Appendix B: Tables for the ROW Model (MC2 Version)
This is an update of Appendix B in the 1994 book. It presents the ROW model that is part of the MC2 model. The date of the ROW model presented here is December 27, 1998. This is the latest version of the ROW model on this site.
Table B.1: The Countries and Variables in the MC Model
Table B.2: The Variables for a Given Country in Alphabetical Order
Table B.3: The Equations for a Given Country
Table B.4: Equations the Pertain to the Trade and Price Links Among Countries
Table B.5: Links Between the US and ROW Models
Table B.6: The Procedure Used to Create Quarterly Data from Annual Data
Table B.7: Construction of the Balance of Payments Data: Data for S and TT
Table B.1: The Countries and Variables in the MC Model
				Table B.1
                   The Countries and Variables in the MC Model
    
Quarterly Countries Local Currency Trade Share Equations Only
1. US United States U.S. Dollar (mil.) 40. TU Turkey 2. CA Canada Can. Dollar (mil.) 41. PD Poland 3. JA Japan Yen (bil.) 42. RU Russia 4. AU Austria Schilling (bil.) 43. UE Ukraine 5. FR France Fr. Franc (bil.) 44. EG Egypt 6. GE Germany D. Mark (bil.) 45. IS Israel 7. IT Italy Lire (bil.) 46. KE Kenya 8. NE Netherlands Guilder (bil.) 47. BA Bangladesh 9. ST Switzerland Swiss Franc (bil.) 48. HK Hong Kong 10. UK United Kingdom Pound Sterling (mil.) 49. SI Singapore 11. FI Finland Markka (mil.) 50. VI Vietnam 12. AS Australia Aust. Dollar (mil.) 51. NI Nigeria 13. SO South Africa Rand (mil.) 52. AL Algeria 14. KO Rep. of Korea Won (bil.) 53. IA Indonesia 54. IN Iran Annual Countries 55. IQ Iraq 56. KU Kuwait 15. BE Belgium Bel. Franc (bil.) 57. LI Libya 16. DE Denmark Den. Kroner (bil.) 58. UA United Arab Emirates 17. NO Norway Nor. Kroner (bil.) 59. AO All Other 18. SW Sweden Swe. Kroner (bil.) 19. GR Greece Drachmae (bil.) 20. IR Ireland Irish Pound (mil.) 21. PO Portugal Escudos (bil.) 22. SP Spain Pesetas (bil.) 23. NZ New Zealand N.Z. Dollar (mil.) 24. SA Saudi Arabia Riyals (bil.) 25. VE Venezuela Bolivares (bil.) 26. CO Colombia Col. Pesos (bil.) 27. JO Jordan Jor. Dinars (mil.) 28. SY Syria Syr. Pound (mil.) 29. ID India Ind. Rupee (bil.) 30. MA Malaysia Ringgit (mil.) 31. PA Pakistan Pak. Rupee (bil.) 32. PH Philippines Phil. Peso (bil.) 33. TH Thailand Baht (bil.) 34. CH China Yuan (bil.) 35. AR Argentina Arg. Peso (mil.) 36. BR Brazil Reais (mil.) 37. CE Chile Chi. Peso (bil.) 38. ME Mexico New Peso (mil.) 39. PE Peru Nuevos Soles (mil.) Note: The countries that make up the EMU, denoted EU in the model, are AU, FR, GE, IT, NE, FI, BE, IR, PO, SP. (Luxembourg, which is also part of the EMU, is not in the model.)
A Brief Listing of the Variables per Country
Variables Determined by Stochastic Equations: 1. M Merchandise Imports, 90 lc 2. C Consumption, constant lc 3. I Fixed Investment, constant lc 4. Y Real GDP, constant lc 5. PY GDP Deflator, base year = 1.0 6. M1 Money Supply, lc 7. RS Three Month Interest Rate, percentage points 8. RB Long Term Interest Rate, percentage points 9. E or H Exchange Rate, lc per $ or lc per DM 10. F Three Month Forward Rate, lc per $ 11. PX Export Price Index, 1990 = 1.0 12. W Nominal Wage Rate, base year = 1.0 13. J Employment, thousands 14. L1 Labor Force---men, thousands 15. L2 Labor Force---women, thousands Variables Determined by Identities: I-1. IM Total Imports (NIPA), constant lc I-2. EX Total Exports (NIPA), constant lc I-3. X Final Sales, constant lc I-4. V1 Inventory Investment, constant lc I-5. V Inventory Stock, constant lc I-6. S Current Account Balance, lc I-7. A Net Stock of Foreign Security and Reserve Holdings, lc I-8. M90$A Merchandise Imports from the Trade Share Calculations, 90 $ I-9. EE Exchange Rate, end of period, lc per $ I-12. UR Unemployment Rate I-13. JMIN Minimum Required Employment, thousands I-14. JJ Employment Population Ratio I-15. JJS Peak to Peak Interpolation of JJ I-16. Z Labor Constraint Variable I-17. YS Potential Y I-18. ZZ Demand Pressure Variable I-19. PM Import Price Index, 1990 = 1.0 I-20. E or H Exchange Rate, lc per $ or lc per DM I-21. NW National Wealth, constant lc I-22. PX$ Export Price Index, $/90$ Variables Determined by the Trade Share Calculations: L-1 aij Trade Share Coefficients L-2 XX90$ij Merchandise Exports from i to j, 90$ L-3. X90$ Total Merchandise Exports, 90$ L-4. PMP Import Price Index, 1990 = 1.0 L-5. PW$ World Price Index, $/90$ Exogenous Variables: AF Level of the Armed Forces, thousands EXDS Export Discrepancy, 90 lc E90 E in 1990, 90 lc per 90 $ G Government Expenditures, constant lc IMDS Import Discrepancy, 90 lc JJP Peak to Peak Interpolation of JJ LAM Peak to Peak Interpolation of Y/J MS Non Merchandise Imports, 90 lc M90$B Merchandise Imports from Countries other than the 58 in the Trade Share Matrix, 90 $ PM90 PM in Base Year divided by PM in 1990 POP Population, millions POP1 Population of men, thousands POP2 Population of women, thousands PSI1 Ratio of (EE + EE-1)/2 to E PSI2 Ratio of PM to PMP PX90 PX in Base Year divided by PX in 1990 STAT NIPA Statistical Discrepancy T Time Trend TT Total Net Transfers, lc XS Non Merchandise Exports, 90 lc
Notation: lc local currency 90 lc 1990 local currency constant lc local currency in the NIPA base year
Table B.2: The Variables for a Given Country in Alphabetical Order
                              Table B.2  
            The Variables for a Given Country in Alphabetical Order 
    
Variable Eq.No. Description
aij L-1 Share of i's merchandise exports to j out of total merchandise imports of j. [See below.] A I-7 Net stock of foreign security and reserve holdings, end of quarter, in lc. [A-1 + S. Base value of zero used for the quarter prior to the beginning of the data.] AF exog Level of the armed forces in thousands. [OECD data.] C 2 Personal consumption in constant lc. [OECD data or IFS96F/CPI.] E 9 Exchange rate, average for the period, lc per $ . [IFSRF.] EE I-9 Exchange rate, end of period, lc per $ . [IFSAE.] EX I-2 Total exports (NIPA) in constant lc. [OECD data or (IFS90C or IFS90N)/ PX.] EXDS exog Discrepancy between NIPA export data and other export data in 90 lc. [ EX - PX90(E90*X90$ + XS).] E90 exog E in 1990, 90 lc per 90 $. [IFSRF in 1990.] F 10 Three month forward exchnage rate, lc per $. [IFSB.] G exog Government purchases of goods and services in constant lc. [OECD data or (IFS91F or IFS91FF)/PY.] (Denoted GZ for countries CO and TH.) H 9 Exchange rate, average for the period, lc per DM. [E/EGE] I 3 Gross fixed investment in constant lc. [OECD data or IFS93/PY.] IM I-1 Total imports (NIPA) in constant lc. [OECD data or IFS98C/PM.] IMDS exog Discrepancy between NIPA import data and other import data in 90 lc. [IM - PM90(M + MS).] J 13 Total employment in thousands. [OECD data or IFS67.] JJ I-14 Employment population ratio. [J/POP.] JJP exog Peak to peak interpolation of JJ. [See Section 3.3.3.] JJS I-15 Ratio of JJ to JJP. [JJ/JJP.] JMIN I-13 Minimum amount of employment needed to produce Y in thousands. [Y/LAM.] LAM exog Peak to peak interpolation of Y/J . [See Section 3.3.3.] L1 14 Labor force of men in thousands. [OECD data.] L2 15 Labor force of women in thousands. [OECD data.] M 1 Total merchandise imports (fob) in 90 lc. [IFS71V/PM.] MS exog Other goods, services, and income (debit) in 90 lc, BOP data. [(IFS77AED*E)/PM.] M90$A I-8 Merchandise imports (fob) from the trade share matrix in 90 $ . [See below.] M90$B exog Difference between total merchandise imports and merchandise imports from the trade share matrix in 90 $ (i.e., imports from countries other than the 44 in the trade share matrix). [M/E90 - M90$A.] M1 6 Money supply in lc. [IFS34 or IFS34..B.] NW I-21 National Wealth in constant lc. [NW-1 + I + V1 + EX - IM. Base value of zero used for the quarter prior to the beginning of the data.] PM I-19 Import price index, 1990 = 1.0. [IFS75/100.] PMP L-4 Import price index from DOT data, 1990 = 1.0. [See below.] PM90 exog PM in the NIPA base year divided by PM in 1990. POP exog Population in millions. [IFS99Z.] POP1 exog Population of labor-force-age men in thousands. [OECD data.] POP2 exog Population of labor-force-age women in thousands. [OECD data.] PSI1 exog [(EE + EE-1)/2]/E.] PSI2 exog [PM/PMP.] PW$ L-5 World price index, $/90$. [See below.] PX 11 Export price index, 1990 = 1.0. [IFS74/100. If no IFS74 data for t, then PXt = PX$t*(Et/E90t, where PX$t is defined next.] PX$ I-22 Export price index, $/90$, 1990 = 1.0. [(E90*PX)/E. If no IFS74 data at all, then PX$t = PXUSt for all t. If IFS74 data only from t through t+h, then for i>0, PX$t-i = PX$t*(PXUSt-i/PXUSt and PX$t+h+i = PX$t+h*(PXUSt+k+i/PXUSt. PX90 exog PX in the NIPA base year divided by PX in 1990. PY 5 GDP or GNP deflator, equals 1.0 in the NIPA base year. [OECD data or (IFS99B/IFS99B.P.] RB 8 Long term interest rate, percentage points. [IFS61 or IFS61A.] RS 7 Three month interest rate, percentage points. [IFS60 or IFS60B or IFS60C or IFS60X.] S I-6 Total net goods, services, and transfers in lc. Current account balance. [See Table B.7.] (Denoted SZ for countries CO and TH.) STAT exog Statistical discrepancy in constant lc. [Y-C-I-G-EX+IM-V1.] T exog Time trend. [For quarterly data, 1 in 1952.1, 2 in 1952.2, etc.; for annual data, 1 in 1952, 2 in 1953, etc.] TT exog Total net transfers in lc. [See Table B.7.] UR I-12 Unemployment rate. [(L1 + L2 - J)/(L1 + L2 - AF).] V I-5 Stock of inventories, end of period, in constant lc. [V-1 + V1. Base value of zero was used for the period (quarter or year) prior to the beginning of the data.] V1 I-4 Inventory investment in constant lc. [OECD data or IFS93I/PY.] W 12 Nominal wage rate. [IFS65 or IFS65EY.] X I-3 Final sales in constant lc. [Y - V1.] (Denoted XZ for country PE.) XS exog Other goods, services, and income (credit) in 90 lc. BOP data. [(IFS77ADD*E)/PX.] X90$ L-3 Merchandise exports from the trade share matrix in 90 $. [See below.] XX90$ijL-2 Merchandise exports from i to j in 90$. [See below.] Y 4 Real GDP or GNP in constant lc. [OECD data or IFS99A.P or IFS99B.P or IFS99A.R or IFS99B.R.] YS I-17 Potential value of Y. [LAM*JJP*POP.] Z I-16 Labor constraint variable. [min(0, 1 - JJP/JJ).] ZZ I-18 Demand pressure variable. [(YS-Y)/YS .] Construction of variables related to the trade share matrix: The raw data are: XX$ij Merchandise exports from i to j in $, i,j = 1,...,58 [DOT data. 0 value used if no data.] X$i Total merchandise exports (fob) in $. i = 1,...,39 [IFS70/E or IFS70D.] The constructed variables are: XX$i59 = X$i - Sum58j=1XX$ij, i = 1,...,39 XX90$ij = XX$ij/PX$i, i = 1,...,39, j = 1,...,59 and i = 40,...,58, j = 1,...,58 M90$Ai = Sum58j=1XX90$ji, i = 1,...,58; M90$A59 = Sum39j=1XX90$j59 aij = XX90$ij/M90$Aj, i = 1,...,39, j = 1,...,59 and i= 40,...,58,j= 1,...,58 X90$i = Sum59j=1XX90$ij, i = 1,...,39; X90$i = Sum58j=1XX90$ij, i=40,...,58 PMPi = (Ei/E90i)Sum58j=1ajiPX$j, i = 1,...,39 PW$i = (Sum58j=1PX$jX90$j)/(Sum58j=1X90$j), i = 1,...,39 An element in this summation is skipped if j = i. This summation also excludes the oil exporting countries, which are SA, VE, NI, AL, IA, IN, IQ, KU, LI, UA.
lc = local currency. NIPA = national income and product accounts. IFS xx = variable number xx from the IFS data. Note: Variables available for trade share only countries are M90$A, PX$, X90$.
Table B.3 The Equations for a Given Country
				Table B.3 
                       The Equations for a Given Country
    
Stochastic Equations LHS Var. Explanatory Variables
1. log(M/POP) cnst, log(M/POP)-1, log(PY/PM), RS or RB, log(Y/POP), [A/(PY*YS)]-1 2. log(C/POP) cnst, log(C/POP)-1, RS or RB, log(Y/POP), [A/(PY*YS)]-1 3. logI cnst, logI-1, logY, RS or RB 4. Y cnst, Y-1, X, V-1 5. logPY cnst, logPY-1, logW, logPM, DP, T 6. log[M1/(POP*PY)] cnst, log[M1/(POP*PY)]-1 or log[M1-1/(POP-1*PY)], RS, log(Y/POP) 7. RS cnst, RS-1, 100[(PY/PY-1 )4 - 1], ZZ or JJS, [S/(PY*YS)]-1, RSGE, RSUS 8. RB - RS-2 cnst, RB-1 - RS-2, RS - RS-2, RS-1 - RS-2 9. <>logE cnst, logp - logE-1, logr, logs-1 9. <>logH cnst, logp' - logH-1, logr', logs'-1 10. logF logEE, logr 11. log[PX/(PW$*(E/E90))] logPY - log(PW$*(E/E90)) 12. logW cnst, logW-1, logPY, DW, T, logPY-1, 13. <>logJ cnst, T, log(J/JMIN)-1, <>logY, <>logY-1 14. log(L1/POP1) cnst, T, log(L1/POP1)-1, log(W/PY), Z 15. log(L2/POP2) cnst, T, log(L2/POP2)-1, log(W/PY), Z Notes: i) The variables in equation 9 are: p = PY/PYUS, p' = PY/PYGE, r = [(1 + RS/100)/(1 + RSUS/100)].25, r' = [(1 + RS/100)/(1 + RSGE/100)].25, s = [1 + S/(PY*YS)]/[1 + SUS/PX*YSUS)], where PX is the US price index of firm sales s' = [1 + S/(PY*YS)]/[1 + SGE/PYGE*YSGE)] ii) In equation 10 r is the same variable as in equation 9. iii) In equations 5 and 12 DP and DW are demand pressure variables. See the discussion in Chapter 6 in Macroeconometric Modeling. The following gives for each country the type of gap variable used for DP and the functional form: CA (gap2, linear), JA (gap2, nonlinear), AU (gap1, nonlinear), FR (gap2, linear), GE (gap1, linear), IT (gap2, linear), ST (gap2, linear), UK (UKUR, linear), FI (gap1, linear), AS (gap2, nonlinear), BE (gap2, nonlinear), DE (gap2, nonlinear), NO (gap2, linear), SW (gap2, nonlinear), GR (gap1, nonlinear), IR (gap2, linear), PO (gap2, linear), SP (gap1, linear), NE (gap1, linear), CO (gap1, linear), JO (gap2, linear), MA (gap2, linear), Th (gap2, linear), CH (gap2, linear). The following gives for each country the type of gap variable used for DW and the functional form: GE (gap1, linear), FI (gap1, nonlinear), KO (gap2, linear), DE (gap2, nonlinear), NO (gap2, linear), SW (gap2, nonlinear), SP (gap1, linear), NZ (gap1, nonlinear).
Identities I-1. IM = PM90(M + MS) + IMDS I-2. EX = PX90(E90*X90$ + XS) + EXDS I-3. X = C + I + G + EX - IM + STAT I-4. V1 = Y - X I-5. V = V-1 + V1 I-6. S = PX(E90*X90$ + XS) - PM(M + MS) + TT I-7. A = A-1 + S I-8. M90$A = M/E90 - M90$B I-9. EE = 2*PSI1*E - EE-1 I-12. UR = (L1 + L2 - J)/(L1 + L2 - AF) I-13. JMIN = Y/LAM I-14. JJ = J/POP I-15. JJS = JJ/JJP I-16. Z = min(0, 1 - JJP/JJ) I-17. YS = LAM*JJP*POP I-18. ZZ = (YS - Y)/YS I-19. PM = PSI2*PMP I-20. E = H*EGE or H = E/EGE I-21. NW = NW-1 + I + V1 + EX - IM I-22. PX$ = (E90/E)*PX Note: PX$ and M90$A are exogenous for trade share only countries
Variables Explained When the Countries are Linked Together (Table B.4) L-1. aij L-2 XX90$ij L-3. X90$ L-4. PMP L-5. PW$
Table B.4 Equations that Pertain to the Trade and Price Links Among Countries
				Table B.4 
      Equations that Pertain to the Trade and Price Links Among Countries
    
L-1. aij = computed from trade share equations L-2. XX90$ij= aijM90$Aj, i = 1,...,39, j = 1,...,59 and i = 40,...,58, j = 1,...,58 L-3. X90$i = Sum59j=1XX90$ij, i = 1,...,39 X90$i = Sum58j=1XX90$ij, i = 40,...,58 L-4. PMPi = (Ei/E90i)Sum58j=1ajiPX$j, i = 1,...,39 L-5. PW$i = (Sum58j=1PX$jX90$j)/(Sum58j=1X90$j), i = 1,...,39 An element in this summation is skipped if j = i. This summation also excludes the oil exporting countries, which are SA, VE, NI, AL, IA, IN, IQ, KU, LI, UA. Linking of the Annual and Quarterly Data Quarterly data exist for all the trade share calculations, and all these calculations are quarterly. Feeding into these calculations from the annual models are predicted annual values of PX$i, M90$Ai, and Ei. For each of these three variables the predicted value for a given quarter was taken to be the predicted annual value multiplied by the ratio of the actual quarterly value to the actual annual value. This means in effect that the distribution of an annual value into its quarterly values is taken to be exogenous. Once the quarterly values have been computed from the trade share calculations, the annual values of X90$i that are needed for the annual models are taken to be the sums of the quarterly values. Similarly, the annual values of PMPi and PW$i are taken to be the averages of the quarterly values.
Table B.5 Links Between the US and ROW Models
				Table B.5 
                      Links Between the US and ROW Models
    
The data on the variables for the United States that are needed when the US model is imbedded in the MC model were collected as described in Table B.2. These variables are (with the US subscript dropped): EXDS, IMDS, M, MS, M90$A, M90$B, PM, PMP, PSI2, PW$, PX (= PX$), S, TT, XS, and X90$. The PXUS variable here is not the same as the PX variable for the United States in Appendix A. The variable here is denoted USPX in the MC2 model. The PX variable for the United States is the price index of total sales of the firm sector.
Variable Determination
X90$US Determined in Table B.4 PMPUS Determined in Table B.4 PW$US Determined in Table B.4 PXUS Determined by an equation that is equivalent to equation 11 for the other countries. See the discussion in Section 1.4 of The MC2 Model Workbook. PEX = DEL3*PXUS. In the US model by itself, PEX is determined as PSI1*PX, which is equation 32 in Table A.2. This equation is dropped when the US model is linked to the ROW model. DEL3 is constructed from the data as PEX/PXUS and is taken to be exogenous. PMUS = PSI2US*PMPUS. This is the same as equation I-19 for the other countries. PIM = DEL4*PMUS. PIM is an exogenous variable in the US model by itself. DEL4 is constructed from the data as PIM/PMUS and is taken to be exogenous. EX = (X90$US + XSUS + EXDSUS)/1000. This is the same as equation I-2 for the other countries. EX is an exogenous variable in the US model by itself. EXDSUS is constructed from the data as 1000*EX-X90$US - XSUS and is taken to be exogenous. MUS = 1000*IM - MSUS - IMDSUS. This is the same as equation I-1 for the other countries. IMDSUS is constructed from the data as 1000*IM - MUS - MSUS and is taken to be exogenous. M90$AUS = MUS - M90$BUS. This is the same as equation I-8 for the other countries. SUS = PXUS(X90$US + XSUS) - PMUS(MUS + MSUS) + TTUS. This is the same as equation I-6 for the other countries.
Note: The new exogenous variables for the US model when it is linked to the ROW model are DEL3, DEL4, EXDSUS, IMDSUS, M90$BUS, MSUS, PSI2US, TTUS, and XSUS. EX and PIM are exogenous in the US model by itself, but endogenous when the US model is linked to the ROW model.
Table B.6 The Procedure Used to Create Quarterly Data from Annual Data
                                  Table B.6 
         The Procedure Used to Create Quarterly Data from Annual Data 
    
Let yt be the (observed) average value of the variable for year t, and let yit be the (unobserved) average value of the variable for quarter i of year t (i = 1, 2, 3, 4). Then: (i) y1t + y2t + y3t + y4t = q.yt where 1 for flow variables (at quarterly rates) q = 4 for stock variables and price variables Assume that the annual data begin in year 1, and let q.y1 = a1, q.y2 = a2, q.y3 = a3, ... The key assumption is that the four quarterly changes within the year are the same: d2 for t = 1,2 (ii) y1t - y4t-1 = y2t - y1t = y3t - y2t = y4t - y3t = dt for t = 3,... Given i and ii for t = 1,2, one can solve for y40 and d2 in terms of a1 and a2: y40 = (13/32)a1 - (5/32)a2 d2 = (a2 - a1)/16 Using y40 and d2, one can then construct quarterly data for years 1 and 2 using ii. Given y42 from these calculations and given i and ii for t=3, one can solve for d3 in terms of a3 and y42: d3 = (a3 - 4y42)/10 Using y42 and d3, one can then construct quarterly data for year 3. One can then solve for d4 in terms of y43 and a4, and so on.
Note: The annual population data that were collected for the model are mid year estimates. In order to apply the above procedure to these data, the assumption was made that the average value for the year equals the mid year value.
Table B.7 Construction of the Balance of Payments Data: Data for S and TT
				Table B.7 
      Construction of the Balance of Payments Data:  Data for S and TT
    
The relevant raw data variables are: M$' Goods imports (fob) in $, BOP data. [IFS78ABD.] M$ Goods imports (fob) in $. [IFS71V/E.] X$' Goods exports (fob) in $, BOP data. [IFS78AAD.] X$ Goods exports (fob) in $. [IFS70/E.] MS$ Services and income (debit) in $, BOP data. [IFS78AED + IFS78AHD.] XS$ Services and income (credit) in $, BOP data. [IFS78ADD + IFS78AGD.] XT$ Current transfers, n.i.e., (credit) in $ BOP data. [IFS78AJD.] MT$ Current transfers, n.i.e., (debit) in $ BOP data. [IFS78AKD.]
  • When quarterly data on all the above variables were available, then S$ and TT$ were constructed as: (i) S$ = X$' + XS$ - M$' - MS$ + XT$ - MT$ (ii) TT$ = S$ - X$ - XS$ + M$ + MS$ where S$ is total net goods, services, and transfers in $ (balance of payments on current account) and TT$ is total net transfers in $.
  • When only annual data on M$' were available and quarterly data were needed, interpolated quarterly data were constructed using M$. Similarly for MS$. When only annual data on X$' were available and quarterly data were needed, interpolated quarterly data were constructed using X$. Similarly for XS$, XT$, and MT$. When no data on M$' were available, then M$' was taken to be q.M$, where q is the last observed value of M$'/M$. Similarly for MS$ (where q is the last observed annual value of MS$/M$.) When no data on X$' were available, then X$' was taken to be q.X$, where q is the last observed value of X$'/X$. Similarly for XS$ (where q is the last observed annual value of XS$/X$), for XT$ (where q is the last observed annual value of XT$/X$), and for MT$ (where q is the last observed annual value of MT$/X$). Equations i and ii were then used to construct quarterly data for S$ and TT$.
  • After data on S$ and TT$ were constructed, data on S and TT were constructed as: (iii) S = E*S$ (iv) TT = E*TT$
  • Note from MS and XS in Table B.2 and from MS$ and XS$ above that (v) MS$ = (PM*MS)/E (vi) XS$ = (PX*XS)/E Note also from Table B.2 that (vii) M$ = (PM*M)/E (viii) X$ = (E90*PX*X90$)/E Therefore, from equations ii-vii, the equation for S can be written S = PX(E90*X90$ + XS) - PM(M + MS) + TT which is equation I-6 in Table B.3.