HEDGEFUNDIE's excellent adventure Part II: The next journey

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
typical.investor
Posts: 2026
Joined: Mon Jun 11, 2018 3:17 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by typical.investor » Fri May 15, 2020 5:22 am

GiveTendies wrote:
Fri May 15, 2020 4:20 am
I don't know why you're always going on about rebalancing, I have only ever talked about TMF standalone. The TMF ETF's performance itself is clearly completely uncorrelated to how much or little you do rebalancing in your account.
It’s simply absurd to ignore that removing (adding) money to a 3X ETF when the fund does well (poorly) can counteract the effect that the daily reset of leverage has.

I can’t understand why you insist on looking at one component of a portfolio in isolation and concluding it shouldn’t be used.

If all you are going to talk about is TMF standalone, why are you even posting here In the thread at all? That's what I don't understand.

Here let's look at a period that isn't favorable to TMF where 30 year rates only dropped 0.04% (basically flat) I chose the 30 year since TMF has an average maturity of 25 years. The 20 year rose 0.18% in that time.

[July 2016 to Oct 2019]
30 yr Treasuries 2.23% -> 2.19%
20 yr Treasuries 1.82% -> 2.00%


CAGR
TLT (iShares 20+ Year Treasury) 3.06% <--no leverage drag
TMF ( 20+ Yr Trsy 3X ETF) 1.82% <-- terrible leverage sucks
TMF/UPRO 20.92% <-- Oh, my leverage wins.
TLT/UPRO 20.13%

So, we do see volatility drag in TMF over a basically flat period. And obviously TLT would make a better pairing with UPRO if we judge by the returns of a single fund as you like to do! Right! Wrong.

And TMF isn't that far away from rolling TLT with futures
TLT X3 - CASHX /UPRO 21.63%

Be honest. What does your method guesstimate for TMF in July 2016 when it was priced to return 2.23%? You estimate it would lose what - four of five percent?

Here's another basically flat period

Jul 2014 - Nov 2018
30 yr 3.33% -> 3.36%
20 yr 3.27% -> 3.27%

CAGR

TLT (iShares 20+ Year Treasury) 2.96% <--- again isn't this best? Hint: again no.
TMF ( 20+ Yr Trsy 3X ETF) 1.97%

TMF/UPRO 14.80%
TLT/UPRO 13.69%
TLT X3 - CASHX /UPRO 15.61%

I simply don't see TMF losing 7% from volatility which you have repeatedly and emphatically stated is sure to be the case, and not only that but any protest to the contrary means one is claiming to know more than other market participants which obviously can't be the case.

Talking about TMF standalone, which you point out is all you do, is both misleading and uninformative I think. I'm going to keep talking about how TMF behaves in the way that it is typically used. You can keep talking about a way that no-one uses it, but to what point?

Anyway, sure futures is likely to do better than TMF. I don't doubt that. I still see TMF as viable though despite your doomed to failure predictions. Leveraged treasuries is better to offset leveraged equity I think. A look at any equity crash will show you that.

Stagflation will kill this strategy. I am not saying it is without real and scary risk.
Last edited by typical.investor on Fri May 15, 2020 7:25 am, edited 6 times in total.

Gufomel
Posts: 524
Joined: Sat Feb 14, 2015 9:52 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Gufomel » Fri May 15, 2020 5:26 am

langlands wrote:
Thu May 14, 2020 11:08 pm
Gufomel wrote:
Thu May 14, 2020 7:50 pm
What’s the total expense for UPRO?

Expense Ratio: 0.92% - I’m assuming this does not include borrowing costs, correct?

Borrowing costs: I’m assuming there’s a cost to obtain 3x leverage. Is there a way to see what this amount is, or roughly calculate it?

Dividend: It looks like the dividend yield is low compared to typical S&P ETF dividends. If an S&P fund pays a 2% dividend but UPRO only pays a 0.6% dividend (based on what I saw somewhere), is that dividend “lost” to UPRO investors? I guess what I’m asking is when the underlying companies pay out dividends, but that doesn’t get paid out to UPRO investors, is there still a decrease in UPRO price for the dividend payouts by the underlying companies (and thus UPRO investors lose out on that amount)? That may still not be clear, but hopefully it makes sense.

Any other UPRO expenses that you wouldn’t have in a regular S&P fund?

In total, what’s the implied expense for UPRO and is there a relatively easy way to calculate it at any point in time?
Do people even read the last page of a thread before they post? Honestly not trying to be snarky.
Read Uncorrelated's post on this page.

Basic summary: Expense ratio does not include borrowing costs. Borrow cost of UPRO (as of about half year ago) is about 1% over risk free rate (which is almost zero). So basically 2%-2.5% total cost right now.

Yes, you do get the dividends from your leveraged position. Because the position is taken using swaps, it's not paid out directly and is just part of the share appreciation (which is efficient from a tax perspective). The small amount of dividend that is paid out is presumably from the outright share positions, but that is only a fraction of the "S&P 500 dividend yield" that you get.
Thanks for responding. Huge thread that I’ve jumped around on, and didn’t catch that borrowing costs had been mentioned on this page.

So 1% over RFR plus expense ratio of 0.92%. That makes sense. The overnight LIBOR was around 2-2.5% middle of last year, so total expense would have been around 4-4.5%. That’s pretty hefty. I guess for 3x leverage that could make sense, but just seems like that would be a pretty significant drag. I’ve enjoyed what I’ve seen on this thread, but just trying to get a better understanding of the impact of costs.

kmft
Posts: 78
Joined: Tue Jan 28, 2020 1:17 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by kmft » Fri May 15, 2020 8:03 am

ltdshred wrote:
Thu May 14, 2020 10:33 pm
You can estimate using the VIX index x2 from 1991 to inception of TVIX in 2011. Although it won't exactly 100% mirror, it should be a good rough approximation.
Tried this in PV, does not compute.
https://www.portfoliovisualizer.com/bac ... on3_1=-100

kmft
Posts: 78
Joined: Tue Jan 28, 2020 1:17 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by kmft » Fri May 15, 2020 8:14 am

GiveTendies wrote:
Thu May 14, 2020 11:01 pm
kmft wrote:
Wed May 13, 2020 9:27 am
He said in a few posts up from here that he doesn’t have an alternative hedge and thinks smaller leverage in equities will outperform UPRO/TMF in the short term.

Not saying I necessarily agree with that, but you can do this pretty cheaply with PPLC if it’s something you’re interested in.

https://www.direxion.com/products/portf ... sp-500-etf
I would advise against PPLC. The fund is just 20$M, which implies all kinds of problems. If you want to go 135% SPX, i would just do it in a margin account. E.g. buy 100% VOO and then another 35% on margin. Rebalance monthly or in +/- 5% bands and you're done. At IBKR your margin costs would be lower than the 0.5% fee that PPLC charges and you don't need to trust that some fund manager, who is 10% of his time allocated to a small fund such as PPLC is doing his job correctly.
Slow down there high speed, it's causing you to spew misleading information all over this thread.

1) Know your audience. The guy asking for advice has stated numerous times he only has a $6K account. He doesn't qualify for a margin account on IBKR.
2) The PPLC ER is 0.36%. Rounding up to 0.50% helps your point though, so why not, right?
3) PPLC passively tracks the largest underlying index in the world. This isn't an "active" fund where you have a fund manager picking names out of a hat. A robot could do the job, and probably does.
GiveTendies wrote:
Thu May 14, 2020 11:01 pm
Walkure wrote:
Wed May 13, 2020 11:35 am
What was TLT's YTM in May 2009? Seems to me the whole estimate hinges on the short-term predictive value of the YTM to determine its viability as a diversifier, not the long run expected return.
I'm not saying long-term govies aren't a great diversifier for SPX. They are by far the best- you benefit from both "flight to safety" as well as "FED lowering interest rates" during crashes. But good diversification not only relies on correlation but also reasonable expected returns. For the same reason you don't throw GLD into this, as you probably expect 0% long term returns from gold.
The VIIX/TVIX guys would like to have a word with you...

GiveTendies
Posts: 21
Joined: Tue May 12, 2020 8:17 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by GiveTendies » Fri May 15, 2020 9:14 am

kmft wrote:
Fri May 15, 2020 8:14 am
Slow down there high speed, it's causing you to spew misleading information all over this thread.

1) Know your audience. The guy asking for advice has stated numerous times he only has a $6K account. He doesn't qualify for a margin account on IBKR.
2) The PPLC ER is 0.36%. Rounding up to 0.50% helps your point though, so why not, right?
3) PPLC passively tracks the largest underlying index in the world. This isn't an "active" fund where you have a fund manager picking names out of a hat. A robot could do the job, and probably does.
1) This is a thread about a 3x leveraged strategy
2) The ER is 0.51% https://www.direxion.com/wp-content/upl ... -sheet.pdf
3) You're probably right about the automation

kmft
Posts: 78
Joined: Tue Jan 28, 2020 1:17 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by kmft » Fri May 15, 2020 9:18 am

GiveTendies wrote:
Fri May 15, 2020 9:14 am
kmft wrote:
Fri May 15, 2020 8:14 am
Slow down there high speed, it's causing you to spew misleading information all over this thread.

1) Know your audience. The guy asking for advice has stated numerous times he only has a $6K account. He doesn't qualify for a margin account on IBKR.
2) The PPLC ER is 0.36%. Rounding up to 0.50% helps your point though, so why not, right?
3) PPLC passively tracks the largest underlying index in the world. This isn't an "active" fund where you have a fund manager picking names out of a hat. A robot could do the job, and probably does.
1) This is a thread about a 3x leveraged strategy
2) The ER is 0.51% https://www.direxion.com/wp-content/upl ... -sheet.pdf
3) You're probably right about the automation
I was looking at the net expense ratio. Gross may be more important though.

IndexCore
Posts: 153
Joined: Wed Mar 25, 2020 1:44 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by IndexCore » Fri May 15, 2020 9:25 am

typical.investor wrote:
Tue May 12, 2020 10:26 pm
... the math of leveraged ETFs. In theory, if all holders reset their own positions on a daily basis (adding when the fund is down and withdrawing when the fund is up), fund flows themselves will remove the need for the fund to adjust its leverage. I've read papers on it.

In practice of course, that only happens to a limited degree, but it does mitigate the effects of volatility drag.
If you can recall any of the papers by name, source or website, I'd be interested in taking a look. I've been meaning to calculate how to mitigate volatility drag, and a head start would be helpful.

In trying to calculate it myself, I was a bit surprised how much cash is involved:
S&P 500 drops -10%, UPRO drops -30%, requiring +20% cash added to UPRO until recovery (and then removing the 20% cash)
S&P 500 drops -20%, UPRO drops -60%, requiring +40% cash loaned to UPRO until recovery

User avatar
ltdshred
Posts: 63
Joined: Sun Jul 07, 2019 9:41 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by ltdshred » Fri May 15, 2020 10:13 am

Register44 wrote:
Thu May 14, 2020 11:23 pm
ltdshred wrote:
Thu May 14, 2020 10:33 pm
kmft wrote:
Tue May 12, 2020 10:27 am
ltdshred wrote:
Mon May 11, 2020 8:19 pm
orvanik wrote:
Wed May 06, 2020 12:47 pm

You can add an asset with a negative expected return to a portfolio and as long as it is significantly negatively correlated, the total portfolio return can actually end up being higher assuming reallocation. This is what I believe all the VIX talk is about. We don't know whats going to happen with TMF, but if I had to take a bet, I'd guess the VIX is likely to react more significantly to deflationary or inflationary shocks and therefore be better at protecting the portfolio.
Precisely my point. Adding a small allocation <5-10% to this portfolio may sacrifice compounding power but reduces risk in the long run. Sortino ratio and VaR metrics much improved even though you may end up with less money at the end of the run. In fact, in one of the simulations and backtests I did, it only produced 4 negative annual returns in a period of 30 years.
If I wanted lower volatility, I think I'd prefer adopting a 2X leverage adventure, rather than adding a declining asset as insurance to a 3X portfolio.

I am interested in seeing your 30 year backtest on a VIXY/VXX product though. How did you manage to do that, when VIX futures have only been around since 2004?
You can estimate using the VIX index x2 from 1991 to inception of TVIX in 2011. Although it won't exactly 100% mirror, it should be a good rough approximation. Analysis below is inception of VIX in Jan-91 to Apr-20, rebalance quarterly:

Portfolio 1: 50/40/10 UPRO-TMF-TVIX
Portfolio 2: 52.5/42.5/5 UPRO-TMF-TVIX
Portfolio 3: 55/45 UPRO-TMF
Image

In OP's portfolio, he would have increased his sharp ratio, minimized drawdowns, and reduced downside monthly variance in all scenarios. This is what we mean when we say that introducing a negative correlating asset, even though it is long term negative CAGR, can be a good thing for your portfolio. I observe the same thing when I replace UPRO with TQQQ. In all the testing I did, it seemed optimal to hold around 5% leveraged VIX long term, which reduced a little bit of your short bet on volatility. Introducing long VIX to the portfolio allows risk-seeking investors to gather an appropriate amount of volatility, but not totally screw themselves emotionally when things go haywire.

Below I have the returns for a 52.5/42.5/5 UPRO vs 52.5/42.5/5 TQQQ portfolio. This is what I was referring to only 4-5 negative annual returns in a period of 30 years. Incorporating VIX protects your portfolio against losses better than the normal 55-45, even better than the SP500 in some circumstances. Even though you sacrifice long term compounding power, it can improve meeting personal liquidity, liability, or risk tolerance needs in this aggressive growth strategy. People are so adamant to dismiss this idea when some of us are proposing a solution for those who are heavily invested in this strategy looking to reduce some volatility or hedge against black swan events. The VIX hedge definitely served it's purpose this year as noted by the 2020 YTD number in these scenarios.

Code: Select all

Year	 Portfolio 1	Portfolio 2
1991		59.19%	133.46%
1992		8.32%	26.36%
1993		27.34%	39.66%
1994		-14.27%	-17.77%
1995		101.51%	113.28%
1996		27.14%	30.29%
1997		59.70%	46.88%
1998		70.75%	99.69%
1999		7.14%	135.53%
2000		-9.17%	-47.55%
2001		-16.16%	-13.71%
2002		-11.87%	-24.70%
2003		40.78%	80.96%
2004		18.25%	16.91%
2005		5.18%	2.39%
2006		11.98%	5.63%
2007		11.01%	21.48%
2008		-36.94%	-38.25%
2009		-2.91%	21.63%
2010		45.76%	57.68%
2011		68.02%	60.12%
2012		18.27%	26.50%
2013		18.18%	25.21%
2014		54.42%	66.00%
2015		-10.09%	1.39%
2016		8.05%	0.47%
2017		32.60%	52.66%
2018		-6.36%	1.02%
2019		55.53%	71.63%
2020		43.31%	59.22%
I really like this idea of using the vix. Its like the insurance for when the strategy could blow up. I would accept lower returns for that.

Is there a link to the backtest you can share. I would like to test 50 sso/50edv and see how it stacks up. I expect lower returns, lower drawdown, but it may have a higher Sharpe.
I deleted the link last night. You can just replicate TVIX returns by downloading the 200% ^VIX and -100% CASHX rebalanced monthly returns from PV and then reupload to 2011 and then use monthly TVIX for the rest of the returns

User avatar
ltdshred
Posts: 63
Joined: Sun Jul 07, 2019 9:41 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by ltdshred » Fri May 15, 2020 10:16 am

occambogle wrote:
Fri May 15, 2020 12:49 am
ltdshred wrote:
Thu May 14, 2020 10:33 pm

You can estimate using the VIX index x2 from 1991 to inception of TVIX in 2011. Although it won't exactly 100% mirror, it should be a good rough approximation. Analysis below is inception of VIX in Jan-91 to Apr-20, rebalance quarterly:

Portfolio 1: 50/40/10 UPRO-TMF-TVIX
Portfolio 2: 52.5/42.5/5 UPRO-TMF-TVIX
Portfolio 3: 55/45 UPRO-TMF
Image
Thanks for your thoughts. In my (very amateur) testing only using PV, I've never got good results with TVIX, but I am not sure why. It always seems slightly inferior to using double the amount of VIIX, but maybe my ratios are wrong. In theory it should be better, in that it take up less %-age, leaving more for the "productive" assets, but it doesn't seem to work out that way.

MoneyMarathon originally suggested 60/30/10 UPRO/TMF/VIIX as a way of reducing dependency on TMF, something that appeals to me, which takes it further in lowering STDev:

TVIX vs VIIX with UPRO PV Link
ltdshred wrote:
Thu May 14, 2020 10:33 pm
I observe the same thing when I replace UPRO with TQQQ. In all the testing I did, it seemed optimal to hold around 5% leveraged VIX long term, which reduced a little bit of your short bet on volatility. Introducing long VIX to the portfolio allows risk-seeking investors to gather an appropriate amount of volatility, but not totally screw themselves emotionally when things go haywire.

Below I have the returns for a 52.5/42.5/5 UPRO vs 52.5/42.5/5 TQQQ portfolio.
I've only done basic PV testing:

TQQQ & TVIX/VIIX PV link

..which clearly has its limitations on time period. You obviously see different results if you only run it up to Jan of this year, vs April of this year where all the VIIX/TVIX showed it's benefit. You've clearly backtested much longer with your figures going back to 1991. Do you have any STDev/Sharpe figures available for that testing?

If using TQQQ do you feel that 52.5/42.5/5 TQQQ is the optimal combination? Have you considered mixing UPRO/TQQQ to hedge your bets a bit? Thanks for your insights.... I'm tempted by introducing VIIX/TVIX. My only issue is I'm doing this with a small $6k Roth... and that means only 1.6 shares of TVIX, or 3.7 shares of VIIX at 5%. Not exactly easy come rebalancing time (!)....
It depends on your opinion. QQQ's are not all tech like people make it out to be, so take it with a grain of salt. Any allocation is going to have an implicit active view on markets, so UPRO, TQQQ, MIDU, or whatever market proxy you use as your main equity sleeve will have good compounding power. I used UPRO last year and switched to TQQQ this year, and it's all personal risk tolerance and preference.

User avatar
ltdshred
Posts: 63
Joined: Sun Jul 07, 2019 9:41 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by ltdshred » Fri May 15, 2020 10:17 am

kmft wrote:
Fri May 15, 2020 8:03 am
ltdshred wrote:
Thu May 14, 2020 10:33 pm
You can estimate using the VIX index x2 from 1991 to inception of TVIX in 2011. Although it won't exactly 100% mirror, it should be a good rough approximation.
Tried this in PV, does not compute.
https://www.portfoliovisualizer.com/bac ... on3_1=-100
You have to register on the site and go upload a new data series. Take the returns of 200% ^VIX -100% CASHX and combine the returns with TVIX from 2011-Present in a new .csv excel file. Then reupload

kmft
Posts: 78
Joined: Tue Jan 28, 2020 1:17 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by kmft » Fri May 15, 2020 10:22 am

ltdshred wrote:
Fri May 15, 2020 10:17 am
kmft wrote:
Fri May 15, 2020 8:03 am
ltdshred wrote:
Thu May 14, 2020 10:33 pm
You can estimate using the VIX index x2 from 1991 to inception of TVIX in 2011. Although it won't exactly 100% mirror, it should be a good rough approximation.
Tried this in PV, does not compute.
https://www.portfoliovisualizer.com/bac ... on3_1=-100
You have to register on the site and go upload a new data series. Take the returns of 200% ^VIX -100% CASHX and combine the returns with TVIX from 2011-Present in a new .csv excel file. Then reupload
I am registered. My point was to compare 200% ^VIX and TVIX directly from 2011 onward to demonstrate they do not match at all. Why are you confident that 200% ^VIX approximates returns of TVIX from 1991-2010, when they aren't the same by any measure from 2011 onward?

typical.investor
Posts: 2026
Joined: Mon Jun 11, 2018 3:17 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by typical.investor » Fri May 15, 2020 10:30 am

IndexCore wrote:
Fri May 15, 2020 9:25 am

If you can recall any of the papers by name, source or website, I'd be interested in taking a look. I've been meaning to calculate how to mitigate volatility drag, and a head start would be helpful.
This one explains how fund flows themself reduce the volatility drag.
Empirically, we find that capital flows considerably reduce ETF
rebalancing demand and, therefore, mitigate the potential for ETFs to amplify volatility.
We also show theoretically that flows can completely eliminate ETF rebalancing in the limit
https://www.federalreserve.gov/econresd ... 106pap.pdf
IndexCore wrote:
Fri May 15, 2020 9:25 am
In trying to calculate it myself, I was a bit surprised how much cash is involved:
S&P 500 drops -10%, UPRO drops -30%, requiring +20% cash added to UPRO until recovery (and then removing the 20% cash)
S&P 500 drops -20%, UPRO drops -60%, requiring +40% cash loaned to UPRO until recovery
I read another one suggesting to add/remove when the fund drops/goes up, but I don't remember where. I'm not interested in keeping such a cash position to counter UPRO. That is what TMF is for. Sure it's not exact and not everyday, but then again TMF can really spike up when UPRO drops.

Ah here ... proshares themselves say it:
Investors looking to approximate a fund's multiple for longer than one day may need to rebalance their holdings by increasing or decreasing the investment to maintain a desired exposure. Rebalancing may result in transaction costs and tax consequences. Of course, rebalancing can reduce the negative effects of compounding on performance, but it may reduce the positive effects as well.
https://www.proshares.com/media/documen ... unding.pdf

I know I have seen similar suggestion but never a plan about how much....

Walkure
Posts: 220
Joined: Tue Apr 11, 2017 9:59 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Walkure » Fri May 15, 2020 10:41 am

GiveTendies wrote:
Thu May 14, 2020 11:01 pm
Walkure wrote:
Wed May 13, 2020 11:35 am
What was TLT's YTM in May 2009? Seems to me the whole estimate hinges on the short-term predictive value of the YTM to determine its viability as a diversifier, not the long run expected return.
I'm not saying long-term govies aren't a great diversifier for SPX. They are by far the best- you benefit from both "flight to safety" as well as "FED lowering interest rates" during crashes. But good diversification not only relies on correlation but also reasonable expected returns. For the same reason you don't throw GLD into this, as you probably expect 0% long term returns from gold.

My fear is, that some people here expect great returns from TMF, which I think will not happen. The market thinks it will return less than -6%. You can have an active view, that YTM of 20y bonds will be much higher than 1.35% or that realized volatility will be much lower than 18%. But realize that this is an active view, against the common opinion of the fixed income market.

Rad Dad wrote:
Wed May 13, 2020 12:05 pm
GiveTendies: Can I ask where the -6% return came from? I must have missed that.
The following equation to calculate returns of 3x leveraged ETFs while assuming 18.6% volatility and 1.35% return for the underlying. The numbers come from TLT as a proxy (implied vol from options market on TLT and yield to maturity at current price):
Return (Levered ETF) = exp[ -3 * sigma^2] * (1 + r)^3 - 1
Since you brought it up, I plugged in the current gold volatility index (VGZ) 24.22% with a zero expected return for unleveraged bullion into your formula and find that 3x Long Gold has an expected future return around -16%. :shock:
To answer my own question, I couldn't find YTM for TLT from May 2009, but the Fed's 20-yr constant maturity yield from that month is 4.22%, which I expect is a pretty good proxy for what would have been the YTM at the time of TMF's inception. Using this value and the same 18.6% volatility says that in May 2009 one would have assigned TMF an expected future return of 2.04%. In the event, TMF returned 19.14% to the present. So I think the fair way to state it is that TMF now has an expected future return of -6%, with a likely range of ±18%, which is sobering but not exactly actionable information. But definitely avoid gold!

User avatar
ltdshred
Posts: 63
Joined: Sun Jul 07, 2019 9:41 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by ltdshred » Fri May 15, 2020 10:51 am

kmft wrote:
Fri May 15, 2020 10:22 am
ltdshred wrote:
Fri May 15, 2020 10:17 am
kmft wrote:
Fri May 15, 2020 8:03 am
ltdshred wrote:
Thu May 14, 2020 10:33 pm
You can estimate using the VIX index x2 from 1991 to inception of TVIX in 2011. Although it won't exactly 100% mirror, it should be a good rough approximation.
Tried this in PV, does not compute.
https://www.portfoliovisualizer.com/bac ... on3_1=-100
You have to register on the site and go upload a new data series. Take the returns of 200% ^VIX -100% CASHX and combine the returns with TVIX from 2011-Present in a new .csv excel file. Then reupload
I am registered. My point was to compare 200% ^VIX and TVIX directly from 2011 onward to demonstrate they do not match at all. Why are you confident that 200% ^VIX approximates returns of TVIX from 1991-2010, when they aren't the same by any measure from 2011 onward?
You're being pedantic. They're correlated to each other enough to make my point (less fees/borrowing costs/daily rebalance). Notice when I used the word approximation

Image

kmft
Posts: 78
Joined: Tue Jan 28, 2020 1:17 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by kmft » Fri May 15, 2020 11:33 am

ltdshred wrote:
Fri May 15, 2020 10:51 am
kmft wrote:
Fri May 15, 2020 10:22 am
ltdshred wrote:
Fri May 15, 2020 10:17 am
kmft wrote:
Fri May 15, 2020 8:03 am
ltdshred wrote:
Thu May 14, 2020 10:33 pm
You can estimate using the VIX index x2 from 1991 to inception of TVIX in 2011. Although it won't exactly 100% mirror, it should be a good rough approximation.
Tried this in PV, does not compute.
https://www.portfoliovisualizer.com/bac ... on3_1=-100
You have to register on the site and go upload a new data series. Take the returns of 200% ^VIX -100% CASHX and combine the returns with TVIX from 2011-Present in a new .csv excel file. Then reupload
I am registered. My point was to compare 200% ^VIX and TVIX directly from 2011 onward to demonstrate they do not match at all. Why are you confident that 200% ^VIX approximates returns of TVIX from 1991-2010, when they aren't the same by any measure from 2011 onward?
You're being pedantic. They're correlated to each other enough to make my point (less fees/borrowing costs/daily rebalance). Notice when I used the word approximation

Image
Yeah, okay. Just thought you might have done some meaningful work, and I was eager to see it. Color me disappointed.

RomeoMustDie
Posts: 164
Joined: Sat Mar 14, 2020 6:07 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by RomeoMustDie » Fri May 15, 2020 11:44 am

Not looking good for this portfolio in the short term.

Predicting a bad year for the S&P, interest rates also have nowhere to go but up.

However, I will stay the course.

Walkure
Posts: 220
Joined: Tue Apr 11, 2017 9:59 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Walkure » Fri May 15, 2020 11:57 am

RomeoMustDie wrote:
Fri May 15, 2020 11:44 am
Not looking good for this portfolio in the short term.

Predicting a bad year for the S&P, interest rates also have nowhere to go but up.

However, I will stay the course.
Always glad to hear a pessimistic outlook, it reassures my inner contrarian that this thing still has room to run. Especially the idea that rates have nowhere to go but up.
In fact, it gives me the idea for a new saying: "Bulls make money, Bears make money, and Pigs have nowhere to go but up, up and away!"

mr_mac3
Posts: 4
Joined: Fri Apr 03, 2020 8:09 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by mr_mac3 » Fri May 15, 2020 12:02 pm

Register44 wrote:
Thu May 14, 2020 11:23 pm
I really like this idea of using the vix. Its like the insurance for when the strategy could blow up. I would accept lower returns for that.

Is there a link to the backtest you can share. I would like to test 50 sso/50edv and see how it stacks up. I expect lower returns, lower drawdown, but it may have a higher Sharpe.
I came across this paper the other day, which those of you looking for VIX exposure may appreciate. From the abstract:

"Hedging market downturns without sacrificing upside has long been sought by investors. If VIX was directly investable, adding it as a hedge to the S&P 500 would result in significantly improved performance over the equity only portfolio. However, tradable VIX products do not provide the hedge or returns investors seek over long-term horizons. Alternatively, deconstructing VIX to find the key S&P 500 options which drive VIX movements leads to a synthetic VIX portfolio that provides a more effective hedge. Using these options captures correlations and returns similar to VIX, and combined with the S&P 500, outperforms the buy-and-hold index portfolio."

https://papers.ssrn.com/sol3/papers.cfm ... id=3514141

The idea is that the best way to get VIX exposure is through ITM puts, selling OTM options to fund it. I don't know anything about this topic but thought I'd pass this along.

On a separate topic, I was wondering if the idea of a NTSX-ish strategy with leveraged ETFs has come up for taxable accounts. Suppose I allocated 90% to UPRO or SSO and the remainder to 2-yr treasury futures to get (10/90) x 18 or (10/90) x 27. I'm not advocating this, just brainstorming. You could adjust your equity exposure with SPY/SSO/UPRO to get whatever leverage you want below x2.8. You would have the tax problem of equity futures, you could let LTCG accrue indefinitely. The hard part would be to get the right treasury exposure. Maybe you could use TMF or different treasury futures to try to adjust to the right exposure to interest rate changes. Any thoughts?

https://www.portfoliovisualizer.com/bac ... n5_3=-1610

User avatar
physixfan
Posts: 164
Joined: Wed Apr 03, 2019 4:09 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by physixfan » Fri May 15, 2020 12:05 pm

GiveTendies wrote:
Thu May 14, 2020 11:01 pm
physixfan wrote:
Thu May 14, 2020 12:46 am
Anyone try to do this strategy with LEAPs instead of LETFs?

I feel the expected return vs volatility for TLT is not quite well, so TMF which will suffer from volatility decay may not be a wise thing to hold. Then what about TLT LEAPs options? It seems quite normal to use LEAPs to leverage up SPY. But is it wise to use LEAPs to leverage up TLT?
Not resetting leverage results in leverage drift (down when markets do well and up when markets do that).

Imagine beginning of 2020, you invested 100k from cash + 200k on margin loan into SPY. Your starting leverage is at 3 (assets/equity).
At the beginning of the crisis, Feb 28, SPY is down 8%. Your portfolio is at 276k, loan of 200k, equity 76k. Your leverage already increased to 276/76 = 3.6

On March 17, SPY is down 21.5% since you invested. Portfolio is at 237k, loan of 200k, equity 37k. Your leverage increased to 237/37 = 6.4

On March 23, SPY is down as much as 32%. You're basically wiped out. Most likely your broker will have margin-called you and closed your positions much earlier.

Not resetting leverage negates volatility drag. But there's no free lunch here. The price you pay is massively increased risk in bad markets and lesser upside in good markets.
LEAP options are not margin account, there is no risk of margin call. I agree that the leverage will be drifted higher when the market is down, but as long as we HODL then we will be fine.

jadela
Posts: 47
Joined: Mon Apr 06, 2020 12:32 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by jadela » Fri May 15, 2020 12:22 pm

I have a super basic question — how is this strategy considered risk parity? My extremely naive understanding of a risk parity approach is that it has assets corresponding to the 4 basic economic conditions in the 2x2 of rising/falling + growth/inflation. This strategy seems to only hit 2 of those boxes — what am I missing?

User avatar
physixfan
Posts: 164
Joined: Wed Apr 03, 2019 4:09 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by physixfan » Fri May 15, 2020 1:11 pm

typical.investor wrote:
Fri May 15, 2020 5:22 am
GiveTendies wrote:
Fri May 15, 2020 4:20 am
I don't know why you're always going on about rebalancing, I have only ever talked about TMF standalone. The TMF ETF's performance itself is clearly completely uncorrelated to how much or little you do rebalancing in your account.
It’s simply absurd to ignore that removing (adding) money to a 3X ETF when the fund does well (poorly) can counteract the effect that the daily reset of leverage has.

I can’t understand why you insist on looking at one component of a portfolio in isolation and concluding it shouldn’t be used.

If all you are going to talk about is TMF standalone, why are you even posting here In the thread at all? That's what I don't understand.

Here let's look at a period that isn't favorable to TMF where 30 year rates only dropped 0.04% (basically flat) I chose the 30 year since TMF has an average maturity of 25 years. The 20 year rose 0.18% in that time.

[July 2016 to Oct 2019]
30 yr Treasuries 2.23% -> 2.19%
20 yr Treasuries 1.82% -> 2.00%


CAGR
TLT (iShares 20+ Year Treasury) 3.06% <--no leverage drag
TMF ( 20+ Yr Trsy 3X ETF) 1.82% <-- terrible leverage sucks
TMF/UPRO 20.92% <-- Oh, my leverage wins.
TLT/UPRO 20.13%

So, we do see volatility drag in TMF over a basically flat period. And obviously TLT would make a better pairing with UPRO if we judge by the returns of a single fund as you like to do! Right! Wrong.

And TMF isn't that far away from rolling TLT with futures
TLT X3 - CASHX /UPRO 21.63%

Be honest. What does your method guesstimate for TMF in July 2016 when it was priced to return 2.23%? You estimate it would lose what - four of five percent?

Here's another basically flat period

Jul 2014 - Nov 2018
30 yr 3.33% -> 3.36%
20 yr 3.27% -> 3.27%

CAGR

TLT (iShares 20+ Year Treasury) 2.96% <--- again isn't this best? Hint: again no.
TMF ( 20+ Yr Trsy 3X ETF) 1.97%

TMF/UPRO 14.80%
TLT/UPRO 13.69%
TLT X3 - CASHX /UPRO 15.61%

I simply don't see TMF losing 7% from volatility which you have repeatedly and emphatically stated is sure to be the case, and not only that but any protest to the contrary means one is claiming to know more than other market participants which obviously can't be the case.

Talking about TMF standalone, which you point out is all you do, is both misleading and uninformative I think. I'm going to keep talking about how TMF behaves in the way that it is typically used. You can keep talking about a way that no-one uses it, but to what point?

Anyway, sure futures is likely to do better than TMF. I don't doubt that. I still see TMF as viable though despite your doomed to failure predictions. Leveraged treasuries is better to offset leveraged equity I think. A look at any equity crash will show you that.

Stagflation will kill this strategy. I am not saying it is without real and scary risk.
Thanks for the data. It's reassuring that rebalancing has such a big effect.

NMBob
Posts: 324
Joined: Thu Apr 23, 2015 8:13 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by NMBob » Fri May 15, 2020 1:18 pm

mr_mac3 wrote:
Fri May 15, 2020 12:02 pm
Register44 wrote:
Thu May 14, 2020 11:23 pm
I really like this idea of using the vix. Its like the insurance for when the strategy could blow up. I would accept lower returns for that.

Is there a link to the backtest you can share. I would like to test 50 sso/50edv and see how it stacks up. I expect lower returns, lower drawdown, but it may have a higher Sharpe.
I came across this paper the other day, which those of you looking for VIX exposure may appreciate. From the abstract:

"Hedging market downturns without sacrificing upside has long been sought by investors. If VIX was directly investable, adding it as a hedge to the S&P 500 would result in significantly improved performance over the equity only portfolio. However, tradable VIX products do not provide the hedge or returns investors seek over long-term horizons. Alternatively, deconstructing VIX to find the key S&P 500 options which drive VIX movements leads to a synthetic VIX portfolio that provides a more effective hedge. Using these options captures correlations and returns similar to VIX, and combined with the S&P 500, outperforms the buy-and-hold index portfolio."

https://papers.ssrn.com/sol3/papers.cfm ... id=3514141

The idea is that the best way to get VIX exposure is through ITM puts, selling OTM options to fund it. I don't know anything about this topic but thought I'd pass this along.

On a separate topic, I was wondering if the idea of a NTSX-ish strategy with leveraged ETFs has come up for taxable accounts. Suppose I allocated 90% to UPRO or SSO and the remainder to 2-yr treasury futures to get (10/90) x 18 or (10/90) x 27. I'm not advocating this, just brainstorming. You could adjust your equity exposure with SPY/SSO/UPRO to get whatever leverage you want below x2.8. You would have the tax problem of equity futures, you could let LTCG accrue indefinitely. The hard part would be to get the right treasury exposure. Maybe you could use TMF or different treasury futures to try to adjust to the right exposure to interest rate changes. Any thoughts?

https://www.portfoliovisualizer.com/bac ... n5_3=-1610
a two year futures proposal thread

viewtopic.php?t=290919

langlands
Posts: 260
Joined: Wed Apr 03, 2019 10:05 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by langlands » Fri May 15, 2020 1:43 pm

Gufomel wrote:
Fri May 15, 2020 5:26 am
Thanks for responding. Huge thread that I’ve jumped around on, and didn’t catch that borrowing costs had been mentioned on this page.

So 1% over RFR plus expense ratio of 0.92%. That makes sense. The overnight LIBOR was around 2-2.5% middle of last year, so total expense would have been around 4-4.5%. That’s pretty hefty. I guess for 3x leverage that could make sense, but just seems like that would be a pretty significant drag. I’ve enjoyed what I’ve seen on this thread, but just trying to get a better understanding of the impact of costs.
If I'm not mistaken, the cost was actually higher because you have to double the 2-2.5% overnight LIBOR. You are investing 1 part with your own money and borrowing two parts to get the 3x leverage. So the cost was nearer 6% per year. Of course now since interest rates are much lower, it's much cheaper to leverage.

Gufomel
Posts: 524
Joined: Sat Feb 14, 2015 9:52 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Gufomel » Fri May 15, 2020 2:34 pm

langlands wrote:
Fri May 15, 2020 1:43 pm
Gufomel wrote:
Fri May 15, 2020 5:26 am
Thanks for responding. Huge thread that I’ve jumped around on, and didn’t catch that borrowing costs had been mentioned on this page.

So 1% over RFR plus expense ratio of 0.92%. That makes sense. The overnight LIBOR was around 2-2.5% middle of last year, so total expense would have been around 4-4.5%. That’s pretty hefty. I guess for 3x leverage that could make sense, but just seems like that would be a pretty significant drag. I’ve enjoyed what I’ve seen on this thread, but just trying to get a better understanding of the impact of costs.
If I'm not mistaken, the cost was actually higher because you have to double the 2-2.5% overnight LIBOR. You are investing 1 part with your own money and borrowing two parts to get the 3x leverage. So the cost was nearer 6% per year. Of course now since interest rates are much lower, it's much cheaper to leverage.
Really? If so, that’s crazy high leverage cost and I’m honestly surprised UPRO performed as well as it did over the past decade even with the big bull run. I haven’t spent a long time looking at it, but for the most part all of periods I’ve looked at seemed to show UPRO returning roughly 3x of SPY (obviously there was variation, but it seemed to hover around 3x for the most part). I don’t know how it would even be remotely close to 3x return if the leverage cost was 6%. Am I missing something here?

langlands
Posts: 260
Joined: Wed Apr 03, 2019 10:05 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by langlands » Fri May 15, 2020 2:54 pm

Gufomel wrote:
Fri May 15, 2020 2:34 pm
langlands wrote:
Fri May 15, 2020 1:43 pm
Gufomel wrote:
Fri May 15, 2020 5:26 am
Thanks for responding. Huge thread that I’ve jumped around on, and didn’t catch that borrowing costs had been mentioned on this page.

So 1% over RFR plus expense ratio of 0.92%. That makes sense. The overnight LIBOR was around 2-2.5% middle of last year, so total expense would have been around 4-4.5%. That’s pretty hefty. I guess for 3x leverage that could make sense, but just seems like that would be a pretty significant drag. I’ve enjoyed what I’ve seen on this thread, but just trying to get a better understanding of the impact of costs.
If I'm not mistaken, the cost was actually higher because you have to double the 2-2.5% overnight LIBOR. You are investing 1 part with your own money and borrowing two parts to get the 3x leverage. So the cost was nearer 6% per year. Of course now since interest rates are much lower, it's much cheaper to leverage.
Really? If so, that’s crazy high leverage cost and I’m honestly surprised UPRO performed as well as it did over the past decade even with the big bull run. I haven’t spent a long time looking at it, but for the most part all of periods I’ve looked at seemed to show UPRO returning roughly 3x of SPY (obviously there was variation, but it seemed to hover around 3x for the most part). I don’t know how it would even be remotely close to 3x return if the leverage cost was 6%. Am I missing something here?
I think I'm right but anyone more knowledgeable should correct me if I'm wrong.

Two things to note.

1) Interest rates were close to 0 from 2009-2015 when UPRO was making a huge run, so the leverage cost was quite low during that period.

2) It's not actually true that on an annual basis UPRO returned very close to 3x SPY. The 3x is on a daily basis, and all the volatility drag/compounding effects that have been debated to death cause it to diverge. You can see the annual performance at e.g.

http://performance.morningstar.com/fund ... ion?t=UPRO

UPRO has done extremely well because we were in a secular bull market. Note that since inception (or at least when google finance starts tracking), UPRO has returned 772% whereas VOO has returned 153%. That is closer to 5x, not 3x outperformance. And that's with UPRO's horrendous performance YTD. Leverage performs fantastically when everything roughly goes in the same direction.

Note that UPRO has done very poorly YTD (-45% compared to -9% for SPY) because the market went through a V shaped movement. Choppy markets that go up and down are the cause of volatility drag.

Kbg
Posts: 75
Joined: Thu Mar 23, 2017 11:33 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Kbg » Fri May 15, 2020 3:30 pm

ltdshred wrote:
Fri May 15, 2020 10:13 am
I deleted the link last night. You can just replicate TVIX returns by downloading the 200% ^VIX and -100% CASHX rebalanced monthly returns from PV and then reupload to 2011 and then use monthly TVIX for the rest of the returns
A 1000x no!

If you don't know why, you should find out.

Topic Author
HEDGEFUNDIE
Posts: 4801
Joined: Sun Oct 22, 2017 2:06 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by HEDGEFUNDIE » Fri May 15, 2020 4:09 pm

jadela wrote:
Fri May 15, 2020 12:22 pm
I have a super basic question — how is this strategy considered risk parity? My extremely naive understanding of a risk parity approach is that it has assets corresponding to the 4 basic economic conditions in the 2x2 of rising/falling + growth/inflation. This strategy seems to only hit 2 of those boxes — what am I missing?
You are referring to the “All-Weather” portfolio which I have explicitly disavowed.

Any set of assets can be held in risk parity combination. Just assess the risk contributed by each asset and then hold them in proportions such that the risk from each is equalized.

jadela
Posts: 47
Joined: Mon Apr 06, 2020 12:32 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by jadela » Fri May 15, 2020 4:41 pm

HEDGEFUNDIE wrote:
Fri May 15, 2020 4:09 pm
jadela wrote:
Fri May 15, 2020 12:22 pm
I have a super basic question — how is this strategy considered risk parity? My extremely naive understanding of a risk parity approach is that it has assets corresponding to the 4 basic economic conditions in the 2x2 of rising/falling + growth/inflation. This strategy seems to only hit 2 of those boxes — what am I missing?
You are referring to the “All-Weather” portfolio which I have explicitly disavowed.

Any set of assets can be held in risk parity combination. Just assess the risk contributed by each asset and then hold them in proportions such that the risk from each is equalized.
Ah, thank you very much. That makes sense.

User avatar
privatefarmer
Posts: 695
Joined: Mon Sep 08, 2014 2:45 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by privatefarmer » Fri May 15, 2020 10:44 pm

This thread is so long that the same concerns or questions keep coming up. it’s like wack a mole, someone explains why this strategy has worked and can continue to work but 5-6 pages later someone else asks the same question.

Everyone’s concern is obviously on LTT yields. I am no expert but I did stay at holiday inn last night and my views are this :

- we live in a global economy. When the rest of the world is going deeper into negative rates it puts downward pressure on US yields. Foreign investors invest in US treasuries as our yield is so much more attractive which drives down the yield.
- if the dollar continues to be strong and other currencies weaken, foreign investors will be happy to own treasuries even at negative yields bc they’ll be repaid 30 years from now in green backs which may be worth more than they are today, compared to their home currency
- federal governments for decades or even centuries have had to stimulate their economies to keep things moving and growing. They’re like the oil in an engine, they have to lubricate the economy via stimulus packages to keep things flowing properly and prevent depressions/deflation.
- we have so many deflationary variables that are forever present. Ie technology. Tech advances make it cheaper to extract commodities (ie oil) and build things. This is obvious when seen over a period of several decades. Things actually become cheaper over time bc it’s easier to manufacture them. Therefore, we need an inflationary counter balance such as devaluing of currency and govt stimulus programs to stave off deflation.
- the federal govt never has to actually repay all its debt at any given time “or else”.... what I mean is that the govt has the ability to print money, continually roll over its debt and even continually grow it. The consequence is a devaluation of its currency. However, everything is relative. If we are doing this less than Europe for example then our currency actually strengthens relative to the euro.
- Disney just borrowed billions of dollars at rates that varied between 1.75%-3.8%!! Disney. Great company. Not the US treasury. If The market is willing to loan money to Disney at such low rates it is not surprising whatsoever that the market will loan money to the fed for 30 yrs at 1.77%. I would never predict where interest rates will go from here but obviously when you have huge corporations borrowing money for cheap, the fed has severe downward pressure to keep rates low and stimulate their growth. It would be catastrophic for a company like Disney if they had to suddenly borrow at rates even 1-2% points higher. Same goes for home mortgages. The fed has pressure from many directions to not only keep rates low but lower them further to stimulate corporate borrowing/spending and the housing market.
- bond funds make money not only from yield or lowering rates but also from rolling down the yield curve. We focus so much on starting yield but we should focusing just as much on the steepness of the yield curve. Especially when the yields are low, rolling down the curve becomes an ever growing source of growth for long term treasuries.
- if one looks at the 30+ year return of VUSTX, the real CAGR for the first 15 years is almost identical to the last 15 years, even though yields started high and finished low. This just proves my point that starting yield is irrelevant.
- lastly, what I feel is the most overlooked point, interest rates effect not just treasuries but every single publicly traded security. Low rates stimulate the stock market as well. And not only that but the current low yields are already priced into all securities. My Point is that it is only UNEXPECTED inflation or UNexpected rate increases that will shock the system and hurt this strategy. The market has already priced in low expected inflation/low yields going forward. So the starting yield is actually IRRELEVANT to this strategy. Because it’s already priced in.

My biggest takeaway : Its essentially a 50/50 chance that yields will either go up or go down from here and that’s ALWAYS true no matter where the yield starts at. It is only UNexpected changes in economic growth/inflation/etc etc that moves the market and by definition this is not something that can be predicted. This strategy relies more on the non-correlation between LTT/equities. It does well if one of the two assets is doing well, whether we are in a economic expansion or recession.
Last edited by privatefarmer on Fri May 15, 2020 11:37 pm, edited 1 time in total.

BayStater
Posts: 17
Joined: Sun Mar 29, 2020 11:57 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by BayStater » Fri May 15, 2020 11:15 pm

privatefarmer wrote:
Fri May 15, 2020 10:44 pm
This thread is so long that the same concerns or questions keep coming up. it’s like whack a mole, someone explains why this strategy has worked and can continue to work but 5-6 pages later someone else asks the same question.
At over 8,000 posts between Part I and Part II, we need a wiki. :D

Register44
Posts: 70
Joined: Sun Feb 09, 2020 12:30 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Register44 » Sat May 16, 2020 12:48 am

Really good points in the last couple of posts about long term treasuries. After reading about this strategies and variations the takeaway I have is that mixing long term treasuries with stocks has been the superior asset allocation since the 80's. Going forward there is no evidence this should change, but we cannot guarantee it.

Whether we use VUSTX, EDv, or TMF as long as we are roughly matching exposure to our stock portoflio whether it is spy, vti, sso, upro or pplc it should not matter. The returns and drawdowns can simply be amplified to ones risk tolerance.

For me I don't think I'm suited for the UPRO / TMF. I think Ideally PPLC/EDV is a really nice combo, it just doesn't have close to the upside of UPRO/TMF. But both smack the S*P500.

User avatar
randyharris
Posts: 83
Joined: Wed Mar 18, 2020 7:45 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by randyharris » Sat May 16, 2020 1:17 am

Register44 wrote:
Sat May 16, 2020 12:48 am
Really good points in the last couple of posts about long term treasuries. After reading about this strategies and variations the takeaway I have is that mixing long term treasuries with stocks has been the superior asset allocation since the 80's. Going forward there is no evidence this should change, but we cannot guarantee it.

Whether we use VUSTX, EDv, or TMF as long as we are roughly matching exposure to our stock portoflio whether it is spy, vti, sso, upro or pplc it should not matter. The returns and drawdowns can simply be amplified to ones risk tolerance.

For me I don't think I'm suited for the UPRO / TMF. I think Ideally PPLC/EDV is a really nice combo, it just doesn't have close to the upside of UPRO/TMF. But both smack the S*P500.
Register, I do a slightly toned and alternate down version myself.

Adaptive Allocation Minimum Variance with 2X funds QLD and UBT.

https://www.portfoliovisualizer.com/te ... odWeight=0


Image

If you click the image above to see full size, Risk Parity for me is QLD/UBT, and Risk Parity Max is TQQQ/TMF.
Last edited by randyharris on Sat May 16, 2020 1:35 am, edited 1 time in total.

typical.investor
Posts: 2026
Joined: Mon Jun 11, 2018 3:17 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by typical.investor » Sat May 16, 2020 1:23 am

Register44 wrote:
Sat May 16, 2020 12:48 am
For me I don't think I'm suited for the UPRO / TMF. I think Ideally PPLC/EDV is a really nice combo, it just doesn't have close to the upside of UPRO/TMF. But both smack the S*P500.
If you are more likely to be able to stick with PPLC/EDV, then that's appropriate for you.

I'm reasonably sure that with any amount of inflation this thread will fill up with "time to abandon" posts and many saying "yes" because "risks have changed". Treasuries were the only thing that didn't crash from COVID19. I think I saw posts about capitulation in everything else.

Anyway, it's your money. Do what's best for you personally and pay no heed to any FOMO pangs that might crop up because at the end of the day, you are not a lemming, you shouldn't want to be a lemming, and nobody can say with 100% certainty how any strategy will do going forward.

I guess I shouldn't be so harsh with people criticizing TMF, but I just want to try and get a realistic view about it.

occambogle
Posts: 383
Joined: Thu Dec 12, 2019 4:58 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by occambogle » Sat May 16, 2020 2:43 am

randyharris wrote:
Sat May 16, 2020 1:17 am
If you click the image above to see full size, Risk Parity for me is QLD/UBT, and Risk Parity Max is TQQQ/TMF.
That's interesting regarding QLD/UBT.. I hadn't looked at those two before. Adaptive allocation seems clever but way too much work for me, I assume it needs a lot of hands-on attention.
I assume you are doing this in non-taxable accounts?
I wonder how QLD/UBT or QLD/EDV would fare in a taxable account, as you could do some milder PSLDX-level combinations:

PV Link QLD,UBT,EDV

User avatar
Uncorrelated
Posts: 734
Joined: Sun Oct 13, 2019 3:16 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Uncorrelated » Sat May 16, 2020 4:14 am

langlands wrote:
Fri May 15, 2020 1:43 pm
Gufomel wrote:
Fri May 15, 2020 5:26 am
Thanks for responding. Huge thread that I’ve jumped around on, and didn’t catch that borrowing costs had been mentioned on this page.

So 1% over RFR plus expense ratio of 0.92%. That makes sense. The overnight LIBOR was around 2-2.5% middle of last year, so total expense would have been around 4-4.5%. That’s pretty hefty. I guess for 3x leverage that could make sense, but just seems like that would be a pretty significant drag. I’ve enjoyed what I’ve seen on this thread, but just trying to get a better understanding of the impact of costs.
If I'm not mistaken, the cost was actually higher because you have to double the 2-2.5% overnight LIBOR. You are investing 1 part with your own money and borrowing two parts to get the 3x leverage. So the cost was nearer 6% per year. Of course now since interest rates are much lower, it's much cheaper to leverage.
A higher risk free rate is not a cost. You're borrowing at higher rates, but the stocks that you invest in also earn the same higher risk free rate.

The height of the risk free rate is irrelevant for the decision to leverage. The relevant qualifiers are the spread between the borrow rate and the risk free rate, and the expected equity risk premium.

This is also why most discussions about long term treasuries end up focusing on the wrong arguments. The decision to invest in long term treasuries is not affected by the risk free rate being 10% of -5%. A bet on long term treasuries is a bet on the persistence of a term premium, which can only be indirectly observed. Frankly there is zero reason why the expected term premium is different from 10, 20 years ago. Pay close attention to the word expected here. There is no way the realized term premium the last 20 years is a good predictor of expected term premium, this is also why backtests involving treasuries are completely useless.

IndexCore
Posts: 153
Joined: Wed Mar 25, 2020 1:44 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by IndexCore » Sat May 16, 2020 7:16 am

typical.investor wrote:
Fri May 15, 2020 10:30 am
IndexCore wrote:
Fri May 15, 2020 9:25 am
If you can recall any of the papers by name, source or website, I'd be interested in taking a look. I've been meaning to calculate how to mitigate volatility drag, and a head start would be helpful.
This one explains how fund flows themself reduce the volatility drag.
Empirically, we find that capital flows considerably reduce ETF
rebalancing demand and, therefore, mitigate the potential for ETFs to amplify volatility.
We also show theoretically that flows can completely eliminate ETF rebalancing in the limit
https://www.federalreserve.gov/econresd ... 106pap.pdf
According to the introduction of that paper, it's about market volatility rather than volatility drag on performance:
Leveraged and inverse exchange-traded funds (ETFs) seek to track a multiple of the
performance of an underlying index, commodity, currency, or some other benchmark over a
specified time frame, which is usually one day. These products have been heavily criticized
based on the belief that they exacerbate volatility in financial markets.
I did see some mentions of (1-m) as the "cash position", without specifying more. I would have liked a few specific examples, but overall I don't think I'm looking for the same thing as this paper.
typical.investor wrote:
Fri May 15, 2020 10:30 am
IndexCore wrote:
Fri May 15, 2020 9:25 am
In trying to calculate it myself, I was a bit surprised how much cash is involved:
S&P 500 drops -10%, UPRO drops -30%, requiring +20% cash added to UPRO until recovery (and then removing the 20% cash)
S&P 500 drops -20%, UPRO drops -60%, requiring +40% cash loaned to UPRO until recovery
I read another one suggesting to add/remove when the fund drops/goes up, but I don't remember where. I'm not interested in keeping such a cash position to counter UPRO. That is what TMF is for. Sure it's not exact and not everyday, but then again TMF can really spike up when UPRO drops.

Ah here ... proshares themselves say it:
Investors looking to approximate a fund's multiple for longer than one day may need to rebalance their holdings by increasing or decreasing the investment to maintain a desired exposure. Rebalancing may result in transaction costs and tax consequences. Of course, rebalancing can reduce the negative effects of compounding on performance, but it may reduce the positive effects as well.
https://www.proshares.com/media/documen ... unding.pdf
I know I have seen similar suggestion but never a plan about how much....
Thanks, that was more interesting. It doesn't go as far as I'd like (2x leverage over 2 days, with a -5% then +5% change)... but at least my calculations match theirs. Since for 3X ETFs the sequence matters, I should probably look at not just a sudden -20% drop in the S&P 500, but two drops of -10% followed by -11.1% (which 0.9 x 0.889 = .80) to see how that compares to a single -20% drop.

GiveTendies
Posts: 21
Joined: Tue May 12, 2020 8:17 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by GiveTendies » Sat May 16, 2020 7:58 am

Uncorrelated wrote:
Sat May 16, 2020 4:14 am
This is also why most discussions about long term treasuries end up focusing on the wrong arguments. The decision to invest in long term treasuries is not affected by the risk free rate being 10% of -5%. A bet on long term treasuries is a bet on the persistence of a term premium, which can only be indirectly observed. Frankly there is zero reason why the expected term premium is different from 10, 20 years ago. Pay close attention to the word expected here. There is no way the realized term premium the last 20 years is a good predictor of expected term premium, this is also why backtests involving treasuries are completely useless.
Long term bonds will (over long term) always outperform short term bonds. This is because of their massively higher interest rate and inflation risk. Higher risk = higher reward.
I'm not sure if I agree with the yields. To a lot of people/companies 5% safe vs. 10% equities is a lot better than 0% safe vs. 5% equities.

typical.investor wrote:
Fri May 15, 2020 10:30 am
This one explains how fund flows themself reduce the volatility drag.
Empirically, we find that capital flows considerably reduce ETF
rebalancing demand and, therefore, mitigate the potential for ETFs to amplify volatility.
We also show theoretically that flows can completely eliminate ETF rebalancing in the limit
https://www.federalreserve.gov/econresd ... 106pap.pdf


Ah here ... proshares themselves say it:
Investors looking to approximate a fund's multiple for longer than one day may need to rebalance their holdings by increasing or decreasing the investment to maintain a desired exposure. Rebalancing may result in transaction costs and tax consequences. Of course, rebalancing can reduce the negative effects of compounding on performance, but it may reduce the positive effects as well.
https://www.proshares.com/media/documen ... unding.pdf

I know I have seen similar suggestion but never a plan about how much....
You're spreading misinformation.
The article talks about market implications of funds themselves having to buy/sell in order to reset their leverage at the end of the trading day. Both long and short levered ETFs have to do the following: buy in up-markets and sell in down-markets. This grew concerns, that levered ETFs would move markets too much. But there is the adverse effect, that investors tend to also withdraw/add to those funds when market is up/down, thus the effect is dampened. The ETFs have to do less active rebalancing themselves.
THIS HAS ABSOLUTELY NOTHING TO DO WITH VOLATILITY DECAY. It still exists. It is absolutely not affected by this.

The Proshares article talks about how long-term holding of levered ETFs is risky and how they vastly over/under-perform in up/down-markets. Rebalance counters that risk but again, does not reduce volatility decay of the fund. It indeed does reduce volatility decay for you in your portfolio, but if the fund resets leverage daily and you rebalance quarterly, this effect is close to zero.

typical.investor
Posts: 2026
Joined: Mon Jun 11, 2018 3:17 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by typical.investor » Sat May 16, 2020 9:52 am

GiveTendies wrote:
Sat May 16, 2020 7:58 am
You're spreading misinformation.
Sorry, but that is not correct. You are spreading misinformation about who is spreading misinformation. It's not me.
GiveTendies wrote:
Sat May 16, 2020 7:58 am
The article talks about market implications of funds themselves having to buy/sell in order to reset their leverage at the end of the trading day.
I am 100% sure you have either not read the article, have not read it carefully enough, or just plain don't understand what you have read no matter how carefully you have tried. I'll bet it's some combination of two and three.

typical.investor
Posts: 2026
Joined: Mon Jun 11, 2018 3:17 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by typical.investor » Sat May 16, 2020 9:54 am

GiveTendies wrote:
Sat May 16, 2020 7:58 am
You're spreading misinformation.
Sorry, but the information I have posted is both correct and useful.
GiveTendies wrote:
Sat May 16, 2020 7:58 am
THIS HAS ABSOLUTELY NOTHING TO DO WITH VOLATILITY DECAY.
THIS HAS EVERYTHING TO DO WITH VOLATILITY DECAY

I'm sorry if you don't understand it. Give me a minute or two and I will explain ... see my next post.
GiveTendies wrote:
Sat May 16, 2020 7:58 am
The article talks about market implications of funds themselves having to buy/sell in order to reset their leverage at the end of the trading day.
I am 100% sure you have either not read the article, have not read it carefully enough, or just plain don't understand what you have read no matter how carefully you have tried. I'll bet it's some combination of two and three. Keep with it and you will get there!
Last edited by typical.investor on Sat May 16, 2020 11:06 am, edited 1 time in total.

typical.investor
Posts: 2026
Joined: Mon Jun 11, 2018 3:17 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by typical.investor » Sat May 16, 2020 11:01 am

IndexCore wrote:
Sat May 16, 2020 7:16 am
According to the introduction of that paper, it's about market volatility rather than volatility drag on performance:
Actually, I think it is about performance. Market volatility includes both aspects of performance -- drag and boost. I mean LETFs can do either, under or over perform the multiple of its index. See this for example ... we are talking about price movement.
leveraged and inverse ETFs must rebalance their portfolios in the same direction as the contemporaneous return on their underlying assets in order to maintain a constant leverage ratio. Conventional thinking suggests that... these types of financial products exert additional upward price pressure on the underlying assets following positive returns and additional downward pressure following negative returns
They go on to point out though that actually flows into/out of the fund mitigate price movement away from the multiple of the index. Actually in theory it can eliminate it.

Anyway, if you look at p. 16-18, they clearly are discussing investors who wish to achieve a multiple of the index.
To clearly illustrate the mechanism, we introduce a representative investor who wishes to track m-times the return on the index over a horizon longer than one period and assume that trading by the investor gives rise to capital flows.
If you want to see the formula, look at page 17. They started with the case the case where the investor wishes to track m-times the index return over two periods, and discuss how rebalancing needs to take place:
the investor generally must rebalance his portfolio every period to successfully track m-times the return on the index over a longer time horizon
To obtain the desired exposure to the index, the flow at time t must be such that the return from holding the ETF matches m-times the return on the index
ft = (1 − m)rt

Actually, I think you already had it solved for the first day at least.
IndexCore wrote:
Sat May 16, 2020 7:16 am
In trying to calculate it myself, I was a bit surprised how much cash is involved:
S&P 500 drops -10%, UPRO drops -30%, requiring +20% cash added to UPRO until recovery (and then removing the 20% cash)
flow needed at time t = (1-multiple of index) return at time t
ft = (1 − m)rt,
+20% =(1-3)-10%

The only difference is I don't think you keep the cash in until recovery. Rather you would rebalance the next day. Here is an example:

Image

I believe it's simply an empirical fact that daily rebalancing will eliminate volatility drag. No one would do that though so it's a good thing that it needn't be done daily. Adding $2870 on day three is the same as adding $2k on day. 1, removing $900 on day 2 and adding 1890 on day 3.

I don't mean to suggest that using TMF (3X leveraged treasuries) to rebalance against UPRO (3X S&P500) will eliminate all volatility decay. It would seem though that if they are negatively correlated, that doing so will definitely reduce volatility decay. Thus I think it is pointless to look at volatility decay in each fund separately. It's simply an empirical fact that flows (such as those involved in rebalancing) can mitigate or eliminate decay.

Of course, if both TMF and UPRO were to go sideways such that money from TMF wasn't available to put into UPRO when it was down and money from UPRO wasn't available to put into TMF when it was down, then I think volatility drag could be a threat and there would be real risk of underperforming 3X the index. The only solution I see is to add more money and trust that TMF and UPRO will see better days.

Gufomel
Posts: 524
Joined: Sat Feb 14, 2015 9:52 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Gufomel » Sat May 16, 2020 12:57 pm

Uncorrelated wrote:
Sat May 16, 2020 4:14 am
langlands wrote:
Fri May 15, 2020 1:43 pm
Gufomel wrote:
Fri May 15, 2020 5:26 am
Thanks for responding. Huge thread that I’ve jumped around on, and didn’t catch that borrowing costs had been mentioned on this page.

So 1% over RFR plus expense ratio of 0.92%. That makes sense. The overnight LIBOR was around 2-2.5% middle of last year, so total expense would have been around 4-4.5%. That’s pretty hefty. I guess for 3x leverage that could make sense, but just seems like that would be a pretty significant drag. I’ve enjoyed what I’ve seen on this thread, but just trying to get a better understanding of the impact of costs.
If I'm not mistaken, the cost was actually higher because you have to double the 2-2.5% overnight LIBOR. You are investing 1 part with your own money and borrowing two parts to get the 3x leverage. So the cost was nearer 6% per year. Of course now since interest rates are much lower, it's much cheaper to leverage.
A higher risk free rate is not a cost. You're borrowing at higher rates, but the stocks that you invest in also earn the same higher risk free rate.

The height of the risk free rate is irrelevant for the decision to leverage. The relevant qualifiers are the spread between the borrow rate and the risk free rate, and the expected equity risk premium.

This is also why most discussions about long term treasuries end up focusing on the wrong arguments. The decision to invest in long term treasuries is not affected by the risk free rate being 10% of -5%. A bet on long term treasuries is a bet on the persistence of a term premium, which can only be indirectly observed. Frankly there is zero reason why the expected term premium is different from 10, 20 years ago. Pay close attention to the word expected here. There is no way the realized term premium the last 20 years is a good predictor of expected term premium, this is also why backtests involving treasuries are completely useless.
Thanks. So is it roughly accurate to say that the total cost for UPRO is 0.92% expense ratio + 0.5% spread = ~1.5%?

And if I understand correctly from reading the other thread “Simulating Returns of Leveraged ETFs”, the spread for SSO was about the same as UPRO even though it’s only 2x leveraged?

langlands
Posts: 260
Joined: Wed Apr 03, 2019 10:05 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by langlands » Sat May 16, 2020 1:21 pm

Gufomel wrote:
Sat May 16, 2020 12:57 pm
Uncorrelated wrote:
Sat May 16, 2020 4:14 am
langlands wrote:
Fri May 15, 2020 1:43 pm
Gufomel wrote:
Fri May 15, 2020 5:26 am
Thanks for responding. Huge thread that I’ve jumped around on, and didn’t catch that borrowing costs had been mentioned on this page.

So 1% over RFR plus expense ratio of 0.92%. That makes sense. The overnight LIBOR was around 2-2.5% middle of last year, so total expense would have been around 4-4.5%. That’s pretty hefty. I guess for 3x leverage that could make sense, but just seems like that would be a pretty significant drag. I’ve enjoyed what I’ve seen on this thread, but just trying to get a better understanding of the impact of costs.
If I'm not mistaken, the cost was actually higher because you have to double the 2-2.5% overnight LIBOR. You are investing 1 part with your own money and borrowing two parts to get the 3x leverage. So the cost was nearer 6% per year. Of course now since interest rates are much lower, it's much cheaper to leverage.
A higher risk free rate is not a cost. You're borrowing at higher rates, but the stocks that you invest in also earn the same higher risk free rate.

The height of the risk free rate is irrelevant for the decision to leverage. The relevant qualifiers are the spread between the borrow rate and the risk free rate, and the expected equity risk premium.

This is also why most discussions about long term treasuries end up focusing on the wrong arguments. The decision to invest in long term treasuries is not affected by the risk free rate being 10% of -5%. A bet on long term treasuries is a bet on the persistence of a term premium, which can only be indirectly observed. Frankly there is zero reason why the expected term premium is different from 10, 20 years ago. Pay close attention to the word expected here. There is no way the realized term premium the last 20 years is a good predictor of expected term premium, this is also why backtests involving treasuries are completely useless.
Thanks. So is it roughly accurate to say that the total cost for UPRO is 0.92% expense ratio + 0.5% spread = ~1.5%?

And if I understand correctly from reading the other thread “Simulating Returns of Leveraged ETFs”, the spread for SSO was about the same as UPRO even though it’s only 2x leveraged?
Let's just be very clear on what "cost" means. I agree with Uncorrelated that from an academic finance point of view, it makes sense to separate the risk free rate from the expense ratio and spread. Theoretically, if you believe things like the CAPM model, the return of the s&p 500 is the risk free rate + equity premium. So if you're trying to capture the equity premium, the risk free rate becomes irrelevant. Just be clear though that this way of thinking about things makes the risk free rate the baseline. In other words, if the risk free rate is 2% and the S&P 500 doesn't move, then you are down 2% (plus the expense ratio and the spread) on the amount you borrowed. Maybe I'm wrong, but to most people not steeped in academic finance, that's a cost.

User avatar
Uncorrelated
Posts: 734
Joined: Sun Oct 13, 2019 3:16 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Uncorrelated » Sat May 16, 2020 1:28 pm

Gufomel wrote:
Sat May 16, 2020 12:57 pm
Uncorrelated wrote:
Sat May 16, 2020 4:14 am
langlands wrote:
Fri May 15, 2020 1:43 pm
Gufomel wrote:
Fri May 15, 2020 5:26 am
Thanks for responding. Huge thread that I’ve jumped around on, and didn’t catch that borrowing costs had been mentioned on this page.

So 1% over RFR plus expense ratio of 0.92%. That makes sense. The overnight LIBOR was around 2-2.5% middle of last year, so total expense would have been around 4-4.5%. That’s pretty hefty. I guess for 3x leverage that could make sense, but just seems like that would be a pretty significant drag. I’ve enjoyed what I’ve seen on this thread, but just trying to get a better understanding of the impact of costs.
If I'm not mistaken, the cost was actually higher because you have to double the 2-2.5% overnight LIBOR. You are investing 1 part with your own money and borrowing two parts to get the 3x leverage. So the cost was nearer 6% per year. Of course now since interest rates are much lower, it's much cheaper to leverage.
A higher risk free rate is not a cost. You're borrowing at higher rates, but the stocks that you invest in also earn the same higher risk free rate.

The height of the risk free rate is irrelevant for the decision to leverage. The relevant qualifiers are the spread between the borrow rate and the risk free rate, and the expected equity risk premium.

This is also why most discussions about long term treasuries end up focusing on the wrong arguments. The decision to invest in long term treasuries is not affected by the risk free rate being 10% of -5%. A bet on long term treasuries is a bet on the persistence of a term premium, which can only be indirectly observed. Frankly there is zero reason why the expected term premium is different from 10, 20 years ago. Pay close attention to the word expected here. There is no way the realized term premium the last 20 years is a good predictor of expected term premium, this is also why backtests involving treasuries are completely useless.
Thanks. So is it roughly accurate to say that the total cost for UPRO is 0.92% expense ratio + 0.5% spread = ~1.5%?
You pay the spread approximately twice, because the fund uses swap contracts to cover approximately 250% of it's equity exposure. For this fund I estimated 1.0164% total for the borrow costs, with the expense ratio added it's around 1.94%.

The exact details are documented in this topic: viewtopic.php?p=4884654#p4884654

In times of stress the borrow costs can be much higher, but we currently have no reliable way of observing the swap spread in real time. The number above is an average on 2014-2019.

Gufomel
Posts: 524
Joined: Sat Feb 14, 2015 9:52 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Gufomel » Sat May 16, 2020 2:29 pm

Uncorrelated wrote:
Sat May 16, 2020 1:28 pm
Gufomel wrote:
Sat May 16, 2020 12:57 pm
Uncorrelated wrote:
Sat May 16, 2020 4:14 am
langlands wrote:
Fri May 15, 2020 1:43 pm
Gufomel wrote:
Fri May 15, 2020 5:26 am
Thanks for responding. Huge thread that I’ve jumped around on, and didn’t catch that borrowing costs had been mentioned on this page.

So 1% over RFR plus expense ratio of 0.92%. That makes sense. The overnight LIBOR was around 2-2.5% middle of last year, so total expense would have been around 4-4.5%. That’s pretty hefty. I guess for 3x leverage that could make sense, but just seems like that would be a pretty significant drag. I’ve enjoyed what I’ve seen on this thread, but just trying to get a better understanding of the impact of costs.
If I'm not mistaken, the cost was actually higher because you have to double the 2-2.5% overnight LIBOR. You are investing 1 part with your own money and borrowing two parts to get the 3x leverage. So the cost was nearer 6% per year. Of course now since interest rates are much lower, it's much cheaper to leverage.
A higher risk free rate is not a cost. You're borrowing at higher rates, but the stocks that you invest in also earn the same higher risk free rate.

The height of the risk free rate is irrelevant for the decision to leverage. The relevant qualifiers are the spread between the borrow rate and the risk free rate, and the expected equity risk premium.

This is also why most discussions about long term treasuries end up focusing on the wrong arguments. The decision to invest in long term treasuries is not affected by the risk free rate being 10% of -5%. A bet on long term treasuries is a bet on the persistence of a term premium, which can only be indirectly observed. Frankly there is zero reason why the expected term premium is different from 10, 20 years ago. Pay close attention to the word expected here. There is no way the realized term premium the last 20 years is a good predictor of expected term premium, this is also why backtests involving treasuries are completely useless.
Thanks. So is it roughly accurate to say that the total cost for UPRO is 0.92% expense ratio + 0.5% spread = ~1.5%?
You pay the spread approximately twice, because the fund uses swap contracts to cover approximately 250% of it's equity exposure. For this fund I estimated 1.0164% total for the borrow costs, with the expense ratio added it's around 1.94%.

The exact details are documented in this topic: viewtopic.php?p=4884654#p4884654

In times of stress the borrow costs can be much higher, but we currently have no reliable way of observing the swap spread in real time. The number above is an average on 2014-2019.
Ok makes sense. Appreciate it. I had read your post on the other thread but didn’t catch the nuance of paying the spread twice.

User avatar
firebirdparts
Posts: 1577
Joined: Thu Jun 13, 2019 4:21 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by firebirdparts » Sat May 16, 2020 2:43 pm

IndexCore wrote:
Fri May 15, 2020 9:25 am
In trying to calculate it myself, I was a bit surprised how much cash is involved:
S&P 500 drops -10%, UPRO drops -30%, requiring +20% cash added to UPRO until recovery (and then removing the 20% cash)
S&P 500 drops -20%, UPRO drops -60%, requiring +40% cash loaned to UPRO until recovery
This is close enough for idle chitchat, but UPRO starts over every day. It’s not true at all that it drops 30% when the index drops 10%. FWIW. If it did that in one day, then yeah we’d say it tracked very well that one day.
A fool and your money are soon partners

Gufomel
Posts: 524
Joined: Sat Feb 14, 2015 9:52 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Gufomel » Sat May 16, 2020 2:47 pm

langlands wrote:
Sat May 16, 2020 1:21 pm
Gufomel wrote:
Sat May 16, 2020 12:57 pm
Uncorrelated wrote:
Sat May 16, 2020 4:14 am
langlands wrote:
Fri May 15, 2020 1:43 pm
Gufomel wrote:
Fri May 15, 2020 5:26 am
Thanks for responding. Huge thread that I’ve jumped around on, and didn’t catch that borrowing costs had been mentioned on this page.

So 1% over RFR plus expense ratio of 0.92%. That makes sense. The overnight LIBOR was around 2-2.5% middle of last year, so total expense would have been around 4-4.5%. That’s pretty hefty. I guess for 3x leverage that could make sense, but just seems like that would be a pretty significant drag. I’ve enjoyed what I’ve seen on this thread, but just trying to get a better understanding of the impact of costs.
If I'm not mistaken, the cost was actually higher because you have to double the 2-2.5% overnight LIBOR. You are investing 1 part with your own money and borrowing two parts to get the 3x leverage. So the cost was nearer 6% per year. Of course now since interest rates are much lower, it's much cheaper to leverage.
A higher risk free rate is not a cost. You're borrowing at higher rates, but the stocks that you invest in also earn the same higher risk free rate.

The height of the risk free rate is irrelevant for the decision to leverage. The relevant qualifiers are the spread between the borrow rate and the risk free rate, and the expected equity risk premium.

This is also why most discussions about long term treasuries end up focusing on the wrong arguments. The decision to invest in long term treasuries is not affected by the risk free rate being 10% of -5%. A bet on long term treasuries is a bet on the persistence of a term premium, which can only be indirectly observed. Frankly there is zero reason why the expected term premium is different from 10, 20 years ago. Pay close attention to the word expected here. There is no way the realized term premium the last 20 years is a good predictor of expected term premium, this is also why backtests involving treasuries are completely useless.
Thanks. So is it roughly accurate to say that the total cost for UPRO is 0.92% expense ratio + 0.5% spread = ~1.5%?

And if I understand correctly from reading the other thread “Simulating Returns of Leveraged ETFs”, the spread for SSO was about the same as UPRO even though it’s only 2x leveraged?
Let's just be very clear on what "cost" means. I agree with Uncorrelated that from an academic finance point of view, it makes sense to separate the risk free rate from the expense ratio and spread. Theoretically, if you believe things like the CAPM model, the return of the s&p 500 is the risk free rate + equity premium. So if you're trying to capture the equity premium, the risk free rate becomes irrelevant. Just be clear though that this way of thinking about things makes the risk free rate the baseline. In other words, if the risk free rate is 2% and the S&P 500 doesn't move, then you are down 2% (plus the expense ratio and the spread) on the amount you borrowed. Maybe I'm wrong, but to most people not steeped in academic finance, that's a cost.
Thanks for the clarification. I guess even if you’re investing un-leveraged in the S&P you’re “borrowing” at the risk-free rate (at a minimum) in order to invest because whatever you’re investing in the S&P could have otherwise been invested at the risk-free rate? In fact, as a small scale retail investor the risk-free rate may be higher than the overnight LIBOR because of things like high-yield savings accounts and bank CDs. So when it comes to a 2x leverage S&P fund, you pay the risk-free rate again (and for retail investors that’s maybe a lower risk-free rate than you’re paying to get the initial 1x exposure) to get access to another unit of the S&P, and then pay the risk-free rate a 3rd time for a 3x leverage S&P fund. It’s only any spread and additional expense ratio that’s truly an additional cost incurred in order to obtain leveraged 2x or 3x S&P exposure that you would not normally pay in order to gain 1x S&P exposure.

Is that an appropriate way to frame it for those of us who aren’t used to thinking about the cost of leverage?

langlands
Posts: 260
Joined: Wed Apr 03, 2019 10:05 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by langlands » Sat May 16, 2020 3:19 pm

Gufomel wrote:
Sat May 16, 2020 2:47 pm
langlands wrote:
Sat May 16, 2020 1:21 pm
Let's just be very clear on what "cost" means. I agree with Uncorrelated that from an academic finance point of view, it makes sense to separate the risk free rate from the expense ratio and spread. Theoretically, if you believe things like the CAPM model, the return of the s&p 500 is the risk free rate + equity premium. So if you're trying to capture the equity premium, the risk free rate becomes irrelevant. Just be clear though that this way of thinking about things makes the risk free rate the baseline. In other words, if the risk free rate is 2% and the S&P 500 doesn't move, then you are down 2% (plus the expense ratio and the spread) on the amount you borrowed. Maybe I'm wrong, but to most people not steeped in academic finance, that's a cost.
Thanks for the clarification. I guess even if you’re investing un-leveraged in the S&P you’re “borrowing” at the risk-free rate (at a minimum) in order to invest because whatever you’re investing in the S&P could have otherwise been invested at the risk-free rate? In fact, as a small scale retail investor the risk-free rate may be higher than the overnight LIBOR because of things like high-yield savings accounts and bank CDs. So when it comes to a 2x leverage S&P fund, you pay the risk-free rate again (and for retail investors that’s maybe a lower risk-free rate than you’re paying to get the initial 1x exposure) to get access to another unit of the S&P, and then pay the risk-free rate a 3rd time for a 3x leverage S&P fund. It’s only any spread and additional expense ratio that’s truly an additional cost incurred in order to obtain leveraged 2x or 3x S&P exposure that you would not normally pay in order to gain 1x S&P exposure.

Is that an appropriate way to frame it for those of us who aren’t used to thinking about the cost of leverage?
Yes, that makes perfect sense.

jadela
Posts: 47
Joined: Mon Apr 06, 2020 12:32 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by jadela » Sat May 16, 2020 8:43 pm

I'm about to implement this strategy and wanted to clarify the intended rebalancing — I believe it's quarterly, but are there any other rebalance conditions e.g. 5/25 rebalancing?

typical.investor
Posts: 2026
Joined: Mon Jun 11, 2018 3:17 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by typical.investor » Sun May 17, 2020 1:46 am

jadela wrote:
Sat May 16, 2020 8:43 pm
I'm about to implement this strategy and wanted to clarify the intended rebalancing — I believe it's quarterly, but are there any other rebalance conditions e.g. 5/25 rebalancing?
I believe quarterly works best in backtests and in live data, but I don't want to take my eye off 5/25.

I don't have access to the portfoliovisualizer sims to backtest rebalancing, so am not 100% sure, but if I test periods in live data where rates were basically flat, 5/25 appears to be a slight winner. Maybe that's just for my allocation though.

https://www.portfoliovisualizer.com/bac ... tion2_1=50

https://www.portfoliovisualizer.com/bac ... tion2_1=50

In any case, PV can only rebalance once a month. Even if you put in a ridiculously small rebalancing band, the most frequently you will see is every month. So I don't think PV is really telling us how 5/25 works. For example, I believe 5/25 would have been triggered more than once in March but PV won't show that.

User avatar
privatefarmer
Posts: 695
Joined: Mon Sep 08, 2014 2:45 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by privatefarmer » Sun May 17, 2020 1:50 am

i've started to play around w/ adding the VIX to this strategy and it certainly is appealing. However, the only way to invest in the VIX that I can find is through an ETN like VXX or VIXY, or use the leveraged version TVIX. all these options seem to drastically underperform the VIX index itself. Is there any better way to gain long-term exposure to the VIX? Thanks

Post Reply