I use calendar days, as does TOS! I think you may have solved the mystery! I suspect if I change to trading days, the consistent Monday expiry IV reduction will vanish! Thank you for your insight!
I've been randomly thinking about this off and on for the past several years but have been too lazy to do the backtesting to actually test it out. I'm not sure that there is the same amount of realized volatility in the hours between market close on one weekday and market open the next and the hours between market close on Friday and market open on Monday. If you just look at business days, you basically say there's the same amount of volatility occurring between say Wed and Thursday and between Friday and Monday, or on a 3 day weekend Friday and Tuesday. I get that there's much less business going on during those 3 days, but events other than business impact volatility and there is some business going on then, even if nothing more than the dumping of bad news on Fridays after the close. Or to put it another way, wouldn't you be sweating it more if you sold an ATM option at the close on Friday that expired on Monday at the open, than at the close on Wednesday that expired on Thursday at the open? Like I said, I have been too lazy to even search the literature on this, but you might have motivated me to do a little work.
3 bows toward "LM3886"! You da man! I just added option to use business days instead of calendar days and the Monday Expiration anomaly (consistently lower IV of other expirys) is no more (AKA: Mystery solved)! Personally, I think they use trading days instead of calendar days as a "close enough for government work" shortcut.
You are very welcome! I'm not even sure if pros use business days, but I've been burnt by the calendar days in TOS. Long Monday, short Friday calendars show great PnL graphs in TOS but the "edge" would never materialize.
pros use business days. Calendar days doesn’t make sense because there is no volatility on Saturday and it’s not like mondays have 1.7x the volatility as other days.
Can you point me to more insight data supporting this? (some research data/articles or something to help me better comprehend) My observations support your statements, but I am having trouble with concept that volatility (risk) is absent on non-trading days! Thanks in advance for your help!
Thought I'd make another post on this topic. IMHO: Statement from "destriero" was accurate. Statement from "newwurldmn" also is accurate. Statement from "LM3886" was instrumental in me being able to better comprehend. I appreciate all the responses! For inquiring minds, I am posting an HTLM file that shows all PUT options that are +/-3 standardized moneyness strike ranges for SPX near terms (AM, and Friday PM, as well as the Monday PM) from Thursday open till Friday (yesterday) close in 5-min steps. You can watch the IV shift as we walk thru this interval. The file was zipped, so ET would accept it. -- My time is still in calendar days, which I think is proper to handle forwards, interest rate, etc.
Interesting animation! It seems that the upside IV fluctuates a lot. Maybe it is due to the put being DITM and the bid/ask spread is large. Out of curiosity, what software did you use to plot it?
Those with + moneyness values (DITM) are not very liquid as you noticed with the huge B/A spread. I wrote that in python. -- plotly.express is very nice for 3D scatter plots.