20.11.2012 Views

recent developments in high frequency financial ... - Index of

recent developments in high frequency financial ... - Index of

recent developments in high frequency financial ... - Index of

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

Modell<strong>in</strong>g f<strong>in</strong>ancial transaction price movements: a dynamic <strong>in</strong>teger count data model 193<br />

JBX HAL<br />

Fig. 11 Autocorrelation function <strong>of</strong> the squared density forecast variable ui* for Jack <strong>in</strong> the Box<br />

Inc. (JBX) and Halliburton pCompany<br />

ffiffiffi (HAL). The dashed l<strong>in</strong>es mark <strong>of</strong>f the approximate 99%<br />

confidence <strong>in</strong>terval 2:58 n<br />

Figures 10 and 11 plot the autocorrelation function for a sequence <strong>of</strong> normalized<br />

residuals u i* and <strong>of</strong> squared normalized residuals (u i*) 2 . The plots <strong>in</strong>dicate<br />

that for both stocks there is nearly no significant autocorrelation left <strong>in</strong> the first and<br />

second moments. Table 6 represents summary statistics <strong>of</strong> the normalized residuals.<br />

They are computed as the correspond<strong>in</strong>g sample means based on 1,000<br />

repeated draws <strong>of</strong> the trajectory {ui*, i: 1→N}. For JBX the Jarque-Bera statistic<br />

<strong>in</strong>dicates that we cannot reject the null <strong>of</strong> a normal distribution for u i*, whereas for<br />

HAL we have to reject the null at 1% significance level.<br />

The quantile–quantile (QQ) plot <strong>of</strong> a sequence <strong>of</strong> u i*’s aga<strong>in</strong>st the standard<br />

normal distribution displayed <strong>in</strong> Fig. 12 reveals that the ICH model approximates<br />

the distributional properties <strong>of</strong> the transaction price changes for both stocks<br />

fairly well. However, the deviation from normality <strong>in</strong> the tails <strong>of</strong> the distribution<br />

<strong>of</strong> the ui*’s <strong>in</strong>dicates slight difficulties characteriz<strong>in</strong>g extreme price changes<br />

appropriately.<br />

Table 6 Properties <strong>of</strong> the density forecast variable u i* for Jack <strong>in</strong> the Box Inc. (JBX) and<br />

Halliburton Company (HAL)<br />

JBX HAL<br />

Mean −0.003 [0.0000] −0.001 [0.0000]<br />

Variance 1.000 [0.0000] 1.002 [0.0000]<br />

Skewness 0.027 [0.0000] 0.029 [0.0000]<br />

Kurtosis 2.989 [0.0000] 3.048 [0.0000]<br />

Jarque-Bera 0.793 [0.0006] 9.910 [0.0019]<br />

p-value (0.6727) (0.007)<br />

The values <strong>of</strong> the statistics are sample means based upon 1,000 repeated draws <strong>of</strong> {u i*, i: 1→N}.<br />

Values <strong>in</strong> brackets report the sampl<strong>in</strong>g standard error, due to repeated sampl<strong>in</strong>g <strong>of</strong> {ui*, i:1→N}<strong>in</strong><br />

the density forecast procedure

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!