序列见附件。
序列单位根检验结果如下：Null Hypothesis: Y has a unit root
Exogenous: Constant, Linear Trend
Lag Length: 1 (Fixed)

t-Statistic Prob.*

Augmented Dickey-Fuller test statistic -2.496317 0.3300
Test critical values: 1% level -3.964101
5% level -3.412773
10% level -3.128365

*MacKinnon (1996) one-sided p-values.


Augmented Dickey-Fuller Test Equation
Dependent Variable: D(Y)
Method: Least Squares
Date: 07/22/14 Time: 14:05
Sample (adjusted): 3 1520
Included observations: 1518 after adjustments

Variable Coefficient Std.Error t-Statistic Prob.

Y(-1) -0.002791 0.001118 -2.496317 0.0127
D(Y(-1)) 0.011582 0.025654 0.451473 0.6517
C 0.018793 0.008326 2.257060 0.0241
@TREND(1) -1.40E-06 1.10E-06 -1.275973 0.2022

R-squared 0.010394 Mean dependent var -0.001104
Adjusted R-squared 0.008433 S.D. dependent var 0.006587
S.E. of regression 0.006559 Akaike info criterion -7.213384
Sum squared resid 0.065129 Schwarz criterion -7.199352
Log likelihood 5478.958 Hannan-Quinn criter. -7.208159
F-statistic 5.300632 Durbin-Watson stat 1.998538
Prob(F-statistic) 0.001232


从这个结果来看，应该是非平稳序列，算出来的R-squared接近于0. 这是不是说明预测拟合程度很低？那现在要怎么办？
原本是想用GARCH模型来做预测。


同时我也用SPSS做了一下回归分析，分析结果如下：
模型汇总
模型 R R 方 调整 R 方 标准 估计的误差
1 .937a .877 .877 .1509172
a. 预测变量: (常量), X。

这是怎么回事呢？