Trong tương lai khi những dữ liệu hồn chỉnh và số kỳ nghiên cứu đủ lớn, hướng nghiên cứu tiếp tục được hồn thiện và bổ sung khi được tăng cỡ mẫu. Do đĩ, bài nghiên cứu sẽ tiếp tục hồn thiện về mặt dữ liệu, kỳ quan sát và bổ sung việc cần phải xem xét đến sự chênh lệch khác biệt giữa các quốc gia.
DANH MỤC TÀI LIỆU THAM KHẢO
1. Aizenman, Joshua, and Jaewoo Lee. "International reserves: precautionary versus mercantilist views, theory and evidence." Open Economies Review 18.2 (2007): 191-214.
2. Aizenman, Joshua, and Yi Sun. "Globalization and the sustainability of large current account imbalances: size matters." Journal of Macroeconomics 32.1 (2010): 35-44.
3. Aizenman, Joshua, Yothin Jinjarak, and Donghyun Park. "International reserves and swap lines: Substitutes orcomplements?." International Review of Economics & Finance 20.1 (2011): 5-18.
4. Blanchard, Olivier J., Mitali Das, and Hamid Faruqee. "The initial impact of the crisis on emerging market countries." Brookings papers on economic activity 2010, no. 1 (2010): 263-307.
5. Bussière, Matthieu, Gong Cheng, Menzie D. Chinn, and Noëmie Lisack. "For a few dollars more: Reserves and growth in times of crises." Journal of International Money and Finance 52 (2015): 127-145.
6. Dominguez, K., 2010. International reserves and underdeveloped capital markets. In: Reichlin, L., West, K. (Eds.), NBER International Seminar on Macroeconomics 2009. University of Chicago Press for the NBER, 193 – 221
7. Dominguez, Kathryn ME, Yuko Hashimoto, and Takatoshi Ito. "International reserves and the global financial crisis." Journal of International Economics88.2 (2012): 388-406.
8. Dooley, Michael P., David Folkerts-Landau, and Peter Garber. An essay on the revived Bretton Woods system. No. w9971. National Bureau of Economic Research, 2003.
9. Dwyer, Gerald P., and James R. Lothian. "International and historical dimensions of the financial crisis of 2007 and 2008." Journal of International Money and Finance 31, no. 1 (2012): 1-9.
10.Frankel, Jeffrey A., and George Saravelos. Are leading indicators of financial crises useful for assessing country vulnerability? Evidence from the 2008-09 global crisis. No. w16047. National bureau of economic research, 2010.
11.Hashimoto, Yuko, and Takatoshi Ito. "Global imbalances and Asian foreign reserves." Unpublished manuscript (2007).
12.Horioka, C.Y. and Terada-Hagiwara, A., 2012. The determinants and long- term projections of saving rates in Developing Asia. Japan and the World Economy, 24(2), pp.128-137.
13.Llaudes, Ricardo, Ferhan Salman, and Mali Chivakul. "The impact of the great recession on emerging markets." IMF Working Papers (2010): 1-34.
14.Obstfeld, M., Shambaugh, J., Taylor, A., 2009. Financial instability, reserves, and central bank swap lines in the panic of 2008. American Economic Review: Papers and Proceedings 99 (2), 480 – 486.
15.Obstfeld, M., Shambaugh, J., Taylor, A., 2010. Financial stability, the Trilemma, and international reserves. American Economic Journal: Macroeconomics 2, 57–94 April.Reinhart, C., Reinhart, V., 2008. Capital in flows and reserve accumulation: the recent evidence. NBER Working Paper 13842.
16.Visokaviciené, Biruté. "Monetary Poloci in advanced economies during the global financial crisis: Lessons for Lithuania”. Ekonomika 93, no. 1 (2014): 40.
17.Ye-chi, M. A., Y. U. Shan, and J. I. A. N. G. Xiao-qin. "Financial Market Development, Foreign Exchange Reserve Accumulation and Economic Growth: A Test Based on Threshold Panel Regression Model." On Economic Problems 6 (2015): 011.
PHỤ LỤC ĐỊNH LƯỢNG
Phục lục 1: Thống kê mơ tả
Phụ lục 2: Ma trận tương quan
Phụ lục 3 : Nhân tử phĩng đại phương sai VIF
Phụ lục 4: Kiểm định mơ hình
Kiểm định Pooled với FEM
termoftrade 800 96.32174 26.91358 0 200.4327 exchangerate 792 101.2302 16.62267 66.09628 166.9438 gdpdeflator 800 87.52563 18.92878 0 122.8554 reservesgd~o 796 .307857 .3002063 0 1.207992 gdp_weo 800 860284.7 1480060 0 5954477 reservebop 800 2981.649 9147.552 -34003.64 145966 activemana~t 792 2172.382 8374.488 -42641.6 98226.36 forexr 800 98513.15 163671.2 924.0002 861453.9 ir 800 148625.4 251222.1 .042965 1259495 Variable Obs Mean Std. Dev. Min Max
activemana~c -0.1227 -0.1065 -0.1862 0.1165 -0.0008 -0.0505 0.1028 1.0000 reservegro~c -0.1233 -0.1241 -0.0001 0.0008 -0.1103 0.0443 1.0000 totchanged~c 0.0109 -0.0222 0.0722 -0.3665 -0.4202 1.0000 empduringgfc 0.0327 0.0604 -0.0193 0.3067 1.0000 reservesgd~c 0.1684 0.1282 -0.0917 1.0000 gdpgrowthd~c 0.5775 0.8901 1.0000 gdpgrowthp~c 0.6920 1.0000 gdpgrowtha~c 1.0000 gdp~egfc gdp~ogfc gdp~ggfc res~ogfc empdur~c totcha~c res~ggfc active~c
Mean VIF 2.87 activemana~c 1.08 0.926400 reservegro~c 1.14 0.877170 empduringgfc 1.27 0.786296 totchanged~c 1.32 0.758995 reservesgd~c 1.50 0.665738 gdpgrowthd~c 6.86 0.145674 gdpgrowthp~c 6.89 0.145170 Variable VIF 1/VIF
Kiểm định Pooled với REM
Kiểm định FEM với REM
Phụ lục 5: Kiểm định phương sai thay đổi
Phụ lục 6: Tự tương quan Prob > chibar2 = 1.0000 chibar2(01) = 0.00 Test: Var(u) = 0 u 0 0 e .0012124 .034819 gdpg~egfc .0025918 .05091 Var sd = sqrt(Var) Estimated results:
gdpgrowthafterthegfc[country,t] = Xb + u[country] + e[country,t] Breusch and Pagan Lagrangian multiplier test for random effects
(V_b-V_B is not positive definite) Prob>chi2 = 0.0103
= 16.73
chi2(6) = (b-B)'[(V_b-V_B)^(-1)](b-B) Test: Ho: difference in coefficients not systematic
B = inconsistent under Ha, efficient under Ho; obtained from xtreg b = consistent under Ho and Ha; obtained from xtreg activemana~c -7.71e-14 -3.07e-13 2.30e-13 1.43e-13 reservegro~c .0242538 -.0107466 .0350004 .0439951 totchanged~c -.0028026 .0000882 -.0028909 .0014387 empduringgfc .0000201 -.0001032 .0001234 .0004771 reservesgd~c -.2502425 .0155283 -.2657708 .2763516 gdpgrowthd~c -.0375169 -.1573987 .1198818 .1544086 gdpgrowthp~c .8149862 .8947368 -.0797505 .198451 tenmohinhfe1 tenmohinhre1 Difference S.E. (b) (B) (b-B) sqrt(diag(V_b-V_B)) Coefficients
Prob>chi2 = 0.0000 chi2 (8) = 216.07
H0: sigma(i)^2 = sigma^2 for all i in fixed effect regression model
Modified Wald test for groupwise heteroskedasticity
Prob > F = 0.0505 F( 1, 6) = 5.949 H0: no first-order autocorrelation
Phụ lục 7: Kết quả hồi quy
Mơ hình Pooled
Mơ hình FEM
_cons .0052391 .0695814 0.08 0.940 -.136865 .1473432 activemanagementduringgfc -3.07e-13 5.90e-13 -0.52 0.606 -1.51e-12 8.98e-13 reservegrowthduringgfc -.0107466 .0706499 -0.15 0.880 -.1550329 .1335397 totchangeduringgfc .0000882 .0002197 0.40 0.691 -.0003605 .000537 empduringgfc -.0001032 .0006429 -0.16 0.874 -.0014163 .0012098 reservesgdp_weopriortogfc .0155283 .0279884 0.55 0.583 -.0416315 .0726882 gdpgrowthduringgfc -.1573987 .3509345 -0.45 0.657 -.8741026 .5593052 gdpgrowthpriortogfc .8947368 .3757401 2.38 0.024 .127373 1.662101 gdpgrowthafterthegfc Coef. Std. Err. t P>|t| [95% Conf. Interval] Total .095897576 37 .002591826 Root MSE = .04007
Adj R-squared = 0.3805 Residual .048171378 30 .001605713 R-squared = 0.4977 Model .047726198 7 .006818028 Prob > F = 0.0023 F( 7, 30) = 4.25 Source SS df MS Number of obs = 38
F test that all u_i=0: F(7, 23) = 2.39 Prob > F = 0.0541
rho .90989646 (fraction of variance due to u_i)
sigma_e .03481903 sigma_u .11064752
_cons .3402765 .1637412 2.08 0.049 .0015521 .679001 activemanagementduringgfc -7.71e-14 5.32e-13 -0.14 0.886 -1.18e-12 1.02e-12 reservegrowthduringgfc .0242538 .0755265 0.32 0.751 -.1319846 .1804922 totchangeduringgfc -.0028026 .0014513 -1.93 0.066 -.0058049 .0001996 empduringgfc .0000201 .0007346 0.03 0.978 -.0014996 .0015399 reservesgdp_weopriortogfc -.2502425 .2774197 -0.90 0.376 -.8241287 .3236438 gdpgrowthduringgfc -.0375169 .3418011 -0.11 0.914 -.7445865 .6695526 gdpgrowthpriortogfc .8149862 .3820716 2.13 0.044 .0246108 1.605362 gdpgrowthafterthegfc Coef. Std. Err. t P>|t| [95% Conf. Interval] corr(u_i, Xb) = -0.9128 Prob > F = 0.0005
F(7,23) = 5.86 overall = 0.0130 max = 8 between = 0.0948 avg = 4.8 R-sq: within = 0.6407 Obs per group: min = 2 Group variable: country Number of groups = 8 Fixed-effects (within) regression Number of obs = 38
Mơ hình FGLS
Mơ hình GMM
_cons .0266366 .0208866 1.28 0.202 -.0143005 .0675736 activemanagementduringgfc -1.57e-15 2.35e-13 -0.01 0.995 -4.62e-13 4.59e-13 reservegrowthduringgfc -.0251185 .0352756 -0.71 0.476 -.0942575 .0440205 totchangeduringgfc -.0000387 .000073 -0.53 0.596 -.0001817 .0001043 empduringgfc -.0001958 .0001902 -1.03 0.303 -.0005687 .000177 reservesgdp_weopriortogfc .0174517 .0126476 1.38 0.168 -.0073371 .0422404 gdpgrowthduringgfc .399154 .1767084 2.26 0.024 .0528119 .745496 gdpgrowthpriortogfc .4672189 .1874485 2.49 0.013 .0998265 .8346112 gdpgrowthafterthegfc Coef. Std. Err. z P>|z| [95% Conf. Interval] Prob > chi2 = 0.0000
Wald chi2(6) = 80.13 max = 8 avg = 4.75 Estimated coefficients = 7 Obs per group: min = 2 Estimated autocorrelations = 1 Number of groups = 8 Estimated covariances = 8 Number of obs = 38 Correlation: common AR(1) coefficient for all panels (-0.7306)
Panels: heteroskedastic Coefficients: generalized least squares Cross-sectional time-series FGLS regression
Hansen test of overid. restrictions: chi2(23) = 1.09 Prob > chi2 = 1.000 (Not robust, but not weakened by many instruments.)
Sargan test of overid. restrictions: chi2(23) = 36.04 Prob > chi2 = 0.041 Arellano-Bond test for AR(2) in first differences: z = -0.59 Pr > z = 0.557 Arellano-Bond test for AR(1) in first differences: z = -0.70 Pr > z = 0.482 D.gdpgrowthpriortogfc
GMM-type (missing=0, separate instruments for each period unless collapsed) _cons
L.reservegrowthduringgfc Standard
Instruments for levels equation L(1/8).gdpgrowthpriortogfc
GMM-type (missing=0, separate instruments for each period unless collapsed) D.L.reservegrowthduringgfc
Standard
Instruments for first differences equation
_cons .0056138 .037063 0.15 0.880 -.0670283 .0782559 reservegrowthduringgfc -.0343247 .0250388 -1.37 0.170 -.0833998 .0147504 totchangeduringgfc .0002079 .0001771 1.17 0.240 -.0001392 .0005551 empduringgfc -.0002586 .0004135 -0.63 0.532 -.0010691 .0005518 reservesgdp_weopriortogfc .028146 .0151499 1.86 0.063 -.0015473 .0578393 gdpgrowthduringgfc -.1754293 .5368833 -0.33 0.744 -1.227701 .8768427 gdpgrowthpriortogfc .8694169 .478983 1.82 0.070 -.0693726 1.808206 gdpgrowthafterthegfc Coef. Std. Err. z P>|z| [95% Conf. Interval] Robust
Prob > chi2 = 0.000 max = 7
Wald chi2(6) = 1582.47 avg = 3.75 Number of instruments = 30 Obs per group: min = 1 Time variable : period Number of groups = 8 Group variable: country Number of obs = 30 Dynamic panel-data estimation, one-step system GMM
Mơ hình 2
(Robust, but weakened by many instruments.)
Hansen test of overid. restrictions: chi2(386) = 0.73 Prob > chi2 = 1.000 (Not robust, but not weakened by many instruments.)
Sargan test of overid. restrictions: chi2(386) = 594.56 Prob > chi2 = 0.000 Arellano-Bond test for AR(2) in first differences: z = -0.03 Pr > z = 0.972 Arellano-Bond test for AR(1) in first differences: z = -1.49 Pr > z = 0.135 D.activemanagement
GMM-type (missing=0, separate instruments for each period unless collapsed) L.exchangerate D2.irgrowth L2.activemanagement
_cons Standard
Instruments for levels equation L(1/3).activemanagement
GMM-type (missing=0, separate instruments for each period unless collapsed) D.(L.exchangerate D2.irgrowth L2.activemanagement)
Standard
Instruments for first differences equation
Warning: Uncorrected two-step standard errors are unreliable.
_cons -.1459856 .1750435 -0.83 0.404 -.4890646 .1970934 activemanagement -1.71e-12 9.66e-13 -1.77 0.077 -3.60e-12 1.84e-13 irgrowth -1.08e-07 5.37e-08 -2.01 0.044 -2.13e-07 -2.69e-09 termoftrade .0008327 .002549 0.33 0.744 -.0041633 .0058287 exchangerate .0024325 .0035979 0.68 0.499 -.0046192 .0094842 lreservesgdp_weo -.5778861 .2797664 -2.07 0.039 -1.126218 -.0295539 ouputgap Coef. Std. Err. z P>|z| [95% Conf. Interval] Prob > chi2 = 0.253 max = 98 Wald chi2(4) = 5.35 avg = 98.00 Number of instruments = 392 Obs per group: min = 98 Time variable : period Number of groups = 8 Group variable: country Number of obs = 784 Dynamic panel-data estimation, two-step system GMM
Performing specification tests. > mation.
Using a generalized inverse to calculate optimal weighting matrix for two-step esti Number of instruments may be large relative to number of groups.
Warning: Two-step estimated covariance matrix of moment conditions is singular. Estimating.
2 instrument(s) dropped because of collinearity. Building GMM instruments..
(Robust, but weakened by many instruments.)
Hansen test of overid. restrictions: chi2(379) = 0.71 Prob > chi2 = 1.000 (Not robust, but not weakened by many instruments.)
Sargan test of overid. restrictions: chi2(379) = 564.24 Prob > chi2 = 0.000 Arellano-Bond test for AR(2) in first differences: z = -0.40 Pr > z = 0.689 Arellano-Bond test for AR(1) in first differences: z = -2.11 Pr > z = 0.035 D.activemanagement
GMM-type (missing=0, separate instruments for each period unless collapsed) L.exchangerate D2.irgrowth L2.activemanagement
_cons Standard
Instruments for levels equation L(1/3).activemanagement
GMM-type (missing=0, separate instruments for each period unless collapsed) D.(L.exchangerate D2.irgrowth L2.activemanagement)
Standard
Instruments for first differences equation
Warning: Uncorrected two-step standard errors are unreliable.
_cons -.4189462 .2907867 -1.44 0.150 -.9888777 .1509853 activemanagement -6.65e-12 2.97e-12 -2.24 0.025 -1.25e-11 -8.26e-13 irgrowth -3.12e-07 1.39e-07 -2.25 0.024 -5.84e-07 -4.07e-08 termoftrade .0059455 .0027211 2.18 0.029 .0006123 .0112788 exchangerate .0008399 .0024544 0.34 0.732 -.0039706 .0056505 lreservesgdp_weo -.4312283 .1973751 -2.18 0.029 -.8180765 -.0443802 ouputgap Coef. Std. Err. z P>|z| [95% Conf. Interval] Prob > chi2 = 0.273 max = 96 Wald chi2(4) = 5.15 avg = 93.25 Number of instruments = 385 Obs per group: min = 90 Time variable : period Number of groups = 8 Group variable: country Number of obs = 746 Dynamic panel-data estimation, two-step system GMM
Performing specification tests. > mation.
Using a generalized inverse to calculate optimal weighting matrix for two-step esti Number of instruments may be large relative to number of groups.
Warning: Two-step estimated covariance matrix of moment conditions is singular. Estimating.
3 instrument(s) dropped because of collinearity. Building GMM instruments..
Ghi chú:
Dữ liệu IFS_IR cĩ sẵn cho các nước: Cộng hịa Séc, Bahrain, Guinea, Nepal và Zimbabwe là Tổng trữ trừ vàng. Dữ liệu IFS_IR của Zimbabwe kết thúc vào tháng 2/2002.
Dữ liệu IFS_IR của Đài Loan là từ Ngân hàng Trung Quốc. Cyprus, Malta, Slovenia và Cộng hịa Slovak được loại trừ. New Zealand khơng phải là một quốc gia theo tiêu chuẩn cơng bố dữ liệu SDDS nhưng cung cấp các dữ liệu chi tiết và được đưa vào phân tích. Thơi gian khủng hoảng dựa trên thay đổi phần trăm trong SA GDP thực sau năm 2007 (từ đỉnh tới đáy).
Difference (null H = exogenous): chi2(5) = 2.45 Prob > chi2 = 0.785 Hansen test excluding group: chi2(32) = 0.45 Prob > chi2 = 1.000 iv(L.activemanagement termoftrade L.exchangerate irgrowth activemanagement) Difference (null H = exogenous): chi2(37) = -0.51 Prob > chi2 = 1.000 Hansen test excluding group: chi2(0) = 3.40 Prob > chi2 = . gmm(exchangerate, eq(level) lag(1 .))
Difference (null H = exogenous): chi2(37) = -0.51 Prob > chi2 = 1.000 Hansen test excluding group: chi2(0) = 3.40 Prob > chi2 = . GMM instruments for levels
Difference-in-Hansen tests of exogeneity of instrument subsets: (Robust, but weakened by many instruments.)
Hansen test of overid. restrictions: chi2(37) = 2.89 Prob > chi2 = 1.000 (Not robust, but not weakened by many instruments.)
Sargan test of overid. restrictions: chi2(37) = 64.28 Prob > chi2 = 0.004 Arellano-Bond test for AR(2) in levels: z = -1.50 Pr > z = 0.132 Arellano-Bond test for AR(1) in levels: z = 0.09 Pr > z = 0.926 DL(1/98).exchangerate
GMM-type (missing=0, separate instruments for each period unless collapsed) _cons
L.activemanagement termoftrade L.exchangerate irgrowth activemanagement Standard
Instruments for levels equation activemanagement)
D.(L.activemanagement termoftrade L.exchangerate irgrowth Standard
Instruments for first differences equation
_cons .0231719 .0685803 0.34 0.735 -.111243 .1575868 irgrowth .0981891 .0642047 1.53 0.126 -.0276498 .2240281 termoftrade -.0002805 .0001801 -1.56 0.119 -.0006335 .0000725 exchangerate .0000789 .0006153 0.13 0.898 -.0011271 .0012849 lreservesgdp_weo -.0198775 .0059637 -3.33 0.001 -.0315661 -.008189 ouputgap Coef. Std. Err. z P>|z| [95% Conf. Interval] Robust
Prob > chi2 = 0.001 max = 8 Wald chi2(4) = 17.93 avg = 4.75 Number of instruments = 42 Obs per group: min = 2 Time variable : period Number of groups = 8 Group variable: country Number of obs = 38 Dynamic panel-data estimation, one-step system GMM
Difference-in-Sargan/Hansen statistics may be negative.
Using a generalized inverse to calculate robust weighting matrix for Hansen test. Warning: Two-step estimated covariance matrix of moments is singular.