News

Analyzing Text to Predict Market Volatility

Economy
0 min read

The Complex Relationship Between the Pandemic and Market Volatility

The Coronavirus pandemic has impacted the global and U.S. economies and led to a sharp rise in unemployment. The U.S. unemployment rate was reported at 13.3% for May based on U.S. Department of Labor statistics, representing levels not seen since the great depression (Exhibit 1). There are currently over 40 million Americans unemployed. Following 128 months of expansion, representing the longest on record since 1854, U.S. entered a recession in February. As the markets react to the uncertainty attributed to coronavirus infections and unemployment, the volatility in the markets seems to be inevitable (Exhibit 2).

Exhibit 1. The Percentage of Unemployment in the last 10 years in U.S.

 

Exhibit 2. The Overlay of S&P 500 (SPX) Index and CBOE’s Volatility
Index (VIX)

Source:
indexindicators.com and Noble Capital Markets

Nobel Laureate and NYU Stern professor Robert F. Engle and his co-authors (Ahmet K. Karagozoglu and Nazli Sila Alan) wrote a working paper on a forecasting model for the impact of COVID-19 pandemic on volatility in global equity markets.  The paper recently became available on the Volatility and Risk Institute at NYU Stern School of Business website at https://vlab.stern.nyu.edu/covid19. The authors used a multi-regime forecasting model to investigate the impact on market volatility. The findings include

  1. daily number of active cases are significant predictors of realized volatility,
  2. stricter policy responses by individual countries result in lower stock market volatilities,
  3. higher negative managerial sentiment causes an increase in realized volatilities.

 

Exhibit 3. Daily COVID-19 Cases for selected 7 countries January 22 to
May 1, 2020

Source: Alan, Engle, and Karagozoglu,
“Multi-regime Forecasting Model for the Impact of COVID-19 Pandemic on
Volatility in Global Equity Markets”, Volatility and Risk Institute Working
Paper, Stern School of Business New York University, June 15, 2020

Two different measures of equity market volatility (i) GJR-GARCH volatility based on global stock indices and, ii) realized volatility based on intraday prices of country specific ETFs), were used to differentiate the market uncertainty attributed to the pandemic. The intraday 5-minute returns for 46 country-specific ETFs were used to determine the realized volatilities, and daily GARCH volatilities are estimated using the stock market indices of 88 countries around the world. Dr. Engle received the 2003 Nobel prize in Economic Sciences for his work on analyzing economic time-series data, that formed the basis of the GJR-GARCH model used in this paper.

The largest increase in volatility levels were observed on February 24th, March 9th, and March 16th , then realized volatility (RV) returned to its pre-pandemic levels on April 3rd. GARCH and RV volatility levels for seven of sample countries are shown in Exhibit 4.  The U.S. stock market reached peak volatility level on March 16th, 4 days after the Italian market. The volatility levels for all countries have been declining since then.

Exhibit 4. Daily Realized Volatility (5-min) for selected 7 countries
between January 22 to May 1, 2020

CHN (China), DEU (Germany), ESP (Spain), GBR (United Kingdom), ITA (Italy), KOR (Korea), and USA Source: Alan, Engle, and Karagozoglu,
“Multi-regime Forecasting Model for the Impact of COVID-19 Pandemic on
Volatility in Global Equity Markets”, Volatility and Risk Institute Working
Paper, Stern School of Business New York University, June 15, 2020

In Exhibit 4, the daily median realized volatilities (RV), calculated using 5-minute intraday returns for country specific ETFs and GARCH volatilities are shown for seven countries including China (CHN), Germany (DEU), Spain (ESP), United Kingdom (GBR), Italy (ITA), Korea (KOR), and the USA. A total of 46 country specific ETFs were analyzed and estimated using GARCH model for daily stock index returns of 88 countries as well as the CBOE VIX for the U.S.

The analysis of text of publicly traded firms earnings call transcripts of in various countries (Textual Data Analytics of Transcripts database of the S&P Global Market Intelligence for 6,500 publicly traded firms from 38 countries) was used to correlate managerial negative sentiment due to the pandemic and the broader market volatility. Negative Sentiment were identified as the ratio of total number of negative words to total number of Master words in the same transcript.

Exhibit 5. Daily Median GARCH Volatility for Global Equity Markets vs
Mean Global COVID-19 Curvature and The Comparison to CBOE Volatility Index
between January 22 to May 1, 2020

 

Source: Alan, Engle, and Karagozoglu,
“Multi-regime Forecasting Model for the Impact of COVID-19 Pandemic on
Volatility in Global Equity Markets”, Volatility and Risk Institute Working
Paper, Stern School of Business New York University, June 15, 2020

The empirical analysis suggests a complex relation between the pandemic metrics and stock
market volatility
. The multi-regime forecasting model for the relation between equity market volatility and the coronavirus pandemic hypothesize that the number of active cases and negative sentiments by the management teams are powerful predictors of daily cross section of volatilities, while stricter policy responses by individual countries result in lower stock market volatilities.

 

Suggested Reading:

 

More
Accurate than Polls to Gauge Election Outcomes

Can the Market Continue to Defy Gravity?

Are
we on the Verge of Acquisition-Mania?

 

Research Over 6000 Small and Microcap Companies on  Premium Channelchek Content at No Cost!

 

Sources:

https://vlab.stern.nyu.edu/static/Covid19Volatility_MultiregimeForecasting_VRI_NYU_20200615.pdf


Share

Inbox Intel from Channelchek.

Informed investors make more money. And it’s all about timing. Get it when it happens.

By clicking submit you are agreeing to the Terms of Use and Privacy Policy
© 2018-2024 Noble Financial Group, Inc. All Rights Reserved. Channelchek is provided at no cost to be used for information purposes only and not as investment advisement.