because of its heterogeneity and velocity. If we cannot measure it, we cannot manage it,
and text-based news is hard to quantify.
In this chapter, we describe one initiative aimed at solving this problem. The Thomson
Reuters NewsScope Event Indices Project is an integrated framework for incorporating
real-time news from the Thomson Reuters NewsScope subscription service into sys-
tematic investment and risk management protocols. The framework consists of a set of
real-time event indices—each one taking on numerical values between 0 and 100—
designed to capture the occurrence of unusual events of a particular kind. For example,
theMacroindex measures the real-time quantity of macroeconomic news, and the
NatDistindex measures the real-time quantity of natural disaster news. Each index
is constructed by applying disciplined pattern recognition algorithms to real-time news-
feeds, and calibrated using econometric methods applied to historical data. In this first
release, we construct indices that are calibrated to foreign exchange markets; future
releases will focus on other markets.
In this chapter, we describe the procedures for constructing and validating the
Thomson Reuters/AlphaSimplex Event Indices. We begin with a brief literature review
in Section 3.2, and in Section 3.3 we introduce the historical datasets used to calibrate
the indices. Section 3.4 contains the algorithms used to construct the indices. In Section
3.5, we describe the event study methodology for validating the indices, and in Section
3.6 we explore the connection between realized volatility (our metric for market impact)
and implied volatility. We conclude in Section 3.8.
3.2 Literature review
There is a surprisingly rich literature on the relationship between news and financial
markets going back to Niederhoffer’s (1971) pioneering study of world events and stock
prices, where world events were defined as five- to eight-column headlines in theNew
York Timesand then organized into categories of meaning. Niederhoffer found that
large stock price changes did follow world events more than randomly selected days, but
that a particular category into which a world event falls did not add much incremental
information about future price movements.
Measuring public information by the number of news releases by Reuter’s News
Service per unit of time, Berry and Howe (1994) showed that there is a positive,
moderate relationship between public information and trading volume. Engle and Ng
(1993) defined the ‘‘news impact curve’’ which measures how new information is
incorporated into volatility estimates. However, by studying the number of news
announcements reported daily by Dow Jones & Co., Mitchell and Mulherin (1994)
did not find any strong relation between news and market activity. Hong, Lim, and
Stein (2000) confirmed that firm-specific information, especially negative information,
diffuses only gradually across the investing public.
On the macroeconomic front, Pearce and Roley (1985) showed that on announcement
days surprises related to monetary policy significantly affect stock prices, but only found
limited evidence of an impact from inflation surprises and no evidence of an impact from
real activity surprises.
74 Quantifying news: Alternative metrics