Gyroscopic Investing

MediumTex wrote: Well, if we know that historically the PP asset that is declining will usually be in the process of losing 40-80% of its value from its prior high and the asset that is in the process of increasing will usually be increasing by between 200% and 1000% (or more) from its prior low, do you see how you would be making more money on the winners than you would be losing on the losers?

It is the 'why' of this that I don't really understand. I wonder if this proposition--that gaining assets always go up more than the lagging assets go down--will continue to be true.

christina wrote:

MediumTex wrote: Well, if we know that historically the PP asset that is declining will usually be in the process of losing 40-80% of its value from its prior high and the asset that is in the process of increasing will usually be increasing by between 200% and 1000% (or more) from its prior low, do you see how you would be making more money on the winners than you would be losing on the losers?

It is the 'why' of this that I don't really understand. I wonder if this proposition--that gaining assets always go up more than the lagging assets go down--will continue to be true.

Yes, it will, because of the volatile assets' relationship to the underlying economic conditions.

It's also helpful that historically ALL of the PP assets rise in value if you give them enough time.

MediumTex wrote:

christina wrote:

MediumTex wrote: Well, if we know that historically the PP asset that is declining will usually be in the process of losing 40-80% of its value from its prior high and the asset that is in the process of increasing will usually be increasing by between 200% and 1000% (or more) from its prior low, do you see how you would be making more money on the winners than you would be losing on the losers?

It is the 'why' of this that I don't really understand. I wonder if this proposition--that gaining assets always go up more than the lagging assets go down--will continue to be true.

Yes, it will, because of the volatile assets' relationship to the underlying economic conditions.

It's also helpful that historically ALL of the PP assets rise in value if you give them enough time.

These are related concepts though. If an asset goes up 900% when it rises, then goes down 80%,  it has grown by 100%.  The asymmetry is just another way of saying that the asset is expected to go up in value over the long term and IMO is irrelevant to rebalancing.

The volatility gives you the rebalancing bonus. 

I am curious what this bonus has been over the last 30 years (vs buy and hold 25x4).  Perhaps I'll run the numbers today.

dragoncar wrote: I am curious what this bonus has been over the last 30 years (vs buy and hold 25x4).  Perhaps I'll run the numbers today.

You can use this tool to find that out: http://www.peaktotrough.com/hbpp.cgi

No rebalancing:  CAGR 8.06% and max DD 36.09%
Std rebalancing: CAGR 8.97% and max DD 19.83%

When you guys say rebalancing bonus you're just referring to lower volitility right?

I mean....you gain less money with rebalancing than you would by letting it run but obviously you cut down on the volitility you'd experience after a few years of letting it run.

Kshartle wrote: When you guys say rebalancing bonus you're just referring to lower volitility right?

I mean....you gain less money with rebalancing than you would by letting it run but obviously you cut down on the volitility you'd experience after a few years of letting it run.

No, the overall rate of return is higher with rebalancing because you don't give up all the gains when the outperforming asset starts underperforming.

Kshartle wrote:I mean....you gain less money with rebalancing than you would by letting it run but obviously you cut down on the volitility you'd experience after a few years of letting it run.

A thought experiment: let's say that the long-term returns of all four assets is zero.  That is, gold, bonds, stocks, and cash will all oscillate wildly over the years, but on average always end up back where they started.

In that situation, a PP without rebalancing would not grow, since none of its four assets are, overall, growing.

But with rebalancing: when stocks are high, you sell them to buy bonds and gold.  When gold is high, you sell it to buy stocks.  Etc.  In this scenario, your return is positive.

That's the rebalancing bonus.

People can always recite "buy low, sell high" but rarely do it in practice.  Rebalancing is simply buying low and selling high without selling completely out of a position.  You lock in gains and minimize downside risk while leaving open the possibility that your hot investment will continue to rise.

I love rebalancing.  That many people don't understand the benefit and believe you must either ride a hot position until it dies (at which point you sell it all and buy something else already hitting its peak) explains a lot about typical investing attitudes. 

Pointedstick wrote:

dragoncar wrote: I am curious what this bonus has been over the last 30 years (vs buy and hold 25x4).  Perhaps I'll run the numbers today.

You can use this tool to find that out: http://www.peaktotrough.com/hbpp.cgi

No rebalancing:  CAGR 8.06% and max DD 36.09%
Std rebalancing: CAGR 8.97% and max DD 19.83%

Cool, so we can assume the rebalancing bonus is around 1%.  On top of that, volatility is reduced.

But I'd say the reason the portfolio "works" is primarily because it's assets have historically appreciated an average of 8% per year. 

If rising interest rates (just a what-if) drive down out average appreciation over the next decade, the rebalancing bonus isn't going to save the day.

dragoncar wrote:

Pointedstick wrote:

dragoncar wrote: I am curious what this bonus has been over the last 30 years (vs buy and hold 25x4).  Perhaps I'll run the numbers today.

You can use this tool to find that out: http://www.peaktotrough.com/hbpp.cgi

No rebalancing:  CAGR 8.06% and max DD 36.09%
Std rebalancing: CAGR 8.97% and max DD 19.83%

Cool, so we can assume the rebalancing bonus is around 1%.  On top of that, volatility is reduced.

But I'd say the reason the portfolio "works" is primarily because it's assets have historically appreciated an average of 8% per year. 

If rising interest rates (just a what-if) drive down out average appreciation over the next decade, the rebalancing bonus isn't going to save the day.

If interest rates start rising, wouldn't that mean that t-bill holdings would be doing quite well?

When interest rates stopped rising, treasury bonds might be a very good investment as well.

And depending on why interest rates were rising, gold might be doing well also.

If interest rates were rising because of high inflation, then stocks would probably do okay as well.

The rising interest rates scenario doesn't seem like it would be a big deal, and the last time it happened in the 1970s it wasn't, in fact, a big deal for the PP.

Am I missing something? 

Pointedstick wrote:

dragoncar wrote: I am curious what this bonus has been over the last 30 years (vs buy and hold 25x4).  Perhaps I'll run the numbers today.

You can use this tool to find that out: http://www.peaktotrough.com/hbpp.cgi

No rebalancing:  CAGR 8.06% and max DD 36.09%
Std rebalancing: CAGR 8.97% and max DD 19.83%

If you run it from Jan 1st 1981 through today without rebalancing it's CAGR of 8.16

With rebalance it's 35/15 it's 8.05

That's 32 years and not re-balancing returned higher.

There's no such thing as volatility capture. Rebalancing can hurt returns just as easily as improve them.

It does reduce volatility though. There's a logical reason for that, not for any volatility capture.

You can always data mine any set of returns and find a superior rule that would have worked for that set of data.

Rebalancing or not has the same chance of working going forward. Imagine stocks or gold tripling while everything else stays put. Not rebalancing will return a lot more. It will subject you to risk beyond what you might be able to bear though.......

Kshartle wrote:

Pointedstick wrote:

dragoncar wrote: I am curious what this bonus has been over the last 30 years (vs buy and hold 25x4).  Perhaps I'll run the numbers today.

You can use this tool to find that out: http://www.peaktotrough.com/hbpp.cgi

No rebalancing:  CAGR 8.06% and max DD 36.09%
Std rebalancing: CAGR 8.97% and max DD 19.83%

If you run it from Jan 1st 1981 through today without rebalancing it's CAGR of 8.16

With rebalance it's 35/15 it's 8.05

That's 32 years and not re-balancing returned higher.

There's no such thing as volatility capture. Rebalancing can hurt returns just as easily as improve them.

It does reduce volatility though. There's a logical reason for that, not for any volatility capture.

You can always data mine any set of returns and find a superior rule that would have worked for that set of data.

Rebalancing or not has the same chance of working going forward. Imagine stocks or gold tripling while everything else stays put. Not rebalancing will return a lot more. It will subject you to risk beyond what you might be able to bear though.......

Yes, with dividends and interest reinvested the total CAGR comes out almost the same. What about taxes, though?
Also, the gold is only about 5% of the portfolio at this point without rebalancing. What if there is a financial crisis for which gold is the most important asset?

Ohh I'm not saying rebalancing isn't a good idea. It is. There's just no bonus to the overall return from rebalancing. It's all data-dependent. Rebalancing will result in higher returns when markets are sideways but will lag when some assets move in very long bull or bear markets.

If you go from Jan 1 '71 through Dec 31 '80 (big gold bull, bad bond bear):

no-rebalance = 17.86%
35/15 = 13.72%

Whether re-balancing helps or hurts in the long run or short run depends on the severity of any assets bull or bear market. I hope the reason is obvious to everyone.

However there's no question more frequent re-balancing dampens volatility a lot without potentially sacrificing much return in the long run (15-20 year it appears).

All that being said I personally think we're staring at a huge move up in gold and a disaster for bonds at some point in the next couple years so I think the best returns from this starting point will go to those who choose not to re-balance. If they can take the heat.

Interesting results when you run 1/1/71 through 9/25/13.

I put them in order of more re-balancing to less:

Option StDev CAGR
1 Years 6.72 9.08
2 Years 7.27 9.56
3 Years 7.04 9.31
5 Years 7.96 9.60
None 10.31 8.28


Option StDev CAGR
30/20 6.6 9.03
35/15 6.76 9.31
40/10 6.78 9.48


You can see that more frequent rebalancing leads to lower volatility (exception being 2 years vs. 3). However there is no real relationship to returns. It's all data dependent. It would make sense that the 40/10 returns more because it allows bull runs to run farther but should prevent some of them from collapsing on you like never rebalancing will. Again though, I don't think there's anyway to predict how rebalancing is going to affect returns, even over a period of 42+ years like this.

I think you just need to target the volatility you're comfortable with if you're going to invest strictly mechanically.

Interesting points, Kshartle.

MediumTex wrote:

christina wrote:

MediumTex wrote: Well, if we know that historically the PP asset that is declining will usually be in the process of losing 40-80% of its value from its prior high and the asset that is in the process of increasing will usually be increasing by between 200% and 1000% (or more) from its prior low, do you see how you would be making more money on the winners than you would be losing on the losers?

It is the 'why' of this that I don't really understand. I wonder if this proposition--that gaining assets always go up more than the lagging assets go down--will continue to be true.

Yes, it will, because of the volatile assets' relationship to the underlying economic conditions.

It's also helpful that historically ALL of the PP assets rise in value if you give them enough time.

How can you say with so much certainty that a period of prosperity, for example, will always result in stocks going up way more than the other assets are going down?

What if we enter a period where more money comes OUT of the stock market than goes in? (I don't know if this will happen. I'm just using it for argument's sake.)

christina wrote:

MediumTex wrote:

christina wrote: It is the 'why' of this that I don't really understand. I wonder if this proposition--that gaining assets always go up more than the lagging assets go down--will continue to be true.

Yes, it will, because of the volatile assets' relationship to the underlying economic conditions.

It's also helpful that historically ALL of the PP assets rise in value if you give them enough time.

How can you say with so much certainty that a period of prosperity, for example, will always result in stocks going up way more than the other assets are going down?

What if we enter a period where more money comes OUT of the stock market than goes in? (I don't know if this will happen. I'm just using it for argument's sake.)

This is a very good point. Traditionaly prosperity brought with it lower prices because of increased production. In our pseudo-prosperity world we get higher stock prices along with everything else do to money printing.

Stocks have more than doubled off the '09 lows.

Does anyone think we are more prosperous or it's around the corner.

The economy could really prosper if we had an economic boom. That would mean lower prices as our money gained value and we were richer. Stock prices might not even move at all but they would be worth more as the dollar gained.

christina wrote: How can you say with so much certainty that a period of prosperity, for example, will always result in stocks going up way more than the other assets are going down?

What if we enter a period where more money comes OUT of the stock market than goes in? (I don't know if this will happen. I'm just using it for argument's sake.)

If money was leaving the stock market, where would it be going?

MediumTex wrote:

christina wrote: How can you say with so much certainty that a period of prosperity, for example, will always result in stocks going up way more than the other assets are going down?

What if we enter a period where more money comes OUT of the stock market than goes in? (I don't know if this will happen. I'm just using it for argument's sake.)

If money was leaving the stock market, where would it be going?

Could be to another asset class, or the money supply is just shrinking so all prices are falling and dollars are gaining purchasing power.

Kshartle wrote:
Short EUR/CHF around 1.28.

Well now it's at 1.22, approaching the lower end of the range where the Swiss tend to come in to buy to boost the Euro. I'd rather just play the short end when it gets back up to 1.28 for safety. 

MediumTex wrote:
As craig and I have noted before, in crazy market conditions there is no requirement that you MUST rebalance.  For example, in a true currency crisis where the dollar was losing 50% or more of its value every month, no PP investor would be required to continue selling stocks and gold and buying more treasuries.  It would probably make more sense to just wait until market conditions stabilized and decide what to do from there.

So, at the height of a financial crisis... revert to market timing??

Coffee wrote:

MediumTex wrote:
As craig and I have noted before, in crazy market conditions there is no requirement that you MUST rebalance.  For example, in a true currency crisis where the dollar was losing 50% or more of its value every month, no PP investor would be required to continue selling stocks and gold and buying more treasuries.  It would probably make more sense to just wait until market conditions stabilized and decide what to do from there.

So, at the height of a financial crisis... revert to market timing??

Yes follow the permanent portfolio strategy then at a crisis ditch it. Hah!

I think the point is if you have all four assets going into a crisis as long you just wait you won't get hurt too bad.

I think this is why Browne said take a look once a year and if it's time to re-balance do it. Hopefully when you look you're not in the middle of some collapse and buying into the crappy asset class. Even if you do if you wait another year you should be ok.

The problem would be if, say, gold is skyrocketting due to the dollar becoming worthless and you re-balance everyday until you're broke.

MediumTex wrote: If money was leaving the stock market, where would it be going?

Abroad, but not into gold. Or paying back debt.

Coffee wrote:So, at the height of a financial crisis... revert to market timing??

It's more like assess the situation. If D.C. is burning, there are riots in the street, troops are implementing martial law, capital controls prevent taking money outside the country, and cats and dogs are living together, I'm not rebalancing out of gold into stocks/bonds/cash.

But we are a long way from that happening. It's just that if something very bad were going on, at least keep the option open that rebalancing into a black hole is probably not a good idea.

frommi wrote:

MediumTex wrote: If money was leaving the stock market, where would it be going?

Abroad, but not into gold. Or paying back debt.

Historically (in the U.S. anyway), if money is leaving the stock market it's going into treasuries, gold or cash.

And note that as dollars are leaving the stock market, they don't necessarily have to go directly into treasuries or gold for treasuries or gold to still react to it, primarily because money would typically be leaving the stock market due to a shift in underlying economic conditions, which would set the stage for another one of the PP's asset to take the lead.