Dr.DEBESH BHOWMIK

Dr.DEBESH BHOWMIK

Monday 20 March 2017

Capital inflows and Silver Standard in India

                Capital inflows and Silver Standard in India

THE FOLLOWING ARTICLE IS PUBLISHED IN THE CHANDIGARH CONFERENCE ON 09-11 MARCH,2017,CONDUCTED BY IMRF

VOLUME-3 ,ISSUE-1,73-80



Capital inflows and Silver Standard in India
Dr.Debesh Bhowmik (Retired Principal and Ex.Associate Editor-Arthabeekshan-Journal of Bengal Economic Association)
Abstract
In this paper author tries to relate gold and silver inflows with GDP,GDP per capita,export, import and gold silver price ratio in India during silver standard regime from 1851 to 1893. Author used semi-log,double-log regression models, Johansen cointegration and VAR models (1991,1996) and Bai-Perron model(2003) for structural change taking data from Maddison(2006) and Ambedkar(1923). The paper concludes that gold inflows during 1851-1893 had decreased at the rate of 0.34% per year insignificantly but it was nonstationary, convergent and had no structural breaks. Silver inflows during 1851-1893 had increased at the rate of 1.51% per year insignificantly and found nonstationary and convergent  and had one upward structural break in 1857.No cointegration among gold or silver inflows with GDP,GDP per capita, export, import and gold silver price ratio was found during 1851-1893 where VAR model was unstable and nonstationary and impulse response functions were diverging. Semi-log linear regression model among silver inflows and gold inflows with those variables were also insignificant although GDP, export, import,and gold silver price ratio had been increasing at the rates of 0.52%,9.14% ,5.16% and 0.77% per year significantly. But double-log linear regression model suggested that gold inflows had significant impact from GDP,GDP per capita, export, and gold-silver price ratio but had no significant impact of silver inflows from those variables during 1851-1893 respectively. Yet,there is bidirectional causality among gold inflows, GDP, GDP per capita, export, import and gold silver price ratio significantly during the given period. Even, there were sharp depreciation of rupee sterling rate,falling silver price ,silver production and rising gold price and gold production during the silver standard regime. Thus, gold and silver inflows could not synthesize the silver standard more effective in macro-dynamic adjustment during 1851-1893 although the series of managerial experiments of the commissions and government are equally responsible for instability of the silver standard in India which was equally identical with gold standard in England
Key words-Net gold inflows, net silver inflows, silver standard, GDP, export, import, cointegration, VAR
JEL-E42,F33,N10,N20
I.Introduction
Silver standard in India was introduced in 1835 but the act of XVII and the act of XXI in 1835 declared both silver coin and copper coins as legal tender ,on the other hand , gold coin was not legal tender yet it was circulated. Later on, in 1861 by Act of XIX , gold coin was treated as legal tender. In 1861,the paper currency notes were circulated. The gold:silver was 1:15.5 and rupee sterling rate was fixed at 1s10.5d where exchange was governed by relative values of gold and silver.
During long 400 years from 1493 to 1893 , gold and silver production were more or less uniform but during 1600-1700,index of gold rose from 130 to 176,which rose to 270 during 1700-1800.In 1870,the index of gold production stipulated to 2124 as compared to 450 for silver. Even the rupee sterling rate depreciated and price of gold silver ratio appreciated to a lager extend. India was one of chief producer of the silver and gold but she was the net importer of both gold and silver which were volatile. Although silver standard during 1873-1893 in India was as like as gold standard in England during 1873-1893,yet British government introduced several policies of mints, currency  circulation as well as bimetallism as an experimental basis which made the silver standard unstable .During this period, most of the countries in the world started to introduce gold standard including British colonies. In India, gold supplies and its prices were stipulating compared to silver, but British Government denied to introduce gold standard in spite of numerous positive signals of implementing gold standard by many commissions. In 1893, England declared gold exchange standard in India where gold was not convertible to rupee but rupee was convertible to sterling which was fixed parity with gold. Therefore , success story of silver standard is little yet there is no vital disturbance in working the system of silver standard in India. 
II.Objective of the paper
In this paper author endeavours to analysis the working of capital inflows in the silver standard in India and its impact on the GDP ,GDP per capita and on international trade and even on the gold silver price ratio during 1851-1893.The net gold import and net silver import were considered as capital flows for the specified period.
III.Methodology and data
Net gold import and net silver import were treated as capital flows in India during 1851-1893.The trend lines of gold inflows, silver inflows, export, import, GDP,GDP per capital, ratio of gold and silver price were calculated by semi-log linear model. Stationary was observed through ARIMA model, structural change was shown by Bai-Perron model(2003).Double-log multiple regression model was used for showing relationship among those variables with gold and silver inflows for the specified period. Since there is no cointegration with gold inflows and other variables and silver inflows with other variables ,author used Johansen VAR model(1991,1996) for showing relationship analysing residual tests and impulse response functions. Even,Granger (1969) model was tested for causality.Data for GDP and per capita GDP were collected from Maddison(2006) and data for all other variables were taken from B.R.Ambedkar(1923). Assume,x1=GDP,x2=GDP per capita,X3=export,x4=import,x5=gold silver price ratio,y1=net gold import,y2=net silver import
IV.Some observations of the model
During the silver standard regime in India from 1851 to 1893,net gold inflows had been decreasing at the 0.34 per cent per year which was insignificant.
Log(y1)=14.770-0.003433t
                   (58.54)* (-0.34)
R2=0.0028,F=0.118 ,DW=0.46 , y1= net gold inflows(imports) ,*=significant at 5% level.
Net gold inflow from 1851 to 1893 is convergent but nonstationary because its AR(1) is convergent and stationary but its MA(1) is convergent and nonstationary.
Log(y1t)=14.63648+0.7232log(y1t-1)+εt+0.083569εt-1+0.26009σ2
                   (40.19)*  (3.46)*                  (0.24)            (6.76)*
R2=0.58  ,F=18.57  ,DW=1.97  ,inverted AR root=0.72 , inverted MA root=-0.08 ,*=significant at 5% level.
This series has no structural breaks during the period.
On the other hand, net inflow of silver in India during silver standard from 1851 to 1893 had been stipulating at the rate of 1.51% per year which was insignificant.
Log(y2)=15.034+0.015138t
                   (43.84)*  (1.115)
R2=0.029 ,F=1.24 , DW=1.36 , y2=net inflow of silver ,*=significant at 5%.

The net inflow of silver in India during 1851-1893 is nonstationary but convergent which is shown by ARIMA(1,1,1) model.It is not a good fit yet it is stable.
Log(y2t)=15.363+0.48873log(y2t-1)+εt-0.183616εt-1+1.0639σ2
                   (35.71)*   (0.96)               (-0.34)             (7.32)*
R2=0.11 , F=1.64,DW=1.96,inverted AR root=0.49,inverted MA root=0.18 ,*=significant at 5% level.
Net inflow of silver has one upward structural breaks in 1857 only.This is verified by Bai-Perron test(2003)in which HAC standard errors and covariance was assumed and trimming 0.15 with maximum 5 beaks is assumed.
Table-1: Structural breaks of net inflow of silver
Variables
Coefficient
Standard error
T statistic
Prob

1851-1856=6 obs


c
14.2328
0.3368
42.25
0.00

1857-1893=37 obs


c
15.537
0.218
71.30
0.00
R2=0.17  ,F=8.64*  ,DW=1.62;Source-Computed by author

Double log multivariate regression model showed that One per cent increase in GDP,GDP per capita, export, import, gold silver price ratio and net silver inflow led to 12.68% decrease ,19.27% increase,1.89% increase ,1.47% increase,9.93% decrease and 0.13% increase in net inflows of gold per year respectively where relation between gold inflows and GDP,GDP per capita,export, gold silver price ratio are significant at 5% level.
Log(y1)=-19.359-12.683log(x1)+19.275log(x2)+1.89log(x3)+1.47log(x4)-9.938log(x5)+0.1319log(y2)
                  (-0.56)   (-1.99)*          (2.86)*          (2.54)*          (1.53)       (-3.35)*           (1.35)
R2= 0.48, F=5.67*   ,DW=1.24 , where x1=GDP,x2=GDP per capita,x3=export,x4=import,x5= gold silver price ratio,y2= net inflows of silver
Similarly, one per cent increase in GDP,GDP per capita, export, import, gold silver price ratio and net gold inflow per year led to11.11% fall,12.13% rise,1.86% increase,0.045% rise,1.47% increase and 0.37% increase in net silver inflows in India per year during 1851-1893 in silver standard regime which are all insignificant.
Log(y2)=2.739-11.1162log(x1)+12.139log(x2)+1.86log(x3)-0.045log(x4)+1.47log(x5)+0.37log(y1)
                (0.047) (-1.10)               (0.98)            (1.416)         (-0.027)          (0.25)            (1.35)
R2=0.24 ,F=1.89 , DW=1.59 ,  
To show linear combination of silver inflows with other variables, Johansen Cointegration test suggests that there is no cointegrating vector shown by Trace and Max Eigen Statistic (Table-2).
Table-2:Cointegration test
Hypothesised no of CEs
Eigen value
Trace statistic
0.05 CV
Prob*
None
0.524
113.692
125.615
0.211
At most 1
0.445
83.189
95.753
0.266
At most2
0.412
58.993
69.818
0.267
At most3
0.299
37.219
47.856
0.337
At most4
0.245
22.630
29.797
0.264
At most5
0.236
11.095
15.494
0.205
At most6
0.0006
0.026
3.841
0.87
Hypothesised no of CEs
Eigen value
Max Eigen statistic
0.05 CV
Prob*
None
0.524
30.502
46.231
0.75
At most 1
0.445
24.196
40.077
0.825
At most2
0.412
21.774
33.876
0.625
At most3
0.299
14.588
27.584
0.779
At most4
0.245
11.534
21.131
0.593
At most5
0.236
11.068
14.264
0.150
At most6
0.0006
0.0266
3.841
0.870





*=Mackinnon Haug Michelis(1999) p values
Since there is no cointegration ,The estimated VAR model is given below.
Δx1t=8451.51+0.212Δx1t-1-2.161Δx2t-1+3.10E-06Δx3t-1+1.53E-05Δx4t-1+174.27Δx5t-1-9.07E-06Δy1t-1+7.32E-06Δy2t-1
         (3.6)*       (1.02)        (-0.54)           (0.32)             (1.27)                   (1.45)               (-0.22)                  (0.51)
R2=0.92  ,F=56.29  , AIC=14.47  SC=14.80
Δx2t=313.711-0.025Δx1t-1+0.796Δx2t-1+1.20E-07Δx3t-1+3.32E-07Δx4t-1+7.67Δx5t-1-8.67E-07Δy1t-1+2.37E-07Δy2t-2
           (3.97)*   (-3.53)*           (5.83)*      (0.365)             (0.807)             (1.86)             (-0.86)            (0.48)
R2=0.765  , F=15.81  , AIC=7.72  , SC=8.05
Δx3t=-46481278+4953.34Δx1t-1-137563.9Δx2t-1+0.466Δx3t-1+0.068Δx4t-1+4792340Δx5t-1+1.919Δy1t-1-0.355Δy2t-1
            (-1.08)           (1.27)                (-1.84)         (2.59)*           (0.305)         (2.14)*             (2.57)*         (-1.33)
R2=0.96  ,F=130.0*  , AIC=34.14 , SC=74.47
Δx4t=-1.11E+08+9232.72Δx1t-1-84374.41Δx2t-1+0.084Δx3t-1+0.3377Δx4t-1+3264343Δx5t-1+0.2086Δy1t-1-0.235Δy2t-1
            (-3.92)*   (3.61)*            (-1.72)               (0.713)           (2.29)*             (2.21)*          (0.425)            (-1.33)
R2=0.98  , F=247.81*  , AIC=33.3  , SC=33.63
Δx5t=4.0918-8.24E-05Δx1t-1+0.00176Δx2t-1+4.27E-09Δx3t-1-2.07E-08Δx4t-1+0.716Δx5t-1-6.34E-08Δy1t-1+1.22-E08Δy2t-1
             (1.8)             (-0.401)         (0.439)       (0.45)               (1.75)                (6.06)*               (-1.73)              (0.908)
R2=0.97  , F=212.09*   , AIC=0.629  ,   SC=0.96  
Δy1t=-8108862+636.61Δx1t-1+4042.18Δx2t-1-0.0215Δx3t-1-0.00204Δx4t-1-76501.53Δx5t-1+0.8244Δy1t-1+0.033Δy2t-1
            (-0.81)       (0.708)            (0.23)             (-0.51)           (-0.03)                (-0.14)            (4.78)*             (0.54)
R2=0.67 , F=10.28  , AIC=31.21  , SC=31.54
Δy2t=-63316662+3366.48Δx1t-1+311.52Δx2t-1-0.034Δx3t-1-0.204Δx4t-1+1705435Δx5t-1+1.036Δy1t-1+0.235Δy2t-1
           (-2.56)*      (0.50)                  (0.007)      (-0.33)           (-1.58)              (1.3)            (2.42)*            (1.53)
R2=0.42  , F=3.63  , AIC=33.03  , SC=33.36 , *=significant at 5% level
The estimated VAR model states that [i] change of GDP per capita is negatively related with change of previous period’s GDP and positively related with previous period’s GDP per capita,[ii] change of export is positively related with change of previous period’s export,ratio of gold and silver price,change of gold inflows,[iii] change of import is positively related with change of previous period’s GDP,import and gold silver price ratio,[iv]change of gold and silver inflows are positively related with their previous period .Other relations are insignificant.         
This VAR model is unstable because one of its 7 roots is greater than one ,two roots are imaginary and 4 roots are less than one (,ie 1.012998,0.853108±0.084339i,0.527198, 0.307550,0.272081),so all roots do not lie inside the unit root circle.

The impulse response functions are diverging so that it is nonstationary and unstable.The exogenous shocks could not tend the model into equilibrium.

VII.Conclusion
The paper concludes that gold inflows during 1851-1893 had decreased at the rate of 0.34% per year insignificantly but it was nonstationary,convergent and had no structural breaks.Silver inflows during 1851-1893 had increased at the rate of 1.51% per year insignificantly and found nonstationary and convergent  and had one upward structural break in 1857.No cointegration among gold or silver inflows with GDP,GDP per capita,export,import and gold silver price ratio was found during 1851-1893 where VAR model was unstable and nonstationary and impulse response functions were diverging.Semi-log linear regression model among silver inflows and gold inflows with those variables were also insignificant although GDP,export,import,and gold silver price ratio had been increasing at the rates of 0.52%,9.14% ,5.16% and 0.77% per year significantly.But double-log linear regression model suggested that gold inflows had significant impact from GDP,GDP per capita,export,and gold-silver price ratio but had no significant impact of silver inflows from those variables during 1851-1893 respectively.Yet,there is bidirectional causality among gold inflows,GDP,GDP per capita,export,import and gold silver price ratio significantly during the given period.Even,there were sharp depreciation of rupee sterling rate,silver price and production and gold price increased with production during the silver standard regime.Thus,gold and silver inflows could not synthesize the silver standard more effective in macro-dynamic adjustment during 1851-1893 although the series of managerial experiments of the commissions and government are equally responsible for instability of the silver standard in India which was equally identical with gold standard in England.

References
[1]Ambedkar,B.R.(1923).The problem of rupee:Its origin and solution,P.S.King & sons Ltd,London.
[2]Bhowmik,Debesh.(2016). B.R.Ambedkar and Silver standard in India,paper will be presented in International seminar in Magadh University.
[3]Bhowmik,Debesh.(1992-93).A profile of Dr.B.R.Ambedkar’s view on the concept of Convertibility of Rupee,published in Indian Economic Association 76th Conference volume-1993-94,Bombay,14-16,republished in in Debendra Kumar Das (ed)-Great Indian Economists: Their creative vision for socio-economic development,314-320,(Deep and Deep Publications, New Delhi)
[4] Bhowmik,Debesh(ed).(2016).International Monetary System:Past,Present and Future,Regal Publications, NewDelhi.
[5]Bhowmik,Debesh.(2003).Essays on International Money,Deep and Deep Publications, New Delhi.
[6]Ching,Jao Yu.(1965).The Silver Standard and China,1911-1935,University of Hongkong, Thesis for Master of Arts.
[7]Coyajee,J.C.(1983).Development of Currency System in India,Deep and Deep Publications, NewDelhi
[8]Das,Debendra Kumar.(2004).Relevance of Dr.Ambedkar’s Economic Ideas on the Problems of Indian Currency:Some Observations,published published in Indian Economic Association 76th Conference volume-1993-94,Bombay,7-10,republished in Debendra Kumar Das (ed)-Great Indian Economists:Their creative vision for socio-economic development,299-306,Deep and Deep Publications,NewDelhi).
[9]Fernholz,Ricardo.Mitchener,Kris James &Weidenenmier,Merc.(2013).Pulling up the Transished Anchor:The End of silver as a Global Unit of Account,www.cepr.org
[10]Granger, C.W.J.(1969). Investigating Causal Relations by Econometric Models and Cross-spectral Methods, Econometrica, Vol-37,No-3, 424-438.

[11]Jadhav,Narendra.(2005,April).Exchange Rate Regime and Capital Flows:The Indian experience, Chief Economists’ Workshop,Bank of England.

[12]Johansen,S.(1988).Statistical Analysis of Co-integrating Vectors. Journal of Economic Dynamics and Control, Vol. 12, 231-254.
[13]………….(1991,Nov). Estimation of Hypothesis Testing of Co-integration Vectors in Gaussian Vector Autoregressive Models,  Econometrica  59,1551-80
[14]………….(1994).The role of the constant and linear terms in co-integration analysis of nonstationary variables. Econometric Reviews,13,205-229
[15]Johansen, S.( 1995). Likelihood-Based Inference in Co-integrated Vector Autoregressive
Models. Oxford  University Press.

[16]……….(1996). Likelihood-Based Inference in Co-integrated Vector Autoregressive Models, 2ndEdition,Oxford University Press.

[17]……….. & Juselius,K.(1990).Maximum likelihood estimation and co-integration with application to the demand of money. Oxford Bulletin of Economics and Statistics,Vol-52,No-2,169-210.

[18]Keynes, J.M.(1913).Indian currency and finance. Macmillan and co.
[19]Leavens,Dickson.H.(1939).Silver Money,Principia Press Inc,Bloomington,Indiana
[20]Ma,Debin.(2012).Money and Monetary System in China in the 19-20th century:An Overview, Working Paper No-159/12,London School of Economics.
[21]Maddison, A.( 2006).The World Economy: Vol-1 and 2. OECD: Paris
[22]…………..(2007).Contours of the world economy,1-2030:Essays in Macro Economic History,OUP.
[23]Roy,Alokananda.(2004).Ambedkar versus Keynes-On Indian Currency:In the Light of Contemporary Relevance,published in published in Indian Economic Association 76th Conference volume-1993-94,Bombay,11-13 republished in  Debendra Kumar Das (ed)-Great Indian Economists:Their creative vision for socio-economic development,307-313,Deep and Deep Publications, New Delhi.
[24]Velde,Francois R.& Weber,Warren E.(1998).A Model of Bimetallism,Federal Reserve Bank of Minneapolis,Research Department,Working Paper-588.
[25]Zaraga,Ludovico.(2014).Effects of Exchange Rate Regimes on Trade: the Case of Silver Standard India,Northwestern University,Nov, 24.