Is the Permanent Portfolio The Gold Medal Winner For Investing?

[clacy]

Gold Medal: The rare trading strategy based on Quant Models (maybe just a select few from the top-tier high frequency trading firms like DE Shaw and Susquehanna)  The vast majority of high frequency trading strategies, even those  designed by Math olympiads at these elite firms, are junk IMO.  The Gold Metal, however, strategies will typically have Sharpe Ratios over 3 and have rarely/never seen a single losing month in over a decade.

Silver Medal: Permanent Portfolio

Bronze Metal: Yale University's Endowment

I'm not sure this actually exists, but I want one.

[dualstow]

Whilst the PP may have been a gold medallist over the last 40 odd years, circumstances have been highly favourable for it to be well placed over those 40 odd samples (on a relative valuation basis it might be little different to stocks in the mid to late 1990's). For the next 15+ years ? ? ? Perhaps those alternatives that have endured a relatively poor past decade+ might out shine relatively.

I'm curious, Clive, what your portfolio looks like. It's clear that you are not a pp'er. If you don't mind my asking, are you mostly in stocks? What is your proportion of stocks to bonds? Are you using Decision Moose? I'm sticking with the pp (and my vp). Just curious.

[Gosso]

Just a quick comment on Wellesley.  If you build a portfolio of 70% Wellesley, 20% Gold, 10% Cash, then you get a better return and lower SD than the PP since 1972:
SD = 7.02%
CARG = 10.17%

PP:
SD = 7.99%
CARG = 9.67%

100% Wellesley gives the following:
SD = 9.28%
CARG = 9.98%

I thought it was interesting that gold did such as amazing job at lowering volatility and increasing returns.

[Storm]

Just a quick comment on Wellesley.  If you build a portfolio of 70% Wellesley, 20% Gold, 10% Cash, then you get a better return and lower SD than the PP since 1972:
SD = 7.02%
CARG = 10.17%

PP:
SD = 7.99%
CARG = 9.67%

100% Wellesley gives the following:
SD = 9.28%
CARG = 9.98%

I thought it was interesting that gold did such as amazing job at lowering volatility and increasing returns.

What about max drawdown?  Wellesley is a really good fund, but without any long bonds you don't have much deflation protection.

[Gosso]

Just a quick comment on Wellesley.  If you build a portfolio of 70% Wellesley, 20% Gold, 10% Cash, then you get a better return and lower SD than the PP since 1972:
SD = 7.02%
CARG = 10.17%

PP:
SD = 7.99%
CARG = 9.67%

100% Wellesley gives the following:
SD = 9.28%
CARG = 9.98%

I thought it was interesting that gold did such as amazing job at lowering volatility and increasing returns.

What about max drawdown?  Wellesley is a really good fund, but without any long bonds you don't have much deflation protection.

I'm looking at the Simba database, so I only have annual returns, but the "bad" years look comparable to the PP (either slightly higher or lower).  Below are the five year average real returns.  I added another portfolio of 50% Well, 20% Gold, 10% STT, 20% LTT, which had a SD of 6.99% and a CARG of 9.94%.

              PP   Well Well/Au/STT Well/AU/LTT/STT
1972-1977 4.9% -0.2% 4.4% 4.2%
1977-1982 3.5% -2.8% 2.8% 1.1%
1982-1987 7.9% 15.0% 10.2% 10.5%
1987-1992 2.9% 6.8% 3.8% 3.3%
1992-1997 6.4% 10.0% 7.3% 6.9%
1997-2002 3.2% 8.4% 4.8% 4.4%
2002-2007 7.0% 5.1% 6.8% 6.7%
2007-2011 7.1% 5.0% 6.7% 7.8%

I'm not saying its better than the PP.  I was just messing around with Simba and saw that gold greatly improved Wellesley.

[MediumTex]

Wellesley trailed the PP in inflation-adjusted terms in the 1970s.

That's the weakness in that 35/65 allocation--you don't get strong inflation protection.

[craigr]

I was just messing around with Simba and saw that gold greatly improved Wellesley.

Gold can greatly improve any stock/bond portfolio when those assets are doing poorly. It's the main asset that has enabled the Permanent Portfolio to have consistent real returns no matter what was going on.  Virtually any stock/bond portfolio I've seen will likely be improved on both a real return and risk adjusted basis by adding even just 10% gold to the mix.

[MediumTex]

Whilst the PP may have been a gold medallist over the last 40 odd years, circumstances have been highly favourable for it to be well placed over those 40 odd samples (on a relative valuation basis it might be little different to stocks in the mid to late 1990's). For the next 15+ years ? ? ? Perhaps those alternatives that have endured a relatively poor past decade+ might out shine relatively.

What circumstances have been especially favorable for the PP over the last 40 years?

Gold has seen two bull markets and one bear market.

Long Term Treasuries have seen one bear market and one bull market.

Stocks have seen two bear markets and one bull market.

I would say that "cash" has seen at least one bear market starting in 2007 as the rates on cash have substantially lagged inflation, while the PP has continued to provide positive inflation-adjusted returns.

When I look at the past 40 years, I see a very nice mix of all possible economic conditions. 

What do you think that we might see in the future that we haven't seen in the past 40 years?

Inflation?  Deflation?  Prosperity?  Recession?

[Ad Orientem]

Just a quick comment on Wellesley.  If you build a portfolio of 70% Wellesley, 20% Gold, 10% Cash, then you get a better return and lower SD than the PP since 1972:
SD = 7.02%
CARG = 10.17%

PP:
SD = 7.99%
CARG = 9.67%

100% Wellesley gives the following:
SD = 9.28%
CARG = 9.98%

I thought it was interesting that gold did such as amazing job at lowering volatility and increasing returns.

Do these figures reflect reinvested dividends?

[MediumTex]

Just a quick comment on Wellesley.  If you build a portfolio of 70% Wellesley, 20% Gold, 10% Cash, then you get a better return and lower SD than the PP since 1972:
SD = 7.02%
CARG = 10.17%

PP:
SD = 7.99%
CARG = 9.67%

100% Wellesley gives the following:
SD = 9.28%
CARG = 9.98%

I thought it was interesting that gold did such as amazing job at lowering volatility and increasing returns.

Do these figures reflect reinvested dividends?

It must, given how much of Wellesley's returns are in the form of dividends.

[Gosso]

Just a quick comment on Wellesley.  If you build a portfolio of 70% Wellesley, 20% Gold, 10% Cash, then you get a better return and lower SD than the PP since 1972:
SD = 7.02%
CARG = 10.17%

PP:
SD = 7.99%
CARG = 9.67%

100% Wellesley gives the following:
SD = 9.28%
CARG = 9.98%

I thought it was interesting that gold did such as amazing job at lowering volatility and increasing returns.

Do these figures reflect reinvested dividends?

It must, given how much of Wellesley's returns are in the form of dividends.

Ad,

I used the Simba Database for the backtesting (Here).  I'm 99% sure that it reinvests dividends, and if I'm not mistaken also accounts for the MER.  It rebalances annually as well.

I have to say that I was shocked to see the volatility drop that low.  It is likely from being overweight dividend/mature companies, so the trend may reverse once growth stocks step back into the forefront.

[dragoncar]

If your investment horizon is 15+ years there are safer, more rewarding alternatives (but that endure shorter term volatility).

I'd love to hear what this is.  I'm particularly interested in how you are determining "safeness"

[cowboyhat]

Clive makes a good point. Most of today's PP investors will drop it in the next bull market in stocks. This Permanent Portfolio as The Gold Medal Winner For Investing thread will be painful for anyone self-aware enough to remember it.

[Alanw]

What circumstances have been especially favorable for the PP over the last 40 years?
...
What do you think that we might see in the future that we haven't seen in the past 40 years?

Since 1972 each of stocks, gold, LTT and cash (STT) all have made something like 3% to 5% real. Just perhaps that might not continue into the future (or some degree of mean reversion's could occur)?

Remember that between 1980 and 2000, after gold had last surged (1970's), the PP lagged STT's (cash?) by 1% annualised for TWO DECADES! Persistence for a 60+ year old at the start to remain focused for that length of time might be a big ask of some.

I don't recall many PP'ers during the 1980's/1990's (nor any in the early 2000's come to that). Which implies to me that most PP'ers are relatively new entrants that were attracted by strong recent relative performance/strength. Performance chasers are often the ones that have the poorest overall mid to longer term results.

As a relatively new investor in the PP(1 - 2 years), I'm not sure we should be called performance chasers as much as capital preservation enthusiasts.  Maybe a little of both.  When we enter a new bull stock market, I'm sure some of the PP investors will leave the strategy for more stock heavy portfolios.  Whether I will be part of that group, time will tell.  Being in my 60's, semi-retired, and withdrawing from my investments for living expenses, I am much more concerned with preserving capital than achieving large returns as was the case of the stock run up of the 90's.  In reviewing the PP's annual returns as shown on Craig's Crawling Road web site, I have not found a sustained period of returns that I couldn't live with.  Of course this is no guarantee of future returns.  For now the PP seems to be the proper portfolio as I do not wish to participate in another 2008 meltdown.

[Gosso]

There are some reasonably sound indicators to time precious metals. The most recent case being the obvious suppression of yields (QE) whilst inflation could at any time rise to double digit levels and leave you sitting on sizeable real losses. As per the 1980's and 1990's when real yields were positive, holding gold can be a significant drag factor over such years (such that 1980 to 2000 the PP lagged STT's by 1% annualised for example).

You just described my VP.  Although I'm glad I have the PP to cover my back just in case something goes wrong, or if I am missing some factor that could cause the VP to blow up.

Clive, do you think the PP can maintain its 3-6% real returns into the future?  I agree that cash, LTT, and stocks are likely not going to perform well, but then wouldn't gold go bonkers and pull in 10-12% real returns over the next decade or so.  The money has to flow somewhere...unless global M2 stagnates, which is my greatest fear in regards to my investments.

[MediumTex]

Clive makes a good point. Most of today's PP investors will drop it in the next bull market in stocks. This Permanent Portfolio as The Gold Medal Winner For Investing thread will be painful for anyone self-aware enough to remember it.

It will be an interesting experiment to observe how the PP does in terms of popularity and durability over the next 10 or so years.

As a person who has spent a lot of time advocating and exploring the PP in especially great detail, I do wonder how I will feel about it over longer periods.  I'm sure that its performance will provide the answer to that question.

As of now, I have sort of come to view the PP the way I view brushing my teeth--it's just sort of something I do without really thinking about it because the benefits are so obvious and doing it is so easy.

[Gosso]

Good stuff, Clive.  Lots to digest in that post.

[Alanw]

Clive,
One of the more attractive attributes of the PP is its simplistic approach to investing.  This may come at a slight cost in performance, but again, time will tell.  Not sure I want to deviate from this approach to achieve slightly greater returns.  Your knowledge of investing and alternative investment approaches are greatly appreciated as it gives us an alternative to the PP or other investment styles.  Fresh ideas are always welcome.

BTW, are you anywhere near the Olympics in London?

[dcllee]

Just a quick comment on Wellesley.  If you build a portfolio of 70% Wellesley, 20% Gold, 10% Cash, then you get a better return and lower SD than the PP since 1972:
SD = 7.02%
CARG = 10.17%

PP:
SD = 7.99%
CARG = 9.67%

100% Wellesley gives the following:
SD = 9.28%
CARG = 9.98%

I thought it was interesting that gold did such as amazing job at lowering volatility and increasing returns.

Personally, I have been thinking of adding a bit more volatility to get higher return
Stock - SCV (20%) and EM(20%)
Bond -  LT Bond(40%)
Gold (20%)

Std. Dev. 10.12%
CAGR 12.77%
Sortino 2.54

[Ad Orientem]

Just a quick comment on Wellesley.  If you build a portfolio of 70% Wellesley, 20% Gold, 10% Cash, then you get a better return and lower SD than the PP since 1972:
SD = 7.02%
CARG = 10.17%

PP:
SD = 7.99%
CARG = 9.67%

100% Wellesley gives the following:
SD = 9.28%
CARG = 9.98%

I thought it was interesting that gold did such as amazing job at lowering volatility and increasing returns.

Personally, I have been thinking of adding a bit more volatility to get higher return
Stock - SCV (20%) and EM(20%)
Bond -  LT Bond(40%)
Gold (20%)

Std. Dev. 10.12%
CAGR 12.77%
Sortino 2.54

I am not sure what SCV is and am assuming that EM is emerging markets. That said this looks like an interesting VP. But unless you are wealthy and can afford to take a hit I would be cautious about using this as a substitute for a PP. Emerging markets are very speculative as are LTT's in such a concentrated block. I also note an absence of cash.

How far back does your back testing go?

[dcllee]

Just a quick comment on Wellesley.  If you build a portfolio of 70% Wellesley, 20% Gold, 10% Cash, then you get a better return and lower SD than the PP since 1972:
SD = 7.02%
CARG = 10.17%

PP:
SD = 7.99%
CARG = 9.67%

100% Wellesley gives the following:
SD = 9.28%
CARG = 9.98%

I thought it was interesting that gold did such as amazing job at lowering volatility and increasing returns.

Personally, I have been thinking of adding a bit more volatility to get higher return
Stock - SCV (20%) and EM(20%)
Bond -  LT Bond(40%)
Gold (20%)

Std. Dev. 10.12%
CAGR 12.77%
Sortino 2.54

I am not sure what SCV is and am assuming that EM is emerging markets. That said this looks like an interesting VP. But unless you are wealthy and can afford to take a hit I would be cautious about using this as a substitute for a PP. Emerging markets are very speculative as are LTT's in such a concentrated block. I also note an absence of cash.

How far back does your back testing go?

Small Cap Value.  Backtest goes back to 1972 from Simba's spreadsheet.  The down SD is 3.24% vs 2.33% for PP.  2008 took a -7% return which isn't horrible, Wellesley was -9%

[D1984]

Personally, I have been thinking of adding a bit more volatility to get higher return
Stock - SCV (20%) and EM(20%)
Bond -  LT Bond(40%)
Gold (20%)

Std. Dev. 10.12%
CAGR 12.77%
Sortino 2.54

I am not sure what SCV is and am assuming that EM is emerging markets. That said this looks like an interesting VP. But unless you are wealthy and can afford to take a hit I would be cautious about using this as a substitute for a PP. Emerging markets are very speculative as are LTT's in such a concentrated block. I also note an absence of cash.

How far back does your back testing go?

Small Cap Value.  Backtest goes back to 1972 from Simba's spreadsheet.  The down SD is 3.24% vs 2.33% for PP.   2008 took a -7% return which isn't horrible, Wellesley was -9%

Two caveats:

One, if we ever get a period of ungodly high rates a la Volcker that portfolio may get creamed and would have no cash to provide ammo for rebalancing when stocks and LTTs got knocked down as everyone rushed to short-term investments like T-bills and CDs to take advantage of the high real rates. How did your hypothetical portfolio do in 1981? In 1990? in 1994? Also, how did it do in 1998 and 1999 which weren't exactly great years for SCV but the broader stock market was hitting new highs every month and large cap growth tech stocks were up by upwards of 100% in a year (the reason I ask is because a strategy that stays flat or even loses money for a year or two will be hard to stick with when everyone else is bragging how their stocks doubled in three months)?

Two, any EM backtest (especially using the data from Simba's spreadsheet) from before 1987 or 1988 should be taken with a few grains of salt...no, make that a giant five pound block of salt from a salt lick. SImba's spreadsheet IIRC just subsituted a mixture of devloped market small cap, value, and/or SCV for EM for any years before those. Therei is actually (useful and accurate) EM data back to 1976 using S&P's EM and Investable EM database but it is not AFAIK available without paying them. Still, if you can get access to such data either through S&P or Datastream I would trust it as regards EM performance for any years before 1988 before I'd trust the Simba spreadsheet.

[MachineGhost]

Remember that between 1980 and 2000, after gold had last surged (1970's), the PP lagged STT's (cash?) by 1% annualised for TWO DECADES! Persistence for a 60+ year old at the start to remain focused for that length of time might be a big ask of some.

I don't recall many PP'ers during the 1980's/1990's (nor any in the early 2000's come to that). Which implies to me that most PP'ers are relatively new entrants that were attracted by strong recent relative performance/strength. Performance chasers are often the ones that have the poorest overall mid to longer term results.

Was the PP returns solely dragged down by gold during that time frame?

[MachineGhost]

If the combined 12% stocks, 12% gold counterbalance excepting over some periods when that compensates for negative real yields, leaving a 75% type T-Bill/T Bond type holding that's insured against seizures (negative real yields), and those T-Bills/Bonds average 4% real with such seizures discounted/eliminated, then that's in the 3% real overall ballpark. With some working of the 75% bond side of things, its not difficult or that much riskier to potentially boost that to an overall 4% real.

It seems to me that for capital preservation, you only need a 5-year STT ladder and Gold.  The stock and LTT merely come in play to reduce the size of the gold allocation, because too much gold would drag down the STT returns in economic expansions.

[dcllee]

Two caveats:

One, if we ever get a period of ungodly high rates a la Volcker that portfolio may get creamed and would have no cash to provide ammo for rebalancing when stocks and LTTs got knocked down as everyone rushed to short-term investments like T-bills and CDs to take advantage of the high real rates. How did your hypothetical portfolio do in 1981? In 1990? in 1994? Also, how did it do in 1998 and 1999 which weren't exactly great years for SCV but the broader stock market was hitting new highs every month and large cap growth tech stocks were up by upwards of 100% in a year (the reason I ask is because a strategy that stays flat or even loses money for a year or two will be hard to stick with when everyone else is bragging how their stocks doubled in three months)?

Two, any EM backtest (especially using the data from Simba's spreadsheet) from before 1987 or 1988 should be taken with a few grains of salt...no, make that a giant five pound block of salt from a salt lick. SImba's spreadsheet IIRC just subsituted a mixture of devloped market small cap, value, and/or SCV for EM for any years before those. Therei is actually (useful and accurate) EM data back to 1976 using S&P's EM and Investable EM database but it is not AFAIK available without paying them. Still, if you can get access to such data either through S&P or Datastream I would trust it as regards EM performance for any years before 1988 before I'd trust the Simba spreadsheet.

Like the PP, this portfolio is designed to be held for long term.  There will be years like you said where it will lag the market(same as PP), but SCV and EM are higher risk than TSM, so it helps boost the performance (12% vs 9% CAGR) with just slightly higher volatility.  That 3% is a big impact to overall portfolio.