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

[Kbg]

Had the opportunity to review the thread from its beginning and there was a lot of discussion about volatility decay. So let's check out almost 4 years of performance using SHY, DGP, UBT, SSO and their 1x counterparts SHY, GLD, TLT and SPY.

Since 1/1/11

Annual Rebalance
1x 19.41/-8.90 DD
2x 50.07/-15.03 DD

My oh my where did that volatility decay go? Are my eyes seeing correctly we have a 2.57x leverage gain on the upside instead of 2.00? What the?? And it gets better...we only lost 1.68x instead of 2x on the downside. Sweet mother of pearl what is going on???

Turns out volatility and leverage are funny things and what is often printed isn't entirely correct/complete. Fact...as posted earlier in the thread in a sideways market you can turn a zero loss into a real one due to volatility decay. Fact...once you get beyond about 2.2x leverage in most cases more isn't going to have the desired reciprocal benefit gained up to that point. (In actuality using stock prices it's more like in the range of 1.7-2.2x as a ballpark of when leverage begins to go bad.)

Now for the good part even when it is bad...that same decay (well the math of leverage more accurately) does some funky things with things that trend. If we start out with two assets and one heads north and the other heads south and they both gain/lose the same amount unleveraged, we actually gain money because the downward bad decay becomes smaller every day while the upward OK decay gets larger every day. Kaching...bank the difference. Unleveraged...we make zero, leveraged we make money. Well that's kinda cool don't cha think? And more to the point, this is why our leveraged 2xPP is doing more than 2x better than the 1x on both the gains and the loss side.

So how does one make this kind of magic?

1. Find a portfolio composed of really volatile non-correlated assets that trend pretty dang good up and down (thank you Harry...Craig and Tex too)
2. Add leverage...daily reset actually helps
3. Rebalance to capture it...the more volatile the more often you should rebalance

But never forget...leverage is a knife so let's all keep an eye out for tight money recessions.

[Kbg]

And a follow up to that fantastic last post...

But kbg, I'm afraid of leverage, everyone says it's bad.

Kbg: It is bad when used inappropriately. However, one could do better than a PP by simply using leveraged ETFs at half the weight. A 12.5% allocation to SSO, UBT and DGP = 25% and 7.5% DD over almost four years plus whatever interest you got from cash that is now 62.5% of your portfolio.

Kbg: But what about counter party risk, the ETF issuers could go under and I could lose all my investment? Maybe, maybe not. Read the prospectus. The really long boring one(s). But let's go there for discussion sake. Armageddon happens. In a normal PP you have your gold stashed somewhere. So let's assume that is safe and poof your gold ETF is no more. Where is your cash, bonds and stocks? Cash is a wash. Depending on where you put it, it either is or isn't there. But let's assume it is safe. No leverage 25% percent of your portfolio is safe, while leveraged 62.5% of your portfolio is safe. So under a normal PP 50% is good to go and under a leveraged one 62.5% is good to go. Now we are down to stocks...I think this one is a wash. If stocks survive S&P500 futures and swaps survive or it all crashes down together. So let's throw out stocks as a wash (loss) and we are now isolated down to a 50% mix of cash and gold vs. 62.5% in cash. If gold spikes, chalk one up for the original version if it goes up over 50% from the start of Armageddon to get the combined port of gold and cash up to 62.5% of cash's starting point.

Personally I've never been an Armageddon guy by nature...well actually I am a tiny bit but my version of Armageddon is not gold, but rather my own land, guns, house and food. You know, real tangible things. If we have Armageddon and that gold isn't buried in the back yard, how are you going to get it? I'm thinking the banks are probably closed for quite a while. The stuff in Canada, Australia and Switzerland. How are you getting there to claim it? Thus, I personally don't think gold is going to help all that much in true Armageddon.

And short of that...well, let's assume the world continues to work normally which is what I do...with my land, guns house and food as my true hedge.


P.S. Actually the reason why I wrote the last two posts is to encourage people to really look into their assumptions. Most folks don't even know when they've made one. Generally this is not a good recipe for both investing and life.

[Pointedstick]

This really makes me want to try it out in a VP! However, I worry that all the buying and selling required for the very frequent rebalancing will incur transaction costs that cut the returns down a lot.

[Pointedstick]

FWIW, I'm totally with you on the subject of armageddon. If things really go to shit, our investment portfolios are probably going to be the least of our worries.

[Kbg]

This really makes me want to try it out in a VP! However, I worry that all the buying and selling required for the very frequent rebalancing will incur transaction costs that cut the returns down a lot.

In a tax deferred account quarterly is a good hack, while annually is almost as good and if you throw in taxes annually definitely wins. If there is a serious melt down or up, an opportunistic rebalance is probably not a bad idea either. Of course now we are moving away from the shear simplicity and elegance of a disciplined rebalancing schedule which for most is probably a loser idea.

The fact is that the best way to rebalance in the future is instrument price path dependent and I like the assumption that this is unknowable in advance .  So as is often said, just pick something and stick with it.

[MachineGhost]

This really makes me want to try it out in a VP! However, I worry that all the buying and selling required for the very frequent rebalancing will incur transaction costs that cut the returns down a lot.

I found myself in this position with 3x PP a couple of weeks ago.  Needed to rebalance after just a week and I wasn't willing to throw more money into it, so I eventually sold at breakeven less the commission loss.  Next time I'll plan ahead better!

The last four years is nothing for what will be at least double and triple the MaxDD of the 1x PP.  You will need to rebalance as often as monthly or weekly when the bands are hit.

[Kbg]

This really makes me want to try it out in a VP! However, I worry that all the buying and selling required for the very frequent rebalancing will incur transaction costs that cut the returns down a lot.

I found myself in this position with 3x PP a couple of weeks ago.  Needed to rebalance after just a week and I wasn't willing to throw more money into it, so I eventually sold at breakeven less the commission loss.  Next time I'll plan ahead better!

The last four years is nothing for what will be at least double and triple the MaxDD of the 1x PP.  You will need to rebalance as often as monthly or weekly when the bands are hit.

What are you using for bands?

[MachineGhost]

What are you using for bands?

I was using +/- 5%.

[Kbg]

With 1 mo to go...

Assets: IWM, TLT, SHY, GLD (Classic PP w/small cap) and XIV (PP w/XIV Twist) and rebalanced quarterly
2x as above only 2x ETFs
3x as above only 3x ETFs

1x PP 4.30/-3.70 DD
2x PP 9.14/-6.92 DD
3x PP 13.61/-9.91 DD

1x Twist 7.41/-10.85 DD
2x Twist 11.22/-12.85 DD
3x Twist 14.72/-14.73 DD

S&P500 13.27/-7.7 DD

Just for fun...substituting SPXL for TNA in the 3x w/Twist port meant 22.83/-7.48 instead of 14.72/-14.73. TQQQ 26.83/-12.52. Next year I will stick with the S&P500 for the reports as it is the PP benchmark. For those who don't like the idea of using short volatility and may be enticed by a prudent use of leverage, to me the SHY, SSO, UBT, DGP variant is worth considering. That version has returned 14.40/- 5.01 DD this year. The 3x 21.83/-7.48.

[Reub]

You know that IWM is up 1% YTD while VTI is up 11%. That would increase PP to approximately 7% for the year so far.

Also if we had a fairly new PP and invested directly in US LTT's in the past few years, the longer duration would exceed TLT's, further increasing our gain for the year.

What would a PP consisting of VBR, SHY, GLD, and EDV look like over the past few years?

[MachineGhost]

Size effect is value effect.  There is no size per se.

[Reub]

Then there is no value either?

[MachineGhost]

There is value.  It's just concentrated in small companies.

[Kbg]

What would a PP consisting of VBR, SHY, GLD, and EDV look like over the past few years?

Rebalanced Annually

Yr      %
2008 4.4%
2009 1.6%
2010 14.8%
2011 13.9%
2012 4.7%
2013 -4.4%
2014 10.1%

CAGR: 6.31% Max DD -14.54 %

[MachineGhost]

This is the study that killed the so-called small size effect:

http://onlinelibrary.wiley.com/doi/10.1 ... 2/abstract

It's not as if large cap stocks can't ever be value; they just tend to be glamor stocks because market-cap weighting in the indexes consistently overvalues them.  It is a pseudo-momentum effect.

[Kbg]

It's not as if large cap stocks can't ever be value; they just tend to be glamor stocks because market-cap weighting in the indexes consistently overvalues them.  It is a pseudo-momentum effect.

Since we are in the VP section of this forum and not bound by PP orthodoxy here...it is pretty clear I'm a fan of the PP concept but a tinkerer on the margins. Another tweak for me in addition to a smidge of leverage is the use of momentum for the stock component. Given we do not time between allocations in a PP, I think there is probably a good case to be made for taking the top two or maybe just one of SPY, IWM, EFA and EEM for the 25% stock portion and swapping out using a momentum sort. Band rebalancing would not work as the timing mechanism and certainly one would have to factor in tax effects for anything in a taxable account; but, there is a ton of scholarship out there on momentum now. One thing I've noticed though is people tend to say if four options are available, 10 must be even better. Not the case usually. Distinct, clear segments tend to do better as you are riding fundamentals vice short term mo.

[Kbg]

Given we do not time between allocations in a PP, I think there is probably a good case to be made for taking the top two or maybe just one of SPY, IWM, EFA and EEM for the 25% stock portion and swapping out using a momentum sort.

There is likely long term data that would support as well, but since all I got for now is since 2004 which is a good roller coaster period...smooth would most likely enhance the effect. The results are...

Since 2004 a 63-252d momentum sort would add anywhere from ~2.0-7.0 CAGR rebalanced quarterly depending on what you choose for your momentum sort. Actually, when everything is down (a bear) picking the worst index (return to mean) is the best. I tend to do my stock index momentum sorts by absolute returns. Narrowing down to just SPY/IWM adds just a smidge and could entirely be noise.

[Kbg]

Going toward the wire...

Versions

UWM 12.27
UWM w/XIV 8.98

SSO 17.54
SSO w/XIV 12.99

QLD 22.89
QLD w/XIV 17.13

SPY 10.48

Real Money...17.33

[Kbg]

I believe these results are dividend adjusted...but I'm using an alpha tester database so perhaps not a 100% accurate, but certainly close. Unless specified otherwise all calculations are using ETFs and S&P 500 based ETFs for the stock component.

First Batch - w/XIV Qtrly Rebalance

1x - 6.74
2x - 13.22
3x - 19.23

Second Batch - no XIV Annual Hold

1x - 9.24
2x - SSO 19.03, QLD 24.09, UWM 15.56 (knock off roughly 6-9% for the XIV versions)
3x - 29.46

Clearly XIV was a non-helper this year.

[Kbg]

Out of curiosity I decided to backtest something to see the actual impact of adding more volatility to the PP.

From 2012-2014 Annual Rebalance

SPY, TLT, SHY, GLD - CAGR 4.34/-7.95 DD, CAGR/Max DD ratio .55, Sharpe -.04 (25/25/25/25)
SPXL, TMF, SHY, UGLD - CAGR 6.43/-9.63 DD, CAGR/Max DD ratio .66, Sharpe .2 (11/11/67/11)

Interesting, and particularly so considering 67% of the portfolio is in cash.

[modeljc]

I would love to hear from the folks that have some or all money in 2XPP or 3XPP.  I still can't find the DECAY!  If there is no decay than what is wrong with some VP or big VP with TMP, UPRO, UGLD?

Also would love to hear how many are following on Paper.

I would love to hear from Wong and when he might put some monies in a levered PP.

[modeljc]

I would love to hear from the folks that have some or all money in 2XPP or 3XPP.  I still can't find the DECAY!  If there is no decay than what is wrong with some VP or big VP with TMP, UPRO, UGLD?

Also would love to hear how many are following on Paper.

I would love to hear from Wong and when he might put some monies in a levered PP.

Sorry it is WONK! 

[dragoncar]

I would love to hear from the folks that have some or all money in 2XPP or 3XPP.  I still can't find the DECAY!  If there is no decay than what is wrong with some VP or big VP with TMP, UPRO, UGLD?

Also would love to hear how many are following on Paper.

I would love to hear from Wong and when he might put some monies in a levered PP.

I'm doing a margin PP, but obviously it's not very "legal".  Also obviously, it's been doing quite well this year!  My leverage is low, though, nowhere near 2x.  I'm basically investing my next year's savings a year early at a marginal rate of 0.6% (or thereabouts, haven't checked recently).  I figure that margin rate is much lower than the fees for the 2x ETFs.

[Kbg]

I would love to hear from the folks that have some or all money in 2XPP or 3XPP.  I still can't find the DECAY!  If there is no decay than what is wrong with some VP or big VP with TMP, UPRO, UGLD?

I would love to hear from Wong and when he might put some monies in a levered PP.

I have a good chunk of money in a 2x PP, but getting ready to change it from 2x to 3x and upping the cash amount. I'm going from the standard 25 split at 2x leverage for the non cash positions and move it to 50% cash with 16.66%% at 3x. Overall net leverage of 1.5 stays the same. I can't remember where on the board I posted it, but simply put two words - path dependency. It is a mathematical fact that if the market gets choppy decay will happen and it will be worse for 3x over 2x ETFs. It is also mathematical fact that in a trending market you will amp up the relative change. But, interestingly if you have two instruments that move in exact opposite directions in a trend the magnification of the gain will be greater than that of the loss. In fact, the loss begins to flatten out while the gain gets steeper. So, profits will be intensified and losses mitigated. The decay is happening, but we aren't seeing it because (as advertised) when the three volatile assets move, they tend to really move.

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. I personally would argue this is a safer portfolio so long as you can get past ETFs constructed of futures and swaps. At least for the two assets mentioned I think it is a net nothing in terms of additional instrument risk. Holding physical gold for the reasons mentioned by HB and Craigr...yeah, can't argue with those reasons.

[Mark Leavy]

I have about 10% of my assets in a 2x (unleveraged) PP.

50% directly held long bonds
35% physical gold
15% XIV

180 days and it seems to be performing per theory - even on a daily basis.
Still just a learning experience.  I wouldn't do this with money that I couldn't afford to lose.