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

[clacy]

Very interesting discussion going on here.

I tend to like the 2x version of the HBPP, because as Clive points out, because of the greater volatility, you're going to have more re-balances, but you still get the overall smoothness of having the 4 key asset classes represented.

I am not running a leveraged PP, but have thought about it in the past.  If you put together a portfolio with 50% cash and the other half divided between 2x gold, 2x spy and 3x LTT's is sounding appealing.

With such a portfolio, your max loss would be 50% (pretty much the entire civilized world would have to melt down to induce that scenario).  And that is still much safer than the stock market, yet could easily produce significantly greater rewards and very likely fairly smooth volatility.

[MediumTex]

Clacy,

Make sure you understand the "decay" issue with all of these leveraged ETFs.

They are not long term investments.

[clacy]

Clacy,

Make sure you understand the "decay" issue with all of these leveraged ETFs.

They are not long term investments.

Yes, I understand decay and this is the main concern frankly.  However, decay doesn't appear to be as significant as I once thought (at least for SSO, UGL, TBT).  Also, you could withstand a certain level of decay if the returns were large enough due to the extra re-bal factor.

The biggest problem with a leveraged PP would be the fact that there is little in the way of testing that can be done for levered ETF's.

[stone]

Clive, that whole point of view rests on viewing cash as being somehow "other" than all alternative stores of value. I've seen cash described as "the purest from of speculation". There is nothing to cash apart from the belief that someone will swap it for something else in the future. I just think that it is important to see cash for what it is and not imagine that it has unworldly powers. I think that what matters is viewing the "risk" as being the potential downside of the entire portfolio in terms of real day to day costs. Quite possibly we are in a secular bear market for developed economy currencies. Cash might be the most hazardous part of the PP at the moment. It certainly was for developing country economies for the 1980s and 1990s.
In principle you could recalibrate all the values for the PP assets in terms of gold such that gold always had zero volatility and cash had great volatility. If you look at cash through that looking glass it becomes harder to want so much of it. Perhaps, I'm just laboring under a "grass is greener on the other side" sentiment because my better half has put so much of our savings in cash. I also just put £15k into ILSC .

[stone]

Clive, I have my doubts about how well ILSC really do keep pace with real inflation. Think of real possible costs. Have a basket of say a university education, health insurance, cost of emigrating to Singapore or where-ever -and think what has happened over the past  decade. The denominator of the cost of university education is zero as it used to be free. The cost of health insurance is still free but for how much longer? Did the cost of university education find its way into the RPI? I think cash needs to be seen with some circumspection just like all the other assets.

[stone]

Clive, the only way I can rationalize it is as an arbitrage on the different "cash" types available to you as a retail saver and the bank concocting the 3xETF. You can get RPI+0.5% as a ILSC depositor whilst the TBTF bank can borrow at 0.25% or whatever. Basically your scheme is a way to maximize your holdings of the market beating (in terms of return/risk/liquidity) ILSC whilst still getting exposure to the volatility of stocks,LTT and gold. I guess if ILSC could be traded and held in unlimited amounts, then they would trade much above par value. Otherwise banks would borrow at 0.5% to buy ILSC and get RPI+0.5% in return. Trillions of £ would be held in ILSC if they were not only available to individual people with a £15k per issue ration.

I guess if short term rates get cranked much above inflation, then leveraged ETFs might get withdrawn but perhaps that wouldn't harm you too much.

[Wonk]

2x PP Update:

The leveraged PP hit another rebalance band today.  This is only 2 months after the first rebalance.  This time, the 2x long bond is what hit the 35%, although the 2x equity portion was nearly at 15%.  All funds were rebalanced back to 25%.  Here are the numbers since inception:

1xPP: 25.19% (no rebalance yet)

2xPP: 57.19% (2 rebalancing events)

[stone]

Clive, I'm just being nosey but is this something you have done for real since 4th Jan or is it a hypothetical?

[Storm]

I admit I haven't followed this thread too closely, however, I'm very intrigued by the notion that perhaps the 2x or 3x leveraged PP can actually boost gains over a traditional 1x PP by hitting rebalance bands more often.  It seems that you might be able to counteract the decay effect these leveraged funds have by rebalancing multiple times a year.

For example, I never hit a LTT rebalance band in my PP (although I did buy stocks at a yearly low) so I missed out on some of the upside that the leveraged PP enjoyed.

Good work everyone.

I think even a fairly conservative investor could benefit by setting up their portfolio in such a way that 50-75% was traditional 1x PP and then take the remainder and do a 2x or 3x PP.  Why not juice returns by using the PP principals, while keeping the majority of your retirement in the traditional PP?  This could replace most VP speculation, as most people would be quite pleased with 20-30% annual gains in their VP.

[Wonk]

I think even a fairly conservative investor could benefit by setting up their portfolio in such a way that 50-75% was traditional 1x PP and then take the remainder and do a 2x or 3x PP.  Why not juice returns by using the PP principals, while keeping the majority of your retirement in the traditional PP?  This could replace most VP speculation, as most people would be quite pleased with 20-30% annual gains in their VP.

That was the original idea for me.  The concept that if a 2xPP would work, it would represent the prospect of 20%+ annual returns with average volatility and no work.  That made me ask the question: what is my target return for a VP for the associated risks and time involved?

I don't think a 2xPP can match the risk/reward that a start-up business venture can present, but otherwise the 2XPP offers the potential for a solid return compared to similar VP investments such as rental real estate and stock picking.  For instance, say someone likes the 4-5% real returns of the PP, but wants more potential for growth.  You could run an 80% PP and 20% VP consisting of 2xPP funds.  The blend would likely offer another 200bps to the annual return for a total real return of 6-7% with a mitigated downside risk.  I like that idea much better than trying to buy apartment houses and manage tenants.

I don't run a VP like this at the moment, but I wouldn't rule it out in the future.  So far I haven't seen any evidence of decay when viewed as a whole (over the last 21 months).  The primary concern for me in such a portfolio is counterparty risk in a systemic crisis.  On the other hand, counterparty risk might be dampened somewhat with a margin account through I.B. rather than 2x funds. 

[clacy]

I think even a fairly conservative investor could benefit by setting up their portfolio in such a way that 50-75% was traditional 1x PP and then take the remainder and do a 2x or 3x PP.  Why not juice returns by using the PP principals, while keeping the majority of your retirement in the traditional PP?  This could replace most VP speculation, as most people would be quite pleased with 20-30% annual gains in their VP.

That was the original idea for me.  The concept that if a 2xPP would work, it would represent the prospect of 20%+ annual returns with average volatility and no work.  That made me ask the question: what is my target return for a VP for the associated risks and time involved?

I don't think a 2xPP can match the risk/reward that a start-up business venture can present, but otherwise the 2XPP offers the potential for a solid return compared to similar VP investments such as rental real estate and stock picking.  For instance, say someone likes the 4-5% real returns of the PP, but wants more potential for growth.  You could run an 80% PP and 20% VP consisting of 2xPP funds.  The blend would likely offer another 200bps to the annual return for a total real return of 6-7% with a mitigated downside risk.  I like that idea much better than trying to buy apartment houses and manage tenants.

I don't run a VP like this at the moment, but I wouldn't rule it out in the future.  So far I haven't seen any evidence of decay when viewed as a whole (over the last 21 months).  The primary concern for me in such a portfolio is counterparty risk in a systemic crisis.  On the other hand, counterparty risk might be dampened somewhat with a margin account through I.B. rather than 2x funds. 

Look at a long term chart of SSO vs SPY dating back to SSO's inception.  You'll see the decay factor. 

Maybe decay can be counterbalanced because of volatility capture.  I'm not sure, but I do see decay when I look at leveraged returns.  The past 21 months has treated the 2x PP very well, but I certainly don't think it will always be like that.  I'm not an expert on the reasons for decay and how these 2/3x ETF's are operated, every leveraged fund that I've looked at with a 4+ year track record has under-performed their 1x counterparts in a big way.

I happen to think that futures might be the best way to leverage the PP.  Hedgefunds and so forth, usually turn to futures when they are looking for leverage. 

[Maestro G]

The SSO, SPY comparison looks about right to me Clacy.

2x leveraged funds look to provide twice the DAILY motions. Which means you both double the daily gains (or losses) and double the standard deviation (volatility).

A 0% yearly average with a 20% standard deviation will typically annualise to -2%
A 0% yearly average with a 40% standard deviation will typically annualise to a -8.3%

i.e. the higher volatility induces a lower annualised.

Countering that however are a number of factors. As prices rise so the 2x in effect scales up exposure, so you gain more over a period of time of rising markets (more than 2x the multiple days reward). As prices decline the 2x in effect scales down exposure so you lose less than 2x over a period of time of declining markets (lose less than 2x the multiple days losses).

Another factor is that you rebalance more, which cost averages down quicker so to speak.

i.e. the 'drag factor' is somewhat countered by running winners, cutting losers type adjustments and by the higher levels of better cost averaging (rebalancing more frequently).

The easiest way to perhaps think of it is that if you borrow to leverage, then typically you'll pay around the cash rate (which generally = inflation) for those borrowed funds. If the PP achieves a 4% real (after inflation) then the benefit of having borrowed to invest in a PP is 4%. If inflation (and cash) is 6% a normal PP might yield a 10% total, whilst a borrowed to invest PP might yield a 4% total gain. Combined that 2x = 10% (non borrowed based PP) + 4% (borrowed based PP) = 14%, compared to 10% for the 1x.

Whilst in nominal terms a 14% reward for a 2x compared to a 10% for a 1x isn't twice the reward, in real (after inflation) terms it is much closer.

10% for the 1x less 6% inflation = 4% real
14% for the 2x less 6% inflation = 8% real

and that doesn't include additional/faster cost averaging benefits (more frequent rebalance trading).

Individual 2x holdings do tend to zigzag around the 1x. Sometimes beating it by more than 2x, sometimes just pacing it, other times being down by approaching 2x. Blending a number of low correlated 2x's tends to smooth that out such that the set more closely paces the 2x performance overall.

It appears to me that the 2x PP, especially in view of the low correlations of the individual components, can work as desired.

A reasonable blend for a 2x like PP might be to hold 16% in BGU (3x stocks so like holding 48% stocks), 16% TMF (3x LTT so like holding 48% LTT), 24% UGL (2x gold so like holding 48% gold) and 44% cash. Year to date that's earned just over 25% and triggered 3 rebalance events. The 1x PP (without any rebalancing) year to date is up around 11%.

Whilst there might be greater third party risk from holding leveraged funds, if you used a 2x to emulate a 1x overall, then you might hold 8% BGU, 8% TMF, 12% UGL and 72% cash. In holding so much cash your risk against the leveraged funds is in effect a -28% maximum loss assuming all three leveraged funds went toes-up. Whilst that would be uncomfortable, its not critical. Little different to a conventional PP investor who perhaps held 25% in a gold ETF and who then saw that ETF fail.

If instead of holding 72% in cash, those funds are deployed into an alternative investment - perhaps a classic PP - that earned > cash then !!!

Clive,

The 72/28 ratio of very safe assets to extremely risky/volatile assets that you are suggesting reminds of the strategy that Nassim Taleb touts in his "Black Swan" tome, although his ratio is a little more conservative IIRC. Something like 90/10. Essentially another variation on the "Fat Tail" strategy, yes?

Maestro G

[clacy]

Clive, since you are in the UK I believe, why not spreadbet for leverage?  I don't know a lot about it, but besides the spread, is there any borrowing cost? 

I would suspect that counter-party risk might be a problem that would make this way of leveraging null and void, but I don't know how that works.

[Reub]

Clive, you are an international treasure!

[Reido]

Personally, I'm not worried about the drawdowns one could see with a leveraged PP.  I could tolerate perhaps a 40% drawdown... 

what concerns me is not the severity - its the DURATION.  If you had a 6+ year period, as with stocks, where the account value stayed significantly below it's highs I would find that extremely frustrating.  If I know that in 2 years my account will be restored, then I'm not too concerned.

What concerns me about the 2xPP with leveraged funds is simply that a flat year could easily turn into a down year due to the decay.  I'm not confident that the portfolio would produce 2x the gains consistently enough to be worth the extra losing years...

Just IMO - I know theres really not enough data to say explicitly one way or the other

Since this is part of the VP - I would consider investing in a leveraged PP after a losing year for the regular PP which historically is followed by immense gains!!

[FarmerD]

I constructed a spreadsheet to see how a synthetic 3X HBPP would have performed since 1972.  I used the yearly asset returns data from Craigr’s website then modified the yearly data using the expected etf decay from the Powershares 3X funds prospectus’s. (Page 5 of any of the Proshares 3X funds shows the expected decay)
https://web.archive.org/web/20160324133 ... l-returns/
http://www.proshares.com/funds/#sort=Na ... 2x,ultra3x

Here’s the highlights:
Year                                                  TSM                                                3X HBPP
1972-1979                                        5.6%                                                   63%
1980-2002                                         12.4%                                               7.9%
2002-present                                     4%                                                    19.6%
1980 –present                               11.0%                                                   11.4%
1972 – present                                     9.9%                                                20.2%
Even though the 3X portfolio doubles the return of the TSM, It’s looks like the 3X portfolio goes through extended periods of outperformance and underperformance.  The 3x HBPP returned an incredible 63% CAGR from 1972-1979, then the 3X lagged the TSM next 22 years, then the 3x has again massively outperformed since 2002.  

Note: This analysis assumes annual rebalancing.

[dragoncar]

Consider a 1x that gains 1% each day for 10 days and a 3x that achieves +3% each day for 10 days.

The 1x gains 1.01^10 compared to the 3x gaining 1.03^10 which equals 10.4% versus 34.4%

If the 1x instead declined 1% each day for 10 days and the 3x lost 3% each day for 10 days then that's 0.99^10 for the 1x compared to the 3x 0.97^10 which equals -9.56% compared to -26.25%

I think you may be ignoring that the 26.25% loss of the 3x occurs to a greater base value (after the initial 34.5% gain).  Say you start with:

1x: $100
3x: $100

you get 10 days of 1% gains:

1x: $100 * 1.01^10 = $110.46
3x: $100 * 1.03^10 = $134.39

you get 10 days of 1% losses:

1x: $110.46 * 0.99^10 = $99.90
3x: $134.39 * 0.97^10 = $99.10

Therefore, the 3x has decayed by $0.80.

[FarmerD]

Clive/Dragoncar,
Decay also depends not only on frequency of the gains/losses but their magnitudes.  Consider a 1X versus a 3x ETF that alternates 10% gains or loses for 10 days:

1X: -5%
3X:  -38%

Granted, this is a pretty extreme example but it makes the point.

From the Proshares 3x S&P 500 ETF prospectus (page 5) “The Index’s annualized historical volatility rate for the five year period ended June 30, 2011 was 25.06%. The Index’s highest June to June volatility rate during the five-year period was 45.47% (June 30, 2009).”? Assuming the average volatility over this period (25.06%) would have been about the same historically since 1972, I get the following decay data from their chart:

Index                      3X Fund
-40%                              -82%
-30%                              -72%
-20%                              -58%
-10                                    -40%
0                                      -17%
10%                                  10%
20%                                  43%
30%                                  82%
40%                                  128%

I think this is a reasonable expected return from 3X ETF and is applicable to backtesting historical results using a 3X strategy.  .  Am I missing something?
Statistics was never my strong suit - feel free to correct me. 

[stone]

Clive, do those USD index etfs have an inverse correlation to the gold:stocks:LTT part of the PP? Might a smidgen of that provide a more aggressive emergency balancer to the leveraged PP? This is straying way off what I'd ever contemplate doing myself!

[stone]

Perhaps a way to prevent a total loss for the leveraged PP might be had by having a call option for a USD bull index such that you got money if the USD became much stronger but normally just had the option expire worthless as an "insurance expense".

[stone]

Clive, don't the dividends and bond interest play a large part in deciding whether it is worth holding leveraged etfs versus a conventional PP? The dividends on UK stocks are now not to be sniffed at. 

[Reub]

Clive, what do you think about incorporating FSA and FSG into a variable portfolio. FSA is 2x T-Bond Bull/S&P 500 Bear ETF and tracks the difference between 2X LTT and the S & P 500 (-200%).  FSG is 2x Gold Bull/S&P 500 Bear  and tracks the difference between 2X gold and the S&P500 ( -200%).

Both were up considerably on a terrible day for stocks (FSA up 9.31%; FSG up 6.09%).

Any ideas how these could fit into a possible variable portfolio that balances between gold, stocks, cash and bonds?

[blackomen]

Super interesting!  Thanks once again for the update.

I've also been tracking a "Leveraged PP" and a "Leveraged Super PP" (3x33 of the leveraged ETFs, discarding cash.)

I established both portfolios on January 25, 2011.  The "Leveraged PP" (4x25) has so far returned 14.43%.  The "Leveraged Super PP" (3x33) has returned 19.19% (in about 6 months.)

Also, none of this is real money, just a SmartMoney simulation.  All the fun with none of the heartburn!

Care to share your implementations of the leveraged PP?

[MediumTex]

Super interesting!  Thanks once again for the update.

I've also been tracking a "Leveraged PP" and a "Leveraged Super PP" (3x33 of the leveraged ETFs, discarding cash.)

I established both portfolios on January 25, 2011.  The "Leveraged PP" (4x25) has so far returned 14.43%.  The "Leveraged Super PP" (3x33) has returned 19.19% (in about 6 months.)

Also, none of this is real money, just a SmartMoney simulation.  All the fun with none of the heartburn!

Care to share your implementations of the leveraged PP?

I think that implementation of this leveraged ETF PP strategy is what started this thread.

Take a look on the first couple of pages.

BTW, I would not recommend it due the decay issue with these funds, but I think several variants were discussed a few pages up.

[Lone Wolf]

Care to share your implementations of the leveraged PP?

Sure thing.  Like MT said, I used the ETFs recommended further up in the thread.

25% SHY = Cash
25% SSO = 2x leveraged S&P 500
25% UBT = 2x leveraged 20+ Year Treasury Bonds
25% UGL = 2x leveraged gold

The "super" version simply throws away the 25% in cash and becomes a 3x33% portfolio.

This is all simulated money, believe me.  Apart from a 529 where I can't implement it, I'm 100% in the Permanent Portfolio (and happy to be there.)