I was reviewing the P&L on a paper-traded short strangle and noticed that my AI was calculating theta decay using trading hours. My understanding was that Black–Scholes runs in continuous time, so I searched this sub expecting a straightforward answer. I did not expect the question to be this contentious.
So, uh, yeah, I know this has been asked before: does an option’s time value decay when the market is closed, including overnight and on weekends? Theta is a major part of the reason I put on a short strangle, so the distinction matters to me. From what I’ve observed, it does decay, but I’d appreciate a more rigorous explanation of what happens in the model versus what shows up in actual option prices.
I have a related question. I often see people say that beginners rely too heavily on the Greeks because they’re just model outputs. So what does a more experienced trader rely on?
Does theta decay during non-trading hours?
byu/ginput inoptions
Posted by ginput
7 Comments
“Can time be stopped?!”
Yes it decays overnight, you or “AI” need to calculate it as well.
Depends whether you mean the model or the quotes. Textbook Black-Scholes runs on calendar time, so it decays overnight and over weekends, and most platforms quote theta per calendar day. In actual prices, market makers tend to mark implied vol down going into a weekend because no new trading happens, so the decay often shows up around Friday’s close instead of leaking out smoothly on Saturday and Sunday. So your AI using trading hours is a different convention, not necessarily a mistake, but worth checking which one each tool uses before comparing P&L. Since it’s a short strangle, the short call side has theoretically unlimited loss, and short options can lose far more than the premium collected.
Yes. And no. Theta doesn’t decay linearly, especially around events like earnings, fed announcements, etc. At earnings release you might decrease by several “days” worth of theta in a single second. Otherwise, it depends on if you’re likely to see more movement during or after market hours. Theta is an abstraction of actual information discovery.
Technically, it does decay. But by “technically” what are we really talking about? Black scholes gives us a theoretical price but really what does that even mean when actually trading?
Theoretical price is simply just theoretical. In practice, the theta decay predicted by BSM is baked in the bid-ask of the Market Makers that are quoting the option prices. The theta decay from overnight or weekend are usually priced in during the final quotes during the end of trading of a specific trading day.
There is obviously some decay that gets priced when the options market starts in the morning.
Overall, because the actual price discovery happens during trading hours, the theta decay is baked in accordingly by the MMs during the first 30 and final 30 minutes of each trading day.
Theta decays continuously, but mostly in theory. Think about the implication of applying that in practice though. Say you sell an option 1min before the close on Friday…do you really think the person on the other side of that trade would just happily eat 65hrs of theta decay so you can make free money over the weekend? Typically theta decay is accelerated in trading models to account for ‘downtime’
this is a much debated topic here on this forum (and in thetagang), and while my response will undoubtedly attract many screaming children saying that i am wrong. the true answer to this question is NO. the why is pretty complicated.
but the short summary is,
you dont decide the price, marketmakers do. and market makers use models to determine price. market makers tend to use 252 trading days a year in their models to predict option pricing. and because of that, you as retailed follow. in practice those models also use real integer days specially for long dated options. so theta seems to tick. day by day.
you are free to price options yourself, with any model you please. and you can create a model that uses 365 days, and say that is the reality. but all that will introduce is a small difference between your model and the market every weekend, which then corrects itself during the week. you are just a small fish, and the big players use trading days only.
now if you actually look at the data (real prices on the market over years) youll see that each trading day, loses 1 day of trading, and for short termed options, theta decayse gradually minute by minute. and that means no extra theta decay in weekends either. the cycle friday to monday, yields 1 day of theta decay, just as the cycle monday to tuesday,
source: me, 25 years of experience in banking, markets and securities developing models and administrations,