AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
comeinvest
Posts: 2603
Joined: Mon Mar 12, 2012 6:57 pm

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by comeinvest »

mmohan76 wrote: Wed Jan 31, 2024 4:08 pm Can you really borrow at that rate though? I believe you still need to keep that premium you earn from selling the box as margin at the OCC, no?
Short answer "no". Box spreads have close to zero margin requirement. How do I know? I traded probably more than thousand of them.
comeinvest
Posts: 2603
Joined: Mon Mar 12, 2012 6:57 pm

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by comeinvest »

mmohan76 wrote: Wed Jan 31, 2024 4:12 pm
comeinvest wrote: Wed Jan 31, 2024 3:49 pm
Lawyered_ wrote: Wed Jan 31, 2024 11:07 am
Lawyered_ wrote: Tue Jan 30, 2024 8:16 am Assets under management are growing crazy fast. Up to $916mm. Word is spreading fast on this ETF. The use of it for cash holdings, especially for high earners in high tax states, is very impressive.
This ETF gained another $7mm in AUM over night and is now up to $923mm AUM. My question is what size can this grow to but still allow them deploy this strategy? At some point there needs to be another party on the other side of these Box trades. Is that market for that big enough?
I think the market is quite big. You can head over to the web site boxtrades.com that a friendly bogleheads user created to monitor box trades. You will occasionally see box trades in the $100M range, if I remember right.
The issue would be in large scale redemptions, if that were ever to happen. This is mainly an "upstairs" market. However the chances look relatively low since no holders have concentrated positions.
And at time of expiration / renewal of the options. I can't remember if BOXX uses rolling daily or weekly expiration, or quarterly. I think the quarterly options expirations are the most liquid.
Walkure
Posts: 976
Joined: Tue Apr 11, 2017 9:59 pm

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by Walkure »

mmohan76 wrote: Wed Jan 31, 2024 12:23 pm Thanks, I'm enjoying this discussion :)

So the thesis is that the creations are mostly in cash, while the redemptions might be mostly in-kind. What I didn't get is why buying the box leads to a loss getting recorded in the fund while selling the box by the AP causes a gain?
I may have some of the particulars off about the timing of when exactly the gain or loss is recognized, but broadly, my understanding is that a long box is sold at a discount, and at expiration it is worth the full value, and therefore a capital gain on the whole position. However, in each of the constituent spreads of the box, there will be one leg that is sold for a premium and another leg for which a premium is paid.

This https://optionalpha.com/strategies/long-box-spread describes it thusly:
For example, if a stock is trading at $50, a $45 call is purchased, and a $55 call is sold. Simultaneously, a $55 put is purchased, and a $45 put is sold. Thus, a $10 wide long box spread is created around the stock. If pricing is efficient, the box spread would be bought and sold for $10. However, investors buying a long box spread look to exploit an inefficiency in the bid-ask spread. If the long box spread can be purchased for $9.50, a risk-free profit of $.50 is anticipated, because the spread should sell for $10 at expiration, regardless of where the underlying stock price is trading.
So, if the underlying ends at $56, the $55 call sold expires in the money for a $1 loss. The $45 call expires in the money for an $11 dollar profit, and both puts expire worthless. The call spread goes to the AP for settlement/redemption, and the delta between the premium made from purchasing the $55 put and selling the $45 put is a loss for the Fund.
comeinvest
Posts: 2603
Joined: Mon Mar 12, 2012 6:57 pm

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by comeinvest »

fujiters wrote: Wed Jan 31, 2024 3:55 pm
comeinvest wrote: Wed Jan 31, 2024 3:49 pm
Lawyered_ wrote: Wed Jan 31, 2024 11:07 am
Lawyered_ wrote: Tue Jan 30, 2024 8:16 am Assets under management are growing crazy fast. Up to $916mm. Word is spreading fast on this ETF. The use of it for cash holdings, especially for high earners in high tax states, is very impressive.
This ETF gained another $7mm in AUM over night and is now up to $923mm AUM. My question is what size can this grow to but still allow them deploy this strategy? At some point there needs to be another party on the other side of these Box trades. Is that market for that big enough?
I think the market is quite big. You can head over to the web site boxtrades.com that a friendly bogleheads user created to monitor box trades. You will occasionally see box trades in the $100M range, if I remember right.
I've noticed the rates for 1-4 month box trades has gone down to approximately the rate of treasuries. BOXX may be to blame for the decreasing yield premium. On the flip side, you can borrow money for 1-4 months for about the same rate as the Treasury can now!
Mar 2024 SPX box trades were around 5.55% on average (range ca. 5.45% to 5.65%) for the past week. Mar 14 T-bills are at about 5.5%, although the data from Interactive Brokers is a bit hard for me to read. 1-month Term SOFR, which I think is a bit more stable than T-bills with short expirations, is about 5.33%. The current fed fund ranges is 5.25% to 5.5%.
It looks like the spread to Term SOFR is about 0.22%. In my experience and what is documented in studies, the average spread is about 0.35%, with a range between about 0.2% and 0.5% to T-bills or Term SOFR.
June 2024 SPX boxes trade currently at about 5.55%, slightly increasing from about 5.45% last week. June 13 T-bills are at about 5.3%. Term SOFR (average of 3-month and 6-month) is about 5.25%. The spread of options to T-bills and Term SOFR seems to be close to the historical average.
I didn't check for possible differences in compounding methods in the various rate calculations.

I see SPX box trades of about $100M per day the last few days, and I see multiple single trades of up to $100M during the last week. Most of the bigger trades seem to be floor trades. I don't know how BOXX submits its options orders, electronically or via negotiated floor trades.
On 12/15/2023, the last quarterly expiration, I see about $170M worth of box trades. On Monday 12/18 I see about $500M worth of box trades. I'm not sure how the big guys roll their box spreads; I usually let them expire and open a new position on the quarterly expiration dates. Possibly the big guys could spread the volume of new positions across various expirations when a position expires.
On 01/08 I see a Mar 2024 expiration box trade worth $425M.
The web site does not include SPXW options, which I think have volumes of similar order of magnitude as SPX.
Open interest of Mar 14 2024 SPX options in the range between 4000 and 7000 is up to ca. 800 for the round strike prices divisible by 100. OI is ca. 200 for 4000 and 5000 strike prices. I read somewhere that box trades are only a small fraction of the total options trading volume. But if we assume for the moment that all 200 at strike prices 4000 and 5000 are from box trades, the loan amount would be ca. 200 * $100 * 1000 = $20M. I don't know how that reconciles with the much higher box trade volumes per boxtrades.com, unless the trades don't keep their positions until expiration. Most large box trades use 4000 and 5000 strike prices. My apologies, OI at those strike prices is about 200k, which would equate to a loan amount of about 200k * $100 * 1000 = $20B for those strike prices, if the OI were from 4000-5000 box trades only (which is unlikely). That would be about 20x the current AUM of BOXX.
fujiters
Posts: 543
Joined: Tue Mar 06, 2018 1:17 pm

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by fujiters »

comeinvest wrote: Wed Jan 31, 2024 9:53 pm
fujiters wrote: Wed Jan 31, 2024 3:55 pm
comeinvest wrote: Wed Jan 31, 2024 3:49 pm
Lawyered_ wrote: Wed Jan 31, 2024 11:07 am
Lawyered_ wrote: Tue Jan 30, 2024 8:16 am Assets under management are growing crazy fast. Up to $916mm. Word is spreading fast on this ETF. The use of it for cash holdings, especially for high earners in high tax states, is very impressive.
This ETF gained another $7mm in AUM over night and is now up to $923mm AUM. My question is what size can this grow to but still allow them deploy this strategy? At some point there needs to be another party on the other side of these Box trades. Is that market for that big enough?
I think the market is quite big. You can head over to the web site boxtrades.com that a friendly bogleheads user created to monitor box trades. You will occasionally see box trades in the $100M range, if I remember right.
I've noticed the rates for 1-4 month box trades has gone down to approximately the rate of treasuries. BOXX may be to blame for the decreasing yield premium. On the flip side, you can borrow money for 1-4 months for about the same rate as the Treasury can now!
Mar 2024 SPX box trades were around 5.55% on average (range ca. 5.45% to 5.65%) for the past week. Mar 14 T-bills are at about 5.5%, although the data from Interactive Brokers is a bit hard for me to read. 1-month Term SOFR, which I think is a bit more stable than T-bills with short expirations, is about 5.33%. The current fed fund ranges is 5.25% to 5.5%.
It looks like the spread to Term SOFR is about 0.22%. In my experience and what is documented in studies, the average spread is about 0.35%, with a range between about 0.2% and 0.5% to T-bills or Term SOFR.
June 2024 SPX boxes trade currently at about 5.55%, slightly increasing from about 5.45% last week. June 13 T-bills are at about 5.3%. Term SOFR (average of 3-month and 6-month) is about 5.25%. The spread of options to T-bills and Term SOFR seems to be close to the historical average.
I didn't check for possible differences in compounding methods in the various rate calculations.

I see SPX box trades of about $100M per day the last few days, and I see multiple single trades of up to $100M during the last week. Most of the bigger trades seem to be floor trades. I don't know how BOXX submits its options orders, electronically or via negotiated floor trades.
On 12/15/2023, the last quarterly expiration, I see about $170M worth of box trades. On Monday 12/18 I see about $500M worth of box trades. I'm not sure how the big guys roll their box spreads; I usually let them expire and open a new position on the quarterly expiration dates. Possibly the big guys could spread the volume of new positions across various expirations when a position expires.
On 01/08 I see a Mar 2024 expiration box trade worth $425M.
The web site does not include SPXW options, which I think have volumes of similar order of magnitude as SPX.
Open interest of Mar 14 2024 SPX options in the range between 4000 and 7000 is up to ca. 800 for the round strike prices divisible by 100. OI is ca. 200 for 4000 and 5000 strike prices. I read somewhere that box trades are only a small fraction of the total options trading volume. But if we assume for the moment that all 200 at strike prices 4000 and 5000 are from box trades, the loan amount would be ca. 200 * $100 * 1000 = $20M. I don't know how that reconciles with the much higher box trade volumes per boxtrades.com, unless the trades don't keep their positions until expiration. Most large box trades use 4000 and 5000 strike prices. My apologies, OI at those strike prices is about 200k, which would equate to a loan amount of about 200k * $100 * 1000 = $20B for those strike prices, if the OI were from 4000-5000 box trades only (which is unlikely). That would be about 20x the current AUM of BOXX.
I'm not understanding how you're seeing a 0.2%+ premium.

https://home.treasury.gov/resource-cent ... value=2024 shows me end of day treasury yields. The 1, 2, 3, and 4 month yields on 30 January were:
5.53 5.47 5.42 5.38

https://www.boxtrades.com/ shows trades on 30 January around 5.49 for 20 days, 5.55 for 47 days, 5.52 for 2.5 months (82 days), 5.40 for 3.5 months (110 days).

This looks like a yield premium of less than 0.2%, which is lower than it has been historically (and lower than it still is for longer dated box trades).
“The purpose of the margin of safety is to render the forecast unnecessary.” -Benjamin Graham
comeinvest
Posts: 2603
Joined: Mon Mar 12, 2012 6:57 pm

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by comeinvest »

fujiters wrote: Thu Feb 01, 2024 1:01 am
comeinvest wrote: Wed Jan 31, 2024 9:53 pm
fujiters wrote: Wed Jan 31, 2024 3:55 pm
comeinvest wrote: Wed Jan 31, 2024 3:49 pm
Lawyered_ wrote: Wed Jan 31, 2024 11:07 am

This ETF gained another $7mm in AUM over night and is now up to $923mm AUM. My question is what size can this grow to but still allow them deploy this strategy? At some point there needs to be another party on the other side of these Box trades. Is that market for that big enough?
I think the market is quite big. You can head over to the web site boxtrades.com that a friendly bogleheads user created to monitor box trades. You will occasionally see box trades in the $100M range, if I remember right.
I've noticed the rates for 1-4 month box trades has gone down to approximately the rate of treasuries. BOXX may be to blame for the decreasing yield premium. On the flip side, you can borrow money for 1-4 months for about the same rate as the Treasury can now!
Mar 2024 SPX box trades were around 5.55% on average (range ca. 5.45% to 5.65%) for the past week. Mar 14 T-bills are at about 5.5%, although the data from Interactive Brokers is a bit hard for me to read. 1-month Term SOFR, which I think is a bit more stable than T-bills with short expirations, is about 5.33%. The current fed fund ranges is 5.25% to 5.5%.
It looks like the spread to Term SOFR is about 0.22%. In my experience and what is documented in studies, the average spread is about 0.35%, with a range between about 0.2% and 0.5% to T-bills or Term SOFR.
June 2024 SPX boxes trade currently at about 5.55%, slightly increasing from about 5.45% last week. June 13 T-bills are at about 5.3%. Term SOFR (average of 3-month and 6-month) is about 5.25%. The spread of options to T-bills and Term SOFR seems to be close to the historical average.
I didn't check for possible differences in compounding methods in the various rate calculations.

I see SPX box trades of about $100M per day the last few days, and I see multiple single trades of up to $100M during the last week. Most of the bigger trades seem to be floor trades. I don't know how BOXX submits its options orders, electronically or via negotiated floor trades.
On 12/15/2023, the last quarterly expiration, I see about $170M worth of box trades. On Monday 12/18 I see about $500M worth of box trades. I'm not sure how the big guys roll their box spreads; I usually let them expire and open a new position on the quarterly expiration dates. Possibly the big guys could spread the volume of new positions across various expirations when a position expires.
On 01/08 I see a Mar 2024 expiration box trade worth $425M.
The web site does not include SPXW options, which I think have volumes of similar order of magnitude as SPX.
Open interest of Mar 14 2024 SPX options in the range between 4000 and 7000 is up to ca. 800 for the round strike prices divisible by 100. OI is ca. 200 for 4000 and 5000 strike prices. I read somewhere that box trades are only a small fraction of the total options trading volume. But if we assume for the moment that all 200 at strike prices 4000 and 5000 are from box trades, the loan amount would be ca. 200 * $100 * 1000 = $20M. I don't know how that reconciles with the much higher box trade volumes per boxtrades.com, unless the trades don't keep their positions until expiration. Most large box trades use 4000 and 5000 strike prices. My apologies, OI at those strike prices is about 200k, which would equate to a loan amount of about 200k * $100 * 1000 = $20B for those strike prices, if the OI were from 4000-5000 box trades only (which is unlikely). That would be about 20x the current AUM of BOXX.
I'm not understanding how you're seeing a 0.2%+ premium.

https://home.treasury.gov/resource-cent ... value=2024 shows me end of day treasury yields. The 1, 2, 3, and 4 month yields on 30 January were:
5.53 5.47 5.42 5.38

https://www.boxtrades.com/ shows trades on 30 January around 5.49 for 20 days, 5.55 for 47 days, 5.52 for 2.5 months (82 days), 5.40 for 3.5 months (110 days).

This looks like a yield premium of less than 0.2%, which is lower than it has been historically (and lower than it still is for longer dated box trades).
Re-read my post, it shows how I estimate the spread.
My numbers are consistent with yours. In my experience, short-term T-bills with less than ca. 2 months to maturity can fluctuate a bit more if measured in yield, and/or the spread to SOFR and/or to options implied rates can fluctuate a bit more, to both sides, than longer maturities of for example 3-6 months. This is probably not exploitable because of transaction cost and limits to arbitrage. You can head over to the two box spread threads or the mHFEA thread for more details.
If you look at the T-bills chronologically listed by maturities for example with TWS (Interactive Brokers) or your own broker, you will often see unexplainable fluctuations in yields between "adjacent" bills. Especially between T-bills and notes with almost equal remaining time to maturity; but also between two T-bills with almost the same maturity.
Also verify the compounding method (although probably not very significant for short durations), and whether bid or ask quotes or midpoint values are taken as data source.
Lastly, verify the time of day that each data point (T-bill yield quotes / par yields / Term SOFR) is taken. Morning, mid day, EOD, after-hours, etc.
mmohan76
Posts: 12
Joined: Sun Jan 28, 2024 8:15 pm

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by mmohan76 »

comeinvest wrote: Wed Jan 31, 2024 4:29 pm
mmohan76 wrote: Wed Jan 31, 2024 4:08 pm Can you really borrow at that rate though? I believe you still need to keep that premium you earn from selling the box as margin at the OCC, no?
Short answer "no". Box spreads have close to zero margin requirement. How do I know? I traded probably more than thousand of them.
Would love to understand how you orchestrated that. The CBOE Margin Manual states that a short box position needs to be 100% margined:

Image
mmohan76
Posts: 12
Joined: Sun Jan 28, 2024 8:15 pm

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by mmohan76 »

comeinvest wrote: Wed Jan 31, 2024 4:32 pm
mmohan76 wrote: Wed Jan 31, 2024 4:12 pm
comeinvest wrote: Wed Jan 31, 2024 3:49 pm
Lawyered_ wrote: Wed Jan 31, 2024 11:07 am
Lawyered_ wrote: Tue Jan 30, 2024 8:16 am Assets under management are growing crazy fast. Up to $916mm. Word is spreading fast on this ETF. The use of it for cash holdings, especially for high earners in high tax states, is very impressive.
This ETF gained another $7mm in AUM over night and is now up to $923mm AUM. My question is what size can this grow to but still allow them deploy this strategy? At some point there needs to be another party on the other side of these Box trades. Is that market for that big enough?
I think the market is quite big. You can head over to the web site boxtrades.com that a friendly bogleheads user created to monitor box trades. You will occasionally see box trades in the $100M range, if I remember right.
The issue would be in large scale redemptions, if that were ever to happen. This is mainly an "upstairs" market. However the chances look relatively low since no holders have concentrated positions.
And at time of expiration / renewal of the options. I can't remember if BOXX uses rolling daily or weekly expiration, or quarterly. I think the quarterly options expirations are the most liquid.
Agreed
comeinvest
Posts: 2603
Joined: Mon Mar 12, 2012 6:57 pm

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by comeinvest »

mmohan76 wrote: Thu Feb 01, 2024 6:04 am
comeinvest wrote: Wed Jan 31, 2024 4:29 pm
mmohan76 wrote: Wed Jan 31, 2024 4:08 pm Can you really borrow at that rate though? I believe you still need to keep that premium you earn from selling the box as margin at the OCC, no?
Short answer "no". Box spreads have close to zero margin requirement. How do I know? I traded probably more than thousand of them.
Would love to understand how you orchestrated that. The CBOE Margin Manual states that a short box position needs to be 100% margined:

Image
I'm struggling a bit with the language of CBOE and the rationale behind it; but long story short, boxes are risk-free with respect to movements of the underlying, and their theoretical value changes only with short-term rates, 3-month rates if you have a 3-month box. 3 months is a very short duration, so boxes have very low risk before expiration, and zero risk at expiration.

Perhaps you forgot to highlight the "net proceeds from sale of short options may be applied". I think those proceeds are typically higher than the difference between exercise prices. But I would be glad if someone can break done for me why the risk is described by this language.
The following $400k short box position for example requires $124 initial margin and $112 maintenance margin:

Image
Image
mmohan76
Posts: 12
Joined: Sun Jan 28, 2024 8:15 pm

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by mmohan76 »

comeinvest wrote: Thu Feb 01, 2024 8:14 am
mmohan76 wrote: Thu Feb 01, 2024 6:04 am
comeinvest wrote: Wed Jan 31, 2024 4:29 pm
mmohan76 wrote: Wed Jan 31, 2024 4:08 pm Can you really borrow at that rate though? I believe you still need to keep that premium you earn from selling the box as margin at the OCC, no?
Short answer "no". Box spreads have close to zero margin requirement. How do I know? I traded probably more than thousand of them.
Would love to understand how you orchestrated that. The CBOE Margin Manual states that a short box position needs to be 100% margined:

Image
I'm struggling a bit with the language of CBOE and the rationale behind it; but long story short, boxes are risk-free with respect to movements of the underlying, and their theoretical value changes only with short-term rates, 3-month rates if you have a 3-month box. 3 months is a very short duration, so boxes have very low risk before expiration, and zero risk at expiration.

Perhaps you forgot to highlight the "net proceeds from sale of short options may be applied". I think those proceeds are typically higher than the difference between exercise prices. But I would be glad if someone can break done for me why the risk is described by this language.
The following $400k short box position for example requires $124 initial margin and $112 maintenance margin:

Image
Image
Actually the risk is very clear to me. If I short a box, at maturity I will owe the difference between strikes. It's essentially a loan of that amount. Why would the OCC allow me to take a loan by posting a very small margin against that loan? They are not in the business of taking credit risk, rather, they mitigate credit risk by taking in IM/VM which are usually in EXCESS of what is owed on an MTM basis. Here, because the payoff is known with certainty, they limit the margin to the strike differential rather than asking for something in excess. Makes perfect sense to me.
comeinvest
Posts: 2603
Joined: Mon Mar 12, 2012 6:57 pm

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by comeinvest »

mmohan76 wrote: Thu Feb 01, 2024 8:51 am Actually the risk is very clear to me. If I short a box, at maturity I will owe the difference between strikes. It's essentially a loan of that amount. Why would the OCC allow me to take a loan by posting a very small margin against that loan? They are not in the business of taking credit risk, rather, they mitigate credit risk by taking in IM/VM which are usually in EXCESS of what is owed on an MTM basis. Here, because the payoff is known with certainty, they limit the margin to the strike differential rather than asking for something in excess. Makes perfect sense to me.
What is IM VM?
I cannot make sense of why exactly the short option legs are applied and not netted with the payments for the long option legs. I never thought about it, as the margin requirement is close to zero. But it becomes higher for longer maturity boxes like six months to a year or more.
Impatience
Posts: 720
Joined: Thu Jul 23, 2020 3:15 pm

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by Impatience »

mmohan76 wrote: Thu Feb 01, 2024 6:04 am
comeinvest wrote: Wed Jan 31, 2024 4:29 pm
mmohan76 wrote: Wed Jan 31, 2024 4:08 pm Can you really borrow at that rate though? I believe you still need to keep that premium you earn from selling the box as margin at the OCC, no?
Short answer "no". Box spreads have close to zero margin requirement. How do I know? I traded probably more than thousand of them.
Would love to understand how you orchestrated that. The CBOE Margin Manual states that a short box position needs to be 100% margined:

Image
Go input a box trade order, for example on interactive brokers. It will clearly say the margin requirement is very low, lower than the rate for buying a T-bill even.

One of the main drawbacks of this etf is the higher margin requirement. Maintenance margin on BOXX is 15% vs. -1% for an actual box. Sadly that is typically the low end for any ETFs. Even SHOV requires a whopping 9%.
mmohan76
Posts: 12
Joined: Sun Jan 28, 2024 8:15 pm

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by mmohan76 »

comeinvest wrote: Thu Feb 01, 2024 9:04 am
mmohan76 wrote: Thu Feb 01, 2024 8:51 am Actually the risk is very clear to me. If I short a box, at maturity I will owe the difference between strikes. It's essentially a loan of that amount. Why would the OCC allow me to take a loan by posting a very small margin against that loan? They are not in the business of taking credit risk, rather, they mitigate credit risk by taking in IM/VM which are usually in EXCESS of what is owed on an MTM basis. Here, because the payoff is known with certainty, they limit the margin to the strike differential rather than asking for something in excess. Makes perfect sense to me.
What is IM VM?
I cannot make sense of why exactly the short option legs are applied and not netted with the payments for the long option legs. I never thought about it, as the margin requirement is close to zero. But it becomes higher for longer maturity boxes like six months to a year or more.
IM = initial margin
VM = variation margin

You are getting proceeds by selling options to enter into the short option positions, which you are allowed to use for the 100% margin requirement. The language is simply clarifying that. Of course you will also pay for the long option positions you are entering into. In effect, if you short a box priced at 95 with a 100 strike differential, you get a net of 95 in proceeds from the box sale (net of what you receive by selling the short options and what you pay to buy the long options) - but you have to post 100 as margin. So I don't see how this strategy can be used to "borrow" cash in the traditional sense of the term.
mmohan76
Posts: 12
Joined: Sun Jan 28, 2024 8:15 pm

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by mmohan76 »

Impatience wrote: Thu Feb 01, 2024 9:18 am
mmohan76 wrote: Thu Feb 01, 2024 6:04 am
comeinvest wrote: Wed Jan 31, 2024 4:29 pm
mmohan76 wrote: Wed Jan 31, 2024 4:08 pm Can you really borrow at that rate though? I believe you still need to keep that premium you earn from selling the box as margin at the OCC, no?
Short answer "no". Box spreads have close to zero margin requirement. How do I know? I traded probably more than thousand of them.
Would love to understand how you orchestrated that. The CBOE Margin Manual states that a short box position needs to be 100% margined:

Image
Go input a box trade order, for example on interactive brokers. It will clearly say the margin requirement is very low, lower than the rate for buying a T-bill even.

One of the main drawbacks of this etf is the higher margin requirement. Maintenance margin on BOXX is 15% vs. -1% for an actual box. Sadly that is typically the low end for any ETFs. Even SHOV requires a whopping 9%.
Are you talking about BUYing or SELLing a box spread? You are comparing to BUYing a tbill, so I'm guessing you're looking into the long side. The discussion we were having was for the short side.
comeinvest
Posts: 2603
Joined: Mon Mar 12, 2012 6:57 pm

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by comeinvest »

mmohan76 wrote: Thu Feb 01, 2024 9:23 am Are you talking about BUYing or SELLing a box spread? You are comparing to BUYing a tbill, so I'm guessing you're looking into the long side. The discussion we were having was for the short side.
Margin requirements are near zero for both buy and sell.
But that has nothing to do with the margin requirement of the ETF in a portfolio. Most ETFs have 15%, some 10% at IB. The broker doesn't look what's inside the ETF to determine the risk. If it had 100% cash, it would still be 15% or whatever.
Lawyered_
Posts: 164
Joined: Wed Feb 09, 2022 10:52 am

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by Lawyered_ »

BOXX is now over 1 billion in assets. I think it’s safe to say that it’s here to stay.
jarjarM
Posts: 2489
Joined: Mon Jul 16, 2018 1:21 pm

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by jarjarM »

I think the biggest question is still if there's taxable distribution (none so far) and how will capital gain be reported upon sale (normal CG or 1256 contract style). Hopefully there will be more concrete evidences of how theses are dealt with. I'm buying some now.
comeinvest
Posts: 2603
Joined: Mon Mar 12, 2012 6:57 pm

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by comeinvest »

Why did the fund's NAV underperform its benchmark (1-3 mo. T-bill index) by 0.2% p.a. since inception, when box spreads typically trade with implied yields of ca. 0.3%-0.4% above T-bill yields, and the (currently reduced) expense ratio is 0.1949% ?
Given the above data, I would expect the fund's NAV to outperform T-bills by ca. 0.15% p.a. after expenses.
My estimate of 0.3% to 0.4% spread of box spreads to T-bill is from my own trading experience, and from the two major box spread threads in this forum. The unexplained difference between the realized vs. my theoretically expected return for the fund is a whopping 0.35% p.a. The discrepancy per the fund performance section in the fact sheet seems to be consistent over time.

Image
klaus14
Posts: 949
Joined: Sun Nov 25, 2018 6:43 pm

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by klaus14 »

comeinvest wrote: Mon Feb 12, 2024 2:03 am Why did the fund's NAV underperform its benchmark (1-3 mo. T-bill index) by 0.2% p.a. since inception, when box spreads typically trade with implied yields of ca. 0.3%-0.4% above T-bill yields, and the (currently reduced) expense ratio is 0.1949% ?
Given the above data, I would expect the fund's NAV to outperform T-bills by ca. 0.15% p.a. after expenses.
My estimate of 0.3% to 0.4% spread of box spreads to T-bill is from my own trading experience, and from the two major box spread threads in this forum. The unexplained difference between the realized vs. my theoretically expected return for the fund is a whopping 0.35% p.a. The discrepancy per the fund performance section in the fact sheet seems to be consistent over time.

Image
i see them using booking (bkng) options, seems like a custom deal.
maybe to handle large cash flows they have a deal with some bank based on t-bill rate?
My investment algorithm: https://www.bogleheads.org/forum/viewtopic.php?f=10&t=351899&p=6112869#p6112869
moneyflowin
Posts: 273
Joined: Sat Oct 16, 2021 6:14 am

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by moneyflowin »

comeinvest wrote: Mon Feb 12, 2024 2:03 am Why did the fund's NAV underperform its benchmark (1-3 mo. T-bill index) by 0.2% p.a. since inception, when box spreads typically trade with implied yields of ca. 0.3%-0.4% above T-bill yields, and the (currently reduced) expense ratio is 0.1949% ?
Given the above data, I would expect the fund's NAV to outperform T-bills by ca. 0.15% p.a. after expenses.
My estimate of 0.3% to 0.4% spread of box spreads to T-bill is from my own trading experience, and from the two major box spread threads in this forum. The unexplained difference between the realized vs. my theoretically expected return for the fund is a whopping 0.35% p.a. The discrepancy per the fund performance section in the fact sheet seems to be consistent over time.

Image
Large boxes don't trade 0.4% above T-bills. Looking at the institutional-size trades on boxtrades.com, they yield 0.08~0.12% above T-bills. BOXX's trades would be in 8 or 9 figures. Don't look at $150k, boxes. Those are retail trades.
comeinvest
Posts: 2603
Joined: Mon Mar 12, 2012 6:57 pm

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by comeinvest »

moneyflowin wrote: Mon Feb 12, 2024 6:13 am
comeinvest wrote: Mon Feb 12, 2024 2:03 am Why did the fund's NAV underperform its benchmark (1-3 mo. T-bill index) by 0.2% p.a. since inception, when box spreads typically trade with implied yields of ca. 0.3%-0.4% above T-bill yields, and the (currently reduced) expense ratio is 0.1949% ?
Given the above data, I would expect the fund's NAV to outperform T-bills by ca. 0.15% p.a. after expenses.
My estimate of 0.3% to 0.4% spread of box spreads to T-bill is from my own trading experience, and from the two major box spread threads in this forum. The unexplained difference between the realized vs. my theoretically expected return for the fund is a whopping 0.35% p.a. The discrepancy per the fund performance section in the fact sheet seems to be consistent over time.

Image
Large boxes don't trade 0.4% above T-bills. Looking at the institutional-size trades on boxtrades.com, they yield 0.08~0.12% above T-bills. BOXX's trades would be in 8 or 9 figures. Don't look at $150k, boxes. Those are retail trades.
Large box trades are even more consistent than small ones. I monitored it for years and did probably more than 1000 trades myself. The spread to treasuries is typically between 0.3% and 0.4%. This spread is also consistent with research papers that are referenced in the box spread threads in this forum.
Last Friday 02/09, there was a $300M box trade at 04/21/2024 box trade at $980.3 (5.52%), or 5.61% depending on compounding assumptions; but let's take 5.52%. 4-month par rate was 4.43% https://home.treasury.gov/resource-cent ... nth=202402 . However, Jun 20 expiration T-bills are quoted at ca. 5.3% at Interactive Brokers. Even those current data, possibly a bit worse than average, wouldn't explain the amount of underperformance of BOXX. The spread usually fluctuates a bit, so a spot check doesn't prove anything; the average is ca. 0.35% in my experience and per the research papers. Bid/ask difference may explain ca. 0.005% - 0.1%.
moneyflowin
Posts: 273
Joined: Sat Oct 16, 2021 6:14 am

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by moneyflowin »

comeinvest wrote: Mon Feb 12, 2024 9:57 am
moneyflowin wrote: Mon Feb 12, 2024 6:13 am
comeinvest wrote: Mon Feb 12, 2024 2:03 am Why did the fund's NAV underperform its benchmark (1-3 mo. T-bill index) by 0.2% p.a. since inception, when box spreads typically trade with implied yields of ca. 0.3%-0.4% above T-bill yields, and the (currently reduced) expense ratio is 0.1949% ?
Given the above data, I would expect the fund's NAV to outperform T-bills by ca. 0.15% p.a. after expenses.
My estimate of 0.3% to 0.4% spread of box spreads to T-bill is from my own trading experience, and from the two major box spread threads in this forum. The unexplained difference between the realized vs. my theoretically expected return for the fund is a whopping 0.35% p.a. The discrepancy per the fund performance section in the fact sheet seems to be consistent over time.

Image
Large boxes don't trade 0.4% above T-bills. Looking at the institutional-size trades on boxtrades.com, they yield 0.08~0.12% above T-bills. BOXX's trades would be in 8 or 9 figures. Don't look at $150k, boxes. Those are retail trades.
Large box trades are even more consistent than small ones. I monitored it for years and did probably more than 1000 trades myself. The spread to treasuries is typically between 0.3% and 0.4%. This spread is also consistent with research papers that are referenced in the box spread threads in this forum.
Last Friday 02/09, there was a $300M box trade at 04/21/2024 box trade at $980.3 (5.52%), or 5.61% depending on compounding assumptions; but let's take 5.52%. 4-month par rate was 4.43% https://home.treasury.gov/resource-cent ... nth=202402 . However, Jun 20 expiration T-bills are quoted at ca. 5.3% at Interactive Brokers. Even those current data, possibly a bit worse than average, wouldn't explain the amount of underperformance of BOXX. The spread usually fluctuates a bit, so a spot check doesn't prove anything; the average is ca. 0.35% in my experience and per the research papers. Bid/ask difference may explain ca. 0.005% - 0.1%.
If the $300M box yields 5.52%, and the identical T-bill yields 5.43%, how is that a 0.4% premium?

Why use IB's quote of 5.3%? Instead, look at the Treasury auctions. A June T-bill has an actual yield of 5.36% according to the latest auction. 5.52 is only 0.16% above 5.36%, again nowhere near 0.4% premium. Can you find a single institutional-size trade on boxtrades that yields 0.4% over T-bill? I can't. Nowhere close to it. There are some retail-size trades that yield 0.3-0.4% over, but those are irrelevant.

As for your experience, unless you're doing $100M boxes, I don't see how it relates to BOXX.

BOXX is clearly getting only about 0.1% over T-bills, which explains the underperformance after 0.2% ER
kalarama
Posts: 134
Joined: Mon Oct 03, 2022 4:55 pm

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by kalarama »

seems like BOXX's bigger advantage is to tax defer the income for as long as you want (assuming they continue to not do a distribution) as well as get long term capital gains if held over a year. if both of those turns out to be correct in the long term, then I'm willing to take a small hit vs t-bill rates.
comeinvest
Posts: 2603
Joined: Mon Mar 12, 2012 6:57 pm

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by comeinvest »

moneyflowin wrote: Mon Feb 12, 2024 7:33 pm
comeinvest wrote: Mon Feb 12, 2024 9:57 am
moneyflowin wrote: Mon Feb 12, 2024 6:13 am
comeinvest wrote: Mon Feb 12, 2024 2:03 am Why did the fund's NAV underperform its benchmark (1-3 mo. T-bill index) by 0.2% p.a. since inception, when box spreads typically trade with implied yields of ca. 0.3%-0.4% above T-bill yields, and the (currently reduced) expense ratio is 0.1949% ?
Given the above data, I would expect the fund's NAV to outperform T-bills by ca. 0.15% p.a. after expenses.
My estimate of 0.3% to 0.4% spread of box spreads to T-bill is from my own trading experience, and from the two major box spread threads in this forum. The unexplained difference between the realized vs. my theoretically expected return for the fund is a whopping 0.35% p.a. The discrepancy per the fund performance section in the fact sheet seems to be consistent over time.

Image
Large boxes don't trade 0.4% above T-bills. Looking at the institutional-size trades on boxtrades.com, they yield 0.08~0.12% above T-bills. BOXX's trades would be in 8 or 9 figures. Don't look at $150k, boxes. Those are retail trades.
Large box trades are even more consistent than small ones. I monitored it for years and did probably more than 1000 trades myself. The spread to treasuries is typically between 0.3% and 0.4%. This spread is also consistent with research papers that are referenced in the box spread threads in this forum.
Last Friday 02/09, there was a $300M box trade at 04/21/2024 box trade at $980.3 (5.52%), or 5.61% depending on compounding assumptions; but let's take 5.52%. 4-month par rate was 4.43% https://home.treasury.gov/resource-cent ... nth=202402 . However, Jun 20 expiration T-bills are quoted at ca. 5.3% at Interactive Brokers. Even those current data, possibly a bit worse than average, wouldn't explain the amount of underperformance of BOXX. The spread usually fluctuates a bit, so a spot check doesn't prove anything; the average is ca. 0.35% in my experience and per the research papers. Bid/ask difference may explain ca. 0.005% - 0.1%.
If the $300M box yields 5.52%, and the identical T-bill yields 5.43%, how is that a 0.4% premium?

Why use IB's quote of 5.3%? Instead, look at the Treasury auctions. A June T-bill has an actual yield of 5.36% according to the latest auction. 5.52 is only 0.16% above 5.36%, again nowhere near 0.4% premium. Can you find a single institutional-size trade on boxtrades that yields 0.4% over T-bill? I can't. Nowhere close to it. There are some retail-size trades that yield 0.3-0.4% over, but those are irrelevant.

As for your experience, unless you're doing $100M boxes, I don't see how it relates to BOXX.

BOXX is clearly getting only about 0.1% over T-bills, which explains the underperformance after 0.2% ER
Like I said, if you read carefully what I said
Institutional size trades are not at higher APR on average than small size trades. They are more consistent, less outliers. The average spread is ca. 0.35%. Small trades are sometimes outliers, probably because of bad fills depending on if they were customer buy or customer sell orders. Go the the box trade threads and do some reading if you have not yet done so.
But even the currently observed, possibly worse than average spread, doesn't fully explain underperformance of BOXX.
Kbg
Posts: 349
Joined: Thu Mar 23, 2017 11:33 am

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by Kbg »

kalarama wrote: Mon Feb 12, 2024 8:14 pm seems like BOXX's bigger advantage is to tax defer the income for as long as you want (assuming they continue to not do a distribution) as well as get long term capital gains if held over a year.
This. It's the tax benefit and associated known math we can all do for our personal tax situations.

Lastly, I don't really care what BOXX or our personal box spreads have been achieving. The market is changing due to the fact BOXX is now exploiting the method. Eventually we will arrive (or have arrived) at a new market efficiency that will equalize it all...no free lunches for long.
comeinvest
Posts: 2603
Joined: Mon Mar 12, 2012 6:57 pm

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by comeinvest »

Kbg wrote: Mon Feb 12, 2024 10:05 pm
kalarama wrote: Mon Feb 12, 2024 8:14 pm seems like BOXX's bigger advantage is to tax defer the income for as long as you want (assuming they continue to not do a distribution) as well as get long term capital gains if held over a year.
This. It's the tax benefit and associated known math we can all do for our personal tax situations.

Lastly, I don't really care what BOXX or our personal box spreads have been achieving. The market is changing due to the fact BOXX is now exploiting the method. Eventually we will arrive (or have arrived) at a new market efficiency that will equalize it all...no free lunches for long.
I think the spread of options implied financing cost to treasuries existed for decades, and the market was used by institutions worldwide. I doubt that this ETF will change the market.
Lawyered_
Posts: 164
Joined: Wed Feb 09, 2022 10:52 am

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by Lawyered_ »

comeinvest wrote: Mon Feb 12, 2024 10:51 pm
Kbg wrote: Mon Feb 12, 2024 10:05 pm
kalarama wrote: Mon Feb 12, 2024 8:14 pm seems like BOXX's bigger advantage is to tax defer the income for as long as you want (assuming they continue to not do a distribution) as well as get long term capital gains if held over a year.
This. It's the tax benefit and associated known math we can all do for our personal tax situations.

Lastly, I don't really care what BOXX or our personal box spreads have been achieving. The market is changing due to the fact BOXX is now exploiting the method. Eventually we will arrive (or have arrived) at a new market efficiency that will equalize it all...no free lunches for long.
I think the spread of options implied financing cost to treasuries existed for decades, and the market was used by institutions worldwide. I doubt that this ETF will change the market.
This ETF is going well over a Billion now. What is the size of this market? How do we determine that?
Walkure
Posts: 976
Joined: Tue Apr 11, 2017 9:59 pm

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by Walkure »

Kbg wrote: Mon Feb 12, 2024 10:05 pm
kalarama wrote: Mon Feb 12, 2024 8:14 pm seems like BOXX's bigger advantage is to tax defer the income for as long as you want (assuming they continue to not do a distribution) as well as get long term capital gains if held over a year.
This. It's the tax benefit and associated known math we can all do for our personal tax situations.

Lastly, I don't really care what BOXX or our personal box spreads have been achieving. The market is changing due to the fact BOXX is now exploiting the method. Eventually we will arrive (or have arrived) at a new market efficiency that will equalize it all...no free lunches for long.
Until the Treasury starts funding the national debt by shorting boxes because it's cheaper than auctioning T-bills! :twisted:
Kbg
Posts: 349
Joined: Thu Mar 23, 2017 11:33 am

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by Kbg »

comeinvest wrote: Mon Feb 12, 2024 10:51 pm I think the spread of options implied financing cost to treasuries existed for decades, and the market was used by institutions worldwide. I doubt that this ETF will change the market.
I agree. Sometimes we make things overly complex. Ultimately the key things for BOXX to make sense for an individual are:

1) A tad more risk with now two intermediaries (Options Clearing House and A.Arch) vs. a direct relationship with Uncle Sam

2) Giving a cut of the profits to AA

3) Your tax situation

If #3 is a good trade in swap of #1 and #2, then it makes sense to do. Personally, I think #1/2 are relatively trivial concerns so it really comes down to #3 which is the main marketing pitch AA is using.

One side note...the older I get the more willing I am to take on fees just to reduce hassle and reduce execution errors. A screwed up attempted box trade that was a mistake could be a very expensive mistake. For those who are or have been active traders, if you're honest with yourself you know you've lost some money due to stupid phat phinger moves.
Kaione
Posts: 150
Joined: Sun Aug 07, 2022 10:57 pm

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by Kaione »

Is there a likely chance that the tax benefits advertised here are not true?
Lyrrad
Posts: 835
Joined: Sun Jul 27, 2008 10:59 am

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by Lyrrad »

Kaione wrote: Wed Feb 14, 2024 1:01 am Is there a likely chance that the tax benefits advertised here are not true?
I don't know. The marketing of the fund makes no specific promises about the tax treatment and I believe the annual report specifically notes that an adverse tax decision is a possible risk.

I expect that if the AUM continues to increase, the ETF will be covered more in the press, hopefully with analysis from tax experts on its structure.

I found an interesting thread on Twitter from last week. Two of the messages:

https://twitter.com/Rick_Ferri/status/1 ... 6226536525
Rick Ferri wrote: To confirm, BOXX provides T-bill-like returns from box spread transactions (options) without distributing income or capital gains to investors because it's in an ETF format. Thus, each investor decides when to take capital gains by selling shares.
https://twitter.com/alphaarchitect/stat ... 2401533303
Wes Gray/@alphaarchitect wrote:Hey Rick,
We cannot provide any specific tax advice.
However, here are some facts:
--BOXX is structured to be taxed as a RIC
--In 2023, we distributed $0 in capital gains, income, or dividends.
Kaione
Posts: 150
Joined: Sun Aug 07, 2022 10:57 pm

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by Kaione »

Is "BOXX is structured to be taxed as a RIC" a good thing? Was the guy from Alpha Architect trying to imply he disagreed with Rick Ferri at all?
Dovahkiin
Posts: 243
Joined: Thu Jan 26, 2012 10:36 pm

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by Dovahkiin »

Walkure wrote: Wed Jan 31, 2024 8:15 am
mmohan76 wrote: Wed Jan 31, 2024 7:49 am
Lyrrad wrote: Tue Jan 09, 2024 1:41 pm Looks like the Annual Report (as of October 31, 2023) is now available. Perhaps there's something interesting in there.

https://alphaarchitect.com/wp-content/u ... Report.pdf

I see this:
Federal Income Taxes. The Fund intends to continue to comply with the requirements of subchapter M of the Internal Revenue Code of 1986, as amended, as necessary to qualify as a regulated investment company and distribute substantially all net taxable investment income and net realized gains to shareholders in a manner which results in no tax cost to a Fund. Therefore, no federal income tax provision is required. As of and during the fiscal period ended October 31, 2023, the Fund did not have any tax positions that did not meet the “more-likely-than-not” threshold of being sustained by the applicable tax authority. As of and during the fiscal period ended October 31, 2023, the Fund did not have liabilities for any unrecognized tax benefits. A Fund would/will recognize interest and penalties, if any, related to unrecognized tax benefits on uncertain tax positions as income tax expense in the Statement of Operations. During the fiscal period ended October 31, 2023, the Fund did not incur any interest or penalties. The Fund is subject to examination by U.S. taxing authorities for the tax periods since the Fund’s commencement of operations.
The fund also claimed to have about $50 million in capital losses to carry forward.
Does anyone understand how they accumulated $50mm in capital losses for tax purposes? I've been able to follow most of report and page 8 gives realized gains as around $5mm and unrealized gains as another $5mm, which makes sense. I believe this should've led them to distribute $5mm of realized gains, but instead for tax purposes they are able to show a $50mm capital loss. Is anyone able to connect the dots on that? I lack sufficient tax expertise myself. Thanks!
my guess? Every box has four legs. Two will have a profit, two have a loss. Collectively it's a wash. So, you heartbeat out the gain legs to an authorized participant for cash and then immediately use the cash to close out the losers for a realized loss. Wash, rinse, and repeat to manufacture infinite capital losses with no offsetting realized gains.
On top of that, by technically "breaking the box" into two transactions, they arguably sidestep the "exclusively attributable to the time value of money" issue. It's beautiful :beer

I've been following the BOXX ETF since inception. I decided to wait to invest until I saw they actually really did distribute capital gains.

My understanding with heartbeat trades is it defers capital gains until their AUM goes down. So it delays capital gains until you start having net outflows, then you have the same issue with all kinds of mutual funds - leaving them realize cap gains for the remainders in the funds:

https://businesslawreview.uchicago.edu/ ... r%20shares.

So collectively this investment won't realize capital gains as long as inflows stay strong or you don't have major outflows.

For instance, the ETF TMF had a whopping $6.8605 short-term capital gains distribution december 10, 2020 due to massive outflows from that fund. https://stockanalysis.com/etf/tmf/dividend/

At the time TMF pre-split was $34.80, so 19.5% of the fund got hit with capital gains despite being buy & hold. It was a real shocker for me being invested in HFEA in taxable and tax advantage, it sucked ass. This is probably why the IRS hasn't gone after heartbeat trades yet - there is economic substance that will eventually pass on the same cap gains to the reamining mutual fund bag holders.

https://ycharts.com/companies/TMF/total ... management

You can see their AUM drop shortlly after the covid crash - tons of people took profits or rebalanced HFEA/etc - leaving people still long TMF end of year to take the brunt of realized capital gains. Sadly I spent an hour trying to find inflow/outflows for TMF, so all we can go off is an AUM graph.

So the BOXX ETF is only going to be gravy as long as AUM increases or stays steady. Once it starts decreasing you're going to get hit with a lot of 60/40 long term/short term capital gains at the end of the year.

That's the hidden risk of this ETF - invest wisely, be prepared for massive 20%+ cap gains to hit you at any time. Also per Green - https://greentradertax.com/category/tax ... nt/page/2/
The IRS can’t apply this same look through logic to ETF’s because as a “registered investment company” (RIC), security ETF’s are taxed as securities.
So my takeaway, not a tax lawyer or CPA, so YMMV and may want to consult Green himself, my take away of the BOXX etf is you'll get 100% long-term cap gain treatment if held for 1+ year, and you'll be playing musical chairs of 10-20-30%+ of NAV of realized 60/40 long term/short term capital gains if this ETF ever gets massive outflows.

Even at the risk of possibly 20-30% of cap gains distributed - it sure as hell beats 100% cap gains realized going long on box/etc. Only other risks I can see is it gets TOO much inflows, see municipal bonds.

Box spreads might start becomming muni rates. This might be why they're now tracking treasury rates and each month I watch boxspreads get lower and lower. :twisted:

Doing quick math I wouldn't be surprised if box spreads start trading at 4.27% yields for a 5.5% overnight/3 mo rate. 5.5*(1-.408) = 3.25% after tax yield. 4.27 * (1-.238) = 3.25% after tax yield.

If I do 15/24% brackets: 5.5*(1-.24) = 4.18% after tax yield. 5.48% * (1-.238) = 4.18% after tax yield. So I'm not surprised boxes are trading ~5.48% thanks to this ETF.

So it's clear thanks to this ETF box spread yields have already been pushed down a bit due to the additional AUM, not needing portfolio margin, not needing the options experience to be allowed to sell uncovered calls for portfolio margin, not needing the $$$ taxable acccount for portfolio margin, impossible to get portfolio margin in tax-advantaged accounts w/o self directing it, etc.

Speaking of AUM - we just hit 1 billion in BOXX, which prompted my post tonight. :)

https://ycharts.com/companies/BOXX/tota ... management


TL;DR

[*]You risk 10-30%+ capital gains distributed end of December with this ETF IF it ever has massive outflows of people realizing cap gains.
[*]For a 5.5% overnight/3mo treasury I predict yields on this ETF/Box spreads will go as low as 4.18% if AUM keeps growing a ton/long box spreads get overweighted vs borrowing demand.
kalarama
Posts: 134
Joined: Mon Oct 03, 2022 4:55 pm

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by kalarama »

Dovahkiin wrote: Thu Feb 15, 2024 10:24 pm
Walkure wrote: Wed Jan 31, 2024 8:15 am
mmohan76 wrote: Wed Jan 31, 2024 7:49 am
Lyrrad wrote: Tue Jan 09, 2024 1:41 pm Looks like the Annual Report (as of October 31, 2023) is now available. Perhaps there's something interesting in there.

https://alphaarchitect.com/wp-content/u ... Report.pdf

I see this:
Federal Income Taxes. The Fund intends to continue to comply with the requirements of subchapter M of the Internal Revenue Code of 1986, as amended, as necessary to qualify as a regulated investment company and distribute substantially all net taxable investment income and net realized gains to shareholders in a manner which results in no tax cost to a Fund. Therefore, no federal income tax provision is required. As of and during the fiscal period ended October 31, 2023, the Fund did not have any tax positions that did not meet the “more-likely-than-not” threshold of being sustained by the applicable tax authority. As of and during the fiscal period ended October 31, 2023, the Fund did not have liabilities for any unrecognized tax benefits. A Fund would/will recognize interest and penalties, if any, related to unrecognized tax benefits on uncertain tax positions as income tax expense in the Statement of Operations. During the fiscal period ended October 31, 2023, the Fund did not incur any interest or penalties. The Fund is subject to examination by U.S. taxing authorities for the tax periods since the Fund’s commencement of operations.
The fund also claimed to have about $50 million in capital losses to carry forward.
Does anyone understand how they accumulated $50mm in capital losses for tax purposes? I've been able to follow most of report and page 8 gives realized gains as around $5mm and unrealized gains as another $5mm, which makes sense. I believe this should've led them to distribute $5mm of realized gains, but instead for tax purposes they are able to show a $50mm capital loss. Is anyone able to connect the dots on that? I lack sufficient tax expertise myself. Thanks!
my guess? Every box has four legs. Two will have a profit, two have a loss. Collectively it's a wash. So, you heartbeat out the gain legs to an authorized participant for cash and then immediately use the cash to close out the losers for a realized loss. Wash, rinse, and repeat to manufacture infinite capital losses with no offsetting realized gains.
On top of that, by technically "breaking the box" into two transactions, they arguably sidestep the "exclusively attributable to the time value of money" issue. It's beautiful :beer

I've been following the BOXX ETF since inception. I decided to wait to invest until I saw they actually really did distribute capital gains.

My understanding with heartbeat trades is it defers capital gains until their AUM goes down. So it delays capital gains until you start having net outflows, then you have the same issue with all kinds of mutual funds - leaving them realize cap gains for the remainders in the funds:

https://businesslawreview.uchicago.edu/ ... r%20shares.

So collectively this investment won't realize capital gains as long as inflows stay strong or you don't have major outflows.

For instance, the ETF TMF had a whopping $6.8605 short-term capital gains distribution december 10, 2020 due to massive outflows from that fund. https://stockanalysis.com/etf/tmf/dividend/

At the time TMF pre-split was $34.80, so 19.5% of the fund got hit with capital gains despite being buy & hold. It was a real shocker for me being invested in HFEA in taxable and tax advantage, it sucked ass. This is probably why the IRS hasn't gone after heartbeat trades yet - there is economic substance that will eventually pass on the same cap gains to the reamining mutual fund bag holders.

https://ycharts.com/companies/TMF/total ... management

You can see their AUM drop shortlly after the covid crash - tons of people took profits or rebalanced HFEA/etc - leaving people still long TMF end of year to take the brunt of realized capital gains. Sadly I spent an hour trying to find inflow/outflows for TMF, so all we can go off is an AUM graph.

So the BOXX ETF is only going to be gravy as long as AUM increases or stays steady. Once it starts decreasing you're going to get hit with a lot of 60/40 long term/short term capital gains at the end of the year.

That's the hidden risk of this ETF - invest wisely, be prepared for massive 20%+ cap gains to hit you at any time. Also per Green - https://greentradertax.com/category/tax ... nt/page/2/
The IRS can’t apply this same look through logic to ETF’s because as a “registered investment company” (RIC), security ETF’s are taxed as securities.
So my takeaway, not a tax lawyer or CPA, so YMMV and may want to consult Green himself, my take away of the BOXX etf is you'll get 100% long-term cap gain treatment if held for 1+ year, and you'll be playing musical chairs of 10-20-30%+ of NAV of realized 60/40 long term/short term capital gains if this ETF ever gets massive outflows.

Even at the risk of possibly 20-30% of cap gains distributed - it sure as hell beats 100% cap gains realized going long on box/etc. Only other risks I can see is it gets TOO much inflows, see municipal bonds.

Box spreads might start becomming muni rates. This might be why they're now tracking treasury rates and each month I watch boxspreads get lower and lower. :twisted:

Doing quick math I wouldn't be surprised if box spreads start trading at 4.27% yields for a 5.5% overnight/3 mo rate. 5.5*(1-.408) = 3.25% after tax yield. 4.27 * (1-.238) = 3.25% after tax yield.

If I do 15/24% brackets: 5.5*(1-.24) = 4.18% after tax yield. 5.48% * (1-.238) = 4.18% after tax yield. So I'm not surprised boxes are trading ~5.48% thanks to this ETF.

So it's clear thanks to this ETF box spread yields have already been pushed down a bit due to the additional AUM, not needing portfolio margin, not needing the options experience to be allowed to sell uncovered calls for portfolio margin, not needing the $$$ taxable acccount for portfolio margin, impossible to get portfolio margin in tax-advantaged accounts w/o self directing it, etc.

Speaking of AUM - we just hit 1 billion in BOXX, which prompted my post tonight. :)

https://ycharts.com/companies/BOXX/tota ... management


TL;DR

[*]You risk 10-30%+ capital gains distributed end of December with this ETF IF it ever has massive outflows of people realizing cap gains.
[*]For a 5.5% overnight/3mo treasury I predict yields on this ETF/Box spreads will go as low as 4.18% if AUM keeps growing a ton/long box spreads get overweighted vs borrowing demand.
Now THAT is the most informative post on BOXX I have read across the internet. Thanks!

The folks over at Alpha are pretty helpful via email and are always willing to jump on a call to answer questions. I wouldn't know exactly what questions to ask to confirm things such as this post discussed. Perhaps someone wants to chat with them to understand/confirm the above.
Lyrrad
Posts: 835
Joined: Sun Jul 27, 2008 10:59 am

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by Lyrrad »

Dovahkiin wrote: Thu Feb 15, 2024 10:24 pm
I've been following the BOXX ETF since inception. I decided to wait to invest until I saw they actually really did distribute capital gains.

My understanding with heartbeat trades is it defers capital gains until their AUM goes down. So it delays capital gains until you start having net outflows, then you have the same issue with all kinds of mutual funds - leaving them realize cap gains for the remainders in the funds:

https://businesslawreview.uchicago.edu/ ... r%20shares.

So collectively this investment won't realize capital gains as long as inflows stay strong or you don't have major outflows.
[...]

So the BOXX ETF is only going to be gravy as long as AUM increases or stays steady. Once it starts decreasing you're going to get hit with a lot of 60/40 long term/short term capital gains at the end of the year.

That's the hidden risk of this ETF - invest wisely, be prepared for massive 20%+ cap gains to hit you at any time.
[...]
I disagree.

The annual report states that they had accumulated capital losses of about $50 million as of October 31, 2023 (about 9% of AUM at the time), so it doesn't seem to be at an immediate risk of distributing gains.

There was a post in this thread in May that suggested how a BOXX might be able to avoid accumulating gains.

This post from a couple of weeks ago described how BOXX might have been able to accumulate the losses reported by only redeeming the two box spread legs that gained value.
Dovahkiin
Posts: 243
Joined: Thu Jan 26, 2012 10:36 pm

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by Dovahkiin »

Lyrrad wrote: Thu Feb 15, 2024 11:48 pm
Dovahkiin wrote: Thu Feb 15, 2024 10:24 pm
I've been following the BOXX ETF since inception. I decided to wait to invest until I saw they actually really did distribute capital gains.

My understanding with heartbeat trades is it defers capital gains until their AUM goes down. So it delays capital gains until you start having net outflows, then you have the same issue with all kinds of mutual funds - leaving them realize cap gains for the remainders in the funds:

https://businesslawreview.uchicago.edu/ ... r%20shares.

So collectively this investment won't realize capital gains as long as inflows stay strong or you don't have major outflows.
[...]

So the BOXX ETF is only going to be gravy as long as AUM increases or stays steady. Once it starts decreasing you're going to get hit with a lot of 60/40 long term/short term capital gains at the end of the year.

That's the hidden risk of this ETF - invest wisely, be prepared for massive 20%+ cap gains to hit you at any time.
[...]
I disagree.

The annual report states that they had accumulated capital losses of about $50 million as of October 31, 2023 (about 9% of AUM at the time), so it doesn't seem to be at an immediate risk of distributing gains.

There was a post in this thread in May that suggested how a BOXX might be able to avoid accumulating gains.

This post from a couple of weeks ago described how BOXX might have been able to accumulate the losses reported by only redeeming the two box spread legs that gained value.
Interesting. I'm not sure how they're accumulating losses. I'd like to elaborate my point too, sure, right now its not at risk of accumulating gains because its having massive inflows.

What happens if some billionaire is using it as his cash fund, sees a good deal, sells, and his stake represents 50% of the AUM? That's the example I'm warning others about. On the other hand - I'd gladly would take a 30% cap gains distribution in a year on a strategy that would distribute 100% capital gains every year if executed by myself.

Only idea I have about them accumulating losses is there are other funds out there in the industry known as "swap" funds. Over a long period of 7 years they do things like "swap" a concentrated portfolio of say 100% NVDA stock for a diversified portfolio of many companies that might also have lots of capital losses but the same notional value on the day of the swap.

So with the heartbeat trades they might be running such a fund to others. Active traders and market makers at times can take large temporary capital losses, so they might be willing to absorb capital gains for a fee, which might explain the cost drag difference others are pointing out to the benchmark/etc, and how the fund is accumulating capital losses.

Unfortunately I'm not a CPA, lawyer, or any other legal professional. I'm just an excited investor like everyone else here. I have pointed out a real risk of ETFs given TMF's history that yes, they are not immune to large capital gains distributions. They happen on massive outflows. These ETFs do a ton to prevent it, they try to find "sponsors" that are willing to buy large initial stakes of the ETF to give it AUM, volume, and liquidity. These "sponsors" generally have detailed contracts/etc to maintain a certain $notional of AUM/etc, and over time some sponsors might end up really liking the ETF and invest more on their own accord. So there is a lot the ETFs do behind the scenes to mitigate risk with their sponsors and APs. At the end of the day though no one can force the market to buy anything they don't want.

If I were to run a sportsbook on this scenario - I'd give it 95% odds that they don't distribute any capital gains at all in the life of the fund. It's just too good of a fund to NOT invest short to medium term cash needs. However, its a tail risk I've identified and feel it's my duty to share with everyone here.

If I were to run a sportsbook on the odds the yield drops down to my previous tax-implied yields, I'd give it 50/50 on it happens or doesn't happen. We see the box spread market is so big with tons of insitutional trades. On the other hand - the AUM growth is freaking impressive. Billionaires and the market really likes their tax dodges.

I also built a spreadsheet to identify how much of an advantage this is over 60/40 investing. At 23.8% long term, 40.8% short term, 30.6% blended 60/40 rate, 5.5% interest rate, in one year the after tax values of both strategies the fund is 0.37% higher, with a 1.66% higher notional value. 5 years 2.38%/8.59%. 10 years - 6.06% higher after-tax value/17.91% higher notional value. 15 years - 10.94%/28.03%. 30 years - 32.43%/63.92%

So I'm not sure how many people would hold a 0-3 mo cash allocation for 30 years either. It definitely shows there is tremendous value in their tax deferral strategy though.
Dovahkiin
Posts: 243
Joined: Thu Jan 26, 2012 10:36 pm

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by Dovahkiin »

Walkure wrote: Wed Jan 31, 2024 8:15 am
mmohan76 wrote: Wed Jan 31, 2024 7:49 am
Lyrrad wrote: Tue Jan 09, 2024 1:41 pm Looks like the Annual Report (as of October 31, 2023) is now available. Perhaps there's something interesting in there.

https://alphaarchitect.com/wp-content/u ... Report.pdf

I see this:
Federal Income Taxes. The Fund intends to continue to comply with the requirements of subchapter M of the Internal Revenue Code of 1986, as amended, as necessary to qualify as a regulated investment company and distribute substantially all net taxable investment income and net realized gains to shareholders in a manner which results in no tax cost to a Fund. Therefore, no federal income tax provision is required. As of and during the fiscal period ended October 31, 2023, the Fund did not have any tax positions that did not meet the “more-likely-than-not” threshold of being sustained by the applicable tax authority. As of and during the fiscal period ended October 31, 2023, the Fund did not have liabilities for any unrecognized tax benefits. A Fund would/will recognize interest and penalties, if any, related to unrecognized tax benefits on uncertain tax positions as income tax expense in the Statement of Operations. During the fiscal period ended October 31, 2023, the Fund did not incur any interest or penalties. The Fund is subject to examination by U.S. taxing authorities for the tax periods since the Fund’s commencement of operations.
The fund also claimed to have about $50 million in capital losses to carry forward.
Does anyone understand how they accumulated $50mm in capital losses for tax purposes? I've been able to follow most of report and page 8 gives realized gains as around $5mm and unrealized gains as another $5mm, which makes sense. I believe this should've led them to distribute $5mm of realized gains, but instead for tax purposes they are able to show a $50mm capital loss. Is anyone able to connect the dots on that? I lack sufficient tax expertise myself. Thanks!
my guess? Every box has four legs. Two will have a profit, two have a loss. Collectively it's a wash. So, you heartbeat out the gain legs to an authorized participant for cash and then immediately use the cash to close out the losers for a realized loss. Wash, rinse, and repeat to manufacture infinite capital losses with no offsetting realized gains.
On top of that, by technically "breaking the box" into two transactions, they arguably sidestep the "exclusively attributable to the time value of money" issue. It's beautiful :beer
Sorry but the straddle rules would come into the play the 2nd they put on the other legs. If they're actually doing this strategy then the IRS will want back taxes/distributions as they're making the implied interest rate of 5.5%/pa and so on.

I think the "exclusively attributable to the time value of money" issue is why they're doing BKNG box spreads to be sure to lay any rest to the claims. Box spreads on individual equities cannot ever be "exclusively attributable to the time value of money" due to the early exercise risk and the risks of paying borrow fees depending on which legs get early-exercised. That trade also earns the lending borrow rate of the individual equity, which isn't exclusively attributeable to the time value.

Likewise if they dispose the box before expiration - they can make the same "exclusively attributable to the time value of money" claim - they closed it as interest rates changed, hey it's an interest rate play IRS! Please look the other way.

Then them closing the trade early - through heartbeat trades, is also prob why they're missing the benchmark. The benchmark assumes one holds to expiration. Boxes get really illiquid the last 3 days before exp in my experience trading them on portfolio margin.
comeinvest
Posts: 2603
Joined: Mon Mar 12, 2012 6:57 pm

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by comeinvest »

Dovahkiin wrote: Fri Feb 16, 2024 12:58 pm
Walkure wrote: Wed Jan 31, 2024 8:15 am
mmohan76 wrote: Wed Jan 31, 2024 7:49 am
Lyrrad wrote: Tue Jan 09, 2024 1:41 pm Looks like the Annual Report (as of October 31, 2023) is now available. Perhaps there's something interesting in there.

https://alphaarchitect.com/wp-content/u ... Report.pdf

I see this:
Federal Income Taxes. The Fund intends to continue to comply with the requirements of subchapter M of the Internal Revenue Code of 1986, as amended, as necessary to qualify as a regulated investment company and distribute substantially all net taxable investment income and net realized gains to shareholders in a manner which results in no tax cost to a Fund. Therefore, no federal income tax provision is required. As of and during the fiscal period ended October 31, 2023, the Fund did not have any tax positions that did not meet the “more-likely-than-not” threshold of being sustained by the applicable tax authority. As of and during the fiscal period ended October 31, 2023, the Fund did not have liabilities for any unrecognized tax benefits. A Fund would/will recognize interest and penalties, if any, related to unrecognized tax benefits on uncertain tax positions as income tax expense in the Statement of Operations. During the fiscal period ended October 31, 2023, the Fund did not incur any interest or penalties. The Fund is subject to examination by U.S. taxing authorities for the tax periods since the Fund’s commencement of operations.
The fund also claimed to have about $50 million in capital losses to carry forward.
Does anyone understand how they accumulated $50mm in capital losses for tax purposes? I've been able to follow most of report and page 8 gives realized gains as around $5mm and unrealized gains as another $5mm, which makes sense. I believe this should've led them to distribute $5mm of realized gains, but instead for tax purposes they are able to show a $50mm capital loss. Is anyone able to connect the dots on that? I lack sufficient tax expertise myself. Thanks!
my guess? Every box has four legs. Two will have a profit, two have a loss. Collectively it's a wash. So, you heartbeat out the gain legs to an authorized participant for cash and then immediately use the cash to close out the losers for a realized loss. Wash, rinse, and repeat to manufacture infinite capital losses with no offsetting realized gains.
On top of that, by technically "breaking the box" into two transactions, they arguably sidestep the "exclusively attributable to the time value of money" issue. It's beautiful :beer
Sorry but the straddle rules would come into the play the 2nd they put on the other legs. If they're actually doing this strategy then the IRS will want back taxes/distributions as they're making the implied interest rate of 5.5%/pa and so on.

I think the "exclusively attributable to the time value of money" issue is why they're doing BKNG box spreads to be sure to lay any rest to the claims. Box spreads on individual equities cannot ever be "exclusively attributable to the time value of money" due to the early exercise risk and the risks of paying borrow fees depending on which legs get early-exercised. That trade also earns the lending borrow rate of the individual equity, which isn't exclusively attributeable to the time value.

Likewise if they dispose the box before expiration - they can make the same "exclusively attributable to the time value of money" claim - they closed it as interest rates changed, hey it's an interest rate play IRS! Please look the other way.

Then them closing the trade early - through heartbeat trades, is also prob why they're missing the benchmark. The benchmark assumes one holds to expiration. Boxes get really illiquid the last 3 days before exp in my experience trading them on portfolio margin.
What "exclusively attributable to the time value of money" are you referring to? What tax would it trigger - cap gains or regular?
moneyflowin
Posts: 273
Joined: Sat Oct 16, 2021 6:14 am

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by moneyflowin »

Dovahkiin wrote: Thu Feb 15, 2024 10:24 pm Box spreads might start becomming muni rates. This might be why they're now tracking treasury rates and each month I watch boxspreads get lower and lower. :twisted:

Doing quick math I wouldn't be surprised if box spreads start trading at 4.27% yields for a 5.5% overnight/3 mo rate. 5.5*(1-.408) = 3.25% after tax yield. 4.27 * (1-.238) = 3.25% after tax yield.

If I do 15/24% brackets: 5.5*(1-.24) = 4.18% after tax yield. 5.48% * (1-.238) = 4.18% after tax yield. So I'm not surprised boxes are trading ~5.48% thanks to this ETF.

So it's clear thanks to this ETF box spread yields have already been pushed down a bit due to the additional AUM, not needing portfolio margin, not needing the options experience to be allowed to sell uncovered calls for portfolio margin, not needing the $$$ taxable acccount for portfolio margin, impossible to get portfolio margin in tax-advantaged accounts w/o self directing it, etc.

[*]You risk 10-30%+ capital gains distributed end of December with this ETF IF it ever has massive outflows of people realizing cap gains.
[*]For a 5.5% overnight/3mo treasury I predict yields on this ETF/Box spreads will go as low as 4.18% if AUM keeps growing a ton/long box spreads get overweighted vs borrowing demand.
I don't ever see 3mo box spreads yielding 4.27% when T-bills yield 5.5%. If they did, you could short a box in taxable then buy a T-bill in a Roth IRA for a free lunch. But whenever there's a free lunch, the market prices it out.

Or, using a taxable account, you could sell boxes to refi all your personal debt at a rate of T-bill minus 1.25%-- that's too good to be true for borrowers.

As for BOXX cap gains distributions, if rates go back to 0% and at the same time stocks soar, that could be a catalyst for mass outflows
User avatar
nps
Posts: 1611
Joined: Thu Dec 04, 2014 9:18 am

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by nps »

Bloomberg has an article on BOXX today (paywall)

T-Bills Without Tax Bills? This Fund Says It Cracked the Code
Lyrrad
Posts: 835
Joined: Sun Jul 27, 2008 10:59 am

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by Lyrrad »

nps wrote: Thu Feb 22, 2024 5:46 am Bloomberg has an article on BOXX today (paywall)

T-Bills Without Tax Bills? This Fund Says It Cracked the Code
Interesting.

Article discusses:

The tax-avoidance strategy of the fund: A tax professor called BOXX "the poster child for tax arbitrage". Article suggests that this type of fund could draw attention from policymakers since it uses options to generate losses without losing real money.

Compared BOXX's performance from inception through February 20: 5.07% compared with 4.98% for BIL or 5.14% for a T-Bill Index

Compares tax treatment: No distribution so far, and can get long term capital gains rate if sold after a year. Might be less of a benefit if in a state with high rates or with holding less than a year.

How capital losses are generated: The fund occasionally buys box spreads on Priceline owner Booking Holdings BKNG due to its high share price. These are usually less than 1% of the fund's holdings. It hands off the winning leg to market maker and holds onto losing leg for the loss. For example, one trade in May generated $32 million in losses, more than all its gains last year.

Mentions that Alpha Architect CEO is careful not to promise any specific tax benefits.
comeinvest
Posts: 2603
Joined: Mon Mar 12, 2012 6:57 pm

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by comeinvest »

Lyrrad wrote: Thu Feb 22, 2024 9:48 am
nps wrote: Thu Feb 22, 2024 5:46 am Bloomberg has an article on BOXX today (paywall)

T-Bills Without Tax Bills? This Fund Says It Cracked the Code
Interesting.

Article discusses:

The tax-avoidance strategy of the fund: A tax professor called BOXX "the poster child for tax arbitrage". Article suggests that this type of fund could draw attention from policymakers since it uses options to generate losses without losing real money.

Compared BOXX's performance from inception through February 20: 5.07% compared with 4.98% for BIL or 5.14% for a T-Bill Index

Compares tax treatment: No distribution so far, and can get long term capital gains rate if sold after a year. Might be less of a benefit if in a state with high rates or with holding less than a year.

How capital losses are generated: The fund occasionally buys box spreads on Priceline owner Booking Holdings BKNG due to its high share price. These are usually less than 1% of the fund's holdings. It hands off the winning leg to market maker and holds onto losing leg for the loss. For example, one trade in May generated $32 million in losses, more than all its gains last year.

Mentions that Alpha Architect CEO is careful not to promise any specific tax benefits.
Sorry if this has been discussed before
In the taxpayer aggregate and ignoring tax rate differences, this should not result in a tax loss for the government, right? The market maker eventually has to pay tax on the winning let, after netting with all his other gains and losses, or not? Unless the market maker is in a foreign country; I don't know if it's possible to siphon all derivates gains out of the country, in aggregate.

How would policy makers prevent tax deferral with BOXX. I'm not sure if it should be prevented. We are just talking about deferral, and/or a slight tax rate arbitrage. Thinking of it, short-term rates will probably come down again soon, and someone who (for whatever reason) parks his or her money for the long run in cash with returns near the inflation rate, is arguably already punished as he will not generate tons of wealth and should not present a big tax loss to the government; not to mention it is unfair in the first place to be taxed on phantom gains (before inflation).
Lyrrad
Posts: 835
Joined: Sun Jul 27, 2008 10:59 am

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by Lyrrad »

Today's Money Stuff newsletter discusses the BOXX article. Mostly a detailed explanation of how it is supposed to work to avoid distributions.

Is there anything preventing active ETF (or even mutual funds) from doing something similar to avoid capital gains distributions? Assuming what BOXX does is legal, could any ETF/mutual fund have custom creation/redemption basket to generate capital losses?

I wonder if the IRS will start to look into this, since it is getting more attention.
comeinvest
Posts: 2603
Joined: Mon Mar 12, 2012 6:57 pm

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by comeinvest »

Lyrrad wrote: Thu Feb 22, 2024 2:23 pm Today's Money Stuff newsletter discusses the BOXX article. Mostly a detailed explanation of how it is supposed to work to avoid distributions.

Is there anything preventing active ETF (or even mutual funds) from doing something similar to avoid capital gains distributions? Assuming what BOXX does is legal, could any ETF/mutual fund have custom creation/redemption basket to generate capital losses?

I wonder if the IRS will start to look into this, since it is getting more attention.
What is the tax basis of a redemption basket that is received by the basket recipient? It's similar to an in-kind property transfer to/from a corporation, right? If true, then it's nothing else but the general rule that capital gains are not taxable until realized, correct?
km91
Posts: 1369
Joined: Wed Oct 13, 2021 12:32 pm

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by km91 »

Lyrrad wrote: Thu Feb 22, 2024 2:23 pm Today's Money Stuff newsletter discusses the BOXX article. Mostly a detailed explanation of how it is supposed to work to avoid distributions.

Is there anything preventing active ETF (or even mutual funds) from doing something similar to avoid capital gains distributions? Assuming what BOXX does is legal, could any ETF/mutual fund have custom creation/redemption basket to generate capital losses?

I wonder if the IRS will start to look into this, since it is getting more attention.
What's the significance of the Booking box trade to the ETF strategy? Why can't they do the same tax trades directly on the SPX box positions?
will23
Posts: 140
Joined: Mon Mar 16, 2009 10:32 pm

Re: AlphaArchitect launches BOXX: 1-3 Month Box Spread ETF

Post by will23 »

comeinvest wrote: Thu Feb 22, 2024 12:07 pm
Lyrrad wrote: Thu Feb 22, 2024 9:48 am
nps wrote: Thu Feb 22, 2024 5:46 am Bloomberg has an article on BOXX today (paywall)

T-Bills Without Tax Bills? This Fund Says It Cracked the Code
Interesting.

Article discusses:

The tax-avoidance strategy of the fund: A tax professor called BOXX "the poster child for tax arbitrage". Article suggests that this type of fund could draw attention from policymakers since it uses options to generate losses without losing real money.

Compared BOXX's performance from inception through February 20: 5.07% compared with 4.98% for BIL or 5.14% for a T-Bill Index

Compares tax treatment: No distribution so far, and can get long term capital gains rate if sold after a year. Might be less of a benefit if in a state with high rates or with holding less than a year.

How capital losses are generated: The fund occasionally buys box spreads on Priceline owner Booking Holdings BKNG due to its high share price. These are usually less than 1% of the fund's holdings. It hands off the winning leg to market maker and holds onto losing leg for the loss. For example, one trade in May generated $32 million in losses, more than all its gains last year.

Mentions that Alpha Architect CEO is careful not to promise any specific tax benefits.
Sorry if this has been discussed before
In the taxpayer aggregate and ignoring tax rate differences, this should not result in a tax loss for the government, right? The market maker eventually has to pay tax on the winning let, after netting with all his other gains and losses, or not? Unless the market maker is in a foreign country; I don't know if it's possible to siphon all derivates gains out of the country, in aggregate.

How would policy makers prevent tax deferral with BOXX. I'm not sure if it should be prevented. We are just talking about deferral, and/or a slight tax rate arbitrage. Thinking of it, short-term rates will probably come down again soon, and someone who (for whatever reason) parks his or her money for the long run in cash with returns near the inflation rate, is arguably already punished as he will not generate tons of wealth and should not present a big tax loss to the government; not to mention it is unfair in the first place to be taxed on phantom gains (before inflation).
This ETF converts short terms capital gains to long term capital gain which means the government gets a different rate. So the advantage is tax deferral + a lower rate. An individual with short term taxable gains could easily borrow money with box spreads and purchase this ETF to defer taxes and achieve a lower rate.
Post Reply