Kbg wrote:
So here's another fun one. Same backtest as below but this time 8.33% for everything and cash at 75%...75% cash, think about that. Three assets could get totally wiped out and you are still sitting with 75% in cash. And the numbers are - 4.33% CAGR/6.83% DD, Sharpe .15! Same performance essentially, less drawdown and arguably safer in at least bonds and stocks.
There was a brilliant UK poster here a while back (can't remember his name) that advocated very much the same scenario. He went a little off the deep end, deleted all of his posts and was later banned, but I think his reasoning (and yours) on combining leverage and cash was sound.
I made some very stupid moves a few decades ago and will never touch leverage again - but that is just me. Your ideas seem reasonable.
If MTex's backtest of a 25% drawdown is correct for a normal PP, then I think we can safely conclude that it is nearly impossible if not impossible for the 8.3% at 3x leverage to have that kind of drawdown. That would mean both gold and long bonds would have to tube simultaneously at which point one can conclude the PP failed in its basic premises as an investment strategy.
Kbg wrote:
If MTex's backtest of a 25% drawdown is correct for a normal PP, then I think we can safely conclude that it is nearly impossible if not impossible for the 8.3% at 3x leverage to have that kind of drawdown. That would mean both gold and long bonds would have to tube simultaneously at which point one can conclude the PP failed in its basic premises as an investment strategy.
It was MG not MT. It is correct and you can easily verify it with Peak2Trough's backtester. Do it annually from 1968 and just add the missing 5% to account for different starting and/or rebalancing dates.
I'm not sure how you figure a 3x PP would have less MaxDD. It will have more unless you normalize it to have the same risk or less as the PP. There's no advantage to using leverage per se; it's all smoke and mirrors unless you believe that having more cash with the same net portfolio return is somehow magical. I think it's just a case of cognitive bias. And we yet don't know what the optimal rebalancing frequency or bands are for such leveraged portfolios.
You're in for a surprise some day because all three have declined simultaneously several times in history.
Last edited by MachineGhost on Wed Jan 07, 2015 12:42 am, edited 1 time in total.
"All generous minds have a horror of what are commonly called 'Facts'. They are the brute beasts of the intellectual domain." -- Thomas Hobbes
Disclaimer: I am not a broker, dealer, investment advisor, physician, theologian or prophet. I should not be considered as legally permitted to render such advice!
MachineGhost wrote:
It was MG not MT. It is correct and you can easily verify it with Peak2Trough's backtester. Do it annually from 1968 and just add the missing 5% to account for different starting and/or rebalancing dates.
I'm not sure how you figure a 3x PP would have less MaxDD. It will have more unless you normalize it to have the same risk or less as the PP. There's no advantage to using leverage per se; it's all smoke and mirrors unless you believe that having more cash with the same net portfolio return is somehow magical. I think it's just a case of cognitive bias. And we yet don't know what the optimal rebalancing frequency or bands are for such leveraged portfolios.
You're in for a surprise some day because all three have declined simultaneously several times in history.
Apologies on the MG vs. MT.
The figures I posted are actual backtest results for a period of three years using dividend/split adjusted data from Norgate. So if the data is bad, mea culpa. If the data is correct, those are the numbers. End of story. A portfolio of 3xETFs at 8.33% of the portfolio each combined with 75% cash performed better than the standard 4x25 unleveraged version on a Sharpe measured basis . Actual performance was 1% less while drawdown was 15% better.
To be clear, I'm not advocating folks replace a normal PP with this version and particularly if they want the characteristics of the "actuals." However, if you aren't a hardcore PP purist but like the notion of a 25x4 PP and you don't really care if the solution set is ETFs I suspect that in just about every case the 3x ETF version will do better due to the mathematics of daily rebalancing in assets that trend. It's nothing more cosmic than a positive vs. a negative exponential relationship. And to repeat...every one of the three could go to zero and the performance will not be worse (cannot be worse) than the unleveraged version per your backtest.
HB was pretty clear, you want high non-correlated volatility as package. All I'm doing is extending the idea. Is it all goodness and roses? No it isn't if you have a go nowhere year the loses are highly likely to be worse than the standard version.
Those are some really heinous draw downs in long bonds and gold, but it's the total package right?
Yeah but historically 2013 or 2009 was hardly the worse year for the PP, that's the problem. So if we do take 25% MaxDD as a given, how would you weight a 3x ETF portfolio so that the MaxDD does not go past 10%? That would be my sweet spot for doing it.
Despite what I said early, I can actually think of some advantage to having more cash available for absolute-return or non-secondary market exposure to the assets. Hmm!
"All generous minds have a horror of what are commonly called 'Facts'. They are the brute beasts of the intellectual domain." -- Thomas Hobbes
Disclaimer: I am not a broker, dealer, investment advisor, physician, theologian or prophet. I should not be considered as legally permitted to render such advice!
MachineGhost wrote:
So if we do take 25% MaxDD as a given, how would you weight a 3x ETF portfolio so that the MaxDD does not go past 10%? That would be my sweet spot for doing it.
Despite what I said early, I can actually think of some advantage to having more cash available for absolute-return or non-secondary market exposure to the assets. Hmm!
Based on historical returns, I don't think it's possible. I don't know if you've seen the below but it appears it is very possible to dial down PP volatility using risk parity. OBTW, if you are going to code up PP in AB with a risk parity overlay, please send me the code when you are done.
With regard to what I've been posting, my guess is that a 3x ETF portfolio (8.33% weighting to sub for full a unlevered 25% weighting) would do a bit better at a similar risk profile. One never knows, the only thing we know is there is a + and - exponential curve to the returns using 3xETFs so that the losses will be somewhat mitigated and the gains somewhat magnified. It's a bet on the math pure and simple.
25% / 3 = 8.33%. I see you just did a simple linear scaling for the weights, but I think the LEAP PP thread has more applicability. One advantage with LEAPS is you can use forward-looking volatility. I believe that would be the best way to risk-paritize the PP as I have zero confidence in trailing volatility, returns or correlations, with all due respect to GestaltU (who charge management fees for their complexity). One thing we've all learned from the PP is that simplicity is robust and complex financial-engineering is fragile. I have a great distaste for the latter.
Last edited by MachineGhost on Thu Jan 08, 2015 11:56 pm, edited 1 time in total.
"All generous minds have a horror of what are commonly called 'Facts'. They are the brute beasts of the intellectual domain." -- Thomas Hobbes
Disclaimer: I am not a broker, dealer, investment advisor, physician, theologian or prophet. I should not be considered as legally permitted to render such advice!
Agree, not a fan of complex financial engineering either. I just saw the leap thread last night. I choke on thinking options as a PP sub, but very interesting as a vp.
I'm going to drop 2x reporting for a couple of reasons. 1) I'm using a 3x version personally and, 2) UBT has absolutely insane spreads now...it's a dying ETF. If someone is considering doing this as a VP they should consider 3x ETFs and then reducing by 1/3rd the allocation. Hopefully enough has been posted to illustrate that decay, within a trending asset, is not the concern you normally see on the web.
I am about to pull the trigger in a 3x PP for my VP (UGLD, UPRO, and TMF). How tax efficient are these 3x ETFs? Morningstar apparently has tax data on ETFs, but I'm not sure I understand everything. For example, for TMF, it says "Potential Cap Gains Exposure" 18.89. What does that mean? http://performance.morningstar.com/fund ... tion?t=TMF
Agree with Pugchief. Please read the literature on the Direxion and ProShares websites and the prospectuses. Also go read everything you can about why you shouldn't buy a leveraged ETF. Realize that using PP history you could possibly suffer a north of 37% DD. If after all that, you are comfortable with the decision only then employ your hard earned money. And if you do decide to invest you must not dwell on the individual components, their movements up will be elating and down extremely brutal. You may want to have a little 3x PP VP as a percent of your portfolio starting out to learn if you really can stick with it.
Once you read the prospectuses you will have answers to both your questions.
Last edited by Kbg on Mon Jan 19, 2015 8:37 am, edited 1 time in total.
My plan was to only put 2% of my overall portfolio into the 3x PP, so I am not overly worried about the risks. I happen to have 2% of my net worth in my IRA, so I was thinking about making my entire IRA a 3x PP, depending on the tax treatment of these ETFs. I looked over the prospectuses of these ETFs did not see anything describing the tax consequences.
When I run backtests at ETFReplay.com, the returns of a 3x PP are pretty amazing (33.3%/33.3%/33.3% with no cash). 86% total return over the past 3 years. I don't feel like I am chasing returns, though, because the PP generally goes up consistently, over time.
Kbg wrote:
Agree with Pugchief. Please read the literature on the Direxion and ProShares websites and the prospectuses. Also go read everything you can about why you shouldn't buy a leveraged ETF. Realize that using PP history you could possibly suffer a north of 37% DD. If after all that, you are comfortable with the decision only then employ your hard earned money. And if you do decide to invest you must not dwell on the individual components, their movements up will be elating and down extremely brutal. You may want to have a little 3x PP VP as a percent of your portfolio starting out to learn if you really can stick with it.
Once you read the prospectuses you will have answers to both your questions.
Last edited by jason on Mon Jan 19, 2015 10:11 am, edited 1 time in total.
jason wrote:
My plan was to only put 2% of my overall portfolio into the 3x PP, so I am not overly worried about the risks. I happen to have 2% of my net worth in my IRA, so I was thinking about making my entire IRA a 3x PP, depending on the tax treatment of these ETFs. I looked over the prospectuses of these ETFs did not see anything describing the tax consequences.
When I run backtests at ETFReplay.com, the returns of a 3x PP are pretty amazing (33.3%/33.3%/33.3% with no cash). 86% total return over the past 3 years. I don't feel like I am chasing returns, though, because the PP generally goes up consistently, over time.
Kbg wrote:
Agree with Pugchief. Please read the literature on the Direxion and ProShares websites and the prospectuses. Also go read everything you can about why you shouldn't buy a leveraged ETF. Realize that using PP history you could possibly suffer a north of 37% DD. If after all that, you are comfortable with the decision only then employ your hard earned money. And if you do decide to invest you must not dwell on the individual components, their movements up will be elating and down extremely brutal. You may want to have a little 3x PP VP as a percent of your portfolio starting out to learn if you really can stick with it.
Once you read the prospectuses you will have answers to both your questions.
I'm confused. If you are planning on doing this entirely from within an IRA, what difference does it make what the 'tax consequences' are? You can buy and sell and reinvest or not 'til the cows come home, and the only tax concern is the ordinary income tax rate you will pay when you withdraw any money from the IRA. If it's a traditional IRA, all withdrawals are fully taxable at your marginal rate. If it's a Roth IRA, none of the withdrawals are subject to tax of any kind as long as the contributions have been in it for at least 5 years.
The reason I am asking is because I'm not sure if I am going to do it in my IRA or not. It depends on how tax efficient the ETFs are. If they are tax inefficient, then I would do it in my IRA. If they are tax efficient, then I would do it outside my IRA. So, I can't figure out how to proceed until I know more about the taxes involved.
Last edited by jason on Mon Jan 19, 2015 2:48 pm, edited 1 time in total.
jason wrote:
If they are tax inefficient, then I would do it in my IRA. If they are tax efficient, then I would do it outside my IRA. So, I can't figure out how to proceed until I know more about the taxes involved.
If all goes well, likely you will be rebalancing enough that tax efficiency isn't going to be what is biting your tax ankles.
"Swap structures continue to grow in prominence for index ETFs because they can be more tax efficient versus holding physical securities. In particular with Swaps, the ETFs do not incur any taxable distributions or index replication errors resulting from the buying and selling of constituent securities. The Swaps reduce both the potential tracking error of the ETFs versus their underlying index and tax consequences for investors."
Of course, you exchange firm risk for the above.
Last edited by Kbg on Mon Jan 19, 2015 5:03 pm, edited 1 time in total.
jason wrote:
If they are tax inefficient, then I would do it in my IRA. If they are tax efficient, then I would do it outside my IRA. So, I can't figure out how to proceed until I know more about the taxes involved.
If all goes well, likely you will be rebalancing enough that tax efficiency isn't going to be what is biting your tax ankles.
"Swap structures continue to grow in prominence for index ETFs because they can be more tax efficient versus holding physical securities. In particular with Swaps, the ETFs do not incur any taxable distributions or index replication errors resulting from the buying and selling of constituent securities. The Swaps reduce both the potential tracking error of the ETFs versus their underlying index and tax consequences for investors."
Of course, you exchange firm risk for the above.
Thank you. Does anyone have any suggestions regarding how rebalancing should be done? It's going to be a 33/33/33 (UPRO/UGLD/TMF) portfolio. What rebalancing bands should I use? Should I just rebalanced annually?
Last edited by jason on Mon Jan 19, 2015 5:59 pm, edited 1 time in total.
Yup, that's how I plan on rebalancing it, after 1 year so its long-term capital gains.
I'm doing the 3x ETF portfolio in my taxable brokerage account. Just a small VP to see how it goes this year. So far I'm very impressed.
Yup, that's how I plan on rebalancing it, after 1 year so its long-term capital gains.
I'm doing the 3x ETF portfolio in my taxable brokerage account. Just a small VP to see how it goes this year. So far I'm very impressed.
This one is tough for me and I really need to solidify my approach. (Yeah I know, should have done that before now)...here is the conundrum.
- Annually backtests well and you push off into LT cap gains by going annually +1.
- Catch a good move and things get really out of whack quickly. In just the 20 days of this year I am eight percentage points off now in my cash balance.
True. One approach could be to gradually buy in. For example, purchase the exact same dollar amount of all 3 three ETFS every month. Slightly more commission costs, but that kind of dollar cost averaging may smooth it out. Then you'd sell each tax lot annually+1.
One a separate note, here's a chart which shows the past 3 years of the 3x ETF, along with their corresponding 1x ETF. Notice how closely they mimic each other.
jason wrote:
My plan was to only put 2% of my overall portfolio into the 3x PP, so I am not overly worried about the risks. I happen to have 2% of my net worth in my IRA, so I was thinking about making my entire IRA a 3x PP, depending on the tax treatment of these ETFs. I looked over the prospectuses of these ETFs did not see anything describing the tax consequences.
When I run backtests at ETFReplay.com, the returns of a 3x PP are pretty amazing (33.3%/33.3%/33.3% with no cash). 86% total return over the past 3 years. I don't feel like I am chasing returns, though, because the PP generally goes up consistently, over time.
Kbg wrote:
Agree with Pugchief. Please read the literature on the Direxion and ProShares websites and the prospectuses. Also go read everything you can about why you shouldn't buy a leveraged ETF. Realize that using PP history you could possibly suffer a north of 37% DD. If after all that, you are comfortable with the decision only then employ your hard earned money. And if you do decide to invest you must not dwell on the individual components, their movements up will be elating and down extremely brutal. You may want to have a little 3x PP VP as a percent of your portfolio starting out to learn if you really can stick with it.
Once you read the prospectuses you will have answers to both your questions.
I'm confused. If you are planning on doing this entirely from within an IRA, what difference does it make what the 'tax consequences' are? You can buy and sell and reinvest or not 'til the cows come home, and the only tax concern is the ordinary income tax rate you will pay when you withdraw any money from the IRA. If it's a traditional IRA, all withdrawals are fully taxable at your marginal rate. If it's a Roth IRA, none of the withdrawals are subject to tax of any kind as long as the contributions have been in it for at least 5 years.
Shit, now I'm considering doing this only in my Roth. The rationale being that I don't have to worry about taxes when rebalancing, this is likely to be my highest growth portfolio, and Roth withdrawals have a very long time horizon.
I'm sure this means that the PP is now totally screwed. We must have massive deflation coming.
I've been doing more research on these leveraged ETFs, and I see articles online saying that you are better off shorting the inverse ETF than buying the long ETF. So, one would short TTT, SPXU and DGLD. Here is one of those articles: http://seekingalpha.com/article/2444425 ... raged-etfs
It's not really clear, but I think they are saying the decay will work in your favor, rather than against you, if you short an inverse ETF. I really don't follow how that would work, though, as I don't think the article explains it clearly. Can anyone put that in terms that I can understand?
Also, since you can't short ETFs in an IRA, this would have to be done in a taxable account.
