Media Attention and the Volatility Effect Blitz, et al. SSRN 2019
The news coverage of companies undoubtedly has an effect on their stock price returns. With news travelling much faster since the widespread adoption of the internet, this effect has become an interesting point of study. The team at Robeco Institutional Management has recently taken swing at quantifying this effect using data from news sources, press releases, premium newswires, and over 19,000 web publications.
It seems logical that a company that has been in the news more often would experience more volatility as investors get more visibility of the company name, logo, and details of their operations with little effort.
Instead of searching for all the information on every possible company, investors may only purchase the stocks of companies that are able to grab their attention. Barber and Odean (2008) develop the attention-grabbing hypothesis and find empirical evidence that individual investors are more likely to buy stocks that have been in the news, all other things equal.
Volkswagen is used as an example of this with the emissions scandal, where volatility in their stock price increased nearly in lockstep with the amount of monthly news articles. This is the above mentioned ‘attention grabbing hypothesis’. Where stocks with low return volatility have high risk-adjusted returns, which may be due to the low amount of media attention given to these companies. This is also outlined in the ‘low-risk anomaly’ (or sometimes referred to as ‘low-volatility effect’):
The low-risk anomaly is the empirical observation that stocks with low return volatility or market beta have higher risk-adjusted returns than the market, while stocks with high-risk characteristics have lower risk-adjusted returns.
Given the attention-grabbing hypothesis and the low-risk anomaly, the paper serves to test two hypotheses:
- The low-volatility effect disappears for stocks with high media attention.
- The low returns for high-volatility stocks are caused primarily by stocks of companies that appear most frequently in the news.
The Analysis
The team identifies a set of what they term ‘attention-grabbing stocks’ based on the number of times each stock has appeared in the media over the past year.
The stocks that receive the most media attention are considered ‘attention-grabbing’ stocks. It is important to stress that the attention-grabbing hypothesis does not involve the news sentiment or tone, i.e. whether the news is positive or negative, but the volume of media attention only. Including the sentiment of news articles does not capture attention, but instead makes the signal more similar to that of price momentum.
They also take the important step to adjust their measure of media attention based on the size of the companies:
If we would simply take the total number of news articles per company as a sorting variable, we would have a measure that is tilted to larger companies, as larger companies are more often in the news. Therefore, we use a size-adjusted news measure for our main analyses. This size adjustment follows from cross-sectional regressions:
Unpacking their size adjustment regression, Newsi,t is the total number of news articles for stock i during the year prior to time t. The variable Zi,tLOGMCAP is the normalized logarithm of the market capitalization of stock i at time t performed using robust z-scores. And the residual εi,t is seen as the adjusted news measure used for sorting stocks into portfolios.
The sample portfolios and calculated volatilities are made up of the 3000 largest stocks in the combined MSCI World Index and the S&P Broad Market Index. The paper also tests the hypothesis on a sample of emerging markets (MSCI Emerging Markets Index and S&P IFC Emerging Markets Index) and U.S. only stocks (top 1500 large cap U.S. stocks for the month). The stocks are then divided into five groups based on volatility, and the average size-adjusted media attention measure over time is calculated for each.
Based on these five portfolios, the chart above outlines the relationship found between the media attention per volatility group, and the volatility per media attention group.
(the above chart) shows that there is an increasing pattern between volatility and media attention. Media attention increases for stocks in higher volatility groups, and volatility increases for stocks in higher media attention groups. This figure indicates that ‘glittery’ stocks tend to be the stocks with relatively high volatility, while the ‘boring’ stocks that the media does not write about tend to have relatively low volatility
The paper then goes into more detail of their process in trying to find the driving force behind the volatility effect and to see if it is indeed the media-attention effect.
Testing the First Hypothesis
The two hypotheses identified earlier are now tested:
- The low-volatility effect disappears for stocks with high media attention.
Table 1 contains the average returns, volatilities, Sharpe ratios, and alphas of a number of portfolios. The top left (“All, All”) is the equally-weighted portfolio of all stocks in our sample. The first column (“All”) further contains the five volatility-sorted portfolios without taking into account media attention, and the first row (“All”) contains the five media attention sorted portfolios without taking into account return volatility. The last row and column (“Low-High”) contains the difference in Sharpe ratios and alphas of the low volatility (or low media attention) and high volatility (or high media attention) portfolios.
The paper notes that the average excess return of all sample stocks is 6.89%, with a volatility of 15.05%, and a Sharpe ratio of 0.46. The statistics in the first column also show that portfolios sorted on return volatility have a higher Sharpe ratio for low volatility stocks, and a lower Sharpe ratio for the high volatility stocks.
For media attention to be able to explain the low-volatility effect, we should see that the low-volatility effect disappears for stocks with similar media attention. However, Table 1 shows that for each column with similar media attention, the Sharpe ratios are the highest for low-risk stocks and monotonously decline when volatility increases.
From the table, it is noted that the differences in Sharpe ratios are statistically significant, same with the alphas (bottom row under ‘media attention’). The first hypothesis is rejected under these conditions.
Testing the Second Hypothesis
Moving along with the second hypothesis, looking at the rows in table 1 for ‘media attention’:
- The low returns for high-volatility stocks are caused primarily by stocks of companies that appear most frequently in the news.
The attention-grabbing hypothesis implies that for stocks with similar risk, the stocks with the highest attention should have inflated prices, which leads to lower future returns. The univariate sort based on media-attention shows that stocks with low media attention have low return volatility, while portfolios with high media attention stocks have higher volatility.
The values for the media attention portfolios don’t align with the attention-grabbing hypothesis and the paper finds that return differences are not significantly different from zero. This leads to rejecting the second hypothesis.
Based on the attention-grabbing hypothesis, we would expect higher Sharpe ratios and alphas for the low-media-attention portfolios compared to the high media-attention groups within each of the five volatility groups. However, we see no statistically significant differences for these portfolios, as all t-statistics in the final column are below two.
Wrapping it all up
The authors take the time to do some robustness analysis using an additional sample of emerging market stocks and U.S. only stocks. The analysis help confirm the main findings discussed above. I won’t go into much more detail, but the whole robustness analysis can be found in the final section.
Overall, the paper took a deep look at the driving forces behind the volatility effect to test the relation with media attention given to specific stocks and finalized their results:
Sharpe ratios and alphas for low-volatility portfolios exceed those of high-volatility portfolios for groups with similar media attention. For groups of similar volatility, the Sharpe ratios and alphas are statistically indistinguishable for stocks regardless of the amount of media attention. Based on these findings, we reject that the attention-grabbing hypothesis explains the volatility effect.
This result isn’t much of a surprise. It would be a long shot to expect the attention-grabbing hypothesis to be a sole driving force behind the low-volatility effect. However, the analysis completed in this paper finally gives some empirical evidence that there is more behind the scenes driving stock volatility than pure exposure.