Let's Talk SPX Box Spreads

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
comeinvest
Posts: 2709
Joined: Mon Mar 12, 2012 6:57 pm

Re: Let's Talk SPX Box Spreads

Post by comeinvest »

muffins14 wrote: Sun Jun 05, 2022 2:19 pm For example I think margin interest is tax deductible if used to purchase securities, up to your net investment income total. So a margin loan’s interest cost would be deducted from short term gains and non-qualified dividends, correct?
But margin interest is also an itemized deduction, i.e. you must exceed the threshold over the standard deduction. So it all depends.
I also heard that having a margin interest deduction means greater audit risk. An audit can cost you a lot of time, even if your tax returns are correct, and time is money.
adamhg
Posts: 218
Joined: Sat Apr 10, 2021 8:40 am

Re: Let's Talk SPX Box Spreads

Post by adamhg »

Kevin M wrote: Sat Jun 04, 2022 2:12 pm It was exactly as shown in my post, so if by strikes you mean the 4,000 and 4,300, then yes, same strikes and same everything else.

Thanks,

Kevin
For what its worth, it looks like this might have also been a floor trade. It has the condition code: "Transaction is a late report of the opening trade and is out of sequence; i.e., other transactions have been reported for the particular option contract."

https://www.interactivebrokers.com/en/t ... ntract.,-G

Any traders with better exchange knowledge know what that means exactly?
User avatar
Kevin M
Posts: 15787
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Spreads on IRA tier

Post by Kevin M »

Kevin M wrote: Thu Jun 02, 2022 4:51 pm Spent an hour and 15 minutes on the phone today with Fidelity re: IRA option Spreads on IRA tier.

<snip>

Where we left it was for me to try again when the market was open, and see if I could get to the preview screen with any price. If not, he said to call and ask for the active trader desk, telling them I wanted help with entering an option spread trade. He said they would still need to call the back office, and that they could possibly open up "the limits" so that I could enter the order, or possibly they could enter it for me. And at least they could give us more insight into what was preventing me from entering the order. I obviously don't want to have to call every time I do this. I'm hoping it's a glitch in their system that can be resolved, since none of them seemed to really understand why this was happening.

Kevin
Today I was able to place an order in my Fidelity IRA for the same box spread I tried on Friday.

Order for 16 Dec 2022, strikes 4,000 and 4,300, price 296.65. This is the same order I have open in my taxable account, on which I upped the price to 296.65 today. Neither has filled yet.

I probably will increase price in the IRA first, since that's where I'd rather hold fixed income anyway.

Kevin
If I make a calculation error, #Cruncher probably will let me know.
adamhg
Posts: 218
Joined: Sat Apr 10, 2021 8:40 am

Re: Let's Talk SPX Box Spreads

Post by adamhg »

comeinvest wrote: Sun Jun 05, 2022 5:17 am Hi adamhg,
I think you might want to consider using days to settlement in your APY calculations instead or days to expiration, because as Kevin correctly said, the yields will be more accurate. If both the trade date and the expiration date are not preceding a weekend or holiday, it doesn't matter, as both are settled T+1. But if the trade date is a Tuesday (for example) and the expiration date is a Friday (for example), then I think the duration of the loan would be 2 days longer than the DTE, right? Because settlement of the trade would be T+1, and settlement of the assignment/exercise at expiration would be on Monday, i.e. T+3 (3 calendar days that are subject to interest).
I can see the argument why you might want to use settlement date, but don't most brokerages let you trade on pre-settled funds? At least TDA does for me. So even though at expiry, it technically doesn't settle for 3 days, the funds are available for trading on T-0 so long as I don't withdraw. So I guess if you're looking for when you get back to cash-cash, T+3 would make sense, but if you're just looking for your liquidity back, T-0 feels better. Am I thinking about this right?
adamhg
Posts: 218
Joined: Sat Apr 10, 2021 8:40 am

Re: Box spread yield calc

Post by adamhg »

Kevin M wrote: Sun Jun 05, 2022 4:49 pm Some more observations about calculating the rate (yield) for box spreads.

It appears that boxtrades.com is using this to calculate rate:

(spread/price - 1) * 365.25/DTE

For example:

(300/296.60 - 1) * 365.25/193 = 2.169%, which is what boxtrades.com shows.

If I use 365 instead for days in a year, we get 2.168%. This also is the yield returned by:

= YIELD("6/6/2022", "12/16/2022", 0, 296.60, 300, 1, 1)

The last two parameters are coupon frequency = 1, day_count_convention = 1 (actual/actual).

Treasuries are generally priced using freq = 2, day count = 1. One exception is the auction yield shown for T bills of 26-week or less maturity, in which case the yield is obtained using freq = 1.

I understand using 365.25 as an approximation for days in the year for longer maturities to account for leap years, but Treasury uses actual days in a year in their calculations, so either 365 or 366.

Kevin
That's the formula I'm using. I considered 365/366 originally but couldn't wrap my head around when I should actually use 366 instead. I'd be happy to change this if you could help me with the logic.

Here are the significant events+dates I could think of and the combinations thereof...
  • Trade before/after/on? 2/29 in a leap year/not in a leap year
  • Expiry before/after/on? 2/29 in a leap year/not in a leap year
Here's what I'm thinking, including "?" where I'm not entirely sure what the convention should be. I think, only 366 when the trade -> exp goes directly through a leap day? Otherwise use 365? Does that make sense/reflect how other securities/apps are doing it?

Code: Select all

+-----------------------------------------------------------------------------+
|Trade / Expiry      |3/1/2019|2/28/2020|2/29/2020|3/1/2020|2/28/2021|3/1/2021|
+-----------------------------------------------------------------------------+
|2/28/2019           |365?    |365?     |366      |366     |366      |366     |
+-----------------------------------------------------------------------------+
|3/1/2019            |--      |365?     |366      |366     |366      |366     |
+-----------------------------------------------------------------------------+
|2/28/2020           |--      |--       |366      |366     |366      |366     |
+-----------------------------------------------------------------------------+
|2/29/2020           |--      |--       |--       |366?    |366?     |366?    |
+-----------------------------------------------------------------------------+
|3/1/2020            |--      |--       |--       |--      |365?     |365?    |
+-----------------------------------------------------------------------------+
|2/28/2021           |--      |--       |--       |--      |--       |365     |
+-----------------------------------------------------------------------------+
|3/1/2021            |--      |--       |--       |--      |--       |--      |
+-----------------------------------------------------------------------------+
johnanglemen
Posts: 713
Joined: Thu Oct 01, 2009 1:00 pm

Re: Let's Talk SPX Box Spreads

Post by johnanglemen »

adamhg wrote: Mon Jun 06, 2022 1:14 pm
comeinvest wrote: Sun Jun 05, 2022 5:17 am Hi adamhg,
I think you might want to consider using days to settlement in your APY calculations instead or days to expiration, because as Kevin correctly said, the yields will be more accurate. If both the trade date and the expiration date are not preceding a weekend or holiday, it doesn't matter, as both are settled T+1. But if the trade date is a Tuesday (for example) and the expiration date is a Friday (for example), then I think the duration of the loan would be 2 days longer than the DTE, right? Because settlement of the trade would be T+1, and settlement of the assignment/exercise at expiration would be on Monday, i.e. T+3 (3 calendar days that are subject to interest).
I can see the argument why you might want to use settlement date, but don't most brokerages let you trade on pre-settled funds? At least TDA does for me. So even though at expiry, it technically doesn't settle for 3 days, the funds are available for trading on T-0 so long as I don't withdraw. So I guess if you're looking for when you get back to cash-cash, T+3 would make sense, but if you're just looking for your liquidity back, T-0 feels better. Am I thinking about this right?
Personally I would prefer settlement dates. That's the date that the broker margin debt gets paid down (or drawn upon) and interest stops (or starts) accruing. I also feel that's a "purer" calculation: It represents when you "actually" get (or lose) your cash.

Thank you for the site, it's extremely helpful.
User avatar
Kevin M
Posts: 15787
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Box spread yield calc

Post by Kevin M »

adamhg wrote: Mon Jun 06, 2022 1:47 pm
Kevin M wrote: Sun Jun 05, 2022 4:49 pm Some more observations about calculating the rate (yield) for box spreads.

It appears that boxtrades.com is using this to calculate rate:

(spread/price - 1) * 365.25/DTE

For example:

(300/296.60 - 1) * 365.25/193 = 2.169%, which is what boxtrades.com shows.

If I use 365 instead for days in a year, we get 2.168%. This also is the yield returned by:

= YIELD("6/6/2022", "12/16/2022", 0, 296.60, 300, 1, 1)

The last two parameters are coupon frequency = 1, day_count_convention = 1 (actual/actual).

Treasuries are generally priced using freq = 2, day count = 1. One exception is the auction yield shown for T bills of 26-week or less maturity, in which case the yield is obtained using freq = 1.

I understand using 365.25 as an approximation for days in the year for longer maturities to account for leap years, but Treasury uses actual days in a year in their calculations, so either 365 or 366.

Kevin
That's the formula I'm using. I considered 365/366 originally but couldn't wrap my head around when I should actually use 366 instead. I'd be happy to change this if you could help me with the logic.

Here are the significant events+dates I could think of and the combinations thereof...
  • Trade before/after/on? 2/29 in a leap year/not in a leap year
  • Expiry before/after/on? 2/29 in a leap year/not in a leap year
Here's what I'm thinking, including "?" where I'm not entirely sure what the convention should be. I think, only 366 when the trade -> exp goes directly through a leap day? Otherwise use 365? Does that make sense/reflect how other securities/apps are doing it?
<snip>
For Tbills, the Treasury Department uses the number of days between settlement (issue) and one year after that. So for 6/7/2022 settlement, days = 365, but for 6/7/2023 settlement, days = 366.

Source: Price, Yield and Rate Calculations for a Treasury Bill

For bills with maturity of not more than one half-year to maturity, they use your formula with days either 365 or 366. I get the same results with YIELD using frequency = 1, day_count_convention = 1.

For bills of longer maturity, they use a more complicated formula to calculate coupon-equivalent yield. I get the same results with YIELD using frequency = 2, dcc = 1.

If you want a rate that you can compare to Treasury yield (bond/coupon-equivalent yield), for maturities of more than six months, you need to factor in the semi-annual coupon payment assumption, even though there is no coupon payment.

Kevin
If I make a calculation error, #Cruncher probably will let me know.
User avatar
Kevin M
Posts: 15787
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Spreads on IRA tier

Post by Kevin M »

Kevin M wrote: Mon Jun 06, 2022 1:09 pm
Kevin M wrote: Thu Jun 02, 2022 4:51 pm <snip>
Today I was able to place an order in my Fidelity IRA for the same box spread I tried on Friday.

Order for 16 Dec 2022, strikes 4,000 and 4,300, price 296.65. This is the same order I have open in my taxable account, on which I upped the price to 296.65 today. Neither has filled yet.

I probably will increase price in the IRA first, since that's where I'd rather hold fixed income anyway.

Kevin
Order in IRA was filled at 296.75. Yay!

This is a gross yield of 2.082% and net yield of 2.064%. Best Treasury yield for T maturing about the same time was 1.673% when I checked earlier today. This gives a gross spread of 41 bps and net spread of 39 bps.

Last price update was at 3:46 pm ET, and order was filled at 3:52 pm ET, so about 6 minutes to fill.

Same order in taxable account still is open at 296.65.

Kevin
If I make a calculation error, #Cruncher probably will let me know.
User avatar
Kevin M
Posts: 15787
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Box spread yield calc

Post by Kevin M »

If we want the box spread yield calc to be consistent with time value of money (TVM) calculations, we should use this:

r = (fv/pv)^(1/n) -1

Example using spread = 300, price = 296.75, DTE = 192:

If we want r annualized, then 1/n = 1/(192/365) = 365/192

r = (300/296.75)^(365/192) - 1

= 2.092%

This is the same rate returned the formula I shared earlier using the RATE function, and the same as using XIRR for the actual cash flows.

Using the boxtrades.com calculation with days in year = 365, the result is 2.082%, so one basis point lower. As mentioned before, this same result is returned by YIELD with freq = 1 and dcc = 1, so would be better for comparison to quoted Treasury yields. But it's only a difference of one bp, so no big deal.

Kevin
If I make a calculation error, #Cruncher probably will let me know.
User avatar
Kevin M
Posts: 15787
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Yield calc for 18-month box

Post by Kevin M »

The next box I'm thinking of buying in my IRA is a 15 Dec 23 with 4,000 and 4,500 strikes. If I buy this tomorrow, settlement will be 6/8, so I use 555 for DTE in my yield calculations. I'm thinking of starting at a price of 478.15, for a net price of 478.1772.

Using the boxtrades.com formula (with my day counts), I get a gross yield of 3.005%, and net yield of 3.001%.

Using standard TVM calculations I get 2.978% for net yield.

Interestingly, I get the same 2.978% net yield using YIELD with freq = 1 and dcc = 1. For the shorter maturities I've looked at, YIELD generates a rate lower than TVM calcs.

The highest yield I saw for a Treasury maturing on about the same date was 2.543% for min qty 25.

So at a gross price of 478.15 this is a gross and net yield spread of 46 bps.

Also, best yield for non-callable brokered CD at Fidelity maturing on 12/15/2023 is 2.5%, so about the same as the Treasury.

Kevin
If I make a calculation error, #Cruncher probably will let me know.
adamhg
Posts: 218
Joined: Sat Apr 10, 2021 8:40 am

Re: Box spread yield calc

Post by adamhg »

Kevin M wrote: Mon Jun 06, 2022 5:33 pm If we want the box spread yield calc to be consistent with time value of money (TVM) calculations, we should use this:

r = (fv/pv)^(1/n) -1

Example using spread = 300, price = 296.75, DTE = 192:

If we want r annualized, then 1/n = 1/(192/365) = 365/192

r = (300/296.75)^(365/192) - 1

= 2.092%

This is the same rate returned the formula I shared earlier using the RATE function, and the same as using XIRR for the actual cash flows.

Using the boxtrades.com calculation with days in year = 365, the result is 2.082%, so one basis point lower. As mentioned before, this same result is returned by YIELD with freq = 1 and dcc = 1, so would be better for comparison to quoted Treasury yields. But it's only a difference of one bp, so no big deal.

Kevin
Funny you mention this, and I'm glad you bring it up. I actually had this with the compound interest formula originally, but switched it to simple interest for shorts to better compare against margin rates 1:1:

viewtopic.php?p=6236190#p6236190

I actually forgot all about that whole conversation. So now, using the long side, we're comparing simple interest vs ytm for treasuries. I might need to think about this one.

I'd hate to display two different rates at the same time or different rates for long vs short. I'm not sure what the best approach is. Let me think about it and see if anybody else wants to weigh in.

For what its worth, the OCC also uses simple interest in their example: https://www.optionseducation.org/refere ... -cash.aspx
comeinvest
Posts: 2709
Joined: Mon Mar 12, 2012 6:57 pm

Re: Spreads on IRA tier

Post by comeinvest »

Kevin M wrote: Mon Jun 06, 2022 3:43 pm
Kevin M wrote: Mon Jun 06, 2022 1:09 pm
Kevin M wrote: Thu Jun 02, 2022 4:51 pm <snip>
Today I was able to place an order in my Fidelity IRA for the same box spread I tried on Friday.

Order for 16 Dec 2022, strikes 4,000 and 4,300, price 296.65. This is the same order I have open in my taxable account, on which I upped the price to 296.65 today. Neither has filled yet.

I probably will increase price in the IRA first, since that's where I'd rather hold fixed income anyway.

Kevin
Order in IRA was filled at 296.75. Yay!

This is a gross yield of 2.082% and net yield of 2.064%. Best Treasury yield for T maturing about the same time was 1.673% when I checked earlier today. This gives a gross spread of 41 bps and net spread of 39 bps.

Last price update was at 3:46 pm ET, and order was filled at 3:52 pm ET, so about 6 minutes to fill.

Same order in taxable account still is open at 296.65.

Kevin
May I ask why you invest in short-term money market instruments ("cash") in your IRA?
comeinvest
Posts: 2709
Joined: Mon Mar 12, 2012 6:57 pm

Re: Yield calc for 18-month box

Post by comeinvest »

Kevin M wrote: Mon Jun 06, 2022 6:32 pm The next box I'm thinking of buying in my IRA is a 15 Dec 23 with 4,000 and 4,500 strikes. If I buy this tomorrow, settlement will be 6/8, so I use 555 for DTE in my yield calculations. I'm thinking of starting at a price of 478.15, for a net price of 478.1772.

Using the boxtrades.com formula (with my day counts), I get a gross yield of 3.005%, and net yield of 3.001%.

Using standard TVM calculations I get 2.978% for net yield.

Interestingly, I get the same 2.978% net yield using YIELD with freq = 1 and dcc = 1. For the shorter maturities I've looked at, YIELD generates a rate lower than TVM calcs.

The highest yield I saw for a Treasury maturing on about the same date was 2.543% for min qty 25.

So at a gross price of 478.15 this is a gross and net yield spread of 46 bps.

Also, best yield for non-callable brokered CD at Fidelity maturing on 12/15/2023 is 2.5%, so about the same as the Treasury.

Kevin
Please let us know how it goes with your 18-months box.
comeinvest
Posts: 2709
Joined: Mon Mar 12, 2012 6:57 pm

Re: Let's Talk SPX Box Spreads

Post by comeinvest »

johnanglemen wrote: Mon Jun 06, 2022 2:19 pm
adamhg wrote: Mon Jun 06, 2022 1:14 pm
comeinvest wrote: Sun Jun 05, 2022 5:17 am Hi adamhg,
I think you might want to consider using days to settlement in your APY calculations instead or days to expiration, because as Kevin correctly said, the yields will be more accurate. If both the trade date and the expiration date are not preceding a weekend or holiday, it doesn't matter, as both are settled T+1. But if the trade date is a Tuesday (for example) and the expiration date is a Friday (for example), then I think the duration of the loan would be 2 days longer than the DTE, right? Because settlement of the trade would be T+1, and settlement of the assignment/exercise at expiration would be on Monday, i.e. T+3 (3 calendar days that are subject to interest).
I can see the argument why you might want to use settlement date, but don't most brokerages let you trade on pre-settled funds? At least TDA does for me. So even though at expiry, it technically doesn't settle for 3 days, the funds are available for trading on T-0 so long as I don't withdraw. So I guess if you're looking for when you get back to cash-cash, T+3 would make sense, but if you're just looking for your liquidity back, T-0 feels better. Am I thinking about this right?
Personally I would prefer settlement dates. That's the date that the broker margin debt gets paid down (or drawn upon) and interest stops (or starts) accruing. I also feel that's a "purer" calculation: It represents when you "actually" get (or lose) your cash.

Thank you for the site, it's extremely helpful.

Frankly I think the settlement dates are the only relevant, useful dates for meaningful APY calcs. You can probably trade with non-settled funds if the settlement dates of a buy and sell are the same, but that is irrelevant. What is relevant is the interest that you implicitly pay with the box, that you want to compare to the interest that you would pay using other financing methods like broker interest, all of which are quoted and based on the number of days between the settlement dates of the cash flows.
User avatar
indexfundfan
Posts: 3962
Joined: Tue Feb 20, 2007 10:21 am
Contact:

Re: Spreads on IRA tier

Post by indexfundfan »

Kevin M wrote: Mon Jun 06, 2022 3:43 pm
Kevin M wrote: Mon Jun 06, 2022 1:09 pm
Kevin M wrote: Thu Jun 02, 2022 4:51 pm <snip>
Today I was able to place an order in my Fidelity IRA for the same box spread I tried on Friday.

Order for 16 Dec 2022, strikes 4,000 and 4,300, price 296.65. This is the same order I have open in my taxable account, on which I upped the price to 296.65 today. Neither has filled yet.

I probably will increase price in the IRA first, since that's where I'd rather hold fixed income anyway.

Kevin
Order in IRA was filled at 296.75. Yay!

This is a gross yield of 2.082% and net yield of 2.064%. Best Treasury yield for T maturing about the same time was 1.673% when I checked earlier today. This gives a gross spread of 41 bps and net spread of 39 bps.

Last price update was at 3:46 pm ET, and order was filled at 3:52 pm ET, so about 6 minutes to fill.

Same order in taxable account still is open at 296.65.

Kevin
Thank you for sharing your experience. So to trade box spreads in your IRA, you were approved for Tier 1 + "Spreads on IRA"?
My signature has been deleted.
User avatar
Kevin M
Posts: 15787
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Short-term IRA trade

Post by Kevin M »

indexfundfan wrote: Tue Jun 07, 2022 7:36 am Thank you for sharing your experience. So to trade box spreads in your IRA, you were approved for Tier 1 + "Spreads on IRA"?
Apparently. When I look at options tier status, it still says Options Tier 1, but one of the reps I talked to said I was approved for spreads in IRA too. He said if I wasn't, I wouldn't have even been able to try and enter a spread trade.

I just entered another order in the IRA. I decided to go short-term this time, so I could actually see what happened at expiration sooner for myself. My understanding is that I just let all the options expire.

I entered an order for 15 Jul 22 at 4,000 - 4,500 at 499.30. This would be gross 1.383% (1.392% IRR) and net 1.329% (1.337% IRR) at 37 DTE. Treasury maturing a few days later yields 0.884% for min qty 50. Entered about 15 minutes ago, and still open. Just increased price to 499.35, and it was filled immediately. So 1.284% gross, which is 40 bps over the Treasury yield I am comparing to, and 1.230% net, so 35 bps over.

Kevin
If I make a calculation error, #Cruncher probably will let me know.
User avatar
Kevin M
Posts: 15787
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Let's Talk SPX Box Spreads

Post by Kevin M »

comeinvest wrote: Tue Jun 07, 2022 4:33 am
johnanglemen wrote: Mon Jun 06, 2022 2:19 pm
adamhg wrote: Mon Jun 06, 2022 1:14 pm
comeinvest wrote: Sun Jun 05, 2022 5:17 am Hi adamhg,
I think you might want to consider using days to settlement in your APY calculations instead or days to expiration, because as Kevin correctly said, the yields will be more accurate. If both the trade date and the expiration date are not preceding a weekend or holiday, it doesn't matter, as both are settled T+1. But if the trade date is a Tuesday (for example) and the expiration date is a Friday (for example), then I think the duration of the loan would be 2 days longer than the DTE, right? Because settlement of the trade would be T+1, and settlement of the assignment/exercise at expiration would be on Monday, i.e. T+3 (3 calendar days that are subject to interest).
I can see the argument why you might want to use settlement date, but don't most brokerages let you trade on pre-settled funds? At least TDA does for me. So even though at expiry, it technically doesn't settle for 3 days, the funds are available for trading on T-0 so long as I don't withdraw. So I guess if you're looking for when you get back to cash-cash, T+3 would make sense, but if you're just looking for your liquidity back, T-0 feels better. Am I thinking about this right?
Personally I would prefer settlement dates. That's the date that the broker margin debt gets paid down (or drawn upon) and interest stops (or starts) accruing. I also feel that's a "purer" calculation: It represents when you "actually" get (or lose) your cash.

Thank you for the site, it's extremely helpful.

Frankly I think the settlement dates are the only relevant, useful dates for meaningful APY calcs. You can probably trade with non-settled funds if the settlement dates of a buy and sell are the same, but that is irrelevant. What is relevant is the interest that you implicitly pay with the box, that you want to compare to the interest that you would pay using other financing methods like broker interest, all of which are quoted and based on the number of days between the settlement dates of the cash flows.
If the purpose is to compare to Treasury yield, then we should use purchase settlement and expiration date, since Treasury yield calcs use purchase settlement and maturity date, regardless of whether or not the proceeds are available at maturity.

I once bought some Treasuries because they had slightly higher yields than their neighbors. I later realized that was because they matured on Saturday or Sunday, so proceeds were available one or two days later than their neighbors.

The existing calc works to compare to Treasuries of less than 1-year maturity (maybe six months) if days in year is 365 (not 365.25). For longer maturities, the coupon payments have to be considered, and that requires a yield calculation.

Remember that bond yield is not APY/IRR. It is two times the semi-annual YTM.

At any rate, it's easy enough to put your own rate calculations in a spreadsheet, which is what I've done. Boxtrades.com is extremely useful as is, and gets the rate close enough.

Kevin
If I make a calculation error, #Cruncher probably will let me know.
comeinvest
Posts: 2709
Joined: Mon Mar 12, 2012 6:57 pm

Re: Let's Talk SPX Box Spreads

Post by comeinvest »

Kevin M wrote: Tue Jun 07, 2022 1:13 pm
comeinvest wrote: Tue Jun 07, 2022 4:33 am
johnanglemen wrote: Mon Jun 06, 2022 2:19 pm
adamhg wrote: Mon Jun 06, 2022 1:14 pm
comeinvest wrote: Sun Jun 05, 2022 5:17 am Hi adamhg,
I think you might want to consider using days to settlement in your APY calculations instead or days to expiration, because as Kevin correctly said, the yields will be more accurate. If both the trade date and the expiration date are not preceding a weekend or holiday, it doesn't matter, as both are settled T+1. But if the trade date is a Tuesday (for example) and the expiration date is a Friday (for example), then I think the duration of the loan would be 2 days longer than the DTE, right? Because settlement of the trade would be T+1, and settlement of the assignment/exercise at expiration would be on Monday, i.e. T+3 (3 calendar days that are subject to interest).
I can see the argument why you might want to use settlement date, but don't most brokerages let you trade on pre-settled funds? At least TDA does for me. So even though at expiry, it technically doesn't settle for 3 days, the funds are available for trading on T-0 so long as I don't withdraw. So I guess if you're looking for when you get back to cash-cash, T+3 would make sense, but if you're just looking for your liquidity back, T-0 feels better. Am I thinking about this right?
Personally I would prefer settlement dates. That's the date that the broker margin debt gets paid down (or drawn upon) and interest stops (or starts) accruing. I also feel that's a "purer" calculation: It represents when you "actually" get (or lose) your cash.

Thank you for the site, it's extremely helpful.

Frankly I think the settlement dates are the only relevant, useful dates for meaningful APY calcs. You can probably trade with non-settled funds if the settlement dates of a buy and sell are the same, but that is irrelevant. What is relevant is the interest that you implicitly pay with the box, that you want to compare to the interest that you would pay using other financing methods like broker interest, all of which are quoted and based on the number of days between the settlement dates of the cash flows.
If the purpose is to compare to Treasury yield, then we should use purchase settlement and expiration date, since Treasury yield calcs use purchase settlement and maturity date, regardless of whether or not the proceeds are available at maturity.

I once bought some Treasuries because they had slightly higher yields than their neighbors. I later realized that was because they matured on Saturday or Sunday, so proceeds were available one or two days later than their neighbors.

The existing calc works to compare to Treasuries of less than 1-year maturity (maybe six months) if days in year is 365 (not 365.25). For longer maturities, the coupon payments have to be considered, and that requires a yield calculation.

Remember that bond yield is not APY/IRR. It is two times the semi-annual YTM.

At any rate, it's easy enough to put your own rate calculations in a spreadsheet, which is what I've done. Boxtrades.com is extremely useful as is, and gets the rate close enough.

Kevin
Then I would say the treasury yield calc is off, but what do I know. We probably compare our box yields to the treasury yields from the treasury yield curve, which I assume is mostly based on treasuries that don't expire on a Friday. So we should use the settlement dates if our box expires on a Friday. But as you say, the differences will be very small.
Last edited by comeinvest on Tue Jun 07, 2022 6:08 pm, edited 1 time in total.
User avatar
Kevin M
Posts: 15787
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Let's Talk SPX Box Spreads

Post by Kevin M »

comeinvest wrote: Tue Jun 07, 2022 6:07 pm Then I would say the treasury yield calc is off, but what do I know. We probably compare our box yields to the treasury yields from the treasury yield curve, which I assume is mostly based on treasuries that don't expire on a Friday. So we should use the settlement dates if our box expires on a Friday. But as you say, the differences will be very small.
These are yield calculations used by all brokers I know of (Vanguard, Fidelity, Schwab), and the Treasury department, so I don't think we can fight an existing standard.

I don't use the Treasury CMT yield curve, which generates model yields based on Treasury bid quotes. I use actual Treasury ask quotes from Fidelity. If I find one that matures on the same Friday as the option expiration, I'll use that yield, unless there is a higher yield with slightly different maturity.

Kevin
If I make a calculation error, #Cruncher probably will let me know.
User avatar
Kevin M
Posts: 15787
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Let's Talk SPX Box Spreads

Post by Kevin M »

Had another box fill, this time in my taxable account.

Box:

19 Aug 22
DTE 72
4,000 - 4,500
Filled at 498.5
Gross yield 1.525%
Net yield 1.498%
TEY net 1.589%

Treasury:

Yield 1.069%
TEY 1.214%

Yield spread over Treasury:

Gross: 46 bps
Net: 43 bps
TEY: 0.38 bps

I originally entered order at 498.45 in my IRA--meant to enter it in taxable. It did not fill by the time I had entered the same order in taxable. So, I cancelled the order in the IRA, then bumped up the price in taxable to 498.50 shortly before market close, and it was filled immediately.

So, I'm basically creating a box spread ladder, with original rungs so far at:

1.22 months
2.37 m
3.52 m
6.31 m

I started out using the quarter end expirations (Sep, Dec), because I read in one of these threads that they seemed the get filled better, but I had no problems getting the Jul and Aug expirations filled.

Kevin
If I make a calculation error, #Cruncher probably will let me know.
comeinvest
Posts: 2709
Joined: Mon Mar 12, 2012 6:57 pm

Re: Short-term IRA trade

Post by comeinvest »

Kevin M wrote: Tue Jun 07, 2022 12:45 pm
indexfundfan wrote: Tue Jun 07, 2022 7:36 am Thank you for sharing your experience. So to trade box spreads in your IRA, you were approved for Tier 1 + "Spreads on IRA"?
Apparently. When I look at options tier status, it still says Options Tier 1, but one of the reps I talked to said I was approved for spreads in IRA too. He said if I wasn't, I wouldn't have even been able to try and enter a spread trade.

I just entered another order in the IRA. I decided to go short-term this time, so I could actually see what happened at expiration sooner for myself. My understanding is that I just let all the options expire.

I entered an order for 15 Jul 22 at 4,000 - 4,500 at 499.30. This would be gross 1.383% (1.392% IRR) and net 1.329% (1.337% IRR) at 37 DTE. Treasury maturing a few days later yields 0.884% for min qty 50. Entered about 15 minutes ago, and still open. Just increased price to 499.35, and it was filled immediately. So 1.284% gross, which is 40 bps over the Treasury yield I am comparing to, and 1.230% net, so 35 bps over.

Kevin
Congrats on your trade. I am looking at the trades on boxtrades.com since 05/31. Fed funds and SOFR rates barely changed since then. The range of 07/15/2022 box trades was 1.11% to 1.38%. I think you got a good deal, only slightly above the midpoint. Of course we don't know what the spread between a box purchase and sale at the same time would have been.
comeinvest
Posts: 2709
Joined: Mon Mar 12, 2012 6:57 pm

Re: Let's Talk SPX Box Spreads

Post by comeinvest »

Kevin M wrote: Tue Jun 07, 2022 7:19 pm Had another box fill, this time in my taxable account.

Box:

19 Aug 22
DTE 72
4,000 - 4,500
Filled at 498.5
Gross yield 1.525%
Net yield 1.498%
TEY net 1.589%

Treasury:

Yield 1.069%
TEY 1.214%

Yield spread over Treasury:

Gross: 46 bps
Net: 43 bps
TEY: 0.38 bps

I originally entered order at 498.45 in my IRA--meant to enter it in taxable. It did not fill by the time I had entered the same order in taxable. So, I cancelled the order in the IRA, then bumped up the price in taxable to 498.50 shortly before market close, and it was filled immediately.

So, I'm basically creating a box spread ladder, with original rungs so far at:

1.22 months
2.37 m
3.52 m
6.31 m

I started out using the quarter end expirations (Sep, Dec), because I read in one of these threads that they seemed the get filled better, but I had no problems getting the Jul and Aug expirations filled.

Kevin
This and your other trades suggest that you might not get better fills by waiting. We still don't have conclusive evidence if waiting with more aggressive limit orders yield statistically better results in the long run.
I usually use the calendar quarter end SPXW options (e.g. June 30 expiration) because I can remember the expiration dates better. I think I get equally good fills as with the SPX options.
What is your strategy with long positions in short term options boxes? Parking cash for better market entry points?
User avatar
Kevin M
Posts: 15787
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Let's Talk SPX Box Spreads

Post by Kevin M »

comeinvest wrote: Tue Jun 07, 2022 8:29 pm I usually use the calendar quarter end SPXW options (e.g. June 30 expiration) because I can remember the expiration dates better. I think I get equally good fills as with the SPX options.
At this point, I'm reliant on boxtrades.com, which only shows selected expiration dates for SPX. However, since I calculate my own yields, I guess it wouldn't be too difficult to use SPXW too. Is there any advantage to that other than you being able to remember the expiration dates better?
comeinvest wrote: Tue Jun 07, 2022 8:29 pm What is your strategy with long positions in short term options boxes? Parking cash for better market entry points?
I've been gradually moving cash into Treasuries and TIPS. If I can beat Treasury yields by 40 bps, I'll use boxes instead of Treasuries (nominal) if liquidity is not a concern.

I have been sticking with maturities up to two years with nominals and three years with TIPS, based on the yield curves.

For me at this point, box spreads are kind of what direct CDs were in previous years, although I got an average of more than 100 bps over Treasury yields with those. I really don't want to mess around with opening new accounts and doing more IRA transfers to buy direct CDs at this point, and the rates have only recently started to be competitive with Treasuries and brokered CDs.

Thanks,

Kevin
If I make a calculation error, #Cruncher probably will let me know.
comeinvest
Posts: 2709
Joined: Mon Mar 12, 2012 6:57 pm

Re: Let's Talk SPX Box Spreads

Post by comeinvest »

Kevin M wrote: Tue Jun 07, 2022 7:05 pm
comeinvest wrote: Tue Jun 07, 2022 6:07 pm Then I would say the treasury yield calc is off, but what do I know. We probably compare our box yields to the treasury yields from the treasury yield curve, which I assume is mostly based on treasuries that don't expire on a Friday. So we should use the settlement dates if our box expires on a Friday. But as you say, the differences will be very small.
These are yield calculations used by all brokers I know of (Vanguard, Fidelity, Schwab), and the Treasury department, so I don't think we can fight an existing standard.

I don't use the Treasury CMT yield curve, which generates model yields based on Treasury bid quotes. I use actual Treasury ask quotes from Fidelity. If I find one that matures on the same Friday as the option expiration, I'll use that yield, unless there is a higher yield with slightly different maturity.

Kevin
What is typically the difference between bid and ask yields?
comeinvest
Posts: 2709
Joined: Mon Mar 12, 2012 6:57 pm

Re: Let's Talk SPX Box Spreads

Post by comeinvest »

Kevin M wrote: Tue Jun 07, 2022 8:47 pm
comeinvest wrote: Tue Jun 07, 2022 8:29 pm I usually use the calendar quarter end SPXW options (e.g. June 30 expiration) because I can remember the expiration dates better. I think I get equally good fills as with the SPX options.
At this point, I'm reliant on boxtrades.com, which only shows selected expiration dates for SPX. However, since I calculate my own yields, I guess it wouldn't be too difficult to use SPXW too. Is there any advantage to that other than you being able to remember the expiration dates better?
No advantage. Just trying to make my life easier since I started options and futures after reading on the bogleheads forum. Trying to align my periodic tasks, also with quarterly rebalancing, for example.
A few times I put in box orders for an SPX and a nearby SPXW. The filled spreads to the corresponding points on the treasury yield curve were similar, or at least small enough so that I couldn't see a pattern with my limited data points.
comeinvest
Posts: 2709
Joined: Mon Mar 12, 2012 6:57 pm

Re: Let's Talk SPX Box Spreads

Post by comeinvest »

Kevin M wrote: Tue Jun 07, 2022 8:47 pm
comeinvest wrote: Tue Jun 07, 2022 8:29 pm I usually use the calendar quarter end SPXW options (e.g. June 30 expiration) because I can remember the expiration dates better. I think I get equally good fills as with the SPX options.
At this point, I'm reliant on boxtrades.com, which only shows selected expiration dates for SPX. However, since I calculate my own yields, I guess it wouldn't be too difficult to use SPXW too. Is there any advantage to that other than you being able to remember the expiration dates better?
comeinvest wrote: Tue Jun 07, 2022 8:29 pm What is your strategy with long positions in short term options boxes? Parking cash for better market entry points?
I've been gradually moving cash into Treasuries and TIPS. If I can beat Treasury yields by 40 bps, I'll use boxes instead of Treasuries (nominal) if liquidity is not a concern.

I have been sticking with maturities up to two years with nominals and three years with TIPS, based on the yield curves.

For me at this point, box spreads are kind of what direct CDs were in previous years, although I got an average of more than 100 bps over Treasury yields with those. I really don't want to mess around with opening new accounts and doing more IRA transfers to buy direct CDs at this point, and the rates have only recently started to be competitive with Treasuries and brokered CDs.

Thanks,

Kevin
Makes sense, but perhaps your box maturities are a bit short to count as part of the "government bond" portion of your portfolio? Just saying because I think maturities up to 12 months or so usually count as "cash" or "money market", if I'm not mistaken. You won't have meaningful duration exposure (risk), and won't earn duration (interest rate) premia. Of course, if you believe in positive real short-term risk-free rates reappearing in the future, it might make sense.
kxl19
Posts: 248
Joined: Wed Mar 13, 2013 12:41 am

Re: Let's Talk SPX Box Spreads

Post by kxl19 »

Can anybody explain why there are some box spreads of 2000-3000 on the July expiration, instead of the typical 4000-5000? These 2000-3000 spreads seem to have lower rates...
User avatar
Kevin M
Posts: 15787
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Let's Talk SPX Box Spreads

Post by Kevin M »

Kevin M wrote: Tue Jun 07, 2022 7:19 pm <snip>

So, I'm basically creating a box spread ladder, with original rungs so far at:

1.22 months
2.37 m
3.52 m
6.31 m
Today I tried to add a 9-month rung in my IRA to my box spread ladder. I got the same message I got when I first tried it in my IRA, which is about the price being too far away from the last trade. I even tried entering the ask price, and got the same message (when attempting to preview the order). It was too late in the session to call Fidelity, so I'll try again earlier tomorrow.

What I was trying to buy:

Expiration: 3/17/2023
Strikes: 4,000 - 4,500
Price: 490.15
Gross yield: 2.610%
Treasury yield: 2.096%
Gross yield spread: 51 bps

I did not expect it to get filled at this price, but couldn't even get the order entered (as I said, even if I tried much higher ask price).

Kevin
If I make a calculation error, #Cruncher probably will let me know.
User avatar
Kevin M
Posts: 15787
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Let's Talk SPX Box Spreads

Post by Kevin M »

Question about boxtrades.com. Looking at the 17 Mar 23, the last trades I see were on 5/16/2022. Does this mean I'm less likely to get this filled at a reasonable price?

I entered a similar order in taxable, and am now trying again to enter the order noted in my previous post in my IRA about price being too far from last trade. I am on the phone with Fidelity trying to get it resolved.

Kevin
If I make a calculation error, #Cruncher probably will let me know.
User avatar
Kevin M
Posts: 15787
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Let's Talk SPX Box Spreads

Post by Kevin M »

After waiting on hold for 40 minutes, the Fidelity order desk "opened a window" for me to place the order in my IRA. Rep said that they were unable to identify what was causing the error message, and if I get it again, I'll have to call again to get the order window opened. He said he made detailed notes, so next time I should just ask them to have the order desk open a window for me, and it will be quicker.

I assume this means I won't be able to replace the order online with a higher price. I will try that tomorrow, as my price is a yield spread of 53 bps over the Treasury I'm comparing to, which is the 3/15/2023 @ 2.103% for min qty 1 (2.104 for min qty 500), and I don't expect it to get filled.

Another question. There are many expiration dates that aren't shown in boxtrades.com. What if I choose one that's close, but not shown? For example, March 31 2023 instead of Mar 17 2023? Obviously I have no visibility into trades, but I can do my own yield calculations. Any issues with doing something like this?

Thanks,

Kevin
If I make a calculation error, #Cruncher probably will let me know.
User avatar
Kevin M
Posts: 15787
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Let's Talk SPX Box Spreads

Post by Kevin M »

I was able to enter an order online for March 31 2023 expiration in the IRA. However, I noticed that the volume for all 4 legs was 0.

Strikes 4,000 - 4,500
Price 489

I also was able to replace my order for the Mar 17 2023 expiration with a higher price.

Kevin
If I make a calculation error, #Cruncher probably will let me know.
User avatar
indexfundfan
Posts: 3962
Joined: Tue Feb 20, 2007 10:21 am
Contact:

Re: Let's Talk SPX Box Spreads

Post by indexfundfan »

Kevin M wrote: Thu Jun 09, 2022 3:07 pm I was able to enter an order online for March 31 2023 expiration in the IRA. However, I noticed that the volume for all 4 legs was 0.

Strikes 4,000 - 4,500
Price 489

I also was able to replace my order for the Mar 17 2023 expiration with a higher price.

Kevin
My experience is that there needs to be open interest in each underlying leg for the box to have a "competitive" fill. Comparing the open interest for 3/17/23 (SPX) vs 3/31/23 (SPXW):

Call 4000 : 2628 vs 1253
Call 4500 : 9875 vs 384
Put 4000 : 5784 vs 44
Put 4500 : 4795 vs 159

I'm not saying the box based on SPXW won't fill, but I think SPX would be relatively easier to fill compared to SPXW.
My signature has been deleted.
User avatar
Kevin M
Posts: 15787
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Let's Talk SPX Box Spreads

Post by Kevin M »

indexfundfan wrote: Thu Jun 09, 2022 3:30 pm
Kevin M wrote: Thu Jun 09, 2022 3:07 pm I was able to enter an order online for March 31 2023 expiration in the IRA. However, I noticed that the volume for all 4 legs was 0.

Strikes 4,000 - 4,500
Price 489

I also was able to replace my order for the Mar 17 2023 expiration with a higher price.

Kevin
My experience is that there needs to be open interest in each underlying leg for the box to have a "competitive" fill. Comparing the open interest for 3/17/23 (SPX) vs 3/31/23 (SPXW):

Call 4000 : 2628 vs 1253
Call 4500 : 9875 vs 384
Put 4000 : 5784 vs 44
Put 4500 : 4795 vs 159

I'm not saying the box based on SPXW won't fill, but I think SPX would be relatively easier to fill compared to SPXW.
Thanks. I can see the open interest you're talking about.

I hadn't even noticed that the index was changed to SPXW, but looking at my order status, I see that it was. I had entered .SPX on the order screen.

I'm just trying to find reasonable workarounds to the Fidelity bug that won't let me enter certain orders. I thought changing the expiration date slightly was one thing to try. But I see what you mean--the open interest is much higher for the quarter-end 3rd Friday expirations.

Kevin
If I make a calculation error, #Cruncher probably will let me know.
comeinvest
Posts: 2709
Joined: Mon Mar 12, 2012 6:57 pm

Re: Let's Talk SPX Box Spreads

Post by comeinvest »

indexfundfan wrote: Thu Jun 09, 2022 3:30 pm
Kevin M wrote: Thu Jun 09, 2022 3:07 pm I was able to enter an order online for March 31 2023 expiration in the IRA. However, I noticed that the volume for all 4 legs was 0.

Strikes 4,000 - 4,500
Price 489

I also was able to replace my order for the Mar 17 2023 expiration with a higher price.

Kevin
My experience is that there needs to be open interest in each underlying leg for the box to have a "competitive" fill. Comparing the open interest for 3/17/23 (SPX) vs 3/31/23 (SPXW):

Call 4000 : 2628 vs 1253
Call 4500 : 9875 vs 384
Put 4000 : 5784 vs 44
Put 4500 : 4795 vs 159

I'm not saying the box based on SPXW won't fill, but I think SPX would be relatively easier to fill compared to SPXW.
In my limited experience, my end-of-month SPXW fills were about the same APY as my SPX fills with nearby expiration (adjusted for treasury yield curve differences). A few times I put in orders for both SPX and SPXW, incrementally adjusting the limit for both. Please let me know what differences you noticed with your fills, if you have any data.
I could be wrong, but I think neither volume nor open interest matters for options. I think that market participants use models that interpolate trades of nearby options over a range of strike prices and expirations. I could imagine that maybe with higher OI you get an occasional fill from someone opening or closing a position with marketable orders if you wait long enough, but I'm not sure how likely that is, and if you wait long, you are also subject to adverse selection. I often didn't get any fill of my limit orders close to the midpoint, even when I waited more than one day. When I did get a fill, it's hard to say if it was filled by a marketable opposing order, or if it was the result of market fluctuation. I think most people in this forum are investors, not traders. The people over at Elitetrader are traders, but they don't care about "boring" things like fill quality of box trades.
Last edited by comeinvest on Fri Jun 10, 2022 1:11 am, edited 2 times in total.
User avatar
Kevin M
Posts: 15787
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Let's Talk SPX Box Spreads

Post by Kevin M »

comeinvest wrote: Thu Jun 09, 2022 6:45 pm In my limited experience, my end-of-month SPXW fills were about the same APY as my SPX fills with nearby expiration (adjusted for treasury yield curve differences). A few times I put in orders for both SPX and SPXW, incrementally adjusting the limit for both. Please let me know what differences you noticed with your fills, if you have any data.
I could be wrong, both I think neither volume nor open interest matters for options. I think that market participants use models that interpolate trades of nearby options over a range of strike prices and expirations. I could imagine that maybe with higher OI you get an occasional fill from someone opening or closing a position with marketable orders if you wait long enough, but I'm not sure how likely that is, and if you wait long, you are also subject to adverse selection. I often didn't get any fill of my limit orders close to the midpoint, even when I waited more than one day. When I did, it's hard to say if it was filled by a marketable opposing order, or if it was the result of market fluctuation. I think most people in this forum are investors, not traders. The people over at Elitetrader are traders, but they don't care about "boring" things like fill quality of box trades.
Thanks for your input. I don't want both filled, so I will alternately walk the price of each up to see which gets filled first. I will increase the March 31 price first with each increase, to help test out SPXW vs SPX.

Kevin
If I make a calculation error, #Cruncher probably will let me know.
comeinvest
Posts: 2709
Joined: Mon Mar 12, 2012 6:57 pm

Re: Let's Talk SPX Box Spreads

Post by comeinvest »

Kevin M wrote: Thu Jun 09, 2022 7:52 pm
comeinvest wrote: Thu Jun 09, 2022 6:45 pm In my limited experience, my end-of-month SPXW fills were about the same APY as my SPX fills with nearby expiration (adjusted for treasury yield curve differences). A few times I put in orders for both SPX and SPXW, incrementally adjusting the limit for both. Please let me know what differences you noticed with your fills, if you have any data.
I could be wrong, both I think neither volume nor open interest matters for options. I think that market participants use models that interpolate trades of nearby options over a range of strike prices and expirations. I could imagine that maybe with higher OI you get an occasional fill from someone opening or closing a position with marketable orders if you wait long enough, but I'm not sure how likely that is, and if you wait long, you are also subject to adverse selection. I often didn't get any fill of my limit orders close to the midpoint, even when I waited more than one day. When I did, it's hard to say if it was filled by a marketable opposing order, or if it was the result of market fluctuation. I think most people in this forum are investors, not traders. The people over at Elitetrader are traders, but they don't care about "boring" things like fill quality of box trades.
Thanks for your input. I don't want both filled, so I will alternately walk the price of each up to see which gets filled first. I will increase the March 31 price first with each increase, to help test out SPXW vs SPX.

Kevin
That's what I used to do. Let us know your result. Don't forget to adjust for the slightly different T-bill rates at the slightly different maturities. We are interested in the spread to T-bills. My guess is that the noise and inaccuracies are bigger than the difference in yield that you will realize.
A more conclusive way to learn whether SPX or SPXW yield a statistically measurable better rate, with just a few dollars in additional commission, is to split your loan amount into two equal parts, one SPX and one SPXW. Then everything else is equal, including the timing. Slowly walk up both, until both fill.
User avatar
indexfundfan
Posts: 3962
Joined: Tue Feb 20, 2007 10:21 am
Contact:

Re: Let's Talk SPX Box Spreads

Post by indexfundfan »

To use an analogy, I think using SPXW is like trading a thinly traded ETF (low volume and high spreads). If you are patient, sometimes you can get real lucky and get better than expected fills since there is less competition.

Personally I have not had a SPXW box filled before, but that doesn't say much either since I have only made a handful of box trades so far.
My signature has been deleted.
User avatar
Kevin M
Posts: 15787
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Let's Talk SPX Box Spreads

Post by Kevin M »

Didn't even get a chance to increase prices. The SPX 3/17/2023 order was filled by the time I logged on this morning.

The thing is that Treasury yields increased today, so what was a 50 bps yield spread yesterday ended up being a 40 bps gross yield spread (39 bps net) today. Settlement being extended to Monday added back a few basis points.

The SPXW yield spread was 47 bps gross, and was not filled before I cancelled it.

Strikes on both were 4,000 - 4,500

SPX

Treasury yield used was 3/15/2023 with yield 2.262% for min qty 25.

Filled at 490.10 for gross yield 2.662%, net 2.654%.

SPXW

Treasury yield used was 3/31/2023 with yield 2.354% for min qty 25.

Price was 489.00 for gross yield 2.822%, net 2.814%.

Kevin
If I make a calculation error, #Cruncher probably will let me know.
User avatar
Kevin M
Posts: 15787
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

SPX Box Spread examples

Post by Kevin M »

I thought it might be useful to walk through what happens to a box spread at different SPX expiration values. I don't find the options charts that explain it very informative. In the table below, I summarize the outcomes for three scenarios: SPX below low strike at expiration, between the strikes at expiration, and above high strike at expiration.

Image

Note that the end result is always equal to the strike spread. How ever much less than this we pay will be our profit.

The formula for the buy rows is like this, using cell H3 as an example:

=MAX(0,H2-$F3)

The formula for the sell rows is like this, using cell H4 as an example:

=MIN(H2-$F4,0)

I trust that the more experienced option traders will correct anything I've got wrong.

Kevin
If I make a calculation error, #Cruncher probably will let me know.
comeinvest
Posts: 2709
Joined: Mon Mar 12, 2012 6:57 pm

Re: Let's Talk SPX Box Spreads

Post by comeinvest »

indexfundfan wrote: Fri Jun 10, 2022 5:49 am To use an analogy, I think using SPXW is like trading a thinly traded ETF (low volume and high spreads). If you are patient, sometimes you can get real lucky and get better than expected fills since there is less competition.

Personally I have not had a SPXW box filled before, but that doesn't say much either since I have only made a handful of box trades so far.
Almost everything you said in this post is incorrect. I don't even know where to start.
comeinvest
Posts: 2709
Joined: Mon Mar 12, 2012 6:57 pm

Re: Let's Talk SPX Box Spreads

Post by comeinvest »

Kevin M wrote: Fri Jun 10, 2022 11:45 am Didn't even get a chance to increase prices. The SPX 3/17/2023 order was filled by the time I logged on this morning.

The thing is that Treasury yields increased today, so what was a 50 bps yield spread yesterday ended up being a 40 bps gross yield spread (39 bps net) today. Settlement being extended to Monday added back a few basis points.

The SPXW yield spread was 47 bps gross, and was not filled before I cancelled it.

Strikes on both were 4,000 - 4,500

SPX

Treasury yield used was 3/15/2023 with yield 2.262% for min qty 25.

Filled at 490.10 for gross yield 2.662%, net 2.654%.

SPXW

Treasury yield used was 3/31/2023 with yield 2.354% for min qty 25.

Price was 489.00 for gross yield 2.822%, net 2.814%.

Kevin
If you *bought* boxes, your SPX limit order was filled @ 40 bps, and you cancelled your SPXW order when it was @ 47, then we didn't get any information which type SPX gets better fills. (I still assume it doesn't matter.)

My 09/30/2022 SPXW box was filled @ 1.83% before commission. I waited about 1 minute before changing each limit by the minimum tick.
3-month T-bills were ca. 1.33% today ASK yield. I don't know the BID yields.
That would have been a 0.5% spread to the T-bill ASK yield. The spread to the BID would be slightly less.
The spread to T-bills seems to be a bit higher than it was last year, but the filled buy and sell box orders seem to differ by less than 0.1% implied APY comparing Kevin's and my fills, if we assume the spread is constant with maturity, and even though we are comparing SPX to SPXW.
User avatar
Kevin M
Posts: 15787
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Let's Talk SPX Box Spreads

Post by Kevin M »

comeinvest wrote: Sat Jun 11, 2022 5:19 am My 09/30/2022 SPXW box was filled @ 1.83% before commission. I waited about 1 minute before changing each limit by the minimum tick.
3-month T-bills were ca. 1.33% today ASK yield. I don't know the BID yields.
That would have been a 0.5% spread to the T-bill ASK yield. The spread to the BID would be slightly less.
The spread to T-bills seems to be a bit higher than it was last year, but the filled buy and sell box orders seem to differ by less than 0.1% implied APY comparing Kevin's and my fills, if we assume the spread is constant with maturity, and even though we are comparing SPX to SPXW.
Thanks for sharing this.

Friday closing yields for Treasuries maturing about 9/30/2022 were higher than that. The bill maturing 9/29 shows a yield of 1.360 for min qty 1,000, but a note maturing 9/30 has a higher yield at 1.421 for min qty 300 and 1.398 for min qty 25 or 1. If there is a higher yielding note I see no reason to compare to the lower yielding bill. I compare to the highest yield I can find for comparable maturity and quantity.

Comparing to 1.42% or 1.40% gives you a 41 or 43 bps yield spread, which seems to be about where they usually get filled.

Kevin
If I make a calculation error, #Cruncher probably will let me know.
gougou
Posts: 1317
Joined: Thu Sep 28, 2017 7:42 pm

Re: Let's Talk SPX Box Spreads

Post by gougou »

If the brokerage you use go under, are your box spreads safe?
The sillier the market’s behavior, the greater the opportunity for the business like investor.
calwatch
Posts: 1447
Joined: Wed Oct 02, 2013 1:48 am

Re: Let's Talk SPX Box Spreads

Post by calwatch »

gougou wrote: Sat Jun 11, 2022 1:33 pm If the brokerage you use go under, are your box spreads safe?
They are guaranteed by the Options Clearing Corporation. They have a page about this here: https://www.optionseducation.org/refere ... -cash.aspx

As far as choosing this over Interactive Brokers, note that these rates are fixed for the time you choose. IBKR's margin rates will go up as the Fed fund rates and other risk free rates continue to increase. You also have to deal with IBKR's interface which may not be for you. I can get good rates for margining using my normal brokerage, Ameritrade, through the use of box spreads.
gougou
Posts: 1317
Joined: Thu Sep 28, 2017 7:42 pm

Re: Let's Talk SPX Box Spreads

Post by gougou »

calwatch wrote: Sat Jun 11, 2022 8:48 pm
gougou wrote: Sat Jun 11, 2022 1:33 pm If the brokerage you use go under, are your box spreads safe?
They are guaranteed by the Options Clearing Corporation. They have a page about this here: https://www.optionseducation.org/refere ... -cash.aspx

As far as choosing this over Interactive Brokers, note that these rates are fixed for the time you choose. IBKR's margin rates will go up as the Fed fund rates and other risk free rates continue to increase. You also have to deal with IBKR's interface which may not be for you. I can get good rates for margining using my normal brokerage, Ameritrade, through the use of box spreads.
I’m trying to understand if there are any risks with long SPX boxes. Why does it yield higher than Treasuries? Is it because there’s some risk that I don’t know about. Or maybe the holder of Treasuries can lend out Treasuries for some extra return?
The sillier the market’s behavior, the greater the opportunity for the business like investor.
comeinvest
Posts: 2709
Joined: Mon Mar 12, 2012 6:57 pm

Re: Let's Talk SPX Box Spreads

Post by comeinvest »

Kevin M wrote: Sat Jun 11, 2022 12:44 pm
comeinvest wrote: Sat Jun 11, 2022 5:19 am My 09/30/2022 SPXW box was filled @ 1.83% before commission. I waited about 1 minute before changing each limit by the minimum tick.
3-month T-bills were ca. 1.33% today ASK yield. I don't know the BID yields.
That would have been a 0.5% spread to the T-bill ASK yield. The spread to the BID would be slightly less.
The spread to T-bills seems to be a bit higher than it was last year, but the filled buy and sell box orders seem to differ by less than 0.1% implied APY comparing Kevin's and my fills, if we assume the spread is constant with maturity, and even though we are comparing SPX to SPXW.
Thanks for sharing this.

Friday closing yields for Treasuries maturing about 9/30/2022 were higher than that. The bill maturing 9/29 shows a yield of 1.360 for min qty 1,000, but a note maturing 9/30 has a higher yield at 1.421 for min qty 300 and 1.398 for min qty 25 or 1. If there is a higher yielding note I see no reason to compare to the lower yielding bill. I compare to the highest yield I can find for comparable maturity and quantity.

Comparing to 1.42% or 1.40% gives you a 41 or 43 bps yield spread, which seems to be about where they usually get filled.

Kevin
I usually look at https://www.wsj.com/market-data/bonds/treasuries for treasury yields. Those are ASK yields. If your are BID yields, then there is a ca 0.1% bid/ask spread for those T-bills.
It is also further evidence that end-of-month SPXW boxes are just as good as SPX boxes - at least for financing i.e. selling boxes. We could still use more data points.
comeinvest
Posts: 2709
Joined: Mon Mar 12, 2012 6:57 pm

Re: Let's Talk SPX Box Spreads

Post by comeinvest »

gougou wrote: Sat Jun 11, 2022 9:46 pm
calwatch wrote: Sat Jun 11, 2022 8:48 pm
gougou wrote: Sat Jun 11, 2022 1:33 pm If the brokerage you use go under, are your box spreads safe?
They are guaranteed by the Options Clearing Corporation. They have a page about this here: https://www.optionseducation.org/refere ... -cash.aspx

As far as choosing this over Interactive Brokers, note that these rates are fixed for the time you choose. IBKR's margin rates will go up as the Fed fund rates and other risk free rates continue to increase. You also have to deal with IBKR's interface which may not be for you. I can get good rates for margining using my normal brokerage, Ameritrade, through the use of box spreads.
I’m trying to understand if there are any risks with long SPX boxes. Why does it yield higher than Treasuries? Is it because there’s some risk that I don’t know about. Or maybe the holder of Treasuries can lend out Treasuries for some extra return?
I am not knowledgable enough to answer your question, but here are some thoughts:
- I think I read somewhere that some institutions (banks?) are subject to regulations on maximum leverage on their balance sheets, that may or may nor reflect the "real" risk. (Laws and regulations rarely make sense rationally, do they?) As a result, there might be "competition" for available financing at rates close to T-bill rates due to a limited pool of counterparties.
I'm not sure if that can explain what appears to be a risk-free arbitrage opportunity for unconstrained investors. (How could one implement the arbitrage? Long the box short the T-bill certainly doesn't work for retail investors due to the technicalities of the cost of shorting; I'm not sure if it would work for institutional investors.)
- Investors prefer liquid over illiquid investments everything else being equal, i.e. investors accept lower yields if they can dump the securities faster during a crash. I'm not sure if you can dump treasuries faster then options, I'm not sure if there is any reason for any market participant or counterparty to dump options boxes even when the underlying index crashes, or perhaps when the money market shows dislocations; and I'm not sure if any of this is a factor impacting options box yields.

As I said, I don't claim these ideas to be conclusive at explaining your question. I'm also not sure if the nature of the underlying (the equity index in this case) plays a role or not. (I think it shouldn't, as the index value is not an input to the valuation of an options box.) But I think boxes of some commodity futures options have lower implied yields than SPX boxes at least since the time when I started monitoring them, which is surprising to me.
User avatar
Kevin M
Posts: 15787
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Let's Talk SPX Box Spreads

Post by Kevin M »

comeinvest wrote: Sun Jun 12, 2022 3:26 am I usually look at https://www.wsj.com/market-data/bonds/treasuries for treasury yields. Those are ASK yields. If your are BID yields, then there is a ca 0.1% bid/ask spread for those T-bills.
I almost always quote ask yields, because I am a buyer at this point, so unless I specify otherwise, all yields I quote are ask.
If I make a calculation error, #Cruncher probably will let me know.
comeinvest
Posts: 2709
Joined: Mon Mar 12, 2012 6:57 pm

Re: Let's Talk SPX Box Spreads

Post by comeinvest »

Kevin M wrote: Sun Jun 12, 2022 12:09 pm
comeinvest wrote: Sun Jun 12, 2022 3:26 am I usually look at https://www.wsj.com/market-data/bonds/treasuries for treasury yields. Those are ASK yields. If your are BID yields, then there is a ca 0.1% bid/ask spread for those T-bills.
I almost always quote ask yields, because I am a buyer at this point, so unless I specify otherwise, all yields I quote are ask.
"Friday closing yields for Treasuries maturing about 9/30/2022 were higher than that. The bill maturing 9/29 shows a yield of 1.360 for min qty 1,000, but a note maturing 9/30 has a higher yield at 1.421 for min qty 300 and 1.398 for min qty 25 or 1. -

Then I don't understand why your ASK quotes are higher than that on WSJ:

Image

Image
User avatar
Kevin M
Posts: 15787
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Let's Talk SPX Box Spreads

Post by Kevin M »

comeinvest wrote: Sun Jun 12, 2022 10:51 pm "Friday closing yields for Treasuries maturing about 9/30/2022 were higher than that. The bill maturing 9/29 shows a yield of 1.360 for min qty 1,000, but a note maturing 9/30 has a higher yield at 1.421 for min qty 300 and 1.398 for min qty 25 or 1. -

Then I don't understand why your ASK quotes are higher than that on WSJ:
<snip>
Don't know what to tell you. They were actual quotes at Fidelity. The market was closed when I looked, so I assume they were from the close on Friday.
If I make a calculation error, #Cruncher probably will let me know.
Post Reply