Oddball T+0 line and P/L behavior

status1

Well-known member
I was playing around with different expiration combinations for a calendar spread in SPX and I came across an oddball combination that seems to make a glitch in the TOS projection of the T+0 line and P/L calculation

Looking at the monthly 15 OCT 84 DTE and 29 OCT on the 4350 Put strike the price for the calendar is 14.35 Delta is -.36 Theta is 41.4 Vega is 677 and the p/l is around 6.60 This is after the market is closed

So comparing that to the weekly OCT 85 DTE which is just one extra day and keeping everything else the same I get
Delta is -.33 Theta is 42.9 Vega is 663 and the p/l is around -1.42

So as you can see the greeks and pl are fairly close but now let's see what happens when I advance the date by one

The biggest difference is a huge jump in p/l to 571.82 for the monthly and actually even lower for the weekly to -2.82
It's only after this hiccup start that the p/l would continue to advance with the normal theta value for both

It is also messed up when going backward I get 484 with -1 date for the monthly and -86 for the weekly
Again continuing lower will be at the theta rate

I was wondering if anyone has an explanation for this
 
Did you try resetting the risk graph between the two? I find that I often have to reset it when I'm trying out different types of trades.
 
I am not sure how you would reset the risk graph between the two
Did you mean to click on "Reset parameters and date " or something else ?
As far as I know that only resets the date and the volatility if I was playing with it which I was not
 
While on the Risk Profile tab, in the top right corner of your screen where you see the 3 horizontal lines, click on that and one of the options is to "Reset". It will reset all of your risk graphs and clear the memory.


Hope that helps.
 
I reset it but it did not make any difference
Another odd thing is when I look at the "More parameters" drop down list there is no difference between the 15 OCT monthly and weekly expiration
Just for fun I try adjusting it this morning and it looks like the adjustment is for both so you can't adjust the monthly without affecting the weekly
 
I was playing around with different expiration combinations for a calendar spread in SPX and I came across an oddball combination that seems to make a glitch in the TOS projection of the T+0 line and P/L calculation

Looking at the monthly 15 OCT 84 DTE and 29 OCT on the 4350 Put strike the price for the calendar is 14.35 Delta is -.36 Theta is 41.4 Vega is 677 and the p/l is around 6.60 This is after the market is closed

So comparing that to the weekly OCT 85 DTE which is just one extra day and keeping everything else the same I get
Delta is -.33 Theta is 42.9 Vega is 663 and the p/l is around -1.42

So as you can see the greeks and pl are fairly close but now let's see what happens when I advance the date by one

The biggest difference is a huge jump in p/l to 571.82 for the monthly and actually even lower for the weekly to -2.82
It's only after this hiccup start that the p/l would continue to advance with the normal theta value for both

It is also messed up when going backward I get 484 with -1 date for the monthly and -86 for the weekly
Again continuing lower will be at the theta rate

I was wondering if anyone has an explanation for this
hello status1:

i am looking at it also after market close; actually saturday night. And here is how it looks like..
1627191849957.png

looks ok to me. other than the fact that since its a cal, you cant even really depened on the expiry pnl curve. But you already know that.. probably.

And the am expiry looks ok too.
1627192099841.png
one thing i notice in your greeks is a huge difference in vega - like a magnitude of by 10. maybe you are looking at a 10lot - but your deltas were low for that. so maybe something else.. why not put a pic to show what you mean?
 
You have the setup correctly
Now, advance the day by one and see what happens and compare the two after that

I do have a 10 lot but they are both 10 lot so I am comparing the same lot size
Yes the vega is bigger because I used a 10 lot but comparing the two the difference is very small which would not explain the difference after one day advance

The Delta was smaller at the time I was looking at it but SPX moved a lot since then but again they were both small so there is not much difference between the two

The biggest difference is in P/L after advancing the date by one
 
You have the setup correctly
Now, advance the day by one and see what happens and compare the two after that

I do have a 10 lot but they are both 10 lot so I am comparing the same lot size
Yes the vega is bigger because I used a 10 lot but comparing the two the difference is very small which would not explain the difference after one day advance

The Delta was smaller at the time I was looking at it but SPX moved a lot since then but again they were both small so there is not much difference between the two

The biggest difference is in P/L after advancing the date by one
ah. I see. it's showing a bigger profit in one vs the other. yes.. surely it's a model error.

I think someone said that before already.. it's just a guide. Real results will be different; and since you are modeling with a 10lot you are seeing a 10x error magnified.
 
I noticed the same weird behavior recently and it happens with any trade not just calendars. The P/L is in error when advancing one day from current date and after that it gets in step with theta. It also happens with a trade you opened some time in the past when you advance one day from today.
I contacted support and got some nonsense answer that P/L depends on current prices and not theta.

Tired of TOS, but what else to use?

Don Kaufman who runs TheoTrade said that back in the day he was part of the team developing TOS and the reason TOS got that complex is because they tried to implement every feature enough users asked for.

No wonder now it got to the point where it crashes under its own weight.
 
I noticed the same weird behavior recently and it happens with any trade not just calendars.
I have not noticed that much of a difference in verticals for example

I set up a vertical 100 wide 4200/4100 10 lot one with Oct 15 monthly and the other with weekly
They both start of around the same p/l -4.60 vs -4.30 weekly vs monthly
after one day the monthly is at 18 with theta at 32 so it's less than one theta movement while the weekly only moved to -4.80
so yes there is a difference but nowhere near as much as the calendar where the one day move is 10 times the value of theta

Another characteristic I noticed is that to the monthly always jumps more on the one day advance than the weekly but I can't say I noticed that much of jump for a calendar

Normally I only trade the weeklies so that's why I have not noticed the big difference until now when using the monthlies
 
Here is a condor I entered past Friday at 1.40 credit. Now showing theta=44.99, P/L=-9.99
1627307821059.png
Advancing one day. theta=45.06, P/L=5.99
1627307777094.png
Advancing one more day. theta=47.52, P/L=51.24 (pretty much as it should be).
1627307850363.png
 
I understand that and I have seen that before and I agree it's a problem but it's not as bad as the calendar
The main problem is with the monthly expiration

With the weeklies since the long is in the same expiration the problem is somewhat minimized but when the long is in a different expiration there is nothing to offset the monthly expiration

See what you get if you replace the short call and put to aug monthly and the long call and put to aug 27 weekly in your iron condor
I get theta at 13 with the first day jump in p/l to 127 which is nearly 10 times the theta value A 10% pop in one day That would be nice except it's not real

You can also look at just a single aug 20 monthly compared to a aug 27 weekly and you will see the monthly jumps a lot more than the weekly Actually the weekly looks more like it stumbles or skips one day as it does not make much on the first day

So when you combine the aug monthly that jumps a lot with the weekly that does not move much you get this very exaggerated fake p/l pop
 
The reverse is even worse
If I use the 15 OCT monthly as the long for the same 4375 strike and the 30 Sep for the short on the calendar the theta is a little higher at 62 and the p/l at open is -1.41 than when I advance it one day I get a loss of -595 than you have to hold it for 2 weeks just to get even Good thing that is not real

So it's roughly about a +/- 500 change in p/l on the first day advance when using the monthly in a calendar so the risk graph is virtually useless to see where the p/l might be on x day in advance
I would have to use the nearby weekly to get an idea of where the p/l might be on that day
 
I also contacted them by email a couple of days ago but I have not heard back from them yet
Hopefully if enough users complain they may do something about it
 
I just got a reply to my email and it's basically the same answer answer that you got
"Recently, there have been some technical issues with the way that the Analyze tab is reacting in terms of theoretical price, and our tech team is aware and have been working on a resolution for this. "

The only thing is that I don't think this is something that happened recently but hopefully they can do something about it but I doubt it will be anytime soon
 
I found a new problem or maybe I did not notice it since I was trading mostly weeklies spreads

I have a calendar on the analyze tab and one of the legs is using the 17 Sep weekly expiration The problem is when I close TOS and open it the next time the expiration changes by itself to the monthly which threw me off the first time because it looked like I made a lot of profit in one day but that I saw that it was the monthly so I changed it back to the weekly and got the p/l back to reality

I have seen expirations change by itself when it expires and if you open TOS the next day it moves to the next expiration but this is not past the expiration so I am not sure why it is moving It's kind of annoying because I have to change it every time I open TOS
I sent TD an email to see what they say about it
 
Looks like not all of them are doing it I just happened to stumble maybe on the only oddball one that is doing it
It's the short 17 Sep weekly and the long 30 Sep Quarterly calendar that is doing it
I tried the single leg and it's not doing it and I also tried the 17 Sep weekly and 24 Sep calendar and it's also not doing it
I guess it's some kind of glitch that TOS does not like so it changes it to the monthly for whatever reason
 
It seems like the quarterly is causing the problem because I see the same problem with the 17 Dec weekly and 31 Dec quarterly
calendar At least it's limited to 4 times a year for this calendar combination
 
The problem is still there and here is a refresh: P/L is not shown correctly when advancing one day from current date. P/L should increase by the value of Theta after each day. It doesn't do that after one day, but if you continue it will obey the formula.
Well I found proof that it was correct in the past. In February I made few videos for a friend to show him few things in TOS. It took me a while to go through them and find this.
Here are three consecutive screen shots while advancing the day from current date which was 2/06/2021.
Currently P/L=-0.16 and Theta=3.71
1628736533642.png

After one day P/L=3.55 and Theta=12.97
1628736558002.png

After another day P/L=16.52 and Theta=31.33
1628736581080.png

Each time P/L increases exactly by the value of Theta.
 

Attachments

  • 1628736353682.png
    1628736353682.png
    252.6 KB · Views: 2
Top
Contact Us