Saturday Review For June 25, 2022

This past week was either a solid dead cat bounce or the end of the down turn for stocks. It felt like a very solid week so I am going to lean to the former as the other short term bullish moves in U.S. stock has covered just a couple of days over the past few months. Last week just felt stronger.

A final thought on U.S. stocks, earnings season is quickly sneaking up on us. Earnings and forward statements could have a dramatic impact on equity index performance through the end of the summer, if not for the balance of 2022.

Last week we ran some numbers regarding VIX levels at the end of weeks where the S&P 500 has dropped 5% or more. Since SPX gained over 6% this past week, we ran the opposite numbers. There are 26 instances of 5% or greater one-week SPX gains since 2000. The buckets in the table below are fairly evenly segmented and give the average, highest, and lowest VIX levels at the end of each of these strong weeks for stocks.

For those not indoctrinated into VIX, the elevated levels for VIX in each row above may be surprising. However, several of these strong weeks occurred during periods where stocks were rebounding from large losses and an already elevated VIX level.

The VIX related ETPs put up expected numbers with the relatively new SVIX leading the way rising almost 8%.

Finally, the VIX curve shifted back into contango, a shape that has persisted more than would be expected in the current market environment as VIX moved well below 30.

Weekly readers may feel like we are a broken record repeating the same observations, but the November future at a discount to October continues to standout, possibly a reflection of concerns about the early November mid-term elections. For clarity, October VIX settlement would be based on options expiring after the election, while November settlement will be calculated using December options. Using the VIX curve as a predictor, this means the market will be a bit on edge before the election, with volatility dropping after the election.

The popular press and fintwit have both been making noise about VIX or implied volatility versus realized volatility saying VIX is not properly indicating market volatility. This is typically displayed as realized volatility over the past month versus the current level of VIX.

The better way to measure this is to compare VIX or volatility expectations to the realized volatility that follows a VIX closing price. VIX and implied volatility represent anticipated price behavior from the underlying market so we ran numbers to see how broad based index implied volatility compares to subsequent realized volatility.

For the following tables, we used Bloomberg’s consistent at-the-money (ATM) 30-day implied volatility measure and compared it with realized volatility over the following 21 trading days. The Bloomberg measure is calendar days while our realized volatility measure is trading days.

This first table shows by year, the percentage of observations where ATM implied volatility was higher than subsequent realized volatility for the S&P 500, Nasdaq-100, and Russell 2000. Note 2022 is a big outlier with no similarly low comparison since 2010.

We still have more than half 2022 to go so it is possible for these numbers to fall in line with history, This will require lower realized or a bump higher in volatility expectations. Probably not to surprising, but NDX realized volatility has been greater than expected volatility almost 9 out of 10 observations in 2022. The following three tables breakout more data by year for each index.

We subtracted realized volatility from expected volatility and calculated the average, maximum, and minimum (always negative) figures by year. The first table used S&P 500 data.

On average ATM 30-day SPX implied volatility has understated realized volatility by 2.50 points. This is not as much of an outlier as the figure on the first table with 2018 (thanks Volmageddon) putting up a slightly more negative number.

The Nasdaq-100, which was the worst of the three indices on the first table, as far as realized volatility versus implied, mirrors that in the table below.

On average ATM NDX 30-day expected volatility has understated realized by almost 6 points. No other historical figures matches up with that figure. We assume that number should drop over the balance of 2022.

Finally, of these three major indices, the small cap focused Russell 2000 option market has done the best job of pricing expected volatility this past year. On the first table, expected volatility was overpriced over 40% of the time so far in 2022, lower than average but better than the other two option markets. Note on the table below, RUT expected volatility has fallen in line with realized volatility with the aveage difference only -0.20.

Although nothing is certain in the world these days, we have to assume when we look back at the end of 2022 that the percect of observations where volatility is overpriced should improve. If that assumption is shared, looking to more neutral spreads in NDX or XND may make sense.