20% annual returns over 40 years...interested?

[Cortopassi]

Good example.

What I tried once and made minimal money, not worth the effort, was to short two competing instruments. So you could short DUST and NUGT and take advantage of the decay. But many of these are not available for shorting, or if they are have shorting fees/expenses, so those costs ate into profits.

I did it with UCO and SCO. And found I could make a little. but eventually they get unbalanced, so you still have to watch too closely for my taste.

AND, you also need to start out when the funds are fresh or split to make them more even. Doing this strategy now with DUST at $5 and NUGT at $154, well, they are already severely unbalanced.

Go to the casino or track instead, at least you'll enjoy yourself for a while as you are losing money.

[mukramesh]

Hasn't kbg already shown that for trending non/inversely correlated assets, the decay actually works in your favor? I'd be most worried when all PP assets are in a sideways market.

[iwealth]

Hasn't kbg already shown that for trending non/inversely correlated assets, the decay actually works in your favor? I'd be most worried when all PP assets are in a sideways market.

The problem is that these correlations are not consistent and positive correlations between assets can last for extended periods of time.

I can assure you that there will be extended periods where 2 of 3 or all 3 volatile asset classes rise and fall at the same time.

[Kbg]

Cartopassi, interesting, but not really relevant. A totally different asset. Again, for like the 1000th time, it is the portfolio composition that matters most. Not, repeat not for the 1000th time, the individual assets.

But hey just for fun...let's toss UGLD and go with NUGT as our gold element. In other words, 16.66% goes to NUGT every year instead of UGLD

1/1/12 to yesterday

UGLD: CAGR 9.21/MDD -16.62
NUGT: 15.54/-31.92

2012,13,15 not impressive...this year the NUGT port is up 98.9%. Having noted this, the portfolio performance CAGR was -.5 on 12/31/16. I doubt most people would hang for that kind of performance.

CAGR of NUGT Port based on start date and ending on 12/13/15
2011: 1.95
2012: -.50
2013: -2.84
2014: -2.56
2015: -16.85

PPs 2012-2016 CAGR/TotRtn/MaxDD

1x 4.80/24.20/-7.97
3x 9.21/50.29/-16.62
NUGT 15.54/95.00/-31.92

And my personal favorite: 75% cash and 8.33% to the rest...5.08/25.74/-8.47. I have zero doubt in my mind this version is a better and safer version than the original. It is literally impossible for this portfolio to have a worse annual return than the PP's full backtest history if rebalanced annually. Not to mention 75% could be sitting in STTs.

But to be clear, I've never said this (50%/16.66%) was a safe leveraged version of the PP. I do believe it has a good chance of returning stock like returns with less Max DD. Using PortViz data I have a rolling 10 year (1971 start date) return max of 26.41, avg of 18.16, and min of 12.84. As of 12/31/15 the rolling 10 year return was 13.83%. The worst annual DD was -12.3%. But of course this is not intraday MaxDD. I had a ~18% DD in 2015.

This compares with a 100% stock port that had a max 10 year rolling of 16.90, an average of 9.16, and a min of -1% whose max annual DD was 38.5. Intraday was at least 57%. If we do a 60 stock/40 10Y Ts the rolling stats are 16.46/10.77/3.16. The worst annual Max DD was -14.74%.

Now tell me, objectively, which is the better portfolio of the three?

[iwealth]

Good example.
AND, you also need to start out when the funds are fresh or split to make them more even. Doing this strategy now with DUST at $5 and NUGT at $154, well, they are already severely unbalanced.

Due to the nature of these instruments, this should not matter as long as you buy equivalent monetary amounts.

[Kbg]

The problem is that these correlations are not consistent and positive correlations between assets can last for extended periods of time.

I can assure you that there will be extended periods where 2 of 3 or all 3 volatile asset classes rise and fall at the same time.

Correlation has nothing to do with the decay factor which absolutely is real.

The issue, above all, for decay is trendiness. All three of the PP assets trend well which is why they are in the portfolio in the first place. For the same reason HB stuck them in the port, we are counting on the ability to trend well. I'm not going to post this again (though I just did), but if you look at the actual data vs. the crap you read somewhere you will see that with this particular portfolio the trendiness is such that it can overcome the decay. Read carefully...I'm not discounting decay. I'm saying these assets can overcome/make up for it when they trend. On average, the real portfolio has come very close to it's advertised 2x leverage factor. The data also indicates we ARE getting nicked a bit for decay, but it's quite minor.

I've also mentioned this before, but if one is actually going to attempt this they really should take the eye poking time required to understand the math behind these ETFs and what it means from a practical stand point. It's on the internet and at the ETF sponsoring company websites. If one does this, they will understand why a short 3x ETF and shorting these ETFs in general is not a particularly good idea.

But folks, do me/us all a favor...don't knock this portfolio for attributes you praise on other parts of the board. Leverage, decay, derivatives, fair game. Negative/low correlations, volatility, all-weather attributes...isn't that why you are here and not so much? If the latter are your concerns then the standard PP isn't good for you either.

[Cortopassi]

Cartopassi, interesting, but not really relevant. A totally different asset. Again, for like the 1000th time, it is the portfolio composition that matters most. Not, repeat not for the 1000th time, the individual assets.
Now tell me, objectively, which is the better portfolio of the three?

And from before: If one can't keep their eyes off a component's bottom line vs. the portfolio's bottom line, then absolutely do not put a dime in this.

----------------

I am in the bolded boat. That would kill me. I love to see the day to day interplay of stocks vs. bonds vs. gold. But to watch one or more of those 3x ETFs go to zero while the other zoomed is not for me, regardless if the bottom line is higher.

I appreciate all the math and work you've put into this. 15 years ago, I probably would have tried it. Not now though.

[stuper1]

Kbg,

I apologize, because I think I've asked this before, but could you please tell me again. If someone wanted to run one of these portfolios, like the 75/8/8/8 or the 50/17/17/17, how often would they need to check in on things and potentially do a re-balance? Is this something you have to check on weekly or monthly? Or is it more like the PP where you don't have to worry even if you only check in every year or so?

Also, can you give us a feel for how many rebalancing transactions per year might be required, so we can have a feel for potential transaction costs?

[Kbg]

With regard to decay...as board members may have noticed I mentioned in an earlier post that I'm paper tracking a futures based version of the portfolio. This will effectively eliminate decay as an issue but requires more work to maintain. Currently, this port is up over 2% since mid-July on a min size port of around $48K. I am not accounting for any returns from the pure cash balance in this study. The one issue I've ran into is the MGC mini gold future. Supposedly it expires on the 29th, but it appears to effectively have stopped trading on the 1st.

Anyone ever traded this future? I'm confused.

[Kbg]

Kbg,

I apologize, because I think I've asked this before, but could you please tell me again. If someone wanted to run one of these portfolios, like the 75/8/8/8 or the 50/17/17/17, how often would they need to check in on things and potentially do a re-balance? Is this something you have to check on weekly or monthly? Or is it more like the PP where you don't have to worry even if you only check in every year or so?

Also, can you give us a feel for how many rebalancing transactions per year might be required, so we can have a feel for potential transaction costs?

On the first, I'd spend the same amount of time as you would the standard version. You could pretty much forget about it if that is your style. Personally, I would rebalance annually as that enables you to monetize the volatility of the assets. If you use bands to rebalance then assuming you are going to do the standard 15-35 then of course that means 5 and 11.67.

With the second one that is a little more difficult. I actively manage it and rebalance on major dives in an asset. Whether that ends up being a good idea or not, IDK. Last year it would have and this year thus far it has been a good idea. However, if something is trending down hard for a good fundamental reason I'd probably wait to rebalance...but I'm more of an active vice passive investor. If you are passive, then once a year. Once a year is what I post the stats on for performance and DD.

Please realize, if you go the more aggressive route you are in fact investing/trading a leveraged portfolio with the good and evil that comes with it. A tight money situation is going to suck even worse for this portfolio than a standard PP. Also, as I've posted a couple of times if you can't handle an asset going to zero this is MOST DEFINITELY NOT for you. The history shows this is a relatively frequent occurrence. So the standard this is for information/education purposes only; I am not a registered advisor and investors should seek professional assistance; Historical returns are no guarantee of future returns; apply.

Also, realize you can dial the cash to leverage at whatever level is comfortable/appropriate to your situation. I categorically do not recommend anything above 2x and I heartily recommend a large cash balance for volatility dampening and a pool of reserves for buying. If one wants to dial in leverage more specifically to their risk tolerance then the numbers are as follows:

Lev % 3xETF
1.00 - 8.33
1.25 - 10.42
1.33 - 11.11
1.50 - 12.5
1.67 - 13.92
1.75 - 14.59
2.00 - 16.67

2012-2016 Performance in order of the above
5.12/-8.48
6.20/-10.57
6.55/-11.27
7.25/-12.62
7.96/-14.00
8.29/-14.63
9.29/-16.60

Standard PP 4.84/-7.97

[Kbg]

I'll be the first to say this is not a PP, just a PP inspired VP. I wish I could fast forward 5 more years to see how the 75% cash version does. The end of 2016 will be 5 full years.

[mukramesh]

kbg, does your personal mix still include XIV? If so, can you describe how much what this is supposed to add to the leveraged PP?
I'm sure you've already explained this but I am having a hard time finding the post in this thread...

[Kbg]

M,

I have 5% of my personal mix in XIV and it comes out of the 50% cash. So cash is actually at 45%. Other options would be to nick the other three 1.67% or take it from the stock allocation. The latter would be most apples to apples but really XIV is a completely different risk premium (carry) as it is effectively continuous put selling.

I really debated pulling it from the port as most likely the short VIX selling type of strategies have seen their best days. The market in VIX futures has become WAY more efficient than the early days. XIV/SVXY are highly correlated with stocks and the leverage is roughly around a 3x ETF so one could make some very sound arguments that there isn't much value added over SPXL/UPRO. I would not debate that viewpoint strenuously. I have a small piece of it for the carry premium diversity and because it does very well in a flat choppy market (like now) whereas nothing else really does. So it gives just a tiny bit of diversification. Another more mellow option is ZIV or somewhere in between with both XIV and ZIV.

Note: XIV is highly risky stuff in a market that continues to evolve, be careful out there.

[mukramesh]

My understanding is that in a flat market, XIV will do very well. But in a situation where stocks are dropping, won't holding XIV magnify your losses?

[Kbg]

Yes indeed. I think the simulated Max DD is like 95%. On the flip side it can do 300-600% multi-year runs afterward. One fallacy people often succumb to is looking at the scary figures/internet warnings that say if you go down 95% it will take a 900% rise to recover. (Or whatever the math is) This mathmatically is a hard truth but the real question is, is that all of your money or do you have a means to recharge the shot?

If no, you're screwed. If yes, then a 2-6x return on the initial loss over the next couple of years makes it a whole different equation...but you must have the spare cash to recharge. Lose a $100 make it back 50 years later not so great. Lose 100 make 300 a couple of years later, that's a worthy bet.

Obviously the rebound returns are critical to the scheme. Losing 95% and having follow on years of 10/20% isn't going to cut it.

This is NOT a port to do without a nice cash cushion. It's an essential element.

[Kbg]

I've been monitoring the hedged futures version of the PP for a month plus now and for those who are comfortable with and understand futures, I would have no problem recommending this approach in lieu of 3x ETF versions. Indeed, there is a very high probability I may switch to it and ditch the 3x ETF version. Costs will be substantially lower, volatility decay will not be an issue (but that could be bad in a good trend) and tax treatment is vastly superior in a taxable account. Some quick details on the paper portfolio:

My overnight margin is $6300 which is carrying a portfolio value of $49,400 (numbers rounded)

Ticker symbols are...

Gold - MGC

Stocks - Long ES/Short YM

LTTs - Long UB/Short ZB

Precision balancing can be accomplished with GLD, SPY, DIA and TLT or one of their leveraged versions.

Assuming the max historical DD of the PP plus a little fudge factor, a minimum of $15K cash would be needed for this size portfolio. Notch cash up to $24.7K for a 2x VPPP.

The above combination was chosen to determine the minimum size portfolio practicable for doing a leveraged VPPP with futures but there are certainly many other possibilities and combinations that may be more suitable depending on individual circumstances and portfolio size. Additionally, there will be tracking error due to hedging as a way to enable a smaller portfolio. If you have the chops for a much larger portfolio the hedging aspect could be eliminated. See my post a while back on portfolio size using futures. Short version: UB is the long pole in the tent due to contract size. On the positive side, hedging directly will also substantially limit the likelihood of getting blown out of the water during a market spike caused by black swan or grey swan kind of events. If one wants to sleep better at night, I'd definitely hedge but of course that costs a bit more in commission and spreads.

Bottom line: A viable approach with some nice advantages. More work and a good understanding of futures required to pull it off in practical terms.

I will not be posting on this version any longer as the learning objective is completed.

[Cortopassi]

My overnight margin is $6300 which is carrying a portfolio value of $49,400 (numbers rounded)

As one who doesn't understand futures, does this mean you are leveraged 7.8x?

So if/when Yellen comes and opens her mouth and gold, for example, dumps $100 instantly, what happens? Margin call, and then are you at risk for anything not able to be gotten out of fast enough or is all that automated?

I know this is the variable portfolio topic, but I think it should be changed to the balls of steel speculation topic.

If I have this all wrong and there is little risk, then credit all the anecdotal stories I've heard of people getting wiped out in futures. Like those two old guys in Trading Places.

[clacy]

My overnight margin is $6300 which is carrying a portfolio value of $49,400 (numbers rounded)

As one who doesn't understand futures, does this mean you are leveraged 7.8x?

So if/when Yellen comes and opens her mouth and gold, for example, dumps $100 instantly, what happens? Margin call, and then are you at risk for anything not able to be gotten out of fast enough or is all that automated?

I know this is the variable portfolio topic, but I think it should be changed to the balls of steel speculation topic.

If I have this all wrong and there is little risk, then credit all the anecdotal stories I've heard of people getting wiped out in futures. Like those two old guys in Trading Places.

Margin is really nothing more than an exchange/broker minimum nominal threshold for holding a futures contract(s). So without checking the numbers, if $6.3k in margin gives you control of a PP worth approximately $49.4k, then yes that would be correct. However Kbg is not advocating for adding contracts for each $6.3k in ones account.

You can essentially dial in the leverage to suit your needs.

For example, someone with $50k in investable assets, under Kbg's plan you could run use $6.3k in margin to buy those futures, and use the other $43k to place in T-bills. In that scenario you would only be leveraged 1x as a total portfolio.

[dragoncar]

I understand futures in general but have no experience with them. I'd be interested in seeing a play by play for any paper trades you do to understand the mechanics involved

[iwealth]

and tax treatment is vastly superior in a taxable account

This wouldn't necessarily be true if you only rebalance yearly which I believe you've advocated if running a VP with the 3x ETFs...right?

[Kbg]

MGC is the micro gold contract and = to 10 oz, so right now $13K and change per contract. Let's say gold dumps a $100 instantaneously then with this contract you would need at least 1000 in cash and if it is in your account exactly nothing happens. Futures are marked to market daily which means if you have a losing day your brokerage takes out whatever you lost from your account in cash at the end of the day. If you have a winning day the opposite occurs. Additionally, overnight margin/contract margin is set for extreme moves and how it is established is using world class risk analysis algorithms. So at $6300 per 49K that means you are geared for almost a 13% negative intraday move.

If you step back and think about it there is nothing different going on with your account. If you are holding equivalent GLD, real gold or MGC and gold dumps $100 then you are down $1000 period. The only difference with a futures contract is $1000 is going to be taken away at the end of the day from your cash balance...but your EOD overall account balance will be exactly the same.
From a pure monetary stand point there is exactly zero difference between 10oz of gold in a futures contract, an ETF or real gold. Obviously the three have different instrument characteristics.

Futures get a bad wrap due to stupidity but assuming the cash value of the contract is fully backed with that amount of cash they are no more or less dangerous than the physical or ETF version. We can have a debate about the merits of physical vs. a future and one would get no real argument from me on the physical. With regard to the ETF versions I'd argue futures are at least as safe and probably more safe. With regard to cost, there is simply no comparison. The future will be much cheaper,to hold.

[Kbg]

and tax treatment is vastly superior in a taxable account

This wouldn't necessarily be true if you only rebalance yearly which I believe you've advocated if running a VP with the 3x ETFs...right?

Correct for stock and LTTs...physical gold is different of course and someone would need to dig into that one.

In a taxable account ST and LT gains are taxed according to holding length. Futures are always 40ST/60LT regardless of hold length. Hands down futures are my favorite asset class at tax time. You get a form from your broker, it has a gain or loss amount on one line, the 40/60 split on two other lines and that's it. All that hold length and wash sell tracking crap you have to do and report trade by trade is completely N/A. Tax loss harvesting also can't be done as there is no reason for it. The down side, and it is a significant one, is that like your daily brokerage account mark to market, on the 31st of December each year the chalk line will be snapped and if you are up for the year you are going to get taxed, if you are down the loss will be realized as well.

Good question...forced out a good issue to consider.

[Kbg]

For a 100K portfolio...

2.5x means 250K equivalent/25% each to SHY, TMF, UGLD, and SPXL/25K SHY and 225K (75K x3) equivalent everything else

2x means 200K equivalent/16.667 each to the above and 50% to SHY/50K SHY and 150K (50K x3) equivalent everything else

1.5x means 150K equivalent/8.333% each to the above and 75% to SHY/75K SHY and 75K (25K x3) equivalent everything else

1x means 100K/25% each to SHY, TLT, GLD, SPY

Purchase price was at the close 12/31/15 through the close on 8/31/16

2.5x = 36.70%/-5.79%DD

2x = 24.80%/-4.36%DD

1.5x = 12.94%/-2.59%DD

1x = 12.28%/-1.63%DD

Personal mix = 34.82%

Comment: August was a modestly down month. For my personal port there was an XIV rebalance and I'm now above my tracking bogey by 7.7% this year. Unless something goes crazy in the last hour of today's market, today will mark another new yearly high in the portfolio. Such a nice change from 2015 which pretty much sucked.

In watching my paper trading futures version of this, I think a good rule would be to roll the futures a month before expiration. There was an automatic force out of a position in UB due to not meeting IB's check on my ability to deliver actual LT treasuries and the MGC future and how it trades is still a bit of a mystery to me. Definitely more work to do in the futures version of this. Even if one is experienced in futures I recommend paper trading in your brokerage's paper trading account first to get a feel for the daily changes and mechanics of it.

Finally, while I'm trading a variation of the 2x version my favorite is still the 1.5x version. I predict here and now it is going to outperform the standard PP when/if interest rates begin to rise assuming one is in STTs for cash. It will receive the benefit of 3x the cash over a traditional PP.

[clacy]

I may have misunderstood you, but shouldn't a 1.5x portfolio be 12.5% each SPXL/TMF/UGLD (with 62.5% SHY)?

That portfolio is up 19.9% YTD.

[Kbg]

This question came up at the end of last year which is why I explicitly post the math of the leverage computations. You can leverage the non cash component (what you did) or look at the total notional portfolio value from the base point of 100K as the leverage (how I'm tracking).

Using your numbers my method, the port is a 1.75x as the total notional value is 175K. In your method the active component is levered up to 1.5x as you note. Regardless, everyone should understand the results/risk are going to be driven by the amount of leverage employed to the active risk components.

There probably is a "correct" method for this and I may not be using it, but so long as everyone can see the math with transparency that is what matters most. In any event, your broker is going to report performance stats based on the starting value of the portfolio and not whatever we consider our portfolio to be leveraged to. All of the numbers I post are based on a 100K starting point. So assuming my starting point this year was 100K my personal port is now valued at $134820 for example. Percent performance would be 69.6% if I was only counting the risk assets...which doesn't make intuitive sense to me at a portfolio level.

Either way works, but for the reason noted in the last two sentences of the above para, I'm going to stick with the current method.