Hey all, I am trying to strategically calculate some strategies that incorporate's the actual time decay during the trading day. I am using the live Theta number IB posts for the options I have in my ticker window. I understand that if it is -0.10 it should decay 0.10 per day, right? But if you look at the SPY options expiring today, Oct 4th. The 252.50 put that is out of the money, a bid premium of 0.07, and is showing a Theta of-0.02. This can only go to 0 today at the current price, so I don't know how it can show anything less than 0.07. (or more than -0.07) Does anyone have any insight into how these theta numbers can be interpreted from an open to close day? Thanks
Please beware that from open to close of a day, time is NOT the only thing that changes! --- Look at the derivation of theta, and realize market conditions are also a primary factor. From the question, it seems you may not be taking a comprehensive view surrounding your question. (doing some calculations using BSM can help to understand the dynamics). If price of the underlying moves, there will also be an impact to your theta; it could be beneficial to understand this.
Two things: Look at the IV of that puppy -- it will be huge. Why? Lumpy pricing. "End-of-Life" stuff. *Ignore* theta (and IV) on tight-time, way-OTM strikes. Look at the theta of a Oct31 2525 -- and realize that the model (and its inputs) takes things down to the second -- as in, 00d:00h:00m:00s. When you are looking at theta/day, keep in mind that the 'theta'-day has 24 very full hours. And so, cash-market-days being 6.5hrs, you can expect roughly only 25% of theta's effect to show up while you're watching it. (The rest is what drives wacky option prices in the last 30 minutes of the day, and the opening 30 minutes of the next day...)
Interesting. So If I was looking at an ATM option, in the morning, with a week expiry (Fri-Fri), with a theta of 1.0. I would most likely see a decline of 0.25? Assuming nothing changed in the underlying. I remember a previous thread on here debunking that there is time decay overnight. That all traders have priced into the next open at the close.
1) What you'll see, most times, will be bracketed by the "$1.0" of the model's theta and the ~25¢ of a 6.5hour window on the day. 2) There is time decay every single second -- the clock doesn't stop for nobody/nuthin'. Whether Mistress Theta deigns to work her magic in a fine, prescribed curvilinear fashion "According to Hoyle"* -- no -- no more than "You should always sell on Thursdays, as the MMs will factor the weekend in, on Fridays." Yeah. Uh-huh. But, No. Do *not* expect linear behavior in a lumpy, curvilinear world. "Wide markets cover a multitude of sins." (*being, Profs Black, Scholes, & Merton, et.al.) In thinking a little further about this.... by no means is this offered up as "empirical" but, ....... in the last .... maybe year or so.... I would swear that most option decay happens in two windows -- between 11:30am and 12:30pm, and again --from 3:20pm--4:00pm, US Eastern. The second is overlaps with "wide markets" so, take that for what it's worth. The first..... might even be........... maybe 10:30--11:00. "Before Lunch" at any rate. And then a big flat spot in the middle of the day, and then the end-of-day maelstrom. Huh. I bet there's a study out there. It'd be easy enough -- to factor out the market's wash/froth from the option/strike decay. Huh. By 30minute window. Hmmmmmm. [Where's that head-scratching emoji when you need it?]
This doesn't really touch on the main topic, but some neat observations on the tangent the thread is on... I've noticed similar time-specific decay, and have a thought on why it happens that way. I've seen a bit more specifically at exactly 12, 2, and 3:30 there are precipitous drops in premiums--especially on Wednesday and Thursday. It's most visible on less liquid options and those with .05 increments. I suspect this is market makers following each other down. The few times I've seen it, you'll watch the bids dry up, and almost immediately after the last one drops down, the asks will follow with other MMs following the ask almost immediately. They just organically find this time and eventually it becomes habit. And then there's the obvious 2pm FOMC Wednesdays... Another more interesting one is the non-linear nature of decay. Consider MO, PM, DAL, and BABA. MO has almost all of it's decay during trading hours, especially on the put side, because that's when regulatory / litigation news is likely to come out (and in the 3 hours or so afterwards--because of West Coast litigation). They're very unlikely to have an earth-shattering event between 8pm-6am ET. PM on the other hand, has a very linear decay because it's not exposed to the single moment pivot of regulation or litigation because it's diversified around the world, and what exposure it has only has a few brief hours of downtime while the sun transits the Pacific. BABA, heavily exposed to the Chinese regulatory regime and China's economy has almost all of it's decay occur during the Chinese day which is completely outside the US trading hours. So it tends to be flat through the day, with big gaps between days. DAL, which operates around the world and is sensitive to major news event (crashes, terrorist attacks...) has a very linear decay, with a slight bias for the US daytime hours. Nothing scientific about those observations, but it's something I noticed a while back, and I haven't come across anything that leads me to believe they're untrue.
Thanks. Well I didn't mean to suggest the observations were invalid--just not strictly scientific. But it is a hypothesis based on a plausible and logical cause-effect relationship, continuously tested and evaluated for disproofs, and as yet not disproved. I just didn't document it, nor was the sample random, nor large enough to state conclusively. So, call it informed by the scientific method if not in conformance with it. Back to my point about exact times on even hours for the market makers to come down, I suspect the same thing happens at 10am too, but I'm never at my computer to watch it live. But it does make sense that this is in the same series at 12 and 2, and is 30 mins from the bell as 3:30 is. So there's another hypothesis to test to fit into this theory.