Checking SPX Implied Volatility Skew for Bad Quote Data
Summary
The document questions an apparent U-shaped implied-volatility pattern in SPX options, especially at far out-of-the-money strikes. A response compares the displayed pattern with a snapshot of implied volatilities for weekly options, reporting generally declining values across strikes and identifying irregular or zero readings in some quotes. The author attributes the discrepancy to potentially stale or unreliable data rather than to a reversal in the relationship between volatility and the underlying.
The practical lesson is to check how option prices were formed before interpreting a volatility surface. The response recommends calculating implied volatility from bid-ask midpoints rather than last trades and checking whether far out-of-the-money contracts have active markets. Its evidence is a single dated snapshot and a particular market-data source, so it does not establish a general explanation for skew behavior or prove that every unusual curve is a data artifact. It also does not resolve the model-based question about how local-volatility or stochastic-volatility models shape the right side of the skew.
Key ideas
- Unusual implied-volatility curves can result from stale or poor-quality option quotes.
- The response compares a chart with a dated snapshot of SPX option implied volatilities.
- Midpoint prices are preferred to last trades when estimating implied volatility from quotes.
- Far out-of-the-money strikes need sufficient market activity for their implied volatilities to be informative.
- A single snapshot cannot explain all volatility-skew shapes or validate a general model.
Tags
Full text
# Understanding skew of SPX - Why does IV of OTM puts increase with strike? # Understanding skew of SPX - Why does IV of OTM puts increase with strike? I've been trying to understand the skew I see when looking at the skew of SPX. Here is a snapshot today from thinkorswim. I understand why IV increases for ITM puts -- namely because there is a negative correlation between volatility and moves of the underlying. But I don't understand why the OTM puts IV increase with strike or why the ITM calls IV increase with strike. I've been studying the different models that give rise to skew. For example: - Local volatility models (e.g. Derman papers) Riding on a Smile Regimes of Volatility The Local Volatility Surface - Heston stochastic volatility model I understand the increased IV on the left, but it is not obvious to me how these models explain the right portion of the plot above. The skew plots seem to say that volatility is inversely correlated with underlying moves for a while, but after a "big" move, that correlation (between volatility and underlying price) changes to positive. Is that the correct interpretation? Is this explained by the mean reverting nature of the stochastic volatility in the Heston model? ## Answer by Alex C (score 2) https://quant.stackexchange.com/a/21478 When I saw these curves they seemed very strange to me. I believe it is a data-quality issue.I went to Bloomberg and I retrieved the implied vols for 70 near ATM strikes of the weekly SPX options expiring November 27 2015 (I believe that is the yellow curve in your diagrams i.e. November 4th week). This was today 2015-oct-27 at about 15:00 New York time. As you can see from the table below, the vols are monotonic for both calls and puts, completely unlike the U shape seen in the curve you posted. 27 Nov 15 (31d); IDiv 3.22; R .20; FF 2062.47 ``` Strike Call IV Put IV 1975 15.950 17.352 1980 16.484 17.058 1985 0.000 16.795 1990 16.175 16.462 1995 21.817 16.258 2000 16.235 16.202 2005 15.485 15.908 2010 15.730 15.712 2015 15.887 15.481 2020 15.250 15.330 2025 15.148 15.013 2030 14.812 14.895 2035 14.701 14.598 2040 14.341 14.363 2045 13.884 14.156 2050 13.916 13.899 2055 13.774 13.678 2060 13.465 13.358 2065 13.217 13.194 2070 13.041 12.981 2075 12.724 12.638 2080 12.560 12.289 2085 12.307 11.587 2090 11.977 11.123 2095 11.737 11.042 2100 11.606 11.676 2105 11.426 11.275 2110 11.146 7.126 2115 11.081 10.512 2120 10.891 9.904 2125 10.755 0.000 2130 10.574 0.000 2135 10.489 0.000 2140 10.370 0.000 2145 10.243 0.000 2150 10.324 0.000 ``` Make sure that your IV's are computed from the bid ask midpoint, not the last price (which could be several hours old) and that there is real market activity in the way far out of the money strikes that you are displaying. There was no activity at all today in strikes higher than 2120 (for puts) and very little above 2170 for calls. Those prices (and IV's) may not be meaningful.
Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.