SERIAL CORRELATION. For this and the next question. In a time series regression, exchange rate $\backslash ((\backslash$mathrm$\{$Y$\})\backslash )$ is regressed on one$-$month crude oil futures price $($ X $).$ The sample consisted of $338$ weekly observations from January $2010$ to September $2016.$ A likely problem in this type of regression is serial correlation, a case where error terms are correlated across time periods. Based on the following intermediate results, and considering the rule, where does the calculated Durbin$-$Watson statistic lie? Note: sum of the squared residuals: $\backslash ($ e$^\{2\}=0\mathrm{.}3719\backslash )$; sum of squared residual differences: $\backslash (\backslash$left$($e$\_\{$t$\}-$e$\_\{$t$-$i$\}\backslash$right$)^\{2\}=0\mathrm{.}0293\backslash )$

$1)\backslash (0-2\backslash )$

$2)\backslash (2-4\backslash )$

$3)$ Approximately $2$

$4)$ None of the above is correct

Question $41(1$ point$)$

SERIAL CORRELATION. In reference to the above question, what is your conclusion regarding serial correlation?

$1)$ There is evidence of a positive serial correlation

$2)$ There is evidence of a negative serial correlation

$3)$ There is no evidence of serial correlation

$4)$ The DW test yielded an inconclusive result