The Permanent Portfolio is the Best Portfolio

[LFKB]

Let me start by saying I'm not sure the Permanent Portfolio is really the best portfolio, but I'm going to start with that opinion and see if I can be convinced otherwise by other on this forum who are much more experienced and well versed investors than I.

For those that don't know what the Permanent Portfolio is, it is a Portfolio consisting of 4 funds with re-balancing each year:
25% – Stocks (Total Stock Market)
25% – Long Term Treasury Bonds
25% – Gold
25% – Cash (in a Treasury Money Market Fund)

I find it hard to argue with the performance of the Permanent Portfolio over an extended period of time. Since 1972, it has yielded returns of 9-10% and its worst year has been a loss of only 4%. While I know that past performance is not an indication of future performance, given the diversification of the Permanent Portfolio and history of its results it seems like a much less volatile approach than say the Three Fund Approach or others that are championed on this forum (most of which I believe are great approaches btw).

Below are the min and max annualized returns over all of the samples from 1972-2011:
2-year: max return 26.1% / min return 1.7%
3-year: max return 21.2% / min return 2.6%
5-year: max return 16.1% / min return 5.1%
10-year: max return 13.2% / min return 6.3%

Given the Permanent Portfolio has yielded similar returns (over 9%) to riskier portfolios consisting primarily of equities and has never suffered a major drop, it stands to reason that it is superior to other more volatile portfolios and would also be an acceptable approach for shorter investment time horizons of only a few years. For example, in 2008 it fell only 0.7% due to the strength in bonds and to a lesser extent, cash and gold while other equity based portfolios would have lost ~30% of their value.

I've only been on this site for a week and I'm sure that there are plenty of holes in my thinking. It would be great to get others thoughts on this Portfolio and why they have chosen their portfolio over it.

Below is a more information on the Permanent Portfolio for those that would like to read up on it...
http://crawlingroad.com/blog/2008/12/22/permanent-portfolio-historical-returns/

[Elbowman]

I don't think the permanent portfolio is terrible or anything, but you must admit that you are touting a portfolio that is light on stock and heavy on gold (relatively), after a decade that was below average for stocks and far, far above average for gold, not to mention the effect a historic drop in interest rates has had on long term bonds. What was the annual return for the PP 1972-1999?

The PP may in fact be very good, but its hard to make that argument right now without appearing to have a great deal of recency bias.

[WendyW]

At any point in time, one can identify oddball portfolio allocations that have performed well over the past, say, 30- or 40-year period.

In 2012, the Permanent Portfolio is a great example of a portfolio that has performed well in recent memory.

The problem is that the past is not the same as the future. This kind of backtesting is great for predicting the past, but very poor at predicting the future.

Instead of just considering the Permanent Portfolio's performance 1972-2012, also check out how it has performed 1932-1972, etc.

[mpt follower]

PP has been discussed to death! Just use this website search tool and you will be able to get enough opinions to keep you busy for a week.
My opinion is that it looks great, it has a great history in terms of performance and risk, but it has scary assets, each independently rather volatile, making many investors to think twice about the portfolio. Plainly, most do NOT have the guts to own it.
By the way, I think that the cash portion was originally intended to be in Swiss Francs, not in a money market account.

[nisiprius]

In real life, when a real mutual fund tried to implement it with real dollars, this is what happened. Note that Browne was one of the founders of the fund, and I presume would not have allowed the use of the name "Permanent Portfolio" if he didn't think it was a fairly reasonable proxy for the portfolio itself.

Blue: Permanent Portfolio Fund; Orange: Wellesley (a conservative Vanguard fund that invests only in traditional securities); Green, Vanguard Prime Money Market Fund



The last decade for the fund was, in fact, great, because gold did great, but any claim that it does well as a "permanent" holding under all market conditions was not borne out by experience. Over thirty years, the Permanent Portfolio Fund has averaged 6.9% annualized, while Wellesley averaged about 10.4%. It gave a smoother ride, but from 12/31/2007 to the bottom, it lost 15% while Wellesley lost 18%. $125,000--the difference in growth of $10,000 over thirty years--is a lot to pay for that shock absorber.

And while it ultimately trounced a money market fund, investors had to wait two decades for that to happen.

Supporters have alibis. They say, and I'm sure they are correct, that the fund didn't really implement the Permanent Portfolio. That's not my fault. If the Permanent Portfolio is so trivially easy to implement in the real world, why didn't the fund just implement it? Why would I, as a retail investor, think that I could do far better implementing the portfolio myself than a professional fund managers could?

Sure, I could cut out some of the 0.71% expense ratio, but even if I cut out all of it that would still leave a 2.79% performance gap between it and Wellesley.

[Joe S.]

In real life, when a real mutual fund tried to implement it with real dollars, this is what happened. Note that Browne was one of the founders of the fund, and I presume would not have allowed the use of the name "Permanent Portfolio" if he didn't think it was a fairly reasonable proxy for the portfolio itself...

Nisiprius makes some good points. However I must add there have been two major versions of the Permanent Portfolio. Harry Browne created the first in 1981, and the mutual fund was based on this first "PP." He then created a second "PP" in 1998 which is the more famous "PP," with the 25-25-25-25 allocation. It had a much heavier allocation of long bonds than the first PP.

Source for this information:
http://www.harrybrowne.org/Books.htm
Harry Brown books:
Inflation-Proofing Your Investments 1981 First PP
Fail-Safe Investing 1998 Second PP

Addendum: I should say there have been at least two major versions of the PP. There may be others I am unaware of.

[BBL]

"The best" will only ever exist in the rear view mirror [and will be very much period-dependent]. "The best" for each individual investor is the portfolio that is appropriate after numerous personal considerations are factored into the plan - timeline, goals, risk tolerance, etc, etc....

I'm going to start with that opinion and see if I can be convinced otherwise by other on this forum

You go first -

Here is what I say is the best portfolio [tongue firmly in cheek]:

50% US SV
50% Individual US treasuries with 5.5-6.0 years until maturity purchased free in an IRA brokerage account and sold after 1-1.5 years to ride the yield curve.

Can you convince me I'm wrong???

[Call_Me_Op]

There are other portfolios (based upon back-testing) that have performed better than the Permanent Portfolio - and which own no long-term bonds and little if any gold. One example is:

25/10/65 Diversified equity/Gold/5-Year T-notes

If you like the Permanent Portfolio that much, consider the prospect that you had a 1 million dollar windfall today. Would you gladly put 250k into long-term treasuries (at the tail-end of the greatest bond bull-market in US history) and another 250k into gold (an investment with no known method for valuation and following a long run-up)?

[Joe S.]

There are many threads on the Permanent Portfolio here, and I suggest you read them and add comments as you feel necessary. A simplified form of my argument against the PP would be this:

The Permanent Portfolio has a volatility similar to a portfolio of 25% stocks and 75% bonds. So you don't need gold for a low volatility portfolio.
The Permanent Portfolio has a higher return than a 25-75 portfolio from 1971-present, but that is only because gold did abberantly well from 1971-present. Many of us feel that gold will return to a real return of 0% (or even worse). In such a case a 25-75 portfolio will do better.

A full understanding of Boglehead arguments against the PP will require reading many threads dating back over a year.

[stemikger]

I'm neither pro or con for the PP because I just don't know too much about it. However, it will take a lot of work to make that shift in the minds of many people who are saving for retirement. I don't even know any companies that have gold as an option. It may be the best (I'm not qualified to determine that) but look how long it took people to get the idea of indexing and many still don't get it after all these years.

[Calm Man]

I have been intrigued by the PP since I've read about it. There are a few problems that have kept me away from it. First, in a taxable account it doesn't work. Also, it requires a gold ETF and I just don't trust those things for some reason. I just don't see it frankly being much different from 50/50 stocks and bonds but maybe that's just me.

[Call_Me_Op]

I have been intrigued by the PP since I've read about it. There are a few problems that have kept me away from it. First, in a taxable account it doesn't work.

How do you figure? Stock index funds are very tax efficient, gold if purchased through a CEF is quite tax efficient, and bond interest can compound if taxes are paid out of the cash allocation.

[Browser]

The problem with backtesting using compounded returns is that the results are completely dependent on the time period selected. It is a great method to torture the data to make it confess to whatever you want it to.

For example, looking at the classic PP (25% TSM, 50% ITT, 25% Gold) we see that it stays ahead of Wellesley and a 50% TSM/50% ITT portfolio for 20 years if you start in 1972. And It beats 50/50 over the last forty years from 1972-2011 while producing the same return as Wellesley over this period. So I guess that proves that the PP is a fine allocation - it does as well or better than the conventional alternatives over the last four decades and avoids the large bear market drawdowns of the 1970s and since 2000. Case closed.

But on the other hand, if you start in 1982 it underperforms conventional portfolios to a substantial degree, only returning about half the compounded returns of the latter. Whoops, I guess that proves the PP is a rotten allocation, since it was a miserable flop compared to conventional stock/bond portfolios over the last three decades. Case closed.

Backtesting does an excellent job of predicting the past, and also makes various predictions of the past depending on which past period you want to predict. I'm working on a time machine that will let me select the particular period of time in which to invest. I can't predict when it will be on the market.

[rmelvey]

The PP has historically beat the CPI by around 4.5% annual with relatively few large drawdowns.

The PP has not beat every other allocation on a consistent basis. But that is not the point of the PP.

These comparisons to Wellesley are other stock bond funds are simply foolish. The PP offers more consistent real returns so it is obviously going to have times where it underperforms the less consistent strategies. Would you rule out Wellesley because it underperformed an all equity fund in the 90s? Of course not.

If your denominator is real goods and services, the PP returns have been consistent. If your denominator is a less consistent strategy (such as the Wellesley) obviously the returns are not going to be consistent.

[wesleymouch]

The PP would also have protected you during the Great Depression. Cash and bonds would have survived. Gold nearly doubled ($20 to $35 but the price was fixed by government). Most telling if you lived in Iceland it would have preserved your wealth when that currency collpased. A classic 60/40 portfolio would have been wiped out. Hulbert tracks newsletters and the Permanent Portfolio ranks in the top 20% of newsletters if it were one. I guess it is safe to say that the PP is great for capital preservation and keeping up with inflation but underperforms during bull markets. If you have already made your wealth it preseves it. If you need to make a return to get wealth it is not as good as classic 60/40 portfolio.

[EDN]

Let me start by saying I'm not sure the Permanent Portfolio is really the best portfolio, but I'm going to start with that opinion and see if I can be convinced otherwise by other on this forum who are much more experienced and well versed investors than I.

For those that don't know what the Permanent Portfolio is, it is a Portfolio consisting of 4 funds with re-balancing each year:
25% – Stocks (Total Stock Market)
25% – Long Term Treasury Bonds
25% – Gold
25% – Cash (in a Treasury Money Market Fund)

I find it hard to argue with the performance of the Permanent Portfolio over an extended period of time. Since 1972, it has yielded returns of 9-10% and its worst year has been a loss of only 4%. While I know that past performance is not an indication of future performance, given the diversification of the Permanent Portfolio and history of its results it seems like a much less volatile approach than say the Three Fund Approach or others that are championed on this forum (most of which I believe are great approaches btw).

Below are the min and max annualized returns over all of the samples from 1972-2011:
2-year: max return 26.1% / min return 1.7%
3-year: max return 21.2% / min return 2.6%
5-year: max return 16.1% / min return 5.1%
10-year: max return 13.2% / min return 6.3%

Given the Permanent Portfolio has yielded similar returns (over 9%) to riskier portfolios consisting primarily of equities and has never suffered a major drop, it stands to reason that it is superior to other more volatile portfolios and would also be an acceptable approach for shorter investment time horizons of only a few years. For example, in 2008 it fell only 0.7% due to the strength in bonds and to a lesser extent, cash and gold while other equity based portfolios would have lost ~30% of their value.

I've only been on this site for a week and I'm sure that there are plenty of holes in my thinking. It would be great to get others thoughts on this Portfolio and why they have chosen their portfolio over it.

Below is a more information on the Permanent Portfolio for those that would like to read up on it...
http://crawlingroad.com/blog/2008/12/22/permanent-portfolio-historical-returns/

LFKB,

I think a simple traditional portfolio of stocks and bonds is superior to the Permanent Portfolio, assuming the traditional portfolio takes advantage of the expected sources of stock & bond return. More specifically, we know over time that smaller and more value oriented stocks have had higher-than-market returns with less than perfect correlation with large growth companies. The same is true in non-US markets, with even better diversification. In the bond markets, 5yr maturities have been optimal, and sticking with government bonds has been most effective at reducing portfolio risk.

So when I compare a traditional asset class portfolio using the above assumptions and adjust for the risk of the Permanent Portfolio, I come up with the following comparison from 1975-2011 (the first year it was legal to own gold):

40% Stock/60% Bond "traditional asset class portfolio"
Annualized Return = +11.6%
Standard Deviation = 8.0
# of down years = 3 (ave loss of -3.5)

25% S&P 500, 25% Gold 25% LT Government Bonds, 25% 1mo t-bills
Annualized Return = +9.0%
Standard Deviation = 7.8
# of down years = 4 (ave loss of -1.9)

So it looks like there has been about a -2.6% cost to the Permanent Portfolio relative to a similar risk traditional asset class portfolio*. That's not a cost I'm willing to pay.

*6% S&P 500, 8% US large value index, 14% US mid/small value index, 6% Int'l value index, 6% Int'l small index, 60% 5YR T-Notes

[DaveS]

During the period you selected gold went from a historic low to a historic high or near historic. Interest rates on long term debt fell from a high in the 1980's of about 11% to the present historic low. Money market funds now yield less than inflation. The future looks bad for the permanent portfolio. Gold is unlikely to appreciate from it's present high. Interest rates are likely to rise when they do long term bonds will decline by the duration of the fund. Say rates go up 3% then a fund with a duration of 20 will loose more than half of it's value. Money market rates are unlikely to beat inflation. When you consider the future the permanent portfolio does not look good at all. Dave

[wellmoneyed]

Most telling if you lived in Iceland it would have preserved your wealth when that currency collpased. A classic 60/40 portfolio would have been wiped out.

I don't believe that is true. Stocks hold companies with actual property. If you look at what happened in Germany during their currency collapse, within a few years they were back to even. I have not looked into Iceland, but I would assume the same thing would happen. Please let me know if you have different information concerning Iceland. Further more, most people with a 60/40 today have a portion in international stocks. Assuming what I am saying is true, then the permanent portfolio would actually lose more in a currency collapse than a typical 60/40. Am I wrong?

[LFKB]

What was the annual return for the PP 1972-1999?

The PP may in fact be very good, but its hard to make that argument right now without appearing to have a great deal of recency bias.

From 1972 to 1999 the annualized return was 10.4%

"The best" will only ever exist in the rear view mirror [and will be very much period-dependent]. "The best" for each individual investor is the portfolio that is appropriate after numerous personal considerations are factored into the plan - timeline, goals, risk tolerance, etc, etc....

I'm going to start with that opinion and see if I can be convinced otherwise by other on this forum

You go first -

Here is what I say is the best portfolio [tongue firmly in cheek]:

50% US SV
50% Individual US treasuries with 5.5-6.0 years until maturity purchased free in an IRA brokerage account and sold after 1-1.5 years to ride the yield curve.

Can you convince me I'm wrong???

I don't actually think it is the best and I know no portfolio is actually the best. I was starting with a hypothesis and waiting for others here to tell me why I am wrong and some good points have been brought up. The title of the thread was a bit tongue in cheek as well.

If you like the Permanent Portfolio that much, consider the prospect that you had a 1 million dollar windfall today. Would you gladly put 250k into long-term treasuries (at the tail-end of the greatest bond bull-market in US history) and another 250k into gold (an investment with no known method for valuation and following a long run-up)?

I am looking to invest a large amount of money soon and while I may use the PP or some modified version, I would lump sum half and DCA the other half over 12-24 months. I know DCA isn't usually recommended around here but it would help me sleep better at night.

For example, looking at the classic PP (25% TSM, 50% ITT, 25% Gold) we see that it stays ahead of Wellesley and a 50% TSM/50% ITT portfolio for 20 years if you start in 1972. And It beats 50/50 over the last forty years from 1972-2011 while producing the same return as Wellesley over this period. So I guess that proves that the PP is a fine allocation - it does as well or better than the conventional alternatives over the last four decades and avoids the large bear market drawdowns of the 1970s and since 2000. Case closed.

But on the other hand, if you start in 1982 it underperforms conventional portfolios to a substantial degree, only returning about half the compounded returns of the latter. Whoops, I guess that proves the PP is a rotten allocation, since it was a miserable flop compared to conventional stock/bond portfolios over the last three decades. Case closed.

Backtesting does an excellent job of predicting the past, and also makes various predictions of the past depending on which past period you want to predict. I'm working on a time machine that will let me select the particular period of time in which to invest. I can't predict when it will be on the market.

I provided the min and max returns over all of the testable time horizons in my OP. Sure you can find another portfolio that beats the PP over certain years, but if you look collectively at all the data it shows a pretty good picture with low volatility. It's not like I hand selected one time period and used that to frame my argument (that is what you did in your response). Also, I think most everyone in the world essentially relies on backtesting for their AA, we use backtesting to tell us that equities return higher than bonds over long historical periods. It seems like you are referring to selective backtesting in your response, but this is not what I have done.

Here are more min and max returns over all testable time periods from 1972-2011
Min CAGR / Max CAGR
2 Years - 1.7% / 26.1%
3 Years - 2.6% / 21.2%
5 Years - 5.1% / 16.1%
10 Years - 6.3% / 13.2%
15 Years - 6.7% / 13.0%
20 Years - 7.3% / 11.6%
25 Years - 7.8% / 10.7%
30 Years - 8.1% / 9.8%

[LFKB]

LFKB,

I think a simple traditional portfolio of stocks and bonds is superior to the Permanent Portfolio, assuming the traditional portfolio takes advantage of the expected sources of stock & bond return. More specifically, we know over time that smaller and more value oriented stocks have had higher-than-market returns with less than perfect correlation with large growth companies. The same is true in non-US markets, with even better diversification. In the bond markets, 5yr maturities have been optimal, and sticking with government bonds has been most effective at reducing portfolio risk.

So when I compare a traditional asset class portfolio using the above assumptions and adjust for the risk of the Permanent Portfolio, I come up with the following comparison from 1975-2011 (the first year it was legal to own gold):

40% Stock/60% Bond "traditional asset class portfolio"
Annualized Return = +11.6%
Standard Deviation = 8.0
# of down years = 3 (ave loss of -3.5)

25% S&P 500, 25% Gold 25% LT Government Bonds, 25% 1mo t-bills
Annualized Return = +9.0%
Standard Deviation = 7.8
# of down years = 4 (ave loss of -1.9)

So it looks like there has been about a -2.6% cost to the Permanent Portfolio relative to a similar risk traditional asset class portfolio*. That's not a cost I'm willing to pay.

*6% S&P 500, 8% US large value index, 14% US mid/small value index, 6% Int'l value index, 6% Int'l small index, 60% 5YR T-Notes

Thanks for the well informed response. What was the maximum loss in any given year of the 40/60 stock bond portfolio you referenced? How would you compare the downside risk of each portfolio if investing over a shorter time horizons (say 5 years).

I think I know the answer, but most Bogleheads allocate 0% of their portfolio to gold, right?

[craigr]

Let me start by saying I'm not sure the Permanent Portfolio is really the best portfolio, but I'm going to start with that opinion and see if I can be convinced otherwise by other on this forum who are much more experienced and well versed investors than I.

Well I'm pretty big on the portfolio (I co-authored a book about it), but I'm not going to say it's "best", I'll simply list out the pros and cons:

Pros:

- It's a real return allocation.
- Diversification based on economic cycles and is future agnostic.
- Diversifies against non-spreadsheet risks.

Cons

- Doesn't swing for the fences for returns.
- Will lag any single overweighted asset.

It's a real-return allocation.

Meaning it targets real returns over inflation (usually +3-6%) and has had a long track record of achieving that goal. A problem I see with stock/bond allocations is they can leave an investor with flat or even negative real returns for extended periods of time. For instance the 1970s were very bad for just about any stock/bond allocation. Most all I've looked at had serious problems with either 0% real or negative real returns over that decade. The 2000s were also very unfavorable to those allocations. But over rolling 10-year periods of time the Permanent Portfolio has not had protracted periods of these negative real returns.

Below are two charts from Simba's spreadsheet illustrating this idea of the Permanent Portfolio vs. the standard 60/40 portfolio:


Permanent Portfolio Rolling Real Returns


60/40 Rolling Real Returns


The charts shows:

1) During really good years the stock heavy portfolio turned in great real returns (1980-2000). Unfortunately, this is the biggest longest sustained bull market in U.S. history. It may repeat, but then again it might not.

2) The Permanent Portfolio had lower real returns during the heyday of stocks. But the returns are consistent through the entire time. The 60/40 portfolio you see had big problems with real returns in the 1970s and 2000s as discussed. The Permanent Portfolio fluctuated between that +3-6% real return range. So the overall returns are far more consistent.

Diversifies based on economic cycles and is future agnostic.

Next, the portfolio chooses assets that are tied specifically to particular economic situations. Mostly, it works pretty well because there are just basic economic reasons why someone wants to own bonds (falling rates) and why someone wants to own stocks (a good outlook for the economy). I find this approach works much better to diversify against market risks than standard stock/bond allocations that rely on asset class correlations. I think that model of diversification is seriously broken and I don't trust it. I discuss the basic problem with the entire idea in this post:

http://crawlingroad.com/blog/2012/06/03 ... esnt-work/

Lastly, the portfolio really is future agnostic. A stock heavy portfolio assumes that the stock market is going to win over an investor's particular time horizon. However if you look at world markets (not just the U.S.) this is simply not true. Even in the U.S. the idea that stocks always win over time has been seriously challenged. I don't think it is a very good assumption myself so I prefer to hold a more balanced portfolio and not make any assumptions about what will happen. Stocks are much riskier than given credit. IMO. They should be used in a balanced way.

Diversifies against non-spreadsheet risks.

Spreadsheets miss a ton of risks that can show up over an investor's lifetime. Things like Bernie Madoff, MF Global, real estate market panics, natural disasters, government interventions, etc. Think about investing for decades and everything that has or could happen going forward. The risks are limitless. So I think it's a really great idea to diversify your assets even against things you think are unthinkable. This means dividing up your money among more than one brokerage and even keeping some outside the country where you live.

But you know some people think this stuff is paranoid and I get it. My views are colored very much by working in the unpredictable world of start-ups and doing a lot of world-travel (25+ countries at this point). I just got back from a trip to Argentina where inflation is perhaps 25%+ and where the currency exchangers (Travelex) won't even buy Argentine Pesos back from you. I'm not saying that any country is going to go the route of Argentina, but I will simply say this:

Inside every paper currency there is an Argentina fighting to get out.

Showing a red-hot CAGR in a spreadsheet backtest is pretty easy. I can do it right now. But there are other non-spreadsheet risks that I think investors should consider and the Permanent Portfolio does consider them as part of the model.

Now for the cons.

First of all, the portfolio is not swinging for the fences for returns. I think people doing that approach are probably going to fail long term anyway because the volatility will wipe them out emotionally. Also I think that investors should take risks on their careers, and not with their life savings. I discuss this idea here:

http://crawlingroad.com/blog/2012/07/09 ... o-podcast/

You can't go back and re-earn the money you take huge gambles with in your portfolio. This risk gets incredibly higher the longer you go out on the timeline of life. Taking a big loss late in your investing career can be fatal. Even taking it mid-way can be a big problem because it could force you into a far too conservative allocation that won't grow enough. I think the Permanent Portfolio has a great balance of growth with risk protection. This translates to investors being able to stick with the plan no matter what the markets are doing.

The next con is it will lag any particular single asset. So if stocks are doing really well, you'll hear everyone around you bragging about their hot returns when you have comparatively little. Of course the portfolio likely is beating inflation by a decent amount so you are doing OK, but you probably aren't going to pull down 15%. Then again, you probably aren't going to take a -30% hit either. Investors in the strategy need to simply accept that the portfolio is pretty boring and you won't have much to talk about at cocktail parties. But personally, I want my life savings to be pretty boring so I'm OK with that.

I can probably add more, but you get the idea. The Permanent Portfolio is "best" for some people that want:

- Low volatility
- Wide diversification
- Likelihood of real returns
- Protection against external events in the market.

If that's what you want, it may be a good fit.

[Browser]

Let's do a thought experiment, examining two 12-year historical periods:

In the 12-year period 2000-2011, the PP outperformed a 60/40 portfolio by an average of 3.4% per year with half the average annual volatility. It generated a compound annual return 50% higher than 60/40. Human nature being what it is, it is not a complete surprise that it's gotten quite popular. The Permanent Portfolio Fund has grown from $52M in 2001 to $17.1B today. Wearing today's sunglassses, it's quite natural to view the past quite favorably in regard to this investing strategy.

Let's travel back in our time machine to 1993. In the 12 years since 1982, the 60/40 portfolio has had an annual average return that was 4.2% higher than the PP. While the volatility was higher than the PP, it wasn't large - only 9.1% annually averaged. At the end of 1993 the compound annual return was 54% greater than the PP. Almost the reverse mirror image of the 2000-2011 period. Wearing 1993's sunglasses, it was quite natural to view the past quite favorably in regard to investing heavily in equities - maybe bonds too, which had also done quite well.

Back in 1993, do you think we would be having this conversation about the PP? Would there be advocates of the PP pounding the table about the virtues of this investing strategy? Would there be endless threads discussing the PP? Even though advocates do their best to assure us that the PP did just fine during that period in the grand scheme of things, "from a distance", in real return terms, etc. I'd hazard a guess - NOPE. It's all about performance-chasing folks. When something does well, especially if it does well for more than an investing nanosecond, the rationalizers begin coming out of the woodwork to provide us with compelling explanations backed by reams of data why that strategy was destined to have done so well and why it will continue to do well for the indefinite future. But if history does provide any guide at all in these matters, it suggests that this usually happens near the point that the strategy unexpectedly, irrationally, and impossibly goes down the chute. Maybe the PP really is different, though. How much are you willing to bet on it?

[Bungo]

Instead of just considering the Permanent Portfolio's performance 1972-2012, also check out how it has performed 1932-1972, etc.

Wasn't it illegal for private citizens to own gold during 1932-1972?

[KyleAAA]

I have a problem with calling ANY portfolio "the best" because you really can't know that in advance, but the PP is a very fine portfolio. There's probably not much to be gained from discussing it further. Those who like it, like it. Those who don't, won't. For the record, I believe a Swedroe-esque barbell portfolio would be a better choice but that's just me.

[Joe S.]

Instead of just considering the Permanent Portfolio's performance 1972-2012, also check out how it has performed 1932-1972, etc.

Wasn't it illegal for private citizens to own gold during 1932-1972?

It was illegal to own gold from 1933 until Jan 1 1975.
http://www.the-privateer.com/gold2.html

Addendum:
I may also add that the Permanent Portfolio's four components are designed to have a low correlation with each other. Before 1971,we were on the gold standard and the dollar was pegged to gold, so two components, gold and cash were unable to move independently. It is argued you therefore cannot evaluate the Permanent Portfolio outside the period 1971-2012.

However 1971-2012 was a period where gold had an aberrantly high return, so it can also be argued that it is inappropriate to confine analysis of the Permanent Portfolio to the years 1971-2012. So we argue back and forth over what years to use.

[marc515]

Tax question:

If we put together a PP, isn't the gold taxed at a higher "Collectable" rate?

But if we buy the Permanent Portfolio (PRPFX) the fund is taxed at regular equity rates.

[craigr]

Tax question:

If we put together a PP, isn't the gold taxed at a higher "Collectable" rate?

Gold is taxed at a collectibles rate, or your marginal rate, whichever is lower. Most people are not likely to pay the full collectibles rate on gold, and even then you rarely need to rebalance it. So taxes are not what people may think they are.

There are also some closed end funds where you can apply for an exemption to allow you to claim a lower long term capital gains rate on the funds.

With that said, each person is different so it's best to talk to a qualified tax expert about your personal situation.

But if we buy the Permanent Portfolio (PRPFX) the fund is taxed at regular equity rates.

The fund was initially set up as a tax-managed fund. They did/do a lot of things to limit taxes which is why Morningstar has them ranked #1 in their category over the past 10 and 15 years last I checked. I know one of the founders of the fund and they set out specifically to make it as tax efficient as possible, which is quite a rare thing considering it was setup back in 1982. Most funds don't care much about taxable investors…(disclosure: I don't own any of the fund).

[grayfox]

Let me start by saying I'm not sure the Permanent Portfolio is really the best portfolio, but I'm going to start with that opinion and see if I can be convinced otherwise by other on this forum who are much more experienced and well versed investors than I.
...

Most of the time it's silly to argue about who or what is the best. Who's the best baseball player of all time? Babe Ruth? Hank Aaron? Barry Bonds? Ty Cobb? You could argue this forever.

The only logical thing is to come up with 1) a measure of best that can be ranked and 2) the field of candidates for best.

Going back to Simba's spreadsheet, he has a worksheet for 25 lazy portfolios. One of them is HBPP. So I will take that as the field of competitors.

That worksheet page has a ton of metrics. I'll show the winner of the other funds and compare to HBPP in several

Metric..................Best......Value..........HBPP
$1 Portfolio............P15.........$16.23.........$8.21
CAGR...................P15.........10.18%..........8.11%
Avg Return.............P11.........12.26%..........8.19%
Std Dev.................P25..........8.44%..........5.82% *
Sharpe.................P25...........0.75............0.69
Sortino................P25...........1.50............1.39
Last Place.............P2.............5................8 years *
First Place............P3.............4................4 years *

P2 = Wellesley
P11 = Ted Aronson Family Taxable
P15 = David Swenson Yale Endowment
P25 = Larry Swedrow Mininize Fat Tails

The lowest risk portfolio is HBPP.
The highest return is Yale Endowment
The best Sharpe Ratio is Larry Swedroe Minimize Fat Tails. HBPP is second.
HBPP tied for best in the number of 1st place years, 4. But HBPP also had the most last place year. 8. Most of the last place years were when the other lazy prtfolios all did well.

If I had to award "the overall best", I would say it is the one with the highest Sharpe ratio.
The gold medal goes to...

...Larry Swedroe Minimize Fat Tails Portfolio. HBPP gets the silver. Now that's irony.

Does this mean that we should set aside out golden idols, and worship at the altar of small-cap value and TIPS?

All this is just looking in a rear view mirror. Past relative returns are not good forecasts of future relative returns.

[WendyW]

For those that don't know what the Permanent Portfolio is, it is a Portfolio consisting of 4 funds with re-balancing each year:
25% – Stocks (Total Stock Market)
25% – Long Term Treasury Bonds
25% – Gold
25% – Cash (in a Treasury Money Market Fund)

If I had to predict the real returns of these asset classes over the next 20 years, I would estimate the following:

Stocks = 5%+/yr
Bonds = 0%/yr
Gold = 0%/yr
Cash = 0%/yr

For this reason, the permanent portfolio does not appeal to me.

I predict that all of those that are piling into the permanent portfolio in 2012, will regret their decision.

Of course, I've been wrong before, and I'll be wrong again.

[wesleymouch]

If you know what stocks will return then you should invest 100% of your money in stocks. By the way how was your prediction of the performnce of assets for the last 10 yrs? If you knew that then you would have invested in gold and long term bonds

[WendyW]

If you know what stocks will return then you should invest 100% of your money in stocks. By the way how was your prediction of the performnce of assets for the last 10 yrs? If you knew that then you would have invested in gold and long term bonds

I continue to have about 80% of my money in stocks, and am feeling pretty good about this allocation.

I am confident that long bonds won't have the same tailwind (ie. continuously declining interest rates) over the next 30 years that they've had over the past 30 years.

I'm not confident that the current gold bubble won't burst in the near future.

If I could predict asset performance 10 years in advance, I would have invested in Apple and Priceline in 2002, not in gold or long bonds.

Smell ya in 2032.

[wesleymouch]

Apple and Priceline are stocks not asset classes.

[WendyW]

Apple and Priceline are stocks not asset classes.

Apple stock and Priceline stock are assets.

[Joe S.]

The only logical thing is to come up with 1) a measure of best that can be ranked and 2) the field of candidates for best.

Going back to Simba's spreadsheet, he has a worksheet for 25 lazy portfolios. One of them is HBPP. So I will take that as the field of competitors.

You are assuming that we accept Simba's data of 1972 to present as the proper data to use. Many of us feel this is a period of time that gold did abberantly high and therefore gives an unfair advantage to gold-based portfolios. The truth is we can't agree on what data to use.

[wesleymouch]

I think the point is that the PP hedges you against tail risk that other conventional portfolios do not. Imagine you are an Icelandic investor with a classic 60/40 portfolio, an Argentine investor during the 2001 corralito and devaluation. You were wiped out. A PP could have potentially saved you from these crippling losses. Another point that Harry Browne makes that you rarely see written elsewhere is that you should hedge political risk also. He suggested foreign bank accounts or holding some of your gold outside your country of residence. A lot of Venezuelan, Cuban, Argentine, Icelandic and Weimar Germany investors would have been glad to have taken that advice.

[Joe S.]

If I could predict asset performance 10 years in advance, I would have invested in Apple and Priceline in 2002, not in gold or long bonds.

Apple and Priceline are stocks not asset classes.

Apple stock and Priceline stock are assets.

For our portfolio we want only the best. Commodities are an asset class. Let's retrospectively pick the best commodity, gold. Bonds are an asset class. Let's retrospectively pick the best bonds, the long bonds. Similarly we should retrospectively pick the best stocks, like Apple and Priceline. We will call this portfolio the PPPP: Pick the Perfect Parts Portfolio. It's been shown by Simba's data to be the best.

[WendyW]

'
My point is not that we should ignore historical returns.

My point is that if we're going to construct a portfolio based on historical performance, using only data from the past 10 or 20 or 30 years is inadequate: We'll end up with a portfolio (eg. Apple+Gold+Long Bonds) that produced great returns in the recent past, but that is unlikely to do the same in the future.

If we're going to assemble our portfolio based on backtesting, we should use data from some reasonably long historical period. For example:

Annual real returns 1871–2001
Stocks = 6.8%
Gold = -0.1%
Bonds = 2.8%

[Joe S.]

'
]\My point is that if we're going to construct a portfolio based on historical performance, using only data from the past 10 or 20 or 30 years is inadequate

Agreed. I would only add that using only the last 41 years is also inadequate.

[WendyW]

I would only add that using only the last 41 years is also inadequate.

Agreed. The annual real return data that I quoted above covers a 130 year period.

[steve r]

I would say why go back only 130 years that happen to correspond with the Industrial Revolution and a period of tremendous economic advancement unlike any seen in the last milenium or two.

Picking 1871 to 2001 is also odd. Why ignore the stock market disaster that occured the previous decade of the 1860s. Why stop in 2001 and ignore the stock marekt disaster of the subsequent decade? This happens to exclude two good periods for gold.

This is of course the entire point of the permanent portfolio. You are not trying to predict if stocks or gold is going to act like they did in the 1870s & 1990s or 1860s & 2000s. Gold has historically provided a positive real return going back to the 1830s (links below) in dollars and a much better real return for any country that experienced hyperinflation (which includes some of the most developed economies in the world).

I might add to compare the PP with Wellsley (or any other portfolio) which we now know did very well over the last 40 years is simply silly. Hindsight is 20/20.

To me the best reason NOT to use the PP is one needs to accept a period like the 1990s where you will underperform for a decade. This, I suspect, will make it very hard to stay the course.


CPI since 1800 from Fed
http://www.minneapolisfed.org/community_education/teacher/calc/hist1800.cfm?
Gold price
http://www.nma.org/pdf/gold/his_gold_prices.pdf

[LFKB]

'
My point is not that we should ignore historical returns.

My point is that if we're going to construct a portfolio based on historical performance, using only data from the past 10 or 20 or 30 years is inadequate: We'll end up with a portfolio (eg. Apple+Gold+Long Bonds) that produced great returns in the recent past, but that is unlikely to do the same in the future.

If we're going to assemble our portfolio based on backtesting, we should use data from some reasonably long historical period. For example:

Annual real returns 1871–2001
Stocks = 6.8%
Gold = -0.1%
Bonds = 2.8%

Why choose 1871-2001? The dates in my OP are based on when gold has been freely traded. Also, I provided the min and max returns for every 2, 3, 5, 10, 15, 20, 25 and 30 year period in that time frame, so I wasn't selectively picking data as you did.

Btw, I do 't necessarily think the PP is the best portfolio as others have suggested, nor do I think there is a best portfolio. I was putting a premise out there to get reactions and start a discussion. However, I do think the PP serves as a less volatile, suitable strategy that could be implemented over a short or long time frame based on it's low volatility, unlike a portfolio of over 50% equities which could lose a substantial % of value over a short time frame.

[Joe S.]

I would say why go back only 130 years..

How about thousands of years.

When I was a teenager, I read Harry Browne's 1970 book "How You Can Profit from the Coming Devaluation" and his 1974 book "You Can Profit from a Monetary Crisis." In those books he claimed that gold was a substance that had held its value for thousands of years. I do not remember any claims that it was something that increased in value. I believe he said that gold bought roughly the same amount of wheat in ancient Assyria that it does today. ( Of course, my memory may be wrong and I will have to defer to anyone that has a copy of the books.) His disciples now seem to sing a different tune than he did.

I attempted to research ancient prices of gold and find that there is great variations on the estimates of the value of gold in ancient days. For instance gold appears to have bought twice as much wheat in Roman Egypt than it did in Rome itself. However, it doesn't make much difference what estimate you use. The average yearly change in the value of gold is going to be around 0.0% over the last 2-3 thousand years. It may have had a significant change in value from ancient times, but on a yearly basis the average yearly change is going to be very small.

Does anyone else have any data on ancient/medieval/early modern gold prices?

[steve r]

You would have even more difficulty with price indexes (inflation) going further back (a milenium). As I understand it, economic historians attempt to gauge real prices from years gone by with prices of goods that they can find. Gold is often (but not always) a price in that basket ...which only complicates things more.

I said in another post that backtesting a buy and hold portfolio one or two years is bad, looking backing ten or twenty is better looking back 40 years is better still - but better than bad does not mean good. Real return estimates going back as far as I have done (and others) is suspect. It may provide insights, but such insights need to be taking with a grain of salt (salt has a more stable price history per other thread on gold).

[grayfox]

For those that don't know what the Permanent Portfolio is, it is a Portfolio consisting of 4 funds with re-balancing each year:
25% – Stocks (Total Stock Market)
25% – Long Term Treasury Bonds
25% – Gold
25% – Cash (in a Treasury Money Market Fund)

If I had to predict the real returns of these asset classes over the next 20 years, I would estimate the following:

Stocks = 5%+/yr
Bonds = 0%/yr
Gold = 0%/yr
Cash = 0%/yr

For this reason, the permanent portfolio does not appeal to me.

I predict that all of those that are piling into the permanent portfolio in 2012, will regret their decision.

Of course, I've been wrong before, and I'll be wrong again.

The portfolio return is the weighted sum of the components, so you are forecasting 1.25% real return. That doesn't sound so great.
But suppose, like stocks, gold has 5% pa expected real return? Then the portfolio return would be 2.50% real.

Now it's true that since the last ice age receded 12,000 years ago, gold has returned 0%.
But only a few of us have such a long horizon. Over, say the next 5 years, maybe an argument can be made that gold has a postive expected return.

And that is all if you just buy and hold, and don't re-balance.

Gold is uncorrelated with stocks and LT bonds have negative correlation with stocks. So maybe there will be a rebalancing return. How much? It probably depends on the re-balancing strategy and how lucky you get. Maybe 0.5%, 1%, 2%, who knows?

I would say if gold can manage a positive real return, and you time the re-balancing moves correctly, the sky's the limit.

[craigr]

I would say why go back only 130 years..

I believe he said that gold bought roughly the same amount of wheat in ancient Assyria that it does today. ( Of course, my memory may be wrong and I will have to defer to anyone that has a copy of the books.) His disciples now seem to sing a different tune than he did.

Gold is not an investment in what people think of investments should be. It has no internal rate of return. So in that sense it cannot grow like a stock or bond. I certainly don't say anything different.

What gold does do however is hold value over long periods of time. Yes, it may fluctuate in value. But that value has remained remarkably consistent over time. But it will never be worth nothing. And no, I don't think it's going away as a form of money in anyone's lifetime reading this message. That 0% return people quote misses the point. The real point is you have this asset that has matched inflation over very long periods of time through very tumultuous times in history. So you have something that has 0% real returns through wars, famines, depressions, etc. Gold was always worth something and that something largely has remained consistent throughout history. That ain't too shabby when you consider how other assets fared.

So in that sense you have gold in a portfolio serving as a fail-safe asset to complement those assets that are generating return. Your portfolio then consists of:

1) Your contributions to buy assets.
2) The growth of assets like stocks and bonds through their internal rate of return.
3) Gold which is an easily accessed way to lock in the above gains in a fairly robust way against unexpected events.

Investors focus WAY too much on trying to eek out every last percent of growth. Slaving over spreadsheets, reading reports, following gurus, etc. They focus comparatively little on considering threats to their savings that can work in the opposite direction. Meaning market events that can deal huge losses that will wipe out years of excess returns they were chasing. Gold in this aspect then can serve as a useful tool to diversify against those risks because it's the only asset a typical investor can own that has the history of surviving with the value largely intact.

[Browser]

Despite all the arguments in favor of owning gold, none of them to my knowledge have established that one-fourth of your wealth should be invested in it. There's simply no way to do that. In fact, you can't even show that's the case with the data from backtesting. From 1972-2011 a 10% allocation to gold, with the remaining 90% split evenly between TSM and intermediate Treasurys had virtually the same compound return and average annual volatility as the 4x25 PP. I'm among those who believe you just don't need to have that large an investment in gold, and it's probably a bad idea to do so. Such a large allocation to gold introduces a large portfolio tracking error vs. conventional stock/bond portfolios. When negative tracking error is large, as it was for the 20-year period 1981-2000, it becomes very difficult to stick with an unorthodox portfolio allocation. IMO, I think it's much more likely that I can stick with a 5% or 10% gold allocation through the long lean years that have characterized gold returns, which makes it much more likely that I'll actually realize the expected returns from my ability to "stay the course" with my portfolio allocation.

[craigr]

Despite all the arguments in favor of owning gold, none of them to my knowledge have established that one-fourth of your wealth should be invested in it.

The argument is really simple. You want to own enough of an asset to help you when you need it to, but not enough to blow things up when it's having a bad time.

So the idea of owning 5-10% of any asset is just a non-starter. It's not enough to matter if the price goes up sharply.

At the same time, you'll never find me owning 50% in anything. This is because a large decline will be a serious problem in the portfolio dealing out huge losses.

At 25% I get a benefit of large price increases, but not a lot of damage if the bet turns against me. When you start with the nominal 25% allocation to each you not only get targeted returns, but also you get firewalls for your portfolio against the unexpected:

http://crawlingroad.com/blog/2011/08/18 ... firewalls/

[Quasimodo]

If the value of gold is not linked to the dollar, then the price of gold in dollars is eventually going up if the purchasing power of the dollar goes down.

I have some gold in savings because I believe the purchasing power of the dollar may continue to drop, over time.

I don't believe the value of gold is inherently going to change all by itself.

John

[Browser]

The argument is really simple. You want to own enough of an asset to help you when you need it to, but not enough to blow things up when it's having a bad time.

So the idea of owning 5-10% of any asset is just a non-starter. It's not enough to matter if the price goes up sharply.

This sounds simple enough, but I'm not sure the facts support this view. For 1972-1981 here are the annual returns of three porfolios: (1) 25X4 PP, (2) 10% Gold/45% TSM/45% ITT, and (3) 50%TSM/50%ITT.

1972 18.80% 14.59% 10.80%
1973 15.41% 0.86% -6.96%
1974 13.88% -2.66% -10.96%
1975 6.46% 17.92% 22.98%
1976 11.80% 17.13% 19.53%
1977 5.01% 0.75% -1.62%
1978 12.56% 8.38% 5.26%
1979 39.02% 24.53% 13.30%
1980 13.61% 17.71% 18.04%
1981 -4.63% -0.93% 2.62%

Contrary to what you say, it looks to me like just 10% in gold actually did "matter" in the above decade, when gold took off (1972-1981); not as much as 25% gold, of course, but nothing to sneeze at. The improvement over a 50/50 allocation averaged 2.5% per year with annual volatility reduced by 16%. The difference in annual returns ranged from +11.2% in 1979 to -5.1% in 1975. Of course a 25% gold allocation did much better, ranging from +25.7% better in 1979 to -16.5% worse in 1975. The PP was actually just as volatile as the 50/50 portfolio over this period, so it's certainly not true that it is always a low-volatility portfolio. But nobody complains about volatility when returns are good, only when they're not.

Now let's look at how each of these three portfolios did in the next decade, 1982-1991, when gold imploded.

1982 23.15% 23.65% 24.67%
1983 4.84% 11.34% 14.45%
1984 3.05% 6.15% 9.00%
1985 19.30% 23.90% 25.97%
1986 16.58% 15.89% 15.34%
1987 7.12% 4.04% 2.07%
1988 3.46% 9.06% 11.80%
1989 12.86% 18.43% 20.83%
1990 2.70% 1.27% 1.62%
1991 13.50% 21.21% 24.72%

Here we see that the 10% gold allocation underperformed the 50/50 by an annual average of -1.5%, with annual volatility reduced by 8%. The worst year was 1991 (-3.5%) while the best year was 1987 (+2.0%). By comparison, the 4X25 PP underperformed 50/50 by an annual average of -4.4%, with annual volatility reduced by 18%. The worst year was 1991 (-11.2%) and the best year was 1987 (+5.0%). The low volatility of the PP during this period wasn't actually a positive; it resulted from the fact that the PP was a consistent under-achiever during this period. I'm not sure what constitutes "blowing up"; but these returns might be viewed by some as a blow-up in slow motion. Over this decade you wound up with about 1/3 less in your portfolio than 50/50 and 1/4 less than holding just 10% gold. By comparison, the 10% gold portfolio generated 88% of the returns of the 50/50 portfolio with lower volatility, so no burning fuse there.

[Akiva]

Assume that you can't predict what returns will be going forward, but that you expect that, on average, the market will compensate each asset class in proportion to its risk. (Alternatively, you can merely assume that you want each asset class to have an equal dollar impact on your overall portfolio.) If we then take uncorrelated asset classes, the correct portfolio will weight each asset class inversely proportional to its volatility.

This implies that the "correct" weights would be about 25% stocks, 60% fixed income, and 15% gold, which quick calculations show did in fact have a better Sharpe ratio vs PP over recent history. To the extent that the PP over-performed (under-performed) this benchmark, it would be because it over-weighted (under-weighted) certain exposures that happened to have higher returns relative to their volatility (or vice-versa for those exposures that happened to have lower returns relative to volatility).

So, if you are going to adopt a portfolio like this, it seems to me that you should follow normal portfolio construction rules and volatility-weight the independent assets.

As for whether having gold at all makes sense, I think that there are some very good arguments for having a commodities allocation, and that there's a case to be made for preferring physically holding gold to investing in collateralized commodity futures given the options that are available today.

The recommendation to hold long term bonds is iffier. Although it paid off with additional returns over the last 30 years, on average, it seems that the additional risks beyond about 5 years out are not fairly compensated. (There are other considerations that should impact what your duration should be, but this isn't the place to go into it.) So, going forward, intermediate term bonds are probably a safer bet than long-term ones, especially given that the yield curve can't fall much further, but could rise by quite a bit.

If you are willing to get a bit more sophisticated, you could improve on this simple 3 asset-class portfolio in various ways. But I don't think the basic idea is unreasonable, and I would note that most of the PP advocates do encourage generally sound investing habits.

(Also, it's worth reminding people that you can always trade-off returns and volatility, so if this portfolio has lower returns but a higher sharpe ratio than a more aggressive stock allocation, you can still beat the more aggressive allocation's returns by leveraging your exposure to the higher sharpe ratio portfolio.)