Dr. StrangeVol or: How I Learned to Stop Worrying and Love the VIX

Well, 2020 has certainly gotten off to an…interesting start. In less than two months, the new year has seen the spread of the coronavirus throughout China and beyond, an escalation of tensions between the US and Iran over the death of Qassem Solemani, and the third-ever impeachment of a president in the history of the United States. All of these massive news events occurring have been leading to apprehension and worry to anyone closely following global affairs. One might expect financial markets to reflect this uncertainty and tension. However, as of this post the S&P 500 is up around 0.5% and the MSCI ACWI (an index of large- and mid-cap stocks from across the globe) is up around 1.5% since the start of 2020. It’s not massive returns, but it’s not the drawdown one would expect in a time of uncertainty.

That being said, market returns aren’t necessarily the best barometer of market uncertainty. The underlying volatility of the market tends to be the best indicator of uncertainty; a low volatility environment means that most investors are fairly content with the market, while a high volatility environment indicates a lot of investors scrambling to change their positions in the market because they’re not exactly sure where things are going to end up. The market standard for measuring volatility in the S&P 500 is the Chicago Board Options Exchange Volatility Index, more commonly known by it’s ticker symbol: VIX. The VIX is calculated from the prices of near-term S&P 500 index options, making this index a measure of implied expected future volatility in the market. Basically put, VIX looks at the price changes of options over time; if option prices are seeing a lot of movement, then VIX goes up. While it is colloquially referred to as “the fear index”, VIX is really more of a measure of market uncertainty: VIX can go up when prices go up. Still, most major drops in market price happen during highly volatile periods, so VIX more often than not increases when the market is expected to go down.

With that being said, how has the VIX reacted to the major news events of 2020? The VIX has averaged around 14.4 at close, very close to the historical baseline of the index and indicating that the market won’t see more than a 0.75% swing in a day about 2/3rds of the time. Even more, the highest VIX has closed so far this year is 18.84, which would expect a price swing under 1% in a day about 2/3rds of the time. This maximum is a noticeable bump above the baseline, but is still magnitudes below the 150 seen during the 1987 Crash, the 2008 recession’s peak above 80, or even the typical 40-50 seen during a temporary market correction. This period of low movement in VIX indicates one of two things: 1) the markets don’t care too much about these major and rare events that are occurring, or 2) the VIX is a poor indicator of market performance.

In order to test this, I took a look at three sets of monthly data: the price of the S&P 500, the VIX, the Consumer Confidence Index (CCI), and the US Crash Confidence Index from the Yale School of Management from July 2001 to December 2019. The Crash Confidence Index measures the responses of institutions and individuals regarding their thoughts on the probability of a major stock market crash occurring in the next six months. The higher the score, the higher the expected chance of a crash occurring. This information presents a behavioral finance perspective on market uncertainty: if people think that the market is going to crash in the next half-year, we’d expect it to reflect in the market itself. The correlation of all sets of data were compared to one another to see if there was any sort of relationship between the S&P 500 and the VIX/CCI/Crash Confidence Index.

Correlation with monthly S&P 500 price:
Crash Confidence (Institutional): 0.03
Crash Confidence (Individual): -0.10
Consumer Confidence Index: 0.57
CBOE Volatility Index (VIX): -0.48

The correlation with the S&P 500 is strongest with the CCI, followed by a moderate negative correlation with the VIX (which makes sense, since VIX typically rises when the S&P 500 goes down). It’s interesting that the Consumer Confidence Index is the top performer here and not the VIX, as the CCI is from qualitative data (consumer optimism) while the VIX is much more quantitative (option pricing). Quants can tend to be dismissive of qualitative data, but results such as this one and the Behavioral Finance movement show the merits of a combination of quant and qual data. Meanwhile, there’s little-to-no correlation between the Crash Confidence Indexes and the S&P 500; in fact, individuals have a higher correlation than institutions, but in the wrong direction. This indicates that there’s a very slight incorrect assumption by individuals on predicting major declines in the market.

However, these correlations cover these indices within the same month as the S&P 500. Could it be possible that the indices are leading indicators? In other words, does an increase in the VIX in January lead to a decline in the S&P 500 in February?

Correlation with t+1 monthly S&P 500 price:
Crash Confidence (Institutional): 0.02
Crash Confidence (Individual): -0.10
Consumer Confidence Index: 0.57
CBOE Volatility Index (VIX): -0.44

Correlation with t+3 monthly S&P 500 price:
Crash Confidence (Institutional): -0.01
Crash Confidence (Individual): -0.13
Consumer Confidence Index: 0.55
CBOE Volatility Index (VIX): -0.38

Correlation with t+6 monthly S&P 500 price:
Crash Confidence (Institutional): -0.07
Crash Confidence (Individual): -0.18
Consumer Confidence Index: 0.52
CBOE Volatility Index (VIX): -0.32

Correlation with t+12 monthly S&P 500 price:
Crash Confidence (Institutional): -0.14
Crash Confidence (Individual): -0.21
Consumer Confidence Index: 0.50
CBOE Volatility Index (VIX): -0.24

Over time, the correlation between the S&P 500 and CCI and VIX get weaker, while the correlation between the S&P 500 and the Crash Confidence Indexes gets increasingly negative. In fact, the Individual CC and VIX are nearly similar one year out, which is not promising for someone trying to guess a market crash a year in advance!

Finally, what is the correlation between the VIX and the other indices? How often do the indices move in lock-step with one another, and is there any relationship that we can take away from this that might be useful as an indicator?

Correlation with VIX:
Crash Confidence (Institutional): -0.52
Crash Confidence (Individual): -0.47
Consumer Confidence Index: -0.61

The strong inverse relationship with both the Institutional and the Individual Crash Confidence Indexes is interesting. This means that these indices are much more precise but much less accurate when it comes to VIX. In other words, Crash Confidence is more likely to move with VIX than with the S&P 500, but is going to move in the wrong direction. Institutions and individuals are more likely to think a crash is impending when the VIX is declining instead of the other way around. This calls into question the wisdom of the respondents to the survey, and more generally calls into question the intuitive feelings and opinions that we have about the markets themselves. If our gut feeling is uncorrelated to the market itself, then perhaps emotions, opinions, and major events don’t really have as much of an impact as we’d expect them to.

Meanwhile, there is a strong inverse relationship between VIX and the Consumer Confidence Indexes. What does this mean? Well, when the Consumer Confidence Index increases from month-to-month, there’s less uncertainty and more confidence in the market. Consumers are more-inclined to spend and less inclined to believe that the market is going to recede. The relationship isn’t inversely perfect (most likely due to the fact that VIX can also increase when the markets go up), but it’s decent. For fun, I put together a crude amalgamation of the two ( [CCI-100]/VIX = Index217* ) and tested that against the S&P 500.

Correlation between the S&P 500 and Index217:
t+0: 0.58
t+1: 0.57
t+3: 0.57
t+6: 0.56
t+12: 0.55

While it doesn’t reach the highs of CCI at t+0, there’s less decline in correlation over time. There might be something to a more-refined combination of CCI and VIX as an indicator of S&P 500…but I’ll leave that to you (or myself at a future date).

Ultimately, the VIX has a moderate inverse correlation with the S&P 500. It’s not the bullet-proof “fear index” that would lead to effective use as a barometer of market sentiment, but there is some merit to increases in VIX leading to reduced returns in the S&P 500 the following month. However, VIX is much stronger than either Crash Confidence Indexes, indicating that assumptions about potential market movements are nearly completely untethered from actual price movements. In fact, the VIX and the Crash Confidence Indexes have a moderate inverse correlation, indicating that uncertainty in the market is relatively opposite of uncertainty by market participants. This isn’t a knock against qualitative data though, since the Consumer Confidence Index has a much stronger correlation with the S&P 500 than either Crash Confidence Indexes or the VIX. This could be due to it being an indirect measure of the markets, a better collector of true-to-life data, or that “it’s the economy, stupid.” In fact, a crude combination of CCI and VIX shows correlation resilience over time with the S&P 500, possibly indicating a potential useful hybrid indicator in a more-refined amalgamation.

So, the next time a major event occurs, do the following: instead of stressing out about how this is going to tank the markets, go take a look at the VIX and the Consumer Confidence Index. It may not be right all the time, but it will give you a much better idea of what to expect from the market instead of your gut feeling.


*CCI = 201, VIX is Roman numeral gibberish as-written, but can be rearranged to XVI =16. 201+16=217, hence Index217. That’s right, I stuck a bad math pun in this post. Deal with it.