The Permanent Portfolio, Investing, Finance and Random Thoughts.
William Bernstein is Wild About Harry
Investment author William Bernstein from Efficient Frontier recently wrote an article about the Permanent Portfolio:
He states:
In many respects, this allocation is a thing of beauty. Not only does it provide some protection against all but the most dire of scenarios, but its correlation grid is one rarely seen in finance: four non-derivative assets populated entirely by near-zeros…
- William Bernstein – Efficient Frontier “Wild About Harry”
He continues:
And therein lies the real problem with the TPP (ed: Theoretical Permanent Portfolio): because of its huge tracking error relative to more conventional portfolios, it attracts assets and adherents during crises, then sheds them in better times. There’s nothing wrong with Harry’s portfolio – nothing at all – but there’s everything wrong with his followers, who seem, on average, to chase performance the way dogs chase cars.
…
Sadly, this [buying assets before they have gone up in price] is the opposite of what the legions of new TPP adherents and PRPFX owners have been doing recently – effectively increasing their allocations to red-hot long Treasuries and gold. Consider: over the long sweep of financial history, the annual real return of long bonds and gold have been 2% and 0%, respectively; over the decade ending 2009, they were 5% and 11%.
- William Bernstein – Efficient Frontier “Wild About Harry”
His point about people not sticking to an investment plan is well taken, but hardly unique to the Permanent Portfolio. As for whether gold or LT bonds are a good buy now: I learned a long time ago that whenever I think I know what the markets are going to do I’m usually surprised later.
I read articles back in 2007 saying how expensive gold was when it was $600 an ounce and would crash any day now (it’s over $1200 now). I read articles over a decade ago about avoiding LT bonds because of the inevitable rise in interest rates (30 year bonds yielded 5.9% in 2000 – today they are 3.5% giving big profits to long bond holders over this time). In fact, I’ve read investment books going back over 20 years saying the exact same thing about the inevitable destruction of the long bond (30 year bonds in 1990 yielded 8.6% BTW).
The above is to say nothing of the copious amounts of doom and gloom stock predictions in 2009 that proved to be horribly incorrect as the market recovered so sharply it could give you whiplash. All of these predictions have been dead wrong and highly unprofitable if followed.
As it were, Harry Browne actually got this type of question about avoiding this or that asset over the years and I made this audio clip of him answering it:
http://crawlingroad.com/blog/wp-content/uploads/2010/01/HarryBrowne-WhichAssetWillDoBest.mp3
This clip was made in 2004 but could have been recorded last week (Browne even mentions about the Dow being 10,000 back then and it’s there again right now in fact – It’s de ja vu all over again). These debates about what to buy and sell are raging constantly in the market and will do so forever. Investors just need to ignore all of them.
The point is not to be a cheerleader for some particular asset because eventually every investment has a bad spell. It’s simply that we can’t predict the future and the reason we own different assets in the portfolio is to provide protection against the unknown.
Diversifying asset classes, as Harry Browne knew well, can benefit a portfolio. The secret is deploying them before those diversifying assets shoot the lights out.
- William Bernstein – Efficient Frontier “Wild About Harry”
I disagree. Investors can’t possibly position their portfolio in a way to take advantage of an asset before it goes up in price. It just doesn’t happen that way.
What’s the real secret? Simple:
Just own everything at once because we have no way of knowing what is going to do best going forward.
It was investors who owned assets like gold and long bonds all the time, regardless of what was being predicted, that were able to reap these rewards the past few years. They may continue to do well, or maybe not. Maybe stocks will finally come out of their funk and go gangbusters again as they did in the 1980s and 1990s. Nobody knows.
We just don’t know when the markets will move one way or another. Investors therefore must diversify, and must stay diversified, all the time regardless of what their opinion is about the future. We simply cannot know what asset is going to do best ahead of time and allocate accordingly.
The correct strategy then is to hold fast to your asset allocation. If something has gone up enough in price to trigger a rebalancing band, then sell it down and use the profits to buy the laggards. But never time the buying or selling of an asset in the portfolio because of how you feel about it. That’s a sure way to run into big problems. The Permanent Portfolio is designed to hold volatile assets together in the form of 25% each stocks, bonds, cash and gold. If you take out any one piece you lose the protection of the portfolio.
Performance chasers will always performance chase and will always show poor results for their efforts. The Permanent Portfolio is not designed to be a hot rod so the performance chasers will eventually be disappointed. What the portfolio provides however is moderate growth with wide diversification and low volatility even in very bad markets. At the same time the strategy has shown reasonable real after-inflation returns for the level of risk involved with no market timing nor close monitoring required. For people looking for those attributes, the strategy may make you wild about Harry, too.
Podcast: Play in new window | Download
Related posts:
-
#2 written by Marshall Costantino 1 year ago
I’m glad Mr. Bernstein is so supportive of Harry Browne’s 4 by 25% Permanent Portfolio.
However, his “near-zero” correlation grid entries comment is different than the empirical results I obtained by analyzing the Crawling Road annual Browne Permanent Portfolio returns from 1972 through 2009. Below is the correlation grid I obtained:
CORRELATION GRID
TSM STBonds LTBonds Gold TotRet
TSM 1.000
STBonds 0.037 1.000
LTBonds (0.001) 0.312 1.000
Gold (0.231) (0.159) (0.251) 1.000
TotRet 0.362 0.159 0.200 0.708 1.000This correlation grid has only 2 asset class versus asset class entries close to zero (.037 and -.001). The rest show the values .312, -.231, -.159, -.251 – all hardly “near-zero.” Other observations about this correlation grid are these:
* TSM exhibits very low (near-zero) correlation with ST Bonds (.037).
Its best correlation is -.231 with Gold.* ST Bonds and LT Bonds show less than .500 correlation with one another (.312). This surprised me. I thought the correlation was higher.
* Gold bullion is negatively correlated with all the other three asset classes.
* Total Return is most correlated with TSM (.362) and Gold (.708).
These correlation coefficients suggest TSM and Gold “drive” the
Browne Permanent Portfolio annual total returns.* All correlation coefficients are less than .350, regardless of positive or negative sign. This suggests that the Browne asset classes are close to being independent of one another.
I also believe Mr. Bernstein’s thesis that the Browne Permanent Portfolio attracts assets during bad market times and then gives them up when good times recur is off the mark. My wife and I have invested in the Permanent Portfolio Fund (PRPFX) since 2003, after management resolved their SEC and Coxon legal issues. Using the PRPFX Annual Reports, the Total Asset base has increased by an annual average 43% increment since 1/31/1998. Since 1/31/2002, there hasn’t been an annual decline in the Total Asset base. Also, according to Morningstar, the calendar year 2008 was the only down year in the last 10 at -8.36%. [Recall that Harry Browne was a technical consultant to PRPFX from its 1982 inception until his death. The mix of asset classes in this mutual fund adds silver bullion, REITs, Natural Resource Company stocks and Swiss Franc assets in addition to the 4 asset classes of the Browne Permanent Portfolio.]
Curious to see the differences in returns between PRPFX and the Browne 4 by 25% Permanent Portfolio, I ran another analysis to see how much $10,000 invested on December 31, 1997 would be worth using both configurations of the Permanent Portfolio (PRPFX = Morningstar calendar year total returns; Browne 4 by 25% = CrawlingRoad.com calendar year total returns table). The side-by-side comparison looks like this:
Browne
PRPFX 4 by 25%
TOTAL RETURNS Browne Fund: Portfolio:
PRPFX 4 by 25% Value of Value of
Year Fund Portfolio $10,000 $10,000
2000 5.9% 3.2% $10,588 $10,320
2001 3.8% 0.9% $10,991 $10,413
2002 14.4% 7.0% $12,570 $11,142
2003 20.5% 14.0% $15,140 $12,702
2004 12.0% 6.5% $16,963 $13,527
2005 7.6% 8.1% $18,256 $14,623
2006 13.8% 11.0% $20,779 $16,231
2007 12.4% 12.9% $23,362 $18,325
2008 -8.4% 2.0% $21,408 $18,692
2009 19.1% 7.8% Ann. $25,493 $20,150
TR= 9.81% 7.26%While the Permanent Portfolio Fund adds an annualized 35% increment over the Browne 4 by 25% portfolio, this annualized increment comes at a very high price – a 4.26 times annualized increase increment in Operating Expense Ratio (.82% versus .1925%).
These results suggest three approaches to investing in and to maintaining a Permanent Portfolio for the money we can’t afford to lose:
1. Best Return vs. Operating Expense tradeoff = Browne 4 by 25% Permanent Portfolio
2. Lazy Person’s tradeoff = Permanent Portfolio Fund (PRPFX)
3. For those who can’t bring themselves to rely on one or the other =
Weighted PRPFX & Browne 4 by 25% portfolio blend
(Say 1/3 PRPFX + 2/3 Browne for .40% overall operating expense ratio & 8.1% annualized TR)I believe Craigr’s comments are spot on:
The Permanent Portfolio, regardless of its configuration (PRPFX or Browne), protects the investor from every common risk they might incur except the unexpected calamity of catastrophic event risk. The “secret” is to buy all 4 asset classes or PRPFX, hold them long-term, and watch the steady accumulation of real wealth – the Crawling Road.
I look forward to our future dialogue.
Marshall Costantino
-
Excellent points Marshall. Bernstein’s different results could be from the use of a slightly different data set than mine. He also started his analysis going back into the 1960s. This is normally fine, but the problem is that gold was price controlled until 1971. So any portfolio analysis of gold prior to that date is going to have problems because the price of gold was not allowed to freely float.
The fund holds a higher allocation to inflation sensitive assets like gold, silver and Swiss francs. So the last 10 years have favored the fund over the 4×25 split. If the markets stabilize and no serious inflation happens you’ll probably see the fund start to lag the DIY version though.
-
#4 written by Marshall Costantino 1 year ago
Since the world’s currencies were pegged to the gold standard before 1971, I agree that using any gold price data prior to 1971 will likely lead to an erroneous conclusion or to a biased analysis (unknowingly), at best. The key here seems to be gold’s negative correlation with every other asset class.
My point in comparing and contrasting PRPFX to the Browne 4 by 25 Permanent Portfolio was to demonstrate there are 2 Permanent Portfolio approaches that provide investors a way to minimize the negative impact of most all investment risks, except catastrophic event risk. One requires investors to delegate rebalancing and asset class choices to a mutual fund company, at high expense – just throwing money at these problems. The other hands-on approach requires investors to rebalance and to choose the 4 ETFs and to do their own due diligence – at the lowest costs possible.
Your comment about PRPFX allocating more to inflation sensitive assets is well taken. I tend to look at silver as an attempt to capitalize on the industrial demand for the metal. I also consider the Swiss franc assets as an attempt to diversify fixed income assets out of the US dollar and into the world’s historically most stable currency. With the globalization of the economy, maybe it is appropriate to invest in a global stock index ETF instead of a US total market ETF. I believe you addressed the question about how much foreign stock representation should be present in a Permanent Portfolio in a separate blog. But this is another subject for another day.
Then there is also the reversion to the mean, too. Over time, markets of all kinds tend to be zero-sum games.
Could abandoning the gold standard have contributed to the mega-inflation of the 1970s and early 1980s?
I believe Hayek (Nobel Prize in Economics winner) and his buddies of the Austrian School demonstrated that if a monetary authority wanted to change the phase of the business cycle, all the authority has to do is to mismanage the money supply. That is, constrict money supply growth when growth should be maintained or increased (resulting in a recession), and loosen money supply growth when growth should be restricted (resulting in a boom, followed by inflation). Hayek not only demonstrated this phenomenon over Austria’s history, but he did so for England, Germany, the United States, Austrailia and Canada, if I remember correctly.
MC :+>)
-
Yes there are two approaches for sure. The fund is a good choice for people that don’t want to DIY. It also had the added advantage of hiding the messy details of rebalancing from the investor so they don’t panic watching everything go up and down. The 4×25 approach is a lot cheaper but investors need to have a certain confidence to implement the entire package without timing the markets.
As for the inflation of the 1970s. Certainly some of it was due to the gold standard breaking. But I think also that much of it was the backlash from the massive spending in social programs from the 1960s and Vietnam War into the early 1970s. Historically these types of spending patterns (welfare and warfare) in many different countries have lead to inflation. The more the government tries to stop it (aside from not spending as much money), the worse they make it.
Overall I think the Austrians are correct. In terms of the monetary cycle I think they are dead on. Browne was a huge backer of the Austrian school of economics and the portfolio design reflects this. His economic cycles seem to match up perfectly with what the expansion and contraction of the money supply can bring in terms of economic havoc. This isn’t a coincidence I’m sure!
-
#6 written by ziggy 1 year ago
You seem to be avoiding the elephant in the room: while gold/stocks can go up without limit, treasury rates cannot (realistically) go below 0. Thus your examples about never knowing which way things will go don’t apply to LTT because at this point they will can only stay the same, decline very slightly or go up by quite possibly a lot. All the risk is on one side now.
Gold, of course, can continue rising though it does seem historically very high.
Thoughts?
-
Excellent article Craig, thanks!
Also my congratiulations to Marshall Costantino. What an excellent post about the correlations. Thank you so much Marshall.
To ziggy:
While bond rates cannot go below zero, other assets like gold and stocks can go much lower and lose 90% of their value in short time, making bonds in such scenario, even with low rates – relatively – a super investment!
-
#8 written by Marshall Costantino 1 year ago
Marc, Ziggy & Craig:
I’m not sure I understand Ziggy’s comment about the elephant in the room or his point about LT Bonds, but here are some thoughts:
1. Harry Browne constructed the 4 by 25% Portfolio to be Permanent so that it could produce a positive real return during all phases of the business cycle – Recession/Deflation (R), Inflation (I), Growth (G) and Stability (S).
2. RIGS also represents the phases of the monetary cycle, and the monetary cycle can be out-of-phase with the business cycle by one or two phases. The monetary cycle phase usually precedes the business cycle phase, sometimes inducing the business cycle phase as Hayek and his buddies demonstrated.
3. From my analysis of the 1972 – 2009 Crawling Road Permanent Portfolio Asset Class and Total Annual Returns, here are some noteworthy summary data, where the coefficient of variation (CV) is the ratio of the standard deviation to the average, expressed as a percentage representing volatility:
TSM has a low return of -37%, a high of +39% & a CV of 160%.
ST Bonds have a low return of -12%, a high of +19% & a CV of 78%.
LT Bonds have a low return of -22%, a high of +40% & a CV of 142%.
Gold has a low return of -33%, a high of 136% & a CV of 260%.
Total Portfolio has a low return of -4%, a high of 42% & a CV of 82%.4. Using the CV as a volatility-of-returns measure, the rank order of the 4 asset classes is:
1st = Gold (260%)
2nd = TSM (160%)
3rd = LT Bonds (142%)
4th = ST Bonds (78%)which we should have expected.
Note that rebalancing the asset classes annually yields a portfolio return CV of 82%, 5% incrementally higher than ST Bonds alone. The annualized portfolio return of 9.51% was slightly higher than the ST Bond return 9.30%, 2% incrementally higher.
Even though the trade-off looks unfavorable (2% incrementally higher return for 5% incrementally higher CV), investing in ST Bonds alone exposes us to seven other kinds of risk:
Stock/Bond [picking a bum stock or bond]
Stockpicker/Bondpicker [picking a bad stockpicker or bondpicker]
Market [investing in a declining market]
Market Timer [picking a bad trend-following market timer]
Market Timing (trend following) [choosing the wrong timing system]
Asset [choosing the wrong asset class]
Catestrophic Event [random event having adverse consequences].To me, it looks like the elephant in the room is the tendency to want to invest in the “silver bullet” asset class that provides the best return at the lowest volatility.
This “silver bullet” asset class is highly unlikely to exist, given the RIGS business and monetary cycle phases and the timing of their occurences. We are stuck with having to invest and to protect our investments under huge uncertainty.
So the best for which we can hope is to invest in different, less correlated asset classes that protect us from the ravages of the business & monetary RIGS phases, and whose combination offers the potential to produce a real return beyond inflation and taxes – the 4 by 25% Browne Permanent Portfolio or the Permanent Portfolio Fund (PRPFX) or both.
Does anyone have any other elephants the want to discuss?
Marshall Costantino :+>)
-
#9 written by ziggy 1 year ago
The point is simply the one clearly made in the article in question: LT bonds have nowhere to go but up sooner or later, so the worst possible time to buy them for a long term portfolio. You definitely know you can’t make much on them at this point but can lose a lot and seriously affect your long term returns.
A 2nd point given all the discussion of correlation: Brown aligns gold with inflation but did anyone see the recent WSJ article where a study showed that gold barely correlated with inflation and instead was highly correlated to the dollar?
Thoughts and implications for PP theory?
————————-
I recently asked research firm Ibbotson Associates to run a correlation study to determine how closely inflation and gold-price movements track each other. You would expect gold, as a purported commodity, and inflation to move in tandem.The data, going back to 1978 and capturing an inflationary spike, shows a correlation of, at most, 0.08.
That is low. Really low. Perfect correlation is plus 1; at minus 1 two assets move in perfect opposition. Near 0 implies gold and inflation barely acknowledge one another, and moves in unison are largely happenstance.
So if inflation doesn’t push and pull at gold prices, what might it be? If you believe correlation studies, the answer is the U.S. dollar. (NOTE: I prefer to call it Federal Reserve Notes.)
Going back to 1973—a period that defines the modern, non-gold-backed dollar—the greenback’s movements closely track gold’s direction. The correlation between month-end gold prices and the Major Currencies Dollar Index, as reported by the Federal Reserve, is minus 0.45.
That clearly is a stronger correlation than you find with inflation. But let’s take this a bit further. Let’s shorten the time frame to the period from gold’s 1980 peak to today.
The result: Over the past 30 years, the correlation between the dollar and gold is minus 0.65—a high negative correlation. It means the dollar and gold are effectively on opposite ends of a seesaw. When the dollar is in favor, gold retreats. When it is under pressure, gold prices swell.
————————-
http://online.wsj.com/article/SB10001424052748703908704575433670771742884.html?mod=WSJ_newsreel_personalFinance#articleTabs%3Darticle -
As stated in the post above, I read the same *exact* arguments about LT bonds over a decade ago. They were all wrong. So far in 2010 they are wrong again.
Now this is not to say that these people won’t be right eventually (we’ve been waiting for over a generation, but what’s 30 years between friends?). The point is, as it is with all asset classes, is we don’t know what is going to happen so just own them all at once. LT bond prices sunk to around 1% in the Great Depression if I recall. So yields can certainly go further down.
But even if the bad should happen and yields spike, chances are the gold and/or stocks will rise in price to offset the losses. This is exactly what happened in the 1970s and indeed the exact thing that happened last year if you were unfortunate enough to buy LT Bonds on January 2009 at the near peak of their prices.
Re: gold and WSJ
This is old news. Terry Coxon and Harry Browne postulated over 30 years ago that gold’s main benefit is that it serves as the worlds second most popular form of money behind the dollar. Browne uses gold for inflation protection, but this is clearly in context of protecting against the fall in value of the dollar as the world’s reserve currency. He is clear that gold will really go up in price when the dollar is threatened. If a world event is not bad for the dollar, gold will do pretty much nothing.
If the dollar experiences high inflation then gold prices will go up sharply. If not, then expect the price to fall. This is why the portfolio design is such that you must hold all the assets all the time. It is designed to work in a future that is not predictable.
-
#11 written by Marshall Costantino 1 year ago
Gentlemen:
As we are all aware, the future never looks like the past.
The WSJ assessment that gold prices aren’t correlated with inflation seems irrelevant, and this brings up the issue of economic globalization from the perspectives of currencies, stocks, bonds and commodities.
Historically, the US dollar has been the world’s store of value, but when the dollar was weak, gold prices have increased. So what if the US dollar doesn’t serve as the world’s store of value and is replaced by the Euro or the Chinese currency?
How does this affect the Permanent Portfolio’s composition?
Would TSM be replaced by two stock ETFs – a US Broad Market + a Foreign Broad Market? 70% US/30% Foreign?
Would ST Bonds be replaced by two bond ETFs – a US ST Treasury + a ST Foreign Bond? In what proportion? 70% US/30% Foreign?
Would LT Bonds be replaced by two bond ETFs – a US LT Treasury + an iShares LT Foreign Bond? In what proportion? 70% US/30% Foreign?
Would Gold be replaced by a currency index?
Given these changes, we now balloon to 7 positions from 4, destroying one of the most profound features of the Harry Browne 4 by 25% Permanent Portfolio – its simplicity.
MC :+>)
-
#12 written by upside 1 year ago
ziggy, you said, “The point is simply the one clearly made in the article in question: LT bonds have nowhere to go but up sooner or later, so the worst possible time to buy them for a long term portfolio. You definitely know you can’t make much on them at this point but can lose a lot and seriously affect your long term returns.”
Your last sentence is incorrect. The 30 year treasury rate is at 3.66% today. If the rate goes to zero, a 30 year bond (29-30 years until maturity) will increase a bit over 100%. So there is plenty of room for gains.
-
If the dollar is no longer the world reserve currency I wouldn’t change anything with the portfolio. Gold isn’t going anywhere as a form of money no matter what is on top at the current time. The stocks and bonds you always want tied to your home country currency because that’s the money you use for day to day living and you rely on that economy for your livelihood.
-
#14 written by Bdahl 1 year ago
OMG! The whole point of the portfolio is to have real diversification and not have to watch your money. Build wealth over time. You guys should go out and play with your kids, or go to a baseball game. Go Twins! or get girlfriends if needed. That is his point, don’t waste your time trying to time the market. Let the portfolio protect your assets and grow steady over time. I love it and no longer spend hours a day in front of my computer trying to guess what stock to buy.
-
#15 written by Bdahl 1 year ago
BTW i think the LTT argument is BS. I get it, bonds have had there day. They will not do well in the future. So what, if you can predict the future, let me know. For now I am sticking to it, when the market went down in May -12% LT bonds went up 22%. I am guessing everyone in the world with this argument the last 2,5,10 years missed the 22% gain.
-
Comments are closed
- Comment Feed for this Post
I agree with his statment that it’s remarkable that the components all have near-zero correlation.
In fact, thats an amazing fact, since, over the long-haul, all of these investments will appreciate… which suggests that their movement during short periods of time, is actually negatively correlated, and not just 0.
If that’s the case, then their movement isn’t just random, but somewhat opposite – an enormous advantage when creating a low-risk investment strategy!