the O.L.S. reports to short run relationships. For these reasons we test the data also with the cointegration method.

3.4 Cointegration estimation

Since the series is non stationary, we estimate the long-run relationships between Wir turnover percentage change, Annual Gross Domestic Product growth rate and Annual Unemployment Rate using the cointegration systems.

Johansen and Engle-Granger tests are shown below.

a.Johansen test:

Number of equations = 3 Lag order = 1

Estimation period: 1981 - 2003 (T = 23) Coefficients, VAR in differences (1 x 3) -0.24244 -0.23078 0.15813 Coefficients, eqns in lagged levels (1 x 3) 9.4770 1463.4 2.0415

Sample variance-covariance matrices for residuals VAR system in first differences (S00)

38.733 -1627.5 -0.76868

10% 5% 2,5% 1%

Critical Values -2,89 3,19 -3,46 -3,77

(20)

-1627.5 2.8312e+006 -52.174 -0.76868 -52.174 0.53786

System with levels as dependent variable (S11) 168.01 3391.6 -14.763

3391.6 2.3961e+006 -569.35 -14.763 -569.35 2.6515

Cross-products (S01)

-17.187 -719.97 -2.6374 2703.7-1.4153e+006 350.05 2.0096 -528.86 -0.28555 Case 3: Unrestricted constant

Log-likelihood = -215.382 (including constant term: -280.653) Rank Eigenvalue Trace test p-value Lmax test p-value 0 0.73700 44.703 [0.0004] 30.719 [0.0010]

1 0.44233 13.984 [0.0826] 13.432 [0.0659]

2 0.023723 0.55219 [0.4574] 0.55219 [0.4574]

Corrected for sample size (df = 19) Rank Trace test p-value

0 44.703 [0.0021]

1 13.984 [0.1131]

2 0.55219 [0.4887]

eigenvalue 0.73700 0.44233 0.023723 beta (cointegrating vectors)

Wir_turnover__ 0.00010356 -0.095840 0.049718 Annual_GDP__ -0.00064036 8.8194e-005 0.00014850 Annual_Unemp___ -0.29661 -0.78790 -0.21760 alpha (adjustment vectors)

Wir_turnover__ 1.2415 3.6616 -0.38753 Annual_GDP__ 802.73 -659.75 -151.92 Annual_Unemp___ 0.42357 -0.014255 0.083509 renormalized beta

Wir_turnover__ 1.0000 -1086.7 -0.22849 Annual_GDP__ -6.1836 1.0000 -0.00068247 Annual_Unemp___ -2864.2 -8933.7 1.0000 renormalized alpha

Wir_turnover__ 0.00012857 0.00032293 0.084324 Annual_GDP__ 0.083130 -0.058186 33.056 Annual_Unemp___ 4.3864e-005 -1.2572e-006 -0.018171 long-run matrix (alpha * beta')

Wir_turnover__ Annual_GDP__Annual_Unemp___

Wir_turnover__ -0.37007 -0.00052964 -3.1689 Annual_GDP__ 55.760 -0.59478 314.77 Annual_Unemp___ 0.0055620 -0.00026009 -0.13257

b. Engle-Granger test:

Step 1: cointegrating regression Cointegrating regression -

(21)

coefficient std. error t-ratio p-value --- const 20.6219 4.12136 5.004 5.94e-05 ***

Annual_GDP__ 0.000130899 0.00132792 0.09857 0.9224 Annual_Unemp___ -5.58757 1.25437 -4.454 0.0002 ***

Mean dependent var 8.981066 S.D. dependent var 13.18748 Sum squared resid 1975.976 S.E. of regression 9.700212 R-squared 0.505996 Adjusted R-squared 0.458948 Log-likelihood -86.98369 Akaike criterion 179.9674 Schwarz criterion 183.5016 Hannan-Quinn 180.9050 rho 0.653633 Durbin-Watson 0.557522 Step 2: testing for a unit root in uhat

Augmented Dickey-Fuller test for uhat including one lag of (1-L)uhat

sample size 22

unit-root null hypothesis: a = 1

model: (1-L)y = (a-1)*y(-1) + ... + e

1st-order autocorrelation coeff. for e: -0.049 estimated value of (a - 1): -0.427074

test statistic: tau_c(3) = -2.66206 asymptotic p-value 0.3986

There is evidence for a cointegrating relationship if:

(a) The unit-root hypothesis is not rejected for the individual variables.

(b) The unit-root hypothesis is rejected for the residuals (uhat) from the cointegrating regression.

The Cointegration estimation finds about the same results of the Ordinary Least Squares regression, that is again a significant p-value for Unemployment rate and non significant one for Gross Domestic Product. Also the coefficients found by the Engle-Granger method are the same of the regression method.

3.5 Results of the estimation

This model uses two alternative methods of estimation. The first is the normal regression, the Ordinary Least Squares; the second is the cointegration test, the Johansen and the Engle-Granger tests as well. We used the percentage changes of Wir turnover, the annual Gross Domestic Product Growth and the annual Unemployment rate are proxy variables of the business cycles.

The first empirical analysis is used the Ordinary Least Square method. The results of this study shows that unemployment is negative related to Wir turnover percentage and significance, conversely GDP is positive related to Wir percentage change, but statistically is not significant.

(22)

Hence we test the data with the Augmented Dickey-Fuller Test to find whether the variables have a unit root. Since we consider 5% as critical value, we can reject the null hypothesis for the first differences but we cannot for the levels. For this reason we use the cointegration method to avoid spurious regressions and to have a better estimation of the variables’ relationships. The cointegration test confirms the Ordinary Least Squares result that is a significant negative relationship between unemployment rate and Wir turnover percentage change and an insignificant positive relationship between Gross Domestic Product growth rate and Wir turnover percentage change.

(23)

4 Conclusions

This paper tries to explain the reason which make barter trading one of the more surprising economic innovations in modern economies. Although its primitive roots, barter can no longer be considered an old system of exchange, an inefficient way of trading, and an obsolete way of trading which cannot meet the needs of advanced economies. Barter is an argument of great actuality. Barter networks are growing all around the world, more and more people and companies join these systems to get benefits.

The economists studied bartering by different points of view achieving significant results; they showed with empirical evidences that barter is a further opportunity for firms to get into new markets and make profit, that barter is related with business cycles and it brings benefits to the economy.

In particular our statistical estimations find a significant negative relationship between unemployment rate and Wir turnover percentage change and an insignificant positive relationship between Gross Domestic Product growth rate and Wir turnover percentage change.

These estimations allow us to say that barter trading is a positive factor for the economic growth and its spread brings benefits to the overall economy. Bartering can no longer be considered an alternative way of trading; it is a current and effective economic system which could be seriously taken into account as an instrument of macroeconomic policy.

The significant meaning of this achievements is that Wir brings benefits for the economy in the long run, it is positive related to the economic growth (GDP). Yet its biggest benefit is the countercylicity. It means that when the economy is in crisis, and then the GDP decreases, the Wir turnover increases. In other words, Wir gives its best in economic downturns, when hard times deny economic activity.

Indeed in financial crisis, when the lack of liquidity is a huge problem for firms, the Wir spreads more helping the economy carrying on.

(24)

5 References

Aquaro, Dario, (2012) “Baratto e Moneta Complementare: così le Pmi combattono il credit crunch”, Il sole 24 ore, March 16.

Beltrametti, Filippo Maria, (2010) “Monete complementari e capitale sociale, il caso del Wir”, University of Bologna.

Cresti, Barbara, (2005), “US domestic barter: an empirical investigation”, Applied Economics, 37, 1953-1966.

Garbarine, Rachelle, (1997) “Widgets for Credits in a Cashless Marketplace”, New York Times, August 03.

Gordon, Michael R., (1998) “As Ruble Withers, Russians Survive on Barter”, September 6.

Hitz, Cristoph, (2009) “Barter It” ,Corporate Meetings and Incentives, January.

Howell, Jerry with Chmielewski, Tom (2009), “The complete idiot’s guide to Barter and Trade Exchanges”,New York.

Marin, Dalia and Schnitzer, Monika (2002), “Contracts in Trade and Transition. The Resurgence of Barter”, Massachusetts institute of technology, Cambridge, Massachusetts.

Marvasti, Akbar and Smyth, David J. (2011), “Barter and Business Cycles: Further Empirical Evidence”, American Economist, Vol. 56, No. 2.

Marvasti, Akbar and Smyth, David J (1998), “Barter in the US economy: a macroeconomic analysis” , Applied Economics, 30, 1077-88.

(25)

Noguera, José, (2004) “Is Barter a Hobson’s Choice? A theory of barter and credit rationing” ,Cerge-Ei, September.

Sloane, Leonard, (1995) “The Media Business: Advertising; Bartering is big for media companies. And it may get bigger with insurance for trade credits”, New York Times, July 03.

Spitznagel, Eric, (2012) “Rise of the Barter Economy”, BusinessWeek, April 26.

Stodder, James, (2007) “Residual Barter Networks and Macro-Economic Stability:

Switzerland’s Wirtschaftsring” ,Rensselaer Polytechnic Institute at Hartford, USA, December 27.

Stodder, James, (2009) “Complementary Credit Networks and Macro-Economic Stability: Switzerland’s Wirtschaftsring” , Journal of Economic Behaviour &

Organization, 72, October, pp. 79–95.

Stodder, James, “Corporate Barter and Economic Stabilisation”, International Journal of Community Currency Research.

Electronic References

Barter Trade Development, “Prosperare al tempo della crisi. Storia e caratteristiche

di uno strumento alternativo al denaro”,

http://www.bartertrade.it/download/IL_BARTER.pdf

BarterNews, http://www.barternews.com/

International Monetary Systems, https://www.imsbarter.com/

International Reciprocal Trade Association, http://www.irta.com

Oregon Trade Expansion Network, http://oten.capex.com/oten_intro.htm

(26)

The World Bank, http://www.worldbank.org/

Tradebank, http://tradebank.com/

Trading Economics, http://www.tradingeconomics.com/

Young, Melina, “Business-to-business barter exchange: a viable marketplace”, http://issuu.com/helyipenz/docs/business-to-business-barter-exchange---a-viable- ma

Independent Work