News Impact Score | Vixplor Analytics

News Impact Score and Stock Market Quality: Information Content in Business News

Textplor News Analysis Finance

Ashok Banerjee | 21 Aug 2021

News Impact Score and Stock Market Quality: Information Content in Business News
The news impact score (NIS) uses newspaper reports to estimate corporate stress levels. There is a strong negative correlation between NIS and stock returns. Higher NIS is followed by greater volatility in stock returns.

News Frequency and NIS: Basic Features:

We look at firm-level news in this paper and hence ignore any news on economy in general. Since news is sourced from business dailies (e.g Economic Times, Mint, Money Control etc.), NIS does not capture sentiments of any general, value-neutral news. We have looked at top 500 NS-listed companies (NSE 500 Index) and analysed the NIS data for these companies over 11 years (2010 to 2020).We have removed companies for which stock price data were not 1 Geert Bekaert, and Guojun Wu. Asymmetric Volatility and Risk in Finance Markets. The Review of Finance available, leaving us with 399 of the Nifty 500 companies. These companies are sorted on the basis of aggregate news frequency in a particular year and the top decile represents top ten percent companies which are heavily reported in the media in that year.

Data and Methodology:

We have looked at the effect of NIS on stock returns and volatility. Since, NIS are used at monthly frequency, stock prices and hence returns are estimated at monthly rest. Stock prices for firms are collected from Yahoo Finance. Stock returns are estimated as (natural)log of price relatives. In other words, we have defined stock returns as monthly change in prices on a logarithmic scale. Historical volatility (standard deviation) of stock returns is estimated using past 30 months returns on a rolling basis.

News flow was much lower during 2010 through 2013. Also, a large number of companies had very low news volume even after 2013. We have only considered, for our analysis, those companies which had news for at least 80% of the months since July 2014. This filter has reduced our sample size to 86 companies out of about 400 companies. Thus, we have monthly data for the variables since July 2014 till December 2020.

-->

The news impact score (NIS) uses newspaper reports to estimate corporate stress levels. There is a strong negative correlation between NIS and stock returns. Higher NIS is followed by greater volatility in stock returns.

News Frequency and NIS: Basic Features:

We look at firm-level news in this paper and hence ignore any news on economy in general. Since news is sourced from business dailies (e.g Economic Times, Mint, Money Control etc.), NIS does not capture sentiments of any general, value-neutral news. We have looked at top 500 NS-listed companies (NSE 500 Index) and analysed the NIS data for these companies over 11 years (2010 to 2020).We have removed companies for which stock price data were not 1 Geert Bekaert, and Guojun Wu. Asymmetric Volatility and Risk in Financial Markets. The Review of Financial available, leaving us with 399 of the Nifty 500 companies. These companies are sorted on the basis of aggregate news frequency in a particular year and the top decile represents top ten percent companies which are heavily reported in the media in that year.

Data and Methodology:

We have looked at the effect of NIS on stock returns and volatility. Since, NIS are used at monthly frequency, stock prices and hence returns are estimated at monthly rest. Stock prices for firms are collected from Yahoo Finance. Stock returns are estimated as (natural)log of price relatives. In other words, we have defined stock returns as monthly change in prices on a logarithmic scale. Historical volatility (standard deviation) of stock returns is estimated using past 30 months returns on a rolling basis.

News flow was much lower during 2010 through 2013. Also, a large number of companies had very low news volume even after 2013. We have only considered, for our analysis, those companies which had news for at least 80% of the months since July 2014. This filter has reduced our sample size to 86 companies out of about 400 companies. Thus, we have monthly data for the variables since July 2014 till December 2020.

Results and Analysis:

We observe that returns are negatively correlated with NIS, lag of NIS, and lag of volatility whereas volatility is positively correlated with NIS, which is ideally correct since higher NIS indicates bad news and thus lower returns, but high volatility). Also, return and 2 Panel analysis is a statistical method to analyse two-dimensional (cross-section and time) panel data. The data are usually collected over time and over the same individuals and volatility are positively correlated as higher risk demands greater rewards. There is a negative relationship between current and past returns suggesting negative autocorrelations in returns which is a well-established feature of stock returns.

There are two types of effects in a panel model: Fixed or Random effect. A model is a fixed effect model if the variables are constant across individuals, random otherwise. To check which effect is best suited for the given data, we have conducted the Hausman test. Results show that it is better to use random effect models.