6.7 US portfolio simulation using news analytic signals
Event studies are screening methods that show potential value, but they are in ‘‘event
time’’, while portfolios are managed in ‘‘calendar time’’. It is instructive to attempt
realistic portfolio construction using only RNSE event data for extreme sentiment days.
Of course, a professional portfolio manager would incorporate these RNSE event data
into her arsenal of other investment ideas and information sources, but it is instructive
to see if even simple portfolio construction techniques using only news data can
perform.
6.7.1 Investment hypothesis
The investment hypothesis under consideration is that it takes market participants a
long time (days) to process a large amount of novel, strongly polar news, as suggested by
the event studies. News and event ambiguity, fact validation, cognitive dissonance are all
Relating news analytics to stock returns 163
Figure 6.13.Subsequent cumulative return after an extreme sentiment day relative to sector and
market capitalization peers. Daily positive or negative sentiment in the top 5%, with at least two
news items with link count 5¼0 in the last 24 hours for stocks in the S&P 1500 (2003–2009, Q3).