Eleven years ago we were trading through the Swine Flu epidemic. At the time, while I was trading our hedge fund (MarketPsy Long-Short Fund LP), my colleague Frank Murtha penned this humorous advice to relax about the outbreak (perhaps directed at me!)
Our fund's strategy took advantage of fear-driven dislocations, and we created the below image of an actual trade for our investors to understand how the fund timed investor overreaction. (In 2009 trading on social media was seen as seriously goofy, to put it politely, and we needed clear explanations of how our quant systems identified good trades).
Of course, most of our trades were not this successful, but overall the strategy worked well during the high-fear environment of the financial crisis, and our performance in our first 12 months - from Sep 2008 through Aug 2009 - was in the top 1% of hedge funds. After the financial crisis the strategy stopped working, and by the end of 2010 we shut it down due to nonperformance.
Fast forward to this month's corona virus...
Twelve days ago we created a flurry of LinkedIn posts describing the likely effects of coronavirus on the financial markets, using past epidemics - and the MarketPsych indices we produce from news and social media. But to objectively understand the outbreak and its affects on financial market prices, it helps to dissect similar events like a scientist.
In that vein, below is a Rembrandt titled “The Anatomy Lesson of Dr. Nicolaes Tulp." Note the differing emotional expressions of the students, as well as the use of light and subtle facial coloration to emphasize mood - the capturing of subtle emotion is once reason Rembrandt is considered an exceedingly rare genius. And one of my highlights (among many) of living in the Netherlands is the easy access to tremendous cultural treasures like this.
“The Anatomy Lesson of Dr. Nicolaes Tulp" 1632. Rembrandt. Now in The Hague, The Netherlands.
And along the lines of scientific inquiry into emotion, one of our star interns from Vrije Universiteit Amsterdam - Tiago Teodoro - generated a report on the effects of such epidemics on the financial markets (specifically airline stocks). His report reviews several cases, and I share a few highlights below.
Mining the Data
In order to study such events, we looked through the lens of our MarketPsych Indices from Refinitiv (RMI) data feed. Our feed measures and delivers real-time market psychology and macroeconomic trends from thousands of news and social media sites.
It's not only investors like us who look at such data streams. Hotel and airline booking websites also track natural disasters and events that affect travel patterns so they can preemptively adjust prices to anticipate demand: "Hyperdynamic pricing factors in lots of data. Along with historical and seasonal information, the new A.I. systems scan the web for global news events, weather predictions, trending Google searches, social media posts, local event schedules and other factors that could affect demand, Ms. Zutavern said."
So it makes sense that big investment firms are focused on this challenge as well, and of all the triggers to investor panic and recovery, contagious diseases are one of the most consistent. Below is Tiago's image from the Ebola epidemic in 2014. The light blue line is the value of the Airlines ETF (FAA, now delisted). The first dark red-shaded spike is the outbreak of the disease into the US media (our U.S. Human Infectious Disease index, in yellow vertical bars, reached over 1% of media references to the United States). The second region of red-shading is news about the first human ebola case in the US, corresponding with a sharp selloff and rebound in the Airlines (even while ebola was still being discussed). Note that the big rebound may also be attributed to a big drop in the oil price.
Source: Tiago Teodoro
Here is a chart of the Chinese stock market (China Composite) versus our Human Infectious Diseases index for China (also below). Note that the MACD represents the frequency of Human Infectious Disease mentions in the global media about China (it peaks at 3%). The yellow bars are the price of the China Composite 300 (which was closed due to Chinese New Year in late January). Note that a significant flu outbreak was reported in the Chinese media in early January but fell off (was quashed to prevent panic?) until it spiked higher in mid-late January.
Here is the same type of chart for Hong Kong and the Hang Seng. The discussion of coronavirus is beginning to wane now in both regions.
Regarding the current outbreak and Chinese stocks, below is where we stand (as of yesterday) in a Chinese surgical mask manufacturer (Tianjin Teda) and China Southern Airlines. The yellow bars are the bottom represent the percentage of Human Infectious Disease reference in articles referencing China, its cities, and its regions.
Source: Tiago Teodoro
When Will the Mayhem End?
In Tiago's research, he found that during an epidemic, it's best to wait to buy airline stocks until the RMI for sentiment starts to rise and the RMI for fear starts to decline (he used moving average crossovers to see this, as in the lower two charts for ebola above). Here are links where you can keep track of sentiment around two coronavirus-affected airlines - Cathay Pacific (sentiment chart and fear chart) and China Southern Airlines (sentiment chart). Note that the gaps in prices are due to the market closures in China during the Chinese New Year holidays. And note that sentiment is now rising, and fear is now falling (see Disclaimers below).
Housekeeping and Closing
We love everything market psychology-related (which in our opinions, is virtually everything financial market-related), so feel free to reach out with any stories, questions, or comments.
Shameless product plug: Our MarketPsych Indices from Refinitiv (RMI) data feed measures and delivers real-time market psychology and macroeconomic trends from thousands of news and social media sites. The commercial dataset includes granular themes and sentiments for 45 currencies, 62 sovereign bonds and stock indexes, 15,000+ companies and stocks, 36 commodities, 187 countries, and 150+ cryptocurrencies back to 1998 (2009 for crypto). If you're an academic interested in data for research, please reach out for access. If you represent an institution, please contact your Refinitiv account manager for a trial or contact us to get set up with an account manager.
Richard Peterson M.D. and the MarketPsych Team