100 M. Corazza, A. Ellero and A. Zorzi
These results show that the S&P 500 stock market was subject to anomalous activities
after September 11, 2001 rather than before.
4.3 Consecutive rejection days analysis
Here, addressing our attention to returns once more, we analyse the sequences of
consecutive rejection days detected usingα=5%. In Table 3 we report the number
of such sequences having lengths from 1 day to 12 days, respectively (12 days is the
maximumlength detected in the investigation period). For deepening the investigation,
we also report the results obtained usingα=1%.
We observe that the length of the large majority of the sequences of consecutive
rejection days is rather low. This fact can be interpreted as the capability of the S&P
500 stock market to “absorb” anomalous events in short time periods.
On the contrary, given such a capability, the presence of long sequences of con-
secutive rejection days is an indicator of malaise of the market. For instance, this is
the case of a 9-day sequence (September 20, 2001 to October 2, 2001) that started im-
mediately after the Twin Towers attack and of a 6-day sequence (February 27, 2007 to
March 6, 2007) that started on the day of the Wall Street crash. Moreover, analogously
to what we already observed in the previous subsection, since the events/causes as-
sociated to such sequences are not always observable, the consecutive rejection days
analysis might be profitably used for detecting continued anomalous behaviours in
financial markets.
Ta b le 3 .Sequences of consecutive rejection daysSequence length # withα=5% # withα=1%
1 412 366
2 148 105
37152
44812
52 4 7
61 3 2
771
810
930
10 1 1
11 0 0
12 1 0