Gyroscopic Investing

My version of this as of this evening is at about -8% for the year.

As of Oct 9th

1xPP: -1.62%/-7.63%DD (25% SHY)

66.67% SHY and 11.11% 3x ETF equivalents: -3.07%/-11.12%

3xPP: -7.62%/-23.26%DD (25% SHY)

3x w/ XIV Twist: -8.10%/-23.01%DD (20% to SPXL, TMF, UGLD, XIV, SHY)

Commentary: Switched the order from least to most aggressive portfolio. Overall a little more draw down for all the portfolios since my last post and a bit of a comeback as well. Perhaps with the "good part" of the investing year ahead of us we shall get to positive territory. For comparison a 100% SPY has returned somewhere between -.61% and -2.1% for the year and had a max DD of -14.7%. So for the hand wringers out there, what and how is the PP not performing to its historical record? Roughly on order of stock market returns with half the volatility. About as "typical" as the PP can get 2015 has been. Would it be nice to be up, yes. My guess for the end of the year: Unless gold does something really surprising the PP will be positive for the year, but not as good as a 100% stock portfolio. It will be interesting to see what happens vis-a-vis a 60/40 port by year end.

My personal version is -5.56% for the year with a high in late Jan of 7.09% and a low in late Sep of -8.51%.

No thank you.

Risk Parity PP: -1.12% CAGR, -6.10% MaxDD

MG,

What weights do you use and how do you compute them?

Kriegsspiel wrote: No thank you.

I didn't start the thread, I just post results of something similar to what started it. However, I personally believe a leveraged version equivalent to a normal version (way more cash with leveraged ETFs at lower percentages) is both safer and will perform better over time. But then again, I don't believe hard metal gold is any better than/superior to the ETF variety.

I also post here so I don't have to debate the merits of what I post's adherence to PP purity...it's a VP as HB would have defined the term.

Kbg wrote: What weights do you use and how do you compute them?

25% Stock
35% T-Bonds
20% Gold
20% T-Bills

via mean variance optimization back to 1968 delevered to Browne PP volatility.  MaxDD was -18.xx%.

As of Oct 30th

1) 1xPP: -1.24%/-7.63%DD (25% SHY)

2) 66.67% SHY and 11.11% 3x ETF equivalents: -2.52%/-11.12%

3) 3xPP: -6.21%/-23.26%DD (25% SHY)

4) 3x w/ XIV Twist: -4.96%/-23.01%DD (20% to SPXL, TMF, UGLD, XIV, SHY)

Commentary: None.

My personal version is -4.54%, net of all costs for the year with a high in late Jan of 7.09% and a low in late Sep of -8.51%.

KPG, I am about -4% since April, and using TMF, UGLD, SPXL/EDC, with no cash.  For a "risky" leverage, it's not doing to badly.
I have been putting in money every month and allocating based on whichever fund is down, which has helped.

internationalcanuck wrote: KPG, I am about -4% since April, and using TMF, UGLD, SPXL/EDC, with no cash.  For a "risky" leverage, it's not doing to badly.
I have been putting in money every month and allocating based on whichever fund is down, which has helped.

Hopefully there have been no surprises for you, that is the main thing! I have deliberately placed this in the VP section in deference to accurate PP terminology. However, after having done quite a bit of backtesting and looking at the math of 3x ETFs I'm reasonably confident that version 2 will outperform a regular PP. I also like the fact that you know your max DD is capped at 33% which I don't ever see happening (gold, equities and LTTs all have to decline a lot simultaneously). One does need to catch a good trend though and unfortunately 2015 has seen few of those and the ones it has seen have been down.

Overall there have been no surprises for this method in 2015. Unfortunately, it has also has been just about the worst environment for a leveraged PP.

One small suggestion...spend some time looking into how a trend will impact 3x ETF performance and what a non-trend will do to it. As a general rule for this version of a PP VP I recommend fewer transactions rather than more. You want to let natural market forces drive the various instruments to extremes and then act. All my backtesting indicates more rebalancing hurts more than it helps. My personal implementation is one annual rebalance except for the XIV component which I use and then I only rebalance that with the SHY component and leave the rest alone.

Kbg wrote: One small suggestion...spend some time looking into how a trend will impact 3x ETF performance and what a non-trend will do to it. As a general rule for this version of a PP VP I recommend fewer transactions rather than more. You want to let natural market forces drive the various instruments to extremes and then act. All my backtesting indicates more rebalancing hurts more than it helps. My personal implementation is one annual rebalance except for the XIV component which I use and then I only rebalance that with the SHY component and leave the rest alone.

Thanks KBG, you are right about the re-balancing.  As for trends, I do understand leveraged ETFs perform well in trending market, and not well a chopping market due to the "Decay" factor.  But this would happen with a non-leveraged ETF anyways, except this is accelerated due to the leveraging.  The people calling this decay dangerous is based on theory, of for example alternating periods of +/-10% market movements, however I have seen the more in depth research that some investors have done, and looked at real world volatility, and while the leveraged ETF did decay more than the non-leveraged, it was minor and certainly not catastrophic.  Certainly with the leveraged ETFs I have seen quick price swings on the order of 15% within a few days.  But then things settle down.

Kbg wrote: My personal implementation is one annual rebalance except for the XIV component which I use and then I only rebalance that with the SHY component and leave the rest alone.

#1 Can I ask what your rules are for re-balancing the XIV component? (How much does XIV have to go up (or down?) before you take profit?
#2 Do you re-balance between SHY and XIV both ways?

ILoveMoney wrote:

Kbg wrote: My personal implementation is one annual rebalance except for the XIV component which I use and then I only rebalance that with the SHY component and leave the rest alone.

#1 Can I ask what your rules are for re-balancing the XIV component? (How much does XIV have to go up (or down?) before you take profit?
#2 Do you re-balance between SHY and XIV both ways?

1. I suggest somewhere between 25-40% increase/decrease. I do 40%. This range is based on an average XIV pop/drop. Of course an XIV pop/drop can be way better or worse but this range is a good ball park.

2. Yes.

As of Nov 30th

1) 1xPP: -2.50%/-7.63%DD (25% SHY)

2) 66.67% SHY and 11.11% 3x ETF equivalents: -4.89%/-11.12%

3) 3xPP: -11.89%/-23.26%DD (25% SHY)

4) 3x w/ XIV Twist: -10.81%/-23.01%DD (20% to SPXL, TMF, UGLD, XIV, SHY)

Commentary: None.

My personal version is -7.07%, net of all costs for the year with a high in late Jan of 7.09% and a low in late Sep of -8.51%. With today's nice little day of everything being up back to -5.8% for the year. :-)

Kbg wrote: As of Nov 30th

1) 1xPP: -2.50%/-7.63%DD (25% SHY)

2) 66.67% SHY and 11.11% 3x ETF equivalents: -4.89%/-11.12%

3) 3xPP: -11.89%/-23.26%DD (25% SHY)

4) 3x w/ XIV Twist: -10.81%/-23.01%DD (20% to SPXL, TMF, UGLD, XIV, SHY)

Commentary: None.

My personal version is -7.07%, net of all costs for the year with a high in late Jan of 7.09% and a low in late Sep of -8.51%. With today's nice little day of everything being up back to -5.8% for the year. :-)

Why don't you have a 16.67%x 3 + 50% shy (i.e., equal weight or is that #3?)

dragoncar wrote: Why don't you have a 16.67%x 3 + 50% shy (i.e., equal weight or is that #3?)

Don't you mean a 8.3% x 3 + 75% SHY?

Obviously this should be done on a risk parity basis not just leverage.  #3 is actually risk whack!

MG...each of these are supposed to be more aggressive than a standard PP. For the one you mentioned the thought was to recreate/simulate a 100% "non-cash" PP (33.33% stock, gold, LTTs). And, I agree with the risk whack comment. More like bordering on insanity. :-)

DC...I could. Maybe next year. However, what you suggest is pretty close to what I'm doing though I had a pinch of XIV to mine. If you are interested in the 16.67% version the performance is slightly worse than what I post for my actual/live performance. For next year I'm considering altering the amount of XIV by upping it just a bit. Haven't decided yet how I will adjust the other components yet.

If folks have an interest in alternate weightings please post and I will consider doing them next year. If I don't forget, I'll try to add them for an end of year summary.

I got curious about the 8.3%/75% SHY version. -3.58%/7.99% DD thus far this year. My WAG is that the extra percent negative has a good deal to do with volatility drag on SPXL and TMF for most of the year and for all of them around mid year where there was a whole lot of nothing going on.

I'm almost loathe to say it, but lets see what a 3x Risk Parity version would do.  Thrice of everything, inclyuding risk?  Still, it would be fantastic to get 27% CAGR with an all weather portfolio!

MachineGhost wrote:

dragoncar wrote: Why don't you have a 16.67%x 3 + 50% shy (i.e., equal weight or is that #3?)

Don't you mean a 8.3% x 3 + 75% SHY?

Obviously this should be done on a risk parity basis not just leverage.  #3 is actually risk whack!

I'm thinking:

16.6-% 3x gld = 50% exposure
16.6-% 3x spy = 50% exposure
16.6-% 3x tlt = 50% exposure
50% shy = 50% exposure

total exposure = 200% PP.  At least that's what I'm doing in my small VP.

You propose:

8.3-% 3x gld = 25% exposure
8.3-% 3x spy = 25% exposure
8.3-% 3x tlt = 25% exposure
75% shy = 75% exposure

That means SHY is still half your exposure, unless I'm misunderstanding you here

MachineGhost wrote: I'm almost loathe to say it, but lets see what a 3x Risk Parity version would do.  Thrice of everything, inclyuding risk?  Still, it would be fantastic to get 27% CAGR with an all weather portfolio!

Do it, do it do it!

dragoncar wrote: I'm thinking:

16.6-% 3x gld = 50% exposure
16.6-% 3x spy = 50% exposure
16.6-% 3x tlt = 50% exposure
50% shy = 50% exposure

total exposure = 200% PP.  At least that's what I'm doing in my small VP.

Subtract 5% from cash, add it to XIV and that is my VP and what I report on. Also considering switching to a momentum based version.

Kbg wrote: Do it, do it do it!

Just use the Browne Risk Parity weights with the 3x ETF's. <wipes sweat>

Although it probably does need its own risk parity optimization but I know the history on the 3x's is nowhere long enough to be stable, so its pointless to try.  Heck, even a naive risk parity would probably be an improvement, so here's that:

SPXL 35.74%
TMF 34.37%
UGLD 29.99%

This has reduced the portfolio volatility from 18.34% (equal weight) to 17.36% (naive risk parity).

Assuming a 2.5 leverage factor, looks like 23.65% CAGR since 1968 and -62.125% MaxDD.  Whoo boy!  Of course, in reality each rebalancing period would have new weights to correct any misbehaving assets rather than the above fixed.

It might be possible to tolerate this risk if you do your "savings" correctly as I mentioned elsewhere.  Around 25% CAGR a year is a pipe dream to 99.99% of people and yet investors take similar risks already with super-majority amounts invested into the S&P 500 or momentum or value.  The difference is an all weather portfolio is, well, all weather!  You know you won't ever go to -100% like you can in any single asset class. 

Dang, now I'm really tempted to kick things up a notch here...  In theory, one could weight a 3x portfolio with the same maximum drawdown risk as the S&P 500 but with several multiples of the return, though a more sane approach may be to mimick the risk of a more "realistic" cult Boglehead 60/40 or 50/50 portfolio.

On your last para...yep. That's why I showed up here. I was attracted to PP not for the performance but for its lack of volatility combined with OK performance.

MachineGhost wrote: It might be possible to tolerate this risk if you do your "savings" correctly as I mentioned elsewhere.  Around 25% CAGR a year is a pipe dream to 99.99% of people and yet investors take similar risks already with super-majority amounts invested into the S&P 500 or momentum or value. 

And it is very tax efficient, too. There is no single asset class or stock selection strategy that can reach that level after taxes. But of course it can be a bumby ride at times. I have 25% of my portfolio in it to have enough free capital for my stock ideas and hedges.