Improving on the Permanent Portfolio

[Stefan]

I really want something simple, relatively safe, proven and cost-effective for my portfolio as I near retirement. I want to enjoy the above average returns from the PP and be able to sleep well at night, knowing that my future is secure. What I'm trying to say is that I really don't want to improve on the PP, just enjoy it!

The system I showed doesn't have a CAGR > HBPP's CAGR. What it does, and that was my whole point with my post, it has a much lower MaxDD, for people who cannot sleep at night with a HBPP's MaxDD.

And, the only way to get lower drawdowns, to the sleeping point, so to speak,  is to manage risk. There is no other way.

An extremely simple way to manage risk is outlined in Mebane's study I linked in my original post. You may want to read that.

I presented a more elaborate way, leading to more MaxDD reduction. But the caveat is that it can only be acomplished on a programmable platform, to make it simple & automatic to run. If you need a simple, manually run, and sleeping well risk management level, I suggest you use Mebane's monthly timing technique.

MG: what method do you use for your timing?

[Reub]

Where is his monthly timing method published?

[rickb]

Where is his monthly timing method published?

See http://www.mebanefaber.com/timing-model/ .  It's also described in a book called "The Ivy Portfolio" by Faber and Richardson.

The bottom line is you compare the month end price to the average of the past 10 month end prices.  If this month's price is higher you buy (or stay in).  If it's lower, you sell (or stay out).  It's more or less the same as a 200 day moving average, but checking only once a month. 

[craigr]

Market timing is bunk. Investors that just sit back and do nothing are taking market timer's money.  

[MachineGhost]

I'm using proprietary value and trend models, but I am going to take a fresh look at your method and see how it does.  I've looked at a gargantuan number of technical-only systems over the last two decades, and channel breakouts have never stood out to me before, so I want to see if I missed anything.

MG

MG: what method do you use for your timing?

[stone]

Machine Ghost, you say that you want to reduce max draw down  but aren't you just exposing yourself to the risk of consecutively exiting positions with a loss? Say you sell gold with a 10% loss, then sell stocks with a 10% loss etc etc. That could accumulate to any size of loss you dare to imagine. I wonder whether we could be in a long term secular trend towards more generalized random volatility of everything. That would be very bad for any trend following strategy wouldn't it?

[MachineGhost]

I'm not quite sure what you mean by "locking in a loss".  To take a loss on an asset would imply that you've sold the asset when conditions are unfavorable to protect against future reductions in value.  As Keynes allegedly said: "the superiority of stocks isn't inevitable; they own the advantage only when certain conditions prevail".  The same applies to the other PP assets.

I don't see how volatility is bad for trend-following since movement is necessary to make any money.  But if you mean the more irrational up and down moves (i.e. manic depressive) that are net effect sideways or bearish after a fixed period of time (like stocks over the last decade), then that is among the kind of conditions intended to be avoided as much as possible.

MG

Machine Ghost, you say that you want to reduce max draw down  but aren't you just exposing yourself to the risk of consecutively exiting positions with a loss? Say you sell gold with a 10% loss, then sell stocks with a 10% loss etc etc. That could accumulate to any size of loss you dare to imagine. I wonder whether we could be in a long term secular trend towards more generalized random volatility of everything. That would be very bad for any trend following strategy wouldn't it?

[stone]

Machine Ghost, doesn't volatility by its very nature mean random up and down noise? ie a volatile asset will have a price that either increases or decreases in price by 2% per day pretty much at random and a less volatile asset will have a price that either increases or decreases in price by 0.02% per day pretty much at random. eg compare UK 40year index linked treasuries (very volatile) and UK index linked treasuries one year from maturity (not volatile). If you use any sort of (implied) stop loss then the tightness of your stop has to be lax enough to avoid whipsaw losses ie  being stopped out by random price shifts that just reverse rather than being a trend. Measuring relative strength (or price momentum any which way) and switching assets accordingly would cause whipsaw losses if you failed to correctly discern any trend (or lack of one) past the random noise.

Let's imagine that you were using some price momentum method to choose whether to hold Apple or HP stock. If you set your parameters too tightly then you would have ended up repeatedly selling Apple and buying HP even though you would have been much better off just holding Apple all along.

I'm just saying that the way you set your parameters (relative strength over days, weeks, months etc, max drawdown whatever) is all based on backtesting and the future might warrent quite different parameters.

If asset prices followed Brownian motion such as in the Black Scholes model, then no price momentum method would be anything other than harmful. It would be like using the last few coin flip results to choose heads rather than tails when flipping a coin.

[MachineGhost]

I see what you mean.

First of all, in the real world, price movements are not random and are not distributed on a lognormal Gaussian basis.  They're serially correlated due to "the wisdom of crowds", i.e. emotional groupthink.  This effect is fractal in nature.  At the most minimal level price movement would be in response to market structural fundamentals (opening/closing times, etc) or news events released intraday of which the effect would percolate up to other shareholders over time.  All of the market beating excess return "anomalies" the academics have identified are due to human behavior quirks (behavioral economics).  Volatility is essentially one way to observe this human behavior, i.e. volatility decreases in uptrends and increases in downtrends, it is not symmetrical.

System development and backtesting is another issue.  One has to design a method that is robust, so that increases or decreases in volatility do not break the system because of inflexibility.  Alot of this is just naive cluelessness (aside from all the hucksters selling ice to Eskimos) in proper system development, like using absolute values instead of relative, overoptimizing, etc..  Robustness and risk are correlated.  The trick is to decrease the latter while maintaining the former as much as possible.  There is no way to earn more than what the market is offering unless the market's risk is too high (inefficient) or gains are compounded more frequently (that margin nowadays is tiny, about 3 days, it used to be about 3 months or so decades ago).

PP is a very robust strategic method as it has clearly delineated margins between broadly non-correlated assets based on macroeconomic conditions that are rebalanced according to technical criteria, i.e. it is a technofundamentalist strategy.  Yet, it its nowhere near optimal on a risk-reward basis or in its so-called volatility capturing (but which due to the width of the bands, is actually more of upside trend capturing and throwing good money after bad).  I took the approach that rather than optimizing by adjusting the margins between the asset classes or bands' width -- which would surely become sub-optimal as volatility and/or correlations increased or decreased -- it was far better to focus on a "risk overlay" by identifying negative technofundamental return regimes within each asset class.

I guess another way of saying all this is it is a hell of a lot more difficult to forecast future long-term volatility and correlations, i.e. predicting the future state of human behavior outright, than avoiding periods of clear human behavior irrationality.  Any method to succeed has to work with these facts.

MG

Machine Ghost, doesn't volatility by its very nature mean random up and down noise? ie a volatile asset will have a price that either increases or decreases in price by 2% per day pretty much at random and a less volatile asset will have a price that either increases or decreases in price by 0.02% per day pretty much at random. eg compare UK 40year index linked treasuries (very volatile) and UK index linked treasuries one year from maturity (not volatile). If you use any sort of (implied) stop loss then the tightness of your stop has to be lax enough to avoid whipsaw losses ie  being stopped out by random price shifts that just reverse rather than being a trend. Measuring relative strength (or price momentum any which way) and switching assets accordingly would cause whipsaw losses if you failed to correctly discern any trend (or lack of one) past the random noise.

Let's imagine that you were using some price momentum method to choose whether to hold Apple or HP stock. If you set your parameters too tightly then you would have ended up repeatedly selling Apple and buying HP even though you would have been much better off just holding Apple all along.

I'm just saying that the way you set your parameters (relative strength over days, weeks, months etc, max drawdown whatever) is all based on backtesting and the future might warrent quite different parameters.

If asset prices followed Brownian motion such as in the Black Scholes model, then no price momentum method would be anything other than harmful. It would be like using the last few coin flip results to choose heads rather than tails when flipping a coin.

[stone]

Machine Ghost I thought that a lot of the PP profits could actually be attributed to the irrationality of other investors? When they buy high and sell low that falls in the lap of the PP. I imagined that periods of extreme irrationality would actually benefit the PP especially a PP where someone was keeping it at 4x25% (perhaps due to accumulating a lot into the lagging asset or drawing down from what ever was leading).

[Stefan]

If the topic under discussion has now become: does trendfollowing improve returns and reduce risk for each asset class? The empirical evidence shows that, in the long run, it does. Empirical evidence here means the results of the best known trend followers like Ed Seykota and many others.

Do trends exist? And will they continue to exist in the future? If you believe in the efficient market hypothesis, you will probably be in denial, when faced with studies showing momentum as a market inefficiency which has not been arbitraged away, despite being recognized in the academic literature since early 90s. The same goes for irrational bubbles in asset classes, that you know will end up in a massive correction and reversion to the mean, but which, in the meantime, while the bubble is underway, creates a (parabolic) trend. And the same goes for right or left fat tails shown in price distributions.

According to EMH, these inefficiencies should not exist. Because they cannot be accounted for by EMH. They can only be explained by including the human element, as an irrational agent in the system, with his idiosyncratic psychological drivers of greed and fear. This is a subject of behavioral economics which extends the system from a rational & deterministic EMH academic framework to a complex adaptive system which includes irrational agents and inefficent pathways to convey the price driving information to these agents. Not only are the agents not acting rationally (as seen in bubble times) but, also, the key information they would need to act on is diffused in an asymetric way, from those "in the know" to the rest of the agents. Prices will reflect this  asymetrical diffusion by creating a trend, where more and more agents bid the prices up, or down, as the information, traveling slowly, reaches them. These are more or less known and recognised phenomena and theories, which have still to find a larger acceptance in an academia stuck, largely, on the EMH determinism.

The second question is, if trends exist, why would trend following work, in theory? How could it create profits higher than losses, with all these whipsaws you incur? The answer here is that trendfollowing tries to eliminate most of the left tail events, while keeping the right tail wins in play.

And, finally, if trends exist, and are theoretically profitable, how could you capture this profit? Your trend following system has to filter out the noise, follow the trend by adapting to the asset class volatility and make sure the risks you are willing to take in the portfolio do not lead to your ruin. There is a lot of literature showing how this last is accomplished (google Kelly formula, for instance). In the end, your trend following model should have positive mathematical expectancy, to give you confidence it may work in the future.

[craigr]

For me it's not a question of whether EMH explains everything because it doesn't. Bubbles, etc. do exist and mis-pricing can happen. But in terms of theories that apply to the markets, EMH is the most consistently profitable despite the problems it has from time to time. I hear a lot of talk about various timing mechanisms and how they did in the past, etc. But I never come across people that actually can show me real-world profits from those strategies.

The problem is that people are not computers. A spreadsheet may make it seem so simple to buy and sell at pre-defined times. But in reality there is a human that has to make those decisions and humans are emotional. Part of that emotion is the idea that an investor needs to not just decide when to sell out when the strategy says, but also when to buy back in. That's two different decision points. And each time there is a chance they will question their judgment and waver. That's where the problems start to multiply and the profits start to dwindle.

Taking these trading strategies into the real-world never seems to work the way the advocates state. Either because the strategy itself really doesn't work but they've talked themselves into seeing a pattern that doesn't exist. Or it is simply too hard to put into practice due to trading costs and human factors.

And again, if these trading systems did work and the markets were so simple to take advantage of, then why isn't everyone doing it already? The firms on Wall St. spend millions a year looking for the slightest edge on each other. How did they manage to miss all this easy money? 

[stone]

Stefan don't worry, I don't believe the efficient market hypothesis. I do believe that trends typically exist. My point was that in the (near) future they might become less prevelant or buried under random volatility or become shorter lived. I think Mark Cuban said that the entire "inovator, imitator, idiot" cycle has now shortened to seconds simply because everything is computorized. Trying to stay ahead of trends is massively more complex than simply sticking to a 4x25% PP. I also don't agree that it is possible to say that momentum trading cuts risk by cutting losses. Consecutive stop loss losses can accumulate to any level of loss.

[MachineGhost]

That is only true on the upside.  It would make no sense to take gains from one irrational upside asset to place into another irrational upside asset.  But the default modus operandi of the PP is to do that; it relies on extremely long-term negative correlations and mean reversion to correct such a mistake.  Right now that would be a case of rebalancing your gold gains to plow into stocks; gold clearly has positive fundamentals, stocks do not (on a factual basis not pundit/analyst squawking).

Its all a matter of degree in how far you want to go to reduce your risk.  Plenty of PPers seem to be happy suffering 19%-25% nominal drawdowns or even 50% real drawdowns.  I cannot take that kind of risk on faith.

MG

Machine Ghost I thought that a lot of the PP profits could actually be attributed to the irrationality of other investors? When they buy high and sell low that falls in the lap of the PP. I imagined that periods of extreme irrationality would actually benefit the PP especially a PP where someone was keeping it at 4x25% (perhaps due to accumulating a lot into the lagging asset or drawing down from what ever was leading).

[MachineGhost]

Which is why the average investor should not bother and just accept the "Black Swan" risk by assuming the market is efficient.  The irrationality in the market comes from the naive investor not having self-control over their own emotional impulses.  It won't matter what method they use, they will eventually succumb to the Siren's Song.  That is why they are average.  Ego and hubris won't change this fact, only inner work.

I'm not sure why you infer that Wall Street doesn't use technical approaches at the professional level successfully.  It certainly does and has been for decades.  Especially hedge funds which are notoriously secret.  But, whether or not Wall Street is successful at it, there are other behavioral and structural incentive factors that come into play than the ones the average investor deals with.  Do keep in mind that Wall Street and its clients as a collective is not very innovative in thinking outside the box, so it limits what can get accepted by committee and peer pressure.

If the average investor can control his emotional impulses and not grow so large in terms of assets that liquidity and structural issues become an issue, it is a large grassy green field to play in unhindered by significant competition.

I think you answered your own question as to why simple systems don't work.  Its not the systems that don't work per se; its the humans.  Humans love control and complexity and believe it must be necessary to be successful at anything.  Do you know how humbling it is to admit you don't know jack shit about anything and to put your fate in the hands of a simple strategy that outperforms all the smartest guys in the room?  It's not easy at all.  The exact same stuff we all went through in coming to terms with the PP is at play when dealing with simple strategies.

MG

The problem is that people are not computers. A spreadsheet may make it seem so simple to buy and sell at pre-defined times. But in reality there is a human that has to make those decisions and humans are emotional. Part of that emotion is the idea that an investor needs to not just decide when to sell out when the strategy says, but also when to buy back in. That's two different decision points. And each time there is a chance they will question their judgment and waver. That's where the problems start to multiply and the profits start to dwindle.

Taking these trading strategies into the real-world never seems to work the way the advocates state. Either because the strategy itself really doesn't work but they've talked themselves into seeing a pattern that doesn't exist. Or it is simply too hard to put into practice due to trading costs and human factors.

And again, if these trading systems did work and the markets were so simple to take advantage of, then why isn't everyone doing it already? The firms on Wall St. spend millions a year looking for the slightest edge on each other. How did they manage to miss all this easy money? 

[MachineGhost]

Momentum trading doesn't cut risk; in my experience it tends to increase it as Black Swan events become a higher probability.  But momentum trading is not the same thing as trend trading.  Momentum is speed.  Trend is direction.

If you want to cut risk, you have to use fundamental factors so that you're not the bagholder.

MG

Stefan don't worry, I don't believe the efficient market hypothesis. I do believe that trends typically exist. My point was that in the (near) future they might become less prevelant or buried under random volatility or become shorter lived. I think Mark Cuban said that the entire "inovator, imitator, idiot" cycle has now shortened to seconds simply because everything is computorized. Trying to stay ahead of trends is massively more complex than simply sticking to a 4x25% PP. I also don't agree that it is possible to say that momentum trading cuts risk by cutting losses. Consecutive stop loss losses can accumulate to any level of loss.

[craigr]

Which is why the average investor should not bother and just accept the "Black Swan" risk by assuming the market is efficient.  The irrationality in the market comes from the naive investor not having self-control over their own emotional impulses.  It won't matter what method they use, they will eventually succumb to the Siren's Song.  That is why they are average.  Ego and hubris won't change this fact, only inner work.

Around 9/10 of the trades on any given day are between professionals, not individuals. The market average represents the best that even the 9/10 of the pros can do.

I'm not sure why you infer that Wall Street doesn't use technical approaches at the professional level successfully.  It certainly does and has been for decades.  Especially hedge funds which are notoriously secret.

Not saying they don't do these things. I'm simply saying it's worthless. And when you study research on hedge fund performance, they are the worst. Even worse than a typical mutual fund.

http://www.nytimes.com/2009/02/22/your- ... 2stra.html

"THERE’S yet more evidence that it makes sense to invest in simple, plain-vanilla index funds, whose low fees often lead to better net returns than hedge funds and actively managed mutual funds with more impressive performance numbers.

Basic stock market index funds generally aspire to nothing more than matching the returns of a market benchmark. So in a miserable year for stocks, index funds may not look very appealing. But it turns out that, after fees and taxes, it is the extremely rare actively managed fund or hedge fund that does better than a simple index fund.

That, at least, is the finding of a new study by Mark Kritzman, president and chief executive of Windham Capital Management of Boston. He presented his results in the Feb. 1 issue of Economics & Portfolio Strategy, a newsletter for institutional investors published by Peter L. Bernstein Inc."

But, whether or not Wall Street is successful at it, there are other behavioral and structural incentive factors that come into play than the ones the average investor deals with.  Do keep in mind that Wall Street and its clients as a collective is not very innovative in thinking outside the box, so it limits what can get accepted by committee and peer pressure.

I've been to financial firms on Wall St. many times in the past and had a chance to see some of their trading operations. Saying those guys do not think of every possible angle to make a buck is just not true. There is too much money at stake and big bonuses to be had for traders that show promise. I know guys that work on advanced FPGA analytical algorithms. Their colleagues work on FPGA technology for traders. There are actually magazines for the field (http://www.automatedtrader.net). They are looking for every possible edge to squeeze. And the harder they look, the more efficient the markets become. It is almost impossible to get an information edge over anyone else at those levels.

Sitting back and letting these guys slice each other to pieces on trading is the best strategy. You cannot compete against them with a home computer and technical analysis charts. The NSA would have a hard time competing against some of the computing power these firms have.

[stone]

Isn't it possible that the market is both not efficient and not predictable? Machine Ghost says that gold is underpriced relative to stocks. The fact that people have such opinions is what sets prices. BUT people are all trying to second guess what everyone else is thinking and that is what makes it unpredictable. The end result often seems to me just a sort of Brownian motion but a bit skewed into having sharper, less frequent, reversals.

To me the PP is a repudiation of the efficient market hypothesis. If the efficient market hypothesis made sense then wouldn't the optimal strategy be to hold a zoo of assets that tried to reflect the entire world. So hold stocks, land, bonds, bank reserves, art etc in ratios as they occur globally and not rebalance except to reflect when gold gets mined or an art gallery burns down or whatever. BUT if random jittering is reality then the optimal approach is pretty much the PP approach of just rebalancing against the random jitters (I'm classing "economic conditions" as being random jitters).

One thing that does puzzle me though is why the PP doesn't choose to extend its approach to holding fixed values of a few stocks rather than holding a capitalization weighted index. I suppose it might be possible that between stocks the situation is different than that between asset classes.

[Stefan]

PP, in its original form, works because it exposes the asset class prices in the portfolio to 3 upside drivers:

1. Economic growth, driving up stock prices. US GDP grew at an average 7%/year in nominal terms - 3.5%/year real terms, since 1900. PP assumes this growth will continue in the future, forever, as US becomes increasingly wealthier. And this is a very solid assumption, which held for the last 200 years. With potholes of 80% (1929-1933) to 50% (2008-2009) corrections down the road. But, on very long time horizons, you do collect a 7% CAGR risk premium on this asset class.

2. Goverment bonds. If you buy the bonds yourself (as oposed to buying an ETF or Mut Fund of bonds) you do collect the yield on the bonds AND, in 20+ years, the bonds return the principal. You do not lose capital and hopefully, if the inflation rate goes down (as it did from 1980 to now), you stand to gain a lot, as you locked in a higher yield along the way. And, whenver an economic crisis hits, the US treasury represents the most liquid safe heaven for the world's money.

3. Gold. Gold prices are driven up by many forces. US inflation, world inflation, fear. As inflation is perrenial, even if its rate fluctuates, it never goes away (in US,  the rate went down from a high inflation about 15-20% in the 1980 to about 2-3% today), gold is supposed to go up, perenially, as a hedge, as the inflation rate is always >0. Add world inflation and bouts of fear about debasing the dollar, euro, etc. and gold becomes a forever increasing currency, compared to a world currency basket, in the long time.

In conclusion, buy into these 3 asset classes, hold forever, and you achieve 3 key objectives:

1. Collect a perpetual income stream from the actualization of their risk premia, as per above.
2. Enhance the growth of your money by locking in gains and buying at a discount, via rebalancing => sell high & buy low. Collect the excess money accumulated in one asset class (like fear driven money in treasuries in a market crash) an buy the value in the market abandoned by the irrational crowd. This is a great mechanical (i.e. unemotional) method to accelerate your portfolio growth along the way.
3. Mitigate the risk to the portfolio of the deep potholes any one asset class will inevitably have along the way, because the other 2 asset classes are uncorrelated or inversely correlated and will compensate for it.

So, here you have it: an income stream in perpetuity at a lower risk than any separate asset class risk premium can offer. This is the genius of HBPP.

[MediumTex]

PP, in its original form, works because it exposes the asset class prices in the portfolio to 3 upside drivers:

1. Economic growth, driving up stock prices. US GDP grew at an average 7%/year in nominal terms - 3.5%/year real terms, since 1900. PP assumes this growth will continue in the future, forever, as US becomes increasingly wealthier. And this is a very solid assumption, which held for the last 200 years. With potholes of 80% (1929-1933) to 50% (2008-2009) corrections down the road. But, on very long time horizons, you do collect a 7% CAGR risk premium on this asset class.

2. Goverment bonds. If you buy the bonds yourself (as oposed to buying an ETF or Mut Fund of bonds) you do collect the yield on the bonds AND, in 20+ years, the bonds return the principal. You do not lose capital and hopefully, if the inflation rate goes down (as it did from 1980 to now), you stand to gain a lot, as you locked in a higher yield along the way. And, whenver an economic crisis hits, the US treasury represents the most liquid safe heaven for the world's money.

3. Gold. Gold prices are driven up by many forces. US inflation, world inflation, fear. As inflation is perrenial, even if its rate fluctuates, it never goes away (in US,  the rate went down from a high inflation about 15-20% in the 1980 to about 2-3% today), gold is supposed to go up, perenially, as a hedge, as the inflation rate is always >0. Add world inflation and bouts of fear about debasing the dollar, euro, etc. and gold becomes a forever increasing currency, compared to a world currency basket, in the long time.

In conclusion, buy into these 3 asset classes, hold forever, and you achieve 3 key objectives:

1. Collect a perpetual income stream from the actualization of their risk premia, as per above.
2. Enhance the growth of your money by locking in gains and buying at a discount, via rebalancing => sell high & buy low. Collect the excess money accumulated in one asset class (like fear driven money in treasuries in a market crash) an buy the value in the market abandoned by the irrational crowd. This is a great mechanical (i.e. unemotional) method to accelerate your portfolio growth along the way.
3. Mitigate the risk to the portfolio of the deep potholes any one asset class will inevitably have along the way, because the other 2 asset classes are uncorrelated or inversely correlated and will compensate for it.

So, here you have it: an income stream in perpetuity at a lower risk than any separate asset class risk premium can offer. This is the genius of HBPP.

+1

All you have to do is stay out of the way and keep your tinkering impulse in check.

[MachineGhost]

The PP cheerleaders make it sound like any tinkering is harmful.  That would be only the case if someone didn't have any perspective on proper money management skills, like plowing 25% into penny stocks or something stupid like that.  But I hear this anti-tinkering dogma so much, I want to hear about concrete examples where tinkering has destroyed the PP's ability to perform, robust as it is to begin with.

Any takers?

MG

[craigr]

It's a question of proving a negative. I can't prove that doing X, Y or Z won't make things better because the future is always moving and we won't know these answers for years or decades.

So for me, the onus is really to prove that the changes are an actual improvement by the proposer.

This change seems to help from empirical experience and historic data:

1) Holding ST cash instead of T-Bills

From my experience I can be possibly persuaded that these ideas may be a workable improvement. Maybe. With caveats:

1) Doing some value tilting in the stocks if you can handle the tracking error.

However these things have never been an improvement from my own experience:

1) Substituting a volatile component with something less volatile to improve the portfolio.
2) Market timing in any way shape or form.
3) Options strategies, margin investing, etc.
4) Anything that generates lots of transactions in the portfolio or moves away from very infrequent rebalancing.
5) Various other trading schemes.

I am all ears to someone if they can show me a low risk way to improve performance that doesn't rely on the above. But I've been looking and I just don't see it. And believe me, if I thought there was a way I could get an extra 1%+ a year for the same risk/reward and historical real return performance, I'd do it. Trust me.

Lastly, please understand that you are trying to take a strategy that has worked fine, remarkably fine, through a number of good and bad markets and performed well. Sure, it's not as best as the best hindsight tested strategy. But in terms of performance in unknown markets it has worked. So saying you want to take something that works fine and then improve it is going to be met with skepticism.

[HB Reader]

PP, in its original form, works because it exposes the asset class prices in the portfolio to 3 upside drivers:
In conclusion, buy into these 3 asset classes, hold forever, and you achieve 3 key objectives:

1. Collect a perpetual income stream from the actualization of their risk premia, as per above.
2. Enhance the growth of your money by locking in gains and buying at a discount, via rebalancing => sell high & buy low. Collect the excess money accumulated in one asset class (like fear driven money in treasuries in a market crash) an buy the value in the market abandoned by the irrational crowd. This is a great mechanical (i.e. unemotional) method to accelerate your portfolio growth along the way.
3. Mitigate the risk to the portfolio of the deep potholes any one asset class will inevitably have along the way, because the other 2 asset classes are uncorrelated or inversely correlated and will compensate for it.

So, here you have it: an income stream in perpetuity at a lower risk than any separate asset class risk premium can offer. This is the genius of HBPP.

Yes, that is pretty much it.

And you have an intregal cash reserve (for unexpected personal emergencies) automatically built into the structure.  And you don't have to make "market judgement" decisions -- you can accept virtually any long term verdict of the marketplace without fearing you will be permanently crippled.

I have also found this makes Variable Portofolio decisions much easier since they are far less consequential.  Obviously this depends on how much of your wealth you allocate to each approach. 

[MediumTex]

The PP cheerleaders make it sound like any tinkering is harmful.  That would be only the case if someone didn't have any perspective on proper money management skills, like plowing 25% into penny stocks or something stupid like that.  But I hear this anti-tinkering dogma so much, I want to hear about concrete examples where tinkering has destroyed the PP's ability to perform, robust as it is to begin with.

Any takers?

MG

Think of a modern car that rolls off the assembly line ready to be driven and perform according to its specifications.

In my view, such a well designed car shouldn't require its owner to tinker with the engineering of the vehicle in his garage in order to get optimal performance.

To me, the PP is sort of like a well designed car.  You can simply apply it and enjoy the safety and stability it provides without feeling the need to take it apart and re-arrange any of its components.

Now, there will always be that guy who modifies some component of his car in his garage and it does actually result in better performance, but for every one of these guys there are normally 99 other guys who tried to improve their car's engineering and failed because the car was well engineered to start with.

At some point, I think you just say "I like the PP's design and past performance and I'm just going to run it 'stock' for a while and enjoy the low stress investment experience it provides."

When running the PP "stock" stops working, THAT will probably be the time to talk about tweaks or maybe a different approach, but when it's putting together year after year of double digit returns, why not just enjoy the ride?

[MachineGhost]

Below are two performance charts comparing the "adolescent" PP (VTSMX, VUSTX, VFISX, LBMA PM Gold) with the "adult" PP (using the parameters I've outlined previously, but domestic equity only and now including gold timing).  While I don't believe there has been enough time to fully realize the potential risks present in LT bonds and gold, it should give anyone a good pause before they start tweaking or market timing.

[align=center][/align]
[align=center]9.01% annual return, 20.50% or so maximum drawdown.[/align]


[align=center][/align]
[align=center]6.73% annual return, 17.80% or so maximum drawdown.[/align]