My Latest Research on the PP's Big Fat Flaw aka Achilles' Heel

[sophie]

Isn't gold on sale now?

Almost by definition, "on sale" means "no one wants it".  That's why I'd be concerned about trying to shift asset weights according to popularity of an asset; you'll miss the real sales.  The PP depends on this for part of its returns.  This isn't to say that I'm predicting gold will go up.  It may well drop further, or stay mired at the current level for years.  But gold is the asset I am most comfortable buying right now.

Remember back in 2009 when stocks had tanked and no one wanted to buy them?  People were selling stocks, getting out of the market etc.  Gold and bonds, on the other hand, were skyrocketing.  What would your allocation choice have been then?

[Cortopassi]

Isn't gold on sale now?

Almost by definition, "on sale" means "no one wants it".  That's why I'd be concerned about trying to shift asset weights according to popularity of an asset; you'll miss the real sales.  The PP depends on this for part of its returns.  This isn't to say that I'm predicting gold will go up.  It may well drop further, or stay mired at the current level for years.  But gold is the asset I am most comfortable buying right now.

Remember back in 2009 when stocks had tanked and no one wanted to buy them?  People were selling stocks, getting out of the market etc.  Gold and bonds, on the other hand, were skyrocketing.  What would your allocation choice have been then?

Exactly!  If I was in the PP in Feb 2009, I would have been having a cow thinking it was insane that I needed to ADD stocks at that point (when in fact I was stupidly getting out of them).  And it would have been the right choice.  Why gold is considered differently is not in line with the equal splitting concept.

For me, at least, whatever feels 100% like the wrong move is usually (>50%) the right move to make.  90% of the benefit (to me) of the PP is it takes my emotions out of the picture.  Well, not 100% yet... but a lot more than ever before.  And it gets better every month.

[buddtholomew]

Isn't gold on sale now?

Almost by definition, "on sale" means "no one wants it".  That's why I'd be concerned about trying to shift asset weights according to popularity of an asset; you'll miss the real sales.  The PP depends on this for part of its returns.  This isn't to say that I'm predicting gold will go up.  It may well drop further, or stay mired at the current level for years.  But gold is the asset I am most comfortable buying right now.

Remember back in 2009 when stocks had tanked and no one wanted to buy them?  People were selling stocks, getting out of the market etc.  Gold and bonds, on the other hand, were skyrocketing.  What would your allocation choice have been then?

Exactly!  If I was in the PP in Feb 2009, I would have been having a cow thinking it was insane that I needed to ADD stocks at that point (when in fact I was stupidly getting out of them).  And it would have been the right choice.  Why gold is considered differently is not in line with the equal splitting concept.

For me, at least, whatever feels 100% like the wrong move is usually (>50%) the right move to make.  90% of the benefit (to me) of the PP is it takes my emotions out of the picture.  Well, not 100% yet... but a lot more than ever before.  And it gets better every month.

+2 for Sophie and Cortopassi.

I hold A + B + C + D in equal proportions. I don't like A more than C or D less than B. At the end of the day I want these four positions to earn a positive real return.

[Kbg]

Cortopassi...there is a lot to be said oftentimes for George Costanza investing.

[ochotona]

Exactly!  If I was in the PP in Feb 2009, I would have been having a cow thinking it was insane that I needed to ADD stocks at that point (when in fact I was stupidly getting out of them).  And it would have been the right choice.  Why gold is considered differently is not in line with the equal splitting concept.

For me, at least, whatever feels 100% like the wrong move is usually (>50%) the right move to make.  90% of the benefit (to me) of the PP is it takes my emotions out of the picture.  Well, not 100% yet... but a lot more than ever before.  And it gets better every month.

I'm not meaning to brag, because it's hardly anything to brag about, but in Feb 2009 I sat down, took all my free cash, divided it by 10, and programmed to automatically buy equities for a period of ten months. Of course, Mr. Market might've kept going down right? Sometimes I get brave and stare down my fears. Sometimes not... I get awarded a big pot of my ex-employer's shares this year at $79.50/sh, they went up, then started crashing again, I bailed at $83, now a few weeks later they're at $93. I lost $20,000 in that deal, if I had held fast I'd be that much richer today. The point is, everyone has their own temperament, and you have to understand how to not be in conflict with your portfolio, or else you will go nuts.

Even though I'm seeking lower gold prices, I am also working to keep emotions out of it. I sent up alerts at Kitco.com to ring me when gold goes down to $1000, $800, and 20% lower steps... when I get those emails, go out and rebalance... but as it goes lower, slightly increased gold %.

Finally, when Sophie sells me her gold, I'll be at 25% gold allocation 

[ochotona]

I was home sick, and therefore endlessly playing with peaktotrough. I initiated two test PP portfolios... One started Jan 1 1975 when gold was low in price (two bulls, two bears), the other Jan 1 1980 (one bull, two bears), just before gold peaked. Both portfolios ran through yesterday. Then I tweaked the gold %.

The fixed gold allocation I liked best from 1975 to today was 14%. The fixed gold allocation I liked best from 1980 to today was 11%. What strikes me, though, is that 11% and 14% are not worlds apart, even though the gold entry price was vastly different. There really is some something concrete and provable about the "pick an allocation and stick to it" philosophy, it definitely is not made up. Over the decades these portfolios ran, the gold entry price mattered some, but not to an overwhelming degree.

I am thinking my idea of tactically overweighting gold at a low price, underweighting at a high price isn't warranted, not if the spread between underweight and overweight is 3%. But I haven't given up on the idea that 25% gold is too high. Each time I run peaktotrough, for a variety of time periods, I see that conclusion over and over.

[stuper1]

Here's how I look at it.  For run-of-the-mill scenarios, 15% gold works great in a portfolio.  The extra 10% gold that HB recommends is for extra insurance in case of extraordinary end-of-the-world scenarios.  This extra 10% isn't going to have a very negative effect on your returns in normal times, but it is going to help a lot in abnormal times.

If you're sure that the financial world isn't going to come to an end in your lifetime, then just go with 15% gold, and you'll be fine . . . until the day the world ends, and then you'll still be a lot better than most people.

[Gabe]

Those of you who have reduced the gold allocation: where do you stick the balance? Cash like MG?

[MachineGhost]

But if we were in another period of gold doing well would we be having this conversation? I suspect not.

I really don't think we're going to worry about missing out on the gain that a measly 5% in gold would represent, but the increased risk that 5% represents.  It is certainly not symmetrical.

[ochotona]

I have it!

When the pessimistic posts in the Gold Scream Room increase, then buy more gold. When the optimistic posts increase, then sell gold.

As a ranger at Yellowstone National Park told us, tongue-in-cheek... "the best statistical predictor of when Old Faithful is going to go off is the number of people crowded around to watch it".

[Stewardship]

MG,

Thanks for all the work you put into this!  I admit some of it is over my head but I've found it invaluable nonetheless.

Limited historic data on T-bonds aside, what about using a cash/bond bullet strategy to equalize risk contribution?

[MachineGhost]

Limited historic data on T-bonds aside, what about using a cash/bond bullet strategy to equalize risk contribution?

I'm not sure that makes any sense.  CD's are offered in 10-year maturities now.  While they're probably not breakable, there's really no need to take on secondary market principal risk for cash.

[Stewardship]

There was discussion in another thread of intermediate bonds being an acceptable replacement for ST/LT bonds for cash & bonds (bullet vs barbell.)

I'm thinking a bullet could be a good way to increase duration beyond that of 30-year treasury + cash.  But yeah I guess you could accomplish the same thing by just increasing your cash duration on its own.

[MachineGhost]

Now something for everyone to chew over.  Below is the annualized volatility of the PP components from the beginning of 1968 to the end of 1980 during the tail end of the greatest bond bear market and currency crisis ever experienced:

Stocks: 13.37%
Bonds: 8.67%
Gold: 25.82%

And from 1981 to date:

Stocks: 17.85%
Bonds: 13.14%
Gold: 18.26%

I think this is important.  Because this is pointing out that bonds have been riskier despite being in a 34-year bull market.  What is the key differentiating metric?  It's the duration, stupid!  It was A LOT lower in the late 1970's and probably 1987 when HB simplified PRPFX into the PP.  So that is the adjustment that needs to be made to fix another flaw in the PP without a full bond historical bear market available to us.  Just normalizing bonds to a lower level of target volatility isn't the proper way to do it.

[MachineGhost]

Yes, the duration of 30 year bonds is much shorter now than in the 80's.  But looking at your volatility numbers, wouldn't you say that bonds are a bit closer to being volatility-matched with gold and stocks than they were before 1980?

Don't you mean "longer"?

They may be relatively closer but still very far from equal.

[iwealth]

So the volatility of the individual assets has changed and will continue to change in all likelihood. So isn't making allocation adjustments based on volatility a market timing tactic just like making allocation adjustments based on price?

[MachineGhost]

So the volatility of the individual assets has changed and will continue to change in all likelihood. So isn't making allocation adjustments based on volatility a market timing tactic just like making allocation adjustments based on price?

No, enough time has passed that the asset volatilities are very stable now.  Short term gyrations don't matter.

[vnatale]

I take some issue with the title, equating sub-optimal weighting based on historical risk contribution to an Achilles heel.  Even if the PP could theoretically be improved, that doesn't mean the traditional version is dangerously flawed or cause for concern.

That said, I admire the research, MG.  Very interesting.

Tyler,

In the now almost five years since MachineGhost initiated this Topic would any of your subsequent research change any of what you stated above?

Vinny

[Tyler]

I take some issue with the title, equating sub-optimal weighting based on historical risk contribution to an Achilles heel.  Even if the PP could theoretically be improved, that doesn't mean the traditional version is dangerously flawed or cause for concern.

That said, I admire the research, MG.  Very interesting.

Tyler,

In the now almost five years since MachineGhost initiated this Topic would any of your subsequent research change any of what you stated above?

Vinny

No.

One can always tweak the PP to their liking, but the default version is still an excellent choice.