Risk-on? How fear triggers sell-offs and when to catch the falling knife

Risk-on / Risk-off

Investing is an activity of forecasting the yield over the life of the asset; speculation is the activity of forecasting the psychology of the market.
~ John Maynard Keynes

On the morning of May 6, 2010, mayhem erupted on the streets of Athens, Greece.  Later that day, the Flash Crash occurred, sending the U.S. stocks markets down by 9% intraday.  A random coincidence?  The data suggests not.

Like that May day in 2010, the spectre of default risk raised its head this month.  Chinese equities plummeted, Grexit appeared imminent, and even little Puerto Rico contributed a scare.   Institutions have a limited power to save us from ourselves, as this excellent article by Jason Zweig recounts, and investors were again reminded that the surges of emotion through markets - especially fear- can drive prices dramatically.  A global risk-off movement took hold briefly, and given the disquiet produced by events in China and Europe, it is probable that the volatility is only beginning.  

Risk-on behavior is a collective pursuit of financial safety, while risk-off describes the flow of capital to momentum-driven and higher yielding securities.  Periods of risk-off are catalyzed by a growing sense of unpredictability, increasingly negative news flow, and rapidly falling prices.

Today's newsletter looks at how a seed of doubt, once germinated, can begin to choke investors with tendrils of fear.  We examine a sentiment-based sell signal in China before the recent downturn which is similar to a signal seen before the May 2010 Flash Crash.  The question for investors on the sidelines of such panics is when to "catch the falling knife" and buy into the fear.  Evidence shows that Warren Buffett is correct to counsel us to "buy on fear," but ONLY when fundamental valuations are also attractive, which we demonstrate through recent research.

China Chain Reaction

Sentiment in the English-language business news about China has generally predicted the recent Shanghai Composite 300's ups and downs.  In the following image we see the 30 and 90 day sentiment averages of all media reporting on China.



Chinese business sentiment appears to lead the Shanghai Composite somewhat.  The shading between the two averages is pink if the 30-day average is lower than the 90-day average, and it is green if the 30-day average is higher than the 90-day.  When the short-term sentiment rises rapidly, it tends to pull prices upwards, and when it falls, prices tend to follow down.  Such sentiment averages moving average crossovers have appeared a useful indicator for timing bubble tops in historical testing, and we see examples of sentiment driving prices most often during dramatic events (bubbles and crashes).

How Fear Drives Prices Lower:  The Flash Crash

On the morning of May 6, 2010, U.S. stock markets opened down and trended down for most of the day.  Television financial news outlets replayed video footage of Greek police lines confronting Molotov-cocktail hurling protestors amidst burning cars on downtown streets.  50,000 Greeks marched in Athens and clashed with police in pitched street battles.  As worries about the Greek debt crisis swelled, at 2:42 p.m., with the DJIA down more than 300 points (3%) for the day, the stock index began to fall rapidly, dropping an additional 600 points in 5 minutes.   By 2:47 pm the DJIA showed a 1,000-point loss for the day (9%).   Then just as suddenly as it had started, prices reversed.  Twenty minutes later, by 3:07 p.m., U.S. equity markets had regained most of the 600 point drop.  This event became known as the “Flash Crash.”   

The following image portrays a 60-minute rolling average of fear expressed about S&P 500 stocks in both news and social media on the day of the flash crash. A large surge in fear was evident in media 1-2 hours before the crash, with a second spike in fear following the event itself (this is a simple 60-minute average of fear, so it takes some minutes to respond to the event).  



Fear cascades into selling, first gradually and then suddenly.  The Flash Crash was also preceded by a rapid decline in media sentiment as seen in the image below.



The preliminary CFTC-SEC investigation of the Flash Crash noted that on May 6, 2010 markets had been depressed by “unsettling political and economic news” and “growing uncertainty in the financial markets.”  In late 2010 the CFTC-SEC blamed a large sell order by asset manager Waddell and Reed for being the proximate cause of the crash.  Then in 2015 London trader Navinder Sarao was indicted for spoofing activities contributing to the Flash Crash.  The CFTC-SEC's search for a concrete technical cause of the crash overlooked a more significant proximate cause: human behavior.  Watching a Greek hurling a Molotov cocktail at the police causes the release of stress hormones and fear neurotransmitters, and it primes us to seek safety (both financially and otherwise).

In experimental consumer markets, induction of fear in participants leads to lower asking prices and lower bid prices.  So as a result, when fear spikes or is increasing, we expect prices to fall.  Our research on the Thomson Reuters MarketPsych Indices broadly confirms this effect:  Daily jumps in fear on average precede price downturns, while declines in fear precede price rises.

Fear in Value Investing

And if they insist on trying to time their participation in equities, they should try to be fearful when others are greedy and greedy when others are fearful.
~Warren Buffett

Our exceptionally talented Head of Research, Changjie Liu performed a quantitative study to understand the effect of fear on value stocks, and in the process he debunked the folk-wisdom to "buy on fear."  CJ confirmed that value investing really does outperform the S&P 500 in the recent period, and when augmented by news fear, it performs even better.  

First, CJ examined value returns.  The equity curve below depicts simulated long-only value investment performance using a monthly rotation strategy based on the following criteria:
1.  Select the top U.S. stocks in news and social media by Buzz over the past twelve months.
2.  Rank those 100 stocks by their Earnings-to-Price ratio (the inverse of P/E).
3.  Buy the top 25 stocks and hold for 12-months.
4.  Repeat this process every month from Feb 1 1998 through 2014 with one-twelfth of the portfolio.

The equity curve derived from this long-only strategy, excluding transaction costs, is below, and it shows a 12.5x return. Recall, this involves one-year holding periods with monthly updating of 1/12 of the portfolio. No transaction costs or dividends are included. - Ken French's website value factor data shows about a 9x return over the same period for a similar strategy, but one not filtered by media "Buzz").​



In the following equity curve, CJ bought only those stocks ranked in both the top 25 of E/P and in the top 25% of Fear over the past 12 months.  These are value stocks undergoing high fear:​


From this yearly equity curve - showing a 50x returns - it appears that Warren Buffett is right: buying value is good when investors are fearful.  However, this result does not persist in non-value stocks.  Growth stocks don't recover predictably after a fear-based sell-off.  Monthly buying of value stocks on fear is also lucrative, but less so than the yearly turnover strategy above. Fear added into a monthly rolling value model improves returns by approximately 50% versus the 400% excess performance when fear is added to a yearly value screening above.  Buying feared stocks ONLY makes sense if they are also objectively cheap (good values), and it is more lucrative over the yearly (versus monthly) measurement horizon.

Greek Default

Last month's newsletter was written in a hurry (our product launch was time consuming last month), and I'm afraid the letter wasn't clear about my thoughts on Greece.  First of all, I'm not assigning blame.  The reality is that Greece needs debt relief.  The country can't pay what it owes, and the large debt overhang is a psychological barrier that is hurtful to the country's progress.  And it's not only Greece, eventually most of the developed world will similarly be overcome by debt, with examples too numerous to mention.  We can argue about institutional reform, that Greeks don't pay their taxes, that German banks were unfair, etc...  But that's not relevant now.

Longer term, without fiscal as well as monetary union, the Euro simply isn't sustainable.  Maybe we'll get another few years with some band-aids this weekend.  And the Euro could last another century or more if sensible rules around political and fiscal unity were implemented.  Perhaps the latest crisis (unlike the ones before it) will spur such reforms.  

By the nature of many participatory democracies, leaders are rewarded with votes by special-interests in exchange for short-term "sweeteners."  Whether it's public sector union, or more liberal oil & gas extraction rules, every party is supported by some with a financial stake in the outcome.  When the negative long-term consequences of short-term political sweeteners (pork) are not realized until the politician is out of office, there is little accountability.  Implementing real consequences on politicians for bad long-term decision making -such as fines and jail time for those who use catastrophically poor decision-making processes (e.g., relying on wishfully biased forecasts of growth when making pension promises or utilizing objectively inadequate intelligence before launching wars) - may make good sense, but it's unlikely to pass the very political bodies it hopes to limit.

Housekeeping and Closing

When both China and Europe sneeze, it isn't long until the rest of the world catches cold.  The shock of a rapid price decline takes a few days to weeks to settle in, and the fear it generates can prompt a contagious flight to safety.  The best time to buy will be when clear values are to be had.  While there are numerous Chinese value stocks traded in Hong Kong - even now - that are becoming even cheaper on the back of Chinese investor angst, they still may have room to fall.

Please contact us if you'd like to see into the mind of the market using our Thomson Reuters MarketPsych Indices to monitor real-time market psychology and macroeconomic trends for 30 currencies, 50 commodities, 130 countries, 50 equity sectors and indexes, and 8,000 global equities extracted in real-time from millions of social and news media articles daily.

We love to chat with our readers about their experience with psychology in the markets.  Please send us feedback on what you'd like to hear more about in this area.

Be safe out there!
Richard L. Peterson, M.D. and the MarketPsych Team