Ever since June (at $200/share), Tesla stock has been driven by a perpetual motion machine of hype and call option flows -- nothing more. And everyone knows it.
Here's what not everyone knows:
When a stock joins the S&P 500, it becomes part of a massive volatility complex, which is a terrifying web of arbitrage and pseudo-arbitrage relationships. Tesla will join the index as a top-ten component of a cap-weighted index. It's big.
Its bigness will allow all manner of dispersion, relative value, and market-making traders to begin relying on Tesla's newfound correlation to the index. This will invariably cause arbitrageurs to buy SPX options/vol and sell TSLA options/vol to "close the spread."
Blue line is the mean 1-month return associated with DIX. Returns are normalized to VIX to isolate signal. E.g., with VIX at 20, a 0.5 MAD return would be +4.6%.
Red line is the VIX-implied mean absolute deviation (MAD). Above 1.0, VIX was cheap. Below 1.0, VIX was expensive.
Between 38% and 46%, where nearly all the data clusters, a clear, linear relationship is apparent. Higher prints are more bullish.
At the extremes, where there is little data to sample, there are anomalies. These tend to be periods of market stress or extreme option positioning.
(For instance, the huge gains associated with 36% DIX are all from the pre-Volmaggedon (1/2018) S&P 500, where call overwriters were short-squeezed into oblivion.
Owing to the clarity of the relationship where data is *abundant*, this seems like a true anomaly. Grain of salt.)
It's been a while since we've posted a scandalous question, so here's a doozy.
Regarding VIX and the front-month VIX future (VX1), which statement is more accurate?
So, this was a really bad poll, and here's why:
There are going to be two types of respondents, (1) those who implicitly believe that "convergence" refers to TIME, and (2) those who implicitly believe that "convergence" refers to PRICE.
If you think we're talking about TIME, then this devolves quickly into a philosophical debate about whether spot ("now") actually moves through time toward the forward contract, or whether forward contracts move toward the present.
Volatility skew implies gains are more likely than losses.
You have two hours.
Thank you to 1,000 respondents for tolerating this strange, ambiguous poll.
There's a theoretical problem here. In the case of a skewed distribution, either the median or the mean can be zero, but not both.
Do options say the mean or median is zero?
Most respondents, by simply allowing that skew could *possibly* predict median gains, are saying that option prices are centered on a mean anticipated return rather than a median.
This means options are "implying" that index upside is always more likely. That's interesting.
@Barton_options While it implicitly emphasizes the GEX values of stocks like the FAANGs (due to the high open interest in those names), it is dwarfed by the size of the gamma coming from CBOE's SPX index options.
That said, it correlates strongly to SPX index gamma.
@Barton_options Right now, two things have caused Component GEX to reach for the sky.
1. The January option expiration is next week, so the gamma from January LEAPS (lots of OI) is elevated.
2. The market is going berserk, because reasons. So more calls are ATM, so gamma is higher.
Traders will have noticed that $SPX has hardly moved at all this week (+0.45% right now). As usual, GEX is the proximate reason.
The speed with which the market flips from high- to low-volatility still continues to catch traders off-guard. People think that volatility should do what their GARCH models say -- smoothly mean-revert. But it doesn't. It jumps.
And it's not as if the evidence hasn't been there. When we first introduced the GEX white paper, we noted that "15 VIX" doesn't have much practical forecasting value. Which, when looking at the plot of GXV by SPX, makes perfect sense.