<p>问题比较多，大家提点我大致怎么去做就好，希望大家帮帮忙啦</p><p></p><p>数据DTA在网盘中~~：&nbsp; http://g.zhubajie.com/urllink.php?id=4934530arnmlm6wvq87osnh</p><p></p><p></p><p></p><p></p><p>1.&nbsp;文件CONS.DTA&nbsp;包含了时间序列数据对消费者支出的非耐用品（&nbsp;N&nbsp;）的，&nbsp;和总的消费支出（&nbsp;c&nbsp;）项，这两项都衡量于实际值。&nbsp;文件还包括了非耐用品的价格指数的数据（pn）和总消费支出（pc），和相对非耐用品价格（p）</p><p></p><p>i）用1966年至1992年间的抽样检验单位根在变量log(n),&nbsp;log(c),和&nbsp;log(p)。&nbsp;用Breusch-Godfrey&nbsp;LM检验法在每一步骤总检验误差项自相关性，&nbsp;并执行适当的ADF检验&nbsp;（augmented&nbsp;Dickey&nbsp;Fuller&nbsp;test）</p><p></p><p></p><p>ii）表明三个变量log(n),&nbsp;log(c),&nbsp;and&nbsp;log(p)是I(1)</p><p></p><p>iii）用1966-92年期间的抽样表明变量log(n)&nbsp;和&nbsp;log(c)&nbsp;是协整的。</p><p></p><p>iv）从自回归分布滞后模型ARDL两个变量中导出误差修正模型(ECM)&nbsp;并指出短期和长期的乘数。</p><p></p><p>v）用1966-92的抽样和格朗杰两步法估计非耐用品的误差修正模型(ECM)需求模型</p><p></p><p>vi）在ECM等式中加入log(p)的一阶差分并评论得出的结果。</p><p></p><p>vii）直接估计误差校正模型，并与第（v）的结果进行比较&nbsp;（和第（v）题使用相同的时间段抽样）</p><p></p><p></p><p>原英文题图：</p><p>1.&nbsp;&nbsp;The&nbsp;file&nbsp;CONS.dta&nbsp;contains&nbsp;time-series&nbsp;data&nbsp;on&nbsp;consumer&nbsp;expenditure&nbsp;on&nbsp;non-durables&nbsp;(n),&nbsp;and&nbsp;total&nbsp;consumer&nbsp;expenditure&nbsp;(c),&nbsp;both&nbsp;measured&nbsp;in&nbsp;real&nbsp;terms.&nbsp;&nbsp;The&nbsp;file&nbsp;also&nbsp;contains&nbsp;data&nbsp;on&nbsp;price&nbsp;indices&nbsp;for&nbsp;non-durables&nbsp;(pn)&nbsp;and&nbsp;total&nbsp;(pc)&nbsp;consumer&nbsp;expenditure,&nbsp;and&nbsp;the&nbsp;relative&nbsp;non-durable&nbsp;prices&nbsp;(p).&nbsp;&nbsp;&nbsp;&nbsp;</p><p></p><p></p><p>(i).&nbsp;&nbsp;Test&nbsp;for&nbsp;unit&nbsp;root&nbsp;in&nbsp;variables&nbsp;log(n),&nbsp;log(c),&nbsp;and&nbsp;log(p)&nbsp;using&nbsp;the&nbsp;sample&nbsp;period&nbsp;1966-92.&nbsp;&nbsp;In&nbsp;each&nbsp;step&nbsp;test&nbsp;for&nbsp;autocorrelation&nbsp;in&nbsp;errors&nbsp;using&nbsp;the&nbsp;Breusch-Godfrey&nbsp;LM&nbsp;test,&nbsp;and&nbsp;perform&nbsp;the&nbsp;appropriate&nbsp;augmented&nbsp;Dickey&nbsp;Fuller&nbsp;test.</p><p></p><p>(ii).&nbsp;&nbsp;Show&nbsp;that&nbsp;the&nbsp;three&nbsp;variables&nbsp;log(n),&nbsp;log(c),&nbsp;and&nbsp;log(p)&nbsp;are&nbsp;I(1).&nbsp;&nbsp;</p><p></p><p>(iii).&nbsp;&nbsp;Show&nbsp;that&nbsp;the&nbsp;variables&nbsp;log(n)&nbsp;and&nbsp;log(c)&nbsp;are&nbsp;cointegrated&nbsp;using&nbsp;the&nbsp;1966-92&nbsp;sample&nbsp;period.</p><p></p><p>(iv).&nbsp;&nbsp;Derive&nbsp;the&nbsp;error&nbsp;correction&nbsp;model&nbsp;(ECM)&nbsp;from&nbsp;an&nbsp;autoregressive&nbsp;distributed&nbsp;lag&nbsp;model&nbsp;[(ARDL(1,1))&nbsp;with&nbsp;two&nbsp;variables,&nbsp;and&nbsp;specify&nbsp;the&nbsp;short&nbsp;run&nbsp;and&nbsp;long&nbsp;run&nbsp;multipliers.</p><p></p><p>(v)&nbsp;Estimate&nbsp;the&nbsp;ECM&nbsp;model&nbsp;of&nbsp;demand&nbsp;for&nbsp;non-durables&nbsp;using&nbsp;Engle-Granger&nbsp;two-stage&nbsp;procedure,&nbsp;using&nbsp;the&nbsp;1967-92&nbsp;sample.&nbsp;&nbsp;</p><p></p><p>(vi).&nbsp;&nbsp;Add&nbsp;the&nbsp;first&nbsp;difference&nbsp;of&nbsp;log(p)&nbsp;to&nbsp;the&nbsp;ECM&nbsp;equation&nbsp;and&nbsp;comment&nbsp;on&nbsp;the&nbsp;results.</p><p></p><p>(vii).&nbsp;Estimate&nbsp;the&nbsp;Error&nbsp;Correction&nbsp;Model&nbsp;directly&nbsp;and&nbsp;compare&nbsp;the&nbsp;results&nbsp;with&nbsp;part&nbsp;(v)&nbsp;(use&nbsp;the&nbsp;same&nbsp;sample&nbsp;period&nbsp;as&nbsp;in&nbsp;v).</p>
[此贴子已经被作者于2009-4-26 21:32:15编辑过]
发帖
雷达卡
加好友,备注jltj
京公网安备 11010802022788号
论坛法律顾问：王进律师
知识产权保护声明
免责及隐私声明
