In series of posts last week, I looked at Harry Browne’s Permanent Portfolio, which includes a hefty 25% allocation to gold. The reason for holding such a large amount, Browne argued, is that gold protects investors from the ravages of inflation.
The problem with this idea is that there’s no evidence that the price of gold is highly correlated with inflation—at least, not in Canada. This is one of those pieces of conventional wisdom that just doesn’t stand up to scrutiny. Gold has held its value when individual curencies have collapsed, like in Weimer Germany in the 1920s or more recently in Zimbabwe. But those are examples of catastrophic hyperinflation, which leaves people pushing around wheelbarrows of cash to buy a loaf of bread. I don’t think that’s what most people mean when they talk of gold as a hedge against inflation in their portfolio.
You can see how the idea developed in the 1970s. Inflation in Canada averaged almost 9% from 1970 through 1982, and gold would have provided an enormous safety net during this period: its annualized return (in Canadian dollars) was over 25% during those 13 years. The correlation was even higher in the United States and, not coincidentally, it was at the end of this period that Harry Browne introduced the Permanent Portfolio and declared gold a hedge against inflation.
Where’s the correlation gone?
Since the 1970s, however, the correlation between inflation and gold has disappeared. From 1983 through 2004, inflation averaged about 3%. Yet the nominal return on gold in Canadian dollars during this period was –0.3% annualized. That’s a real return of –50% over a period of 22 years. Not only was gold useless as an inflation hedge for more than two decades, it lost half its value along the way.
Of course, since 2005 gold has been the best performing asset class, with annualized returns of about 18% in Canadian dollars. You certainly would have done well if you had held 25% of your portfolio in gold during the last seven years—but this had nothing at all to do with inflation. Since 2005, the Consumer Price Index has gone up just 2% annually, less than the historical average.
How about on a short-term basis? If inflation spikes in a given a year, will gold act as hedge? Sure, there are examples of gold shooting up during calendar years with high inflation, such as 1974, 1980 and 1982. But there are just as many opposite moves. In 1975, inflation was almost 10% and gold lost 22%. From 1980 through 1982, inflation was over 10% annually, yet gold lost value over those three years thanks to a –33% plunge in 1981. In 1989 and 1990, inflation was well over 5%, while the return on gold was again negative in both years.
Clearly the price of gold has just about zero correlation with the Consumer Price Index.
Gold moves in mysterious ways
There is no tidy relationship between inflation and the price of gold, even in the 1970s. The tremendous returns the metal delivered during that period were the result of several factors, notably the United States abandoning the gold standard in August 1971. (Before that, the price of gold had been fixed at $35 USD per ounce since the Bretton Woods agreement of 1944.) People lost their confidence in stocks during the bear market of the early 1970s, much the way many investors feel today. This was also a time when central banks did not have the ability, the mandate, nor the political will to control inflation the way central banks do today. For example, the Bank of Canada set a target rate of 2% to 3% for inflation in 1991 and has done an extremely good job in this respect for the last 20 years.
That’s not to say that we can’t ever see a period of double-digit inflation again—of course, that remains a possibility. The point is that the cause of that inflation, and the tools we use to fight it, will be different. And therefore, if we experience double-digit inflation in the next decade, there’s no reason to expect the price of gold to behave the same way it did during the 1970s.
I don’t object to holding a small allocation of gold in a diversified portfolio, but it’s important for investors to be clear about their expectations. Gold can have periods of negative correlation with the US dollar, the world’s reserve currency. It can be a safe haven during crises of confidence in stock markets. If the global financial system collapses, if the Canadian dollar becomes worthless, or if the Harper government confiscates your land then, sure, you can exchange the bullion in your safety deposit box for canned beans and ammunition. But if you’re expecting gold to provide a hedge against inflation, then you’re likely to de disappointed.
If you take yearly gold gains since 1970 and compare that to CPI, you’ll see something around a 0.27 correlation. Short Term Treasury’s compared to inflation was just under 0.4.
Plot a running graph of the growth rate of twice the yearly CPI rate and compare that to the running price of gold and you’ll see a longer term overlap to some extent. But gold swings around more, as in how stocks tend to swing around bonds, sometimes leading, sometimes lagging.
Stocks generally provide a capital value that grows with inflation, plus a dividend that somewhat matches cash, and that also grows with inflation – so around a 2 x inflation nominal reward.
Both stocks and gold total returns since 1970 have in effect been close to 2 x inflation.
Gold might be considered as a form of commodity index that reflects the overall average of a wide range of commodities. Its also handy to hold in times of great fear or hyperinflation. And it has a low/inverse correlation to stocks.
If you consider gold as just another stock, rather than what’s actually behind that stock then it potentially provides similar longer term rewards, but tends to move counter direction to many other stocks. Having two stocks with similar rewards, are volatile and have low correlation, provides a higher potential compound benefit if periodically rebalanced than holding a bunch of stocks that all move up and down together.
Gold provides a longer term inverse correlation to stocks in a similar manner to how long dated treasury’s perform a shorter term inverse correlation to stocks.
What I’d like to know is that if gold has been a 2 x inflation rewarding asset since having come off the gold standard, but more generally gold might be expected to just pace inflation over the long term, will it revert to 1x inflation by perhaps halving in value, or continue the 2x inflation trend its followed since the 1970’s. (Or could it be that inflation is understated by half) ???
@Clive: Thanks for your analysis. A 0.24 correlation over 40 years is no correlation.
I agree that gold can be a good hedge against stock market declines (it played that role in the 1970s and the last several years), but I have a hard time believing that the long-term expected returns of stocks and gold will be roughly the same. I can’t justify holding equal amounts of each in a long-term portfolio.
I think it’s important to remember that rebalancing the PP would have been extremely difficult during the the two decades that gold was horrendous performer. It’s hard enough to get people to rebalance stocks and bonds during an 18-month bear market. How many people would have sold their stocks in the 1990s (when they were gaining 20% a year) and plowed that money into gold as it sunk like a stone for years? How many people have sold their gold and bought stocks in the last four years?
Of course, the Couch Potato strategy is not immune to this problem. This is all a reminder that sensible long-term investment strategies work, but they require a great amount of discipline. Usually you feel like you’re doing something stupid, and it’s only after many years that you see the payoff.
Thanks Dan.
>”a reminder that sensible long-term investment strategies work, but they require a great amount of discipline”
Exactly.
As an alternative to holding gold directly, perhaps consider the potential benefits of a precious metals equity based holding (PME). Compare for instance a chart of a precious metals fund with gold http://chart.finance.yahoo.com/z?s=DBP&t=5y&q=l&l=off&z=m&c=GLD&a=v&p=s&lang=en-GB®ion=GB and read William Bernstein’s article http://www.efficientfrontier.com/ef/197/preci197.htm that indicates 1969 to 1996 Precious Metal Equity annualised gains of 12.81% compared to 11.25% for the S&P500
and that concludes ….
” 1. While the long term return of precious metal bullion is close to the risk free rate the long term return for precious metal equity appears to be similar to that of common stock.
2. Because the correlation of PME returns and other common stock returns is very low substantial decreases in portfolio risk are available from the judicious use of this asset. For those willing and able to rebalance, significant increments in return are available as well.
3. In order to fully reap the portfolio benefits of PME the investor must be able to ignore its substantial nonsystematic risk. She must also be able to endure long periods in which PME will be an albatross around the portfolio’s neck. Most important of all, she must be able to weather from time to time the jeers of others. “
I totally agree, Gold is good to preserve value when a national currency is losing value or during extreme events (hyperinflation, wars, financial crises).
From 1900 to 2010, inflation adjusted gold market price rarely outperforms inflation. The Dow Jones was a much better inflation hedge in 110 years history.
I’ve used the ”Periodic Table of Annual Returns for Canadians” from Libra Investment Management to compare the Permanent Portfolio vs Complete Couch . The period is 1992-2010 (real return bonds were not available before 1992).
30% all Can bonds – 10% real return bonds – 30% TSX (no datas for Reits) – 15% S&P 500 – 15% EAFE. The returns are similar. Maybe as canadians we don’t really need to add gold to our porfolio or only a small allocation.
But what I like about the PP is it’s very low volatility (SD) and one negative year (1981) in 41 years (1970-2010). Index returns but still very impressive.
CCP thanks for the post and the useful analysis. It’s a shame that as Canadians we don’t have as much information on the PP as we would as Americans.
The stated purpose of the permanent portfolio is to provide reasonable growth with low volatility by attempting to diversify across all economic conditions (prosperity, inflation, deflation, recession). Your analyis shows that inflation and gold are not very correlated (at least not in Canada), and that is something to give pause before rushing headlong into a PP. However, I would note that, at least from my research into the topic, the permanent portfolio’s secondary goal, or perhaps you could refer to it as the natural offshoot of it’s stated goal, is to always have at least one asset doing well. As such I’d be interested to see what the correlation has been between stocks and gold in Canada since 1972. I note from the historical returns provided by Craig on Crawling Road that the only years (between 1972 and 2008) where both the S&P and gold went down were 1981, 1990, 1994, and 2000. In contrast, there were 14 years when both gold and the S&P went up, and 19 years where one assest went up and the other went down. My (rough) calculation comes with a 11.5% CAGR if one held a 50/50 split of gold from 1972-2009, which appears to beat a 100% S&P portfolio. The problem with the 50/50 stocks & gold is that instead of 2 down years like the PP it has 8 down years, the worst being -15.9% in 2008 and a couple of -8% years. Despite the higher CAGR it doesn’t have the low volatility of a PP.
One final though regarding the correlation between gold and inflation is that the gold doesn’t respond just to Canadian inflation as it is international. The reason why it syncs up well with US inflation is likely due to the size of the US economy. If the data was available it would be interesting to see if there were a correlation between gold and “global” inflation (if that could possibly be determined).
Thanks,
Daniel
William Bernstein’s article http://www.efficientfrontier.com/ef/197/preci197.htm indicates 1969 to 1996 Precious Metal Equity annualised gains of 12.81% compared to 11.25% for the S&P500
and that concludes ….
” 1. While the long term return of precious metal bullion is close to the risk free rate the long term return for precious metal equity appears to be similar to that of common stock.
2. Because the correlation of PME returns and other common stock returns is very low substantial decreases in portfolio risk are available from the judicious use of this asset. For those willing and able to rebalance, significant increments in return are available as well.
3. In order to fully reap the portfolio benefits of PME the investor must be able to ignore its substantial nonsystematic risk. She must also be able to endure long periods in which PME will be an albatross around the portfolio’s neck. Most important of all, she must be able to weather from time to time the jeers of others. “
The PP might be a good option for money that I need in 10-15 years.
@Daniel: you might find the following link useful (thanks to Clive for providing it). It’s a very handy asset mix returns calculator that provides Canadian and other data:
http://www.ndir.com/cgi-bin/downside_adv.cgi
@Pat: I’ve seen this tool before and agree that it is extremely useful.
Using Craig’s suggested portfolio which replaces the cash with short bonds a “Canadian” Permanent Portfolio, according to the asset mixer, of 25% TSX, 25% Long Bonds, 25% Short Bonds, and 25% Gold would have a CAGR of 8.6% from 1980 to 2010 with a standard deviation of 7% with 3 down years (worst being -9.3%, then -1.2%).
If you take out the gold and make it 34% TSX, 33% Short Bonds, 33% Long Bonds then you find yourself with a CAGR of 10.2% from 1980 to 2010 with a standard deviation of 8% and 4 down years (-7.5%, then -2.8%). Interestingly this portfolio does slightly better than one with 66% all Canadian bonds, giving credence to Craig’s barbell approach to long and short bonds. In contrast, a 100% TSX portfolio from 1980 to 2010 has a CAGR of 9.7% with a standard deviation of 16.7%.
The issue, and I believe that CCP correctly identifies this in his post, is that most of gold’s enormous growth came in the 1970s. If you put a 100% gold allocation into the asset mixer from 1970 to 2010 you see that gold returns 1755% from 1970 to 1980, and is negative from 1980 to 2000. That’s why the 1980 to 2010 portfolios above benefit from removing gold.
The question is what role gold will play in the future. As CCP notes, gold has been the best performing asset since 2005, will that continue? Harry Browne’s theory is that we can not know, but that as long as gold exists as the de facto back up to US dollars as the world’s reserve currency, then it provides a hedge against future economic conditions (although this might not necessarily be inflation).
A final interesting note is that if you exchange the gold allocation in the PP for real return bonds the rate of return doesn’t change, but the standard deviation rises by 2%.
Daniel
Gold in isolation is a long term store of wealth. Yes if you look at it from the absolute peak in the early 1980s to the early 2000’s it looks horrible. But the NASDAQ looks pretty bad too from 1999 to today. However when used in a diversified portfolio it can counterbalance the swings in stocks and bonds. And there have been serious currency problems as recently as 2008. Someone did an analysis of how the Permanent Portfolio would have done in Iceland when their currency collapsed and the price in Iceland krona was +259% that year vs. USD +5% or so. So there is protection from gold against local currency problems even if you are not a USD investor.
Gold is part of the “have money” portfolio. It is an asset that has a remarkably long track record of resisting destruction of wealth. It is not a “make money” asset like stocks and bonds. But there is no reason why a portfolio can’t have a place for have money and make money assets both. Each has their day in the sun and in the dog house.
Analysis of Permanent Portfolio during Iceland’s 2008 market and currency meltdown:
http://europeanpermanentportfolio.blogspot.com/2009/08/permanent-portfolio-in-iceland.html
In particular, check the chart at the end of the article updated for 2010. The Permanent Portfolio (with gold) withstood some brutal assaults by the Icelandic markets. Gold is not a worthless asset at all!
Gold does not seem to be linked to inflation but rather with interest rates. It makes sense since gold does not “earn” you anything, unlike dividend paying stocks or compounded interest from bonds or GICs, while it has generally done badly in high interest rate periods and often quite well in low interest periods, like now in our almost ZIRP period (actually, possibly negative rates if you factor in inflation and currency fluctuations depending on where you live).
One of the best explanations I have found so far on the movement of gold prices is here:
A Possible Model for the Price of Gold, by Eddy Elfenbein
http://www.crossingwallstreet.com/archives/2010/10/a-model-to-explain-the-price-of-gold.html
Gold would thus be an hedge against low interest rates for your portfolio. Good now, but for how long until interest rates go up???
@Craig: Many thanks again for your thorough answers to these comments. We all appreciate the time you’ve put into this discussion, and your thorough knowledge of the Permanent Portfolio’s subtleties.
For the record, I hope I didn’t imply in my post that gold was a poor investment in general, or that it has no place in a diversified portfolio. My only argument was that “gold is a hedge against inflation” is something you hear all the time, yet there is no data to support that idea, assuming you’re talking about a sharp rise in the cost of living. Clearly holding gold is a huge boon if your home country’s currency collapses completely—fair enough. I just think there is a distinction that investors should make.
http://www.efficientfrontier.com/ef/197/preci197.htm
Gold doesn’t increase in value, it maintains purchasing power as private Central Banks around the world debase/devalue their currency in a race to the bottom. The definition of inflation is an increase in the money supply. Forget the bogus numbers the governments come up with as they manufacture these formulas to suit their own results they want to achieve. If you want real numbers go to http://www.shadowstats.com/ . It shows the real inflation rate in the US is about 11.2% and real unemployment of over 22%. See his background, look at the methodology he uses. He uses what governments used in early 80s.
The correlation of gold to dollars is overwhelming. Greenspan has admitted on more than one occasion that the gold price is manipulated (as is silver). The Chinese have found out about this too. http://www.zerohedge.com/news/wikileaks-discloses-reasons-behind-chinas-shadow-gold-buying-spree
To protect yourself, invest in gold and silver. Gold will preserve your purchasing power and you will make money with silver as it heads back to its historic ratio of around 16:1.
Since 2001, stocks have not really performed well at all considering the lost purchasing power due to inflation (more currency in circulation). Again, forget the govts CPI numbers, they are bogus. Look at M1 and M2 numbers then factor in the banking systems fractional reserve and how they create even more currency with a keystroke.
Great information, great analysis, great reader input. Wow! you have a great forum. Keep the great simplified articles on complex business flowing for the neophytes like myself. The regular contributing commentary by well informed readers who we see week after week completes this great site. Thank you Dan, and all your regular followers.
@Superior John: Thanks, and I couldn’t agree with you more about the value added by the regular readers of this blog. I think investors learn as much from the comments as they do from the posts.
Hi Dan,
Fast forward to 2016, and it seems gold seems to have made somewhat of a recovery after a few years of decimation since the last comment on this article was posted.
My question is, have your thoughts on gold changed at all in the last few years? Does the stuff warrant a slice in a well balanced portfolio? Heck, given the current financial malaise, massive debt, low rates, questionable government policies, and who knows what new devious plans are being contrived in the minds of the Wall Street elite as of late, maybe it’s not such a bad idea. Looking at a long term graph of the S&P 500, seems like a large correction would not be out of place – even though it feels like we’re in one right now.
Anyway, enough rambling. My current portfolio is:
VXC – 60%
VCN – 20%
VAB – 20%
I’m not a huge fan of the yellow metal, but I am a fan of diversification, and these days with international markets moving in tandem, well, maybe I should consider adding the stuff to my portfolio.
I was wondering if, in your opinion, you’d say that since Canada is a commodities based market, I kinda already do own gold in a very roundabout way, and in a very diluted amount through 20% VCN (Barrick, Goldcorp etc). My preference would be to use a more liquid ETF consisting of gold miners, in this case, XGD, but I’m open to a bullion backed ETF.
I was thinking of diversifying my allocation to:
VXC – 60%
VCN – 10%
VAB – 15%
XGD – 15% (iShares S&P TSX Global Gold Index Fund) MER=0.61%
This steers the portfolio away from replicating Canadian sectors in VCN, that are already held internationally in VXC, and replacing them with more of a gold focus via XGD.
Anyway, thought I’d throw this out there, I’m guessing some readers may be thinking of something similar.
Cheers.
Hi Dan :
Recently, investment advisor and writer John DeGoey on BNN recommended holding 5 % or less of XGD in an 6 to 10 ETF portfolio . Does this provide a counterbalance to currency devaluation or down turns in the market. It would seem you don’t think it’s necessary, does it make a difference if you take a shorter ( 2 to 3 years ) or longer term approach ? (More than 5 years) Goey says he’s looking at 10 years or more. thanks
Fast-forward to 2021. I think an overweight portion of the portfolio in gold bars makes sense in the current context with governments all over the world that keep printing money with no end in sight. Gold is a finite currency and I don’t see why it wouldn’t rise in value relatively to other currencies if they drop. I read the post and agree that the data looks shaky on the topic of gold as an inflation hedge depending on your start and end points, but on the long-term (multi-decades) it looks to me as if the argument holds. Besides, gold is not correlated or inversely correlated with stock market movement, which provides a nice hedge against sudden drops. With the current U.S. market being at high historical valuations it could well stagnate or even fall. I see gold as an insurance policy against such scenario.
The “unproductive asset” argument does not ring so true today. A bond at near-zero rate (or negative real rate) is also an unproductive asset, but it is more correlated to the stock market than gold would be. I agree that the argument holds if one is speaking about replacing stocks with a portion of gold, however I don’t see it holding as well if one were to replace bonds instead. Personally I see gold as a portfolio stabilizer and it helps me sleep at night knowing that it will not move together with the market (or lose as much) in the advent of a crash.
I built a portfolio with the following weighing to navigate the coming years, all stocks or gold.
US: 27%
CAN: 13%
EAFE: 14 %
Emerging: 19%
Gold: 27%
In short it’s an aggressive 30/70 portfolio which replaces bonds with gold. Time will tell if it was a wise move…