Should you own some gold? If so, how much? Why own gold? What might make its price rise dramatically? Although the future is uncertain, rational answers to these questions are possible. That means you can answer them. I can help somewhat. To begin with, one should understand the relations among three things: money supply growth, gold prices, and the consumer price index (CPI).
Between 1930 and 2004, gold rose in price from $20.67 an ounce to its present level of about $440. That’s a growth rate of about 4.1 percent a year. Over the same period, the CPI index rose from 16.1 to 190.3 or about 3.3 percent a year. Both gold and the CPI are linked to how much the money supply grew over this period. The M1 money supply was $24.92 billion in 1930 and grew to $1,340.2 billion in 2004. The growth rate was 5.4 percent per year.
I do not discuss here what real money might be in a free economy, which the U.S. does not have, or the pros and cons of various definitions of the printing press (fiat) money we do have. I will mention that the Federal Reserve (Fed) controls the supply of this fiat money. It is the only legalized entity that can write any amount of checks on itself without having anything in its account. When these checks are presented for payment by banks, it credits their account with reserves — a purely electronic or paper transaction. If the bank wants currency, the Fed gets the Treasury to print some and delivers it. This legal counterfeiting operation is how central banking works.
We the people have no possible use for the Fed that I can think of, but the State does. The Fed buys a lot of the debt that the State floats. It also seems to be a convenient way for the State to distribute its printing press money whenever it wants to. However, the U.S. government got along without the Fed until 1913.
Naturally the Constitution nowhere gives the Congress the power to print non—interest-bearing paper and call it money or legal tender. It does prohibit the States to "make any thing but gold and silver coin a tender in payment of debts," which clearly suggests that the Constitution meant gold and silver coin to be money. It does give Congress the power "to coin money, regulate the value thereof, and of foreign coin, and fix the standard of weights and measures," which means to imprint gold coins and determine their weight. Congress also has the power "to provide for the punishment of counterfeiting the securities and current coin of the United States," which again says that money was coin. I suppose that this punishment has been declared to be nil in the case of the Fed.
There is also a demand for money, subject to a myriad of factors that will not detain us. A major one is that as the population grows, the demand tends to grow along with it. Since 1930 the U.S. population grew at almost a 1 percent rate.
We notice that gold’s price growth exceeded that of the CPI by 0.8 percent a year. Although that seems fairly close, over a period of 75 years the difference adds up. There are many difficulties and strange features of how the government computes the CPI, there have been many new sources of supply and demand for gold, and there have been many economic and political changes since 1930. Yet the two growth rates are not too far apart. It seems that gold has risen in price about in line with other common goods and services. Some people say that it takes about the same amount of gold (about an ounce) to buy an off-the-rack average suit of clothes today as it did 75 years ago.
This means that gold has been a good hedge against price rises (as measured by the CPI), or that gold prices kept up with price rises. The natural growth of supply and demand for gold has apparently remained so steady relative to other goods that the price of gold in terms of other goods has not changed very much.
There are some whose assessment differs substantially from this one. They estimate that it takes 3 ounces of gold worth $1,320 to buy a suit today, where it used to cost 1 ounce worth $21. They believe that the gold price today should be higher by a factor of 3 compared to its present price, or that $440 is way too low. The source of these statements is often someone anxious to market gold. I stick with the conventional view that gold has tracked the CPI closely.
The huge rise in the CPI index and in the price of gold means that the value of the dollar fell drastically between 1930 and the present. Since gold rose by a factor of 21.3 (or 440/20.67), the dollar today is worth only 1/21.3 = 4.7 percent of its value in 1930. The dollar has lost over 95 percent of its value in terms of gold. In terms of the CPI index, it has lost 92 percent of its value.
The cause of these declines in value of the dollar is that the Federal Reserve has shifted the supply of dollars upwards, well beyond the amounts that were warranted by the increases in demand for dollars that have occurred over this period. It bought a lot of things (mainly the debt of the U.S. government), created a lot of phantom reserves, and printed a lot of counterfeit currency called legal tender. These busy printing presses drove the value of the dollar down. The more dollars that were sitting in banks (because of cashing the Fed’s checks), the more that people borrowed and spent. Since the printing presses did not create any real goods or services, these dollars simply bid up the prices of the goods and services. That is (mainly) why a t-bone steak that cost $0.29 a pound in the 1920’s costs $7.99 today.
Since the population of the U.S. grew by 1 percent a year, a very rough guess is that the supply of money since 1930 grew by 4.4 percent in excess of what population growth required. That might help explain why gold’s price went up by 4.1 percent a year, which is 1.3 percent less than what M1 rose. Gold’s growth exceeded the CPI’s growth perhaps because the CPI understates inflation. Both gold’s price increase and the excess growth of the money supply are better measures of inflation than the CPI. Measuring "excess money" is a problem, however, and gold’s value is volatile in the short run.
What’s inflation? Some economists define inflation in terms of price rises of goods and services. They use the CPI, but this measure is flawed for a variety of reasons. Alternatively, it may be more clear to say that inflation means excessive growth in the money supply, which then usually translates into rises in prices of various goods and services. Since in our system the State controls money via the Fed, the State is the sole cause of inflation because it controls the supply of fiat money. The Fed, usually doing what the President and Secretary of the Treasury want, causes inflation by increasing the supply of money. The public, however, can affect the rate of increase in prices by how intensively it uses this money (the money velocity).
Suppose that gold prices accurately measure the inflation rate. Its growth rate was 4.1 percent, which was 1.3 percent below M1 growth. Suppose the Fed had made the money supply rise by just 1.3 percent a year, then gold would today be about the same price as in 1930.
As we look ahead, we see that the Fed can keep inflating the money supply indefinitely, or until a hyperinflation occurs that ends the process. To get the gold price to $10,000 an ounce, gold has to rise by a factor of 22.7 (that’s 10,000/440.) That means the dollar has to lose another 95 percent of its value. By then, a candy bar will cost $11. That’s not so outlandish. Today it costs fifty cents and many of us remember when it cost a nickel.
How long might this further depreciation in dollar value take if hyperinflation does not intervene? If it continues on at the same rate as during the past 75 years, then it will take about another 75 years. In the year 2100 a home that today costs $150,000 will cost $3,405,000. Some home buyers are rushing history a bit, it appears. I have a gut feeling that this is not going to happen, that something is going to happen that interrupts this process, for better or for worse.
Back to gold. Gold is an asset that goes up at the rate of inflation in the long run. Its ups and down in the short run depend on many factors, including investor perceptions of money supply growth in the future. More importantly, the ups and downs of the money supply also influence the prices of other portfolio assets like stocks and bonds in ways that are not always easy to decipher and not rigidly connected to what the price of gold does.
Remember the little toy where you roll the balls around trying to get them into the holes? You roll one and it dislodges the others? There must be some video games like this. That’s how stocks, bonds, and gold are. They move in crazy-quilt patterns sometimes; one goes in the hole and the others jump out. The ups and downs of gold do not correlate well with other assets like stocks and bonds. That makes gold a good asset to hold for diversification purposes, because sometimes it gives you a home run when the other assets are striking out.
Most investors do not hold much gold in their portfolios. They should, because gold is a good diversifier. There is no magic number for how much to hold. One source suggests a 6 percent holding, which gives the conventional order of magnitude that is recommended, usually from 3 to 10 percent. Actually, it would be nice if investors could conveniently hold a number of commodities. For example, the lowly metal, copper, has far outperformed gold in the past few years.
If you hold too much gold, you are likely (in many or most scenarios) to get greater volatility and too low a return, since its price is volatile and, although gold’s overall rate of price appreciation keeps up with inflation, it does not give a positive real return. How much is too much? Harry Browne has recommended holding 25 percent in gold, 25 percent in stocks, 25 percent in long-term bonds and 25 percent in short-term bonds. This is actually a conservative portfolio that has shown very steady performance. So you could go up to 25 percent in gold it seems. Since 1930, gold, by itself, has been a worse investment than bonds or common stocks. Over long periods of time, stocks and bonds have kept up with inflation and also provided some additional or real return. Nevertheless, a mix of some gold in a portfolio of other assets typically benefits the portfolio in those time periods when the stocks and bonds are doing badly.
Holding gold via an exchange-traded fund (ETF) currently has a big drawback, namely, gains are taxable at a 28 percent rate for any ETF that holds bullion. The IRS treats gold as a collectible and taxes capital gains accordingly. It is better either to buy gold stocks or a precious metals mutual fund until some ETFs come along that avoid this negative tax feature. An ETF based upon an underlying mutual fund will not have this problem.
Gold will be a superior investment to stocks and bonds in a number of gloomy but possible scenarios that involve rapid depreciation in the country’s paper or fiat money. If you believe strongly in these scenarios or have special insight that they will occur, then you may want to increase the amount of gold you hold. However, before going overboard, it is well to remember that most of us can’t beat the pari-mutual odds in a race with 5 horses.
First. If the Fed speeds up growth in the money supply, that’s greater inflation. Then gold will tend to rise more quickly in price too. Gold fell below $300 in the Clinton-Rubin era because the rate of growth in M1 slowed down from 7.8 percent a year in 1989 to 1994 to —0.01 percent a year over 1994—1999. That’s right. Inflation was essentially zero between 1994 and 1999. This is also one reason that interest rates declined so much and why the stock market rose so much.
Between 1999 and 2004, however, the growth rate of M1 picked up to 4 percent a year. The big spurts came on Bush’s watch as did the big stock market drops. I hold the minority opinion that the bear market in stocks is Bush’s bear market. The market began dropping one month before his election and the drop accelerated thereafter throughout 2001 as investors figured out what he was up to. Gold then jumped from the $300 area to the $440 area where it is now as it became apparent that his administration was up to more inflation. Wars always generate inflation, and Bush’s wars are no exception. Ordinarily, a record as poor as his would have meant a change in Presidents but the Democrats put up an extraordinarily weak ticket and ran a terrible campaign.
If the Fed in the future increases M1 at the 5—7 percent rates that were common from 1964 to 1994, what is likely to happen? A sustained 6 percent rate of increase in M1 will get gold up to $10,000 in just 52 years if gold rises at the 6 percent rate too. However, a steady rise in gold is very unlikely. Investor expectations could change sharply, as they have in the past, and gold might overshoot and jump to near $800 within several years of an M1 spurt of this size. Of course, the market would be highly speculative and constantly be re-assessing what the government (and thus the Fed) was up to and how long the high rates of M1 growth might last.
This scenario has some likelihood, anywhere from a 5 percent chance (20-1 odds against) to a 20 percent chance (5-1 odds against). These numbers start in my gut and progress to my pen. The chances rise if the Bush administration starts another costly war or decides to increase the military budget sharply. The administration is planning a war against Iran but "on the cheap," by using massive bombardment and local revolution rather than ground troops. However, these plans could go awry and lead to bigger expenditures. If the U.S. leads the world back into another arms race, as it seems to be trying to do, that also makes this scenario more likely.
Second. If the Fed increases the money supply faster and investors flee from the dollar or from dollar-denominated assets into commodities and real assets, then gold will skyrocket. In other words, if investor expectations of inflation alter from expecting a relatively benign (steady) and manageable process toward expecting a malign (wild) and uncontrollable process, then the velocity of money will increase. This will set off much higher price inflation even if the Fed keeps money growth at 6 percent. This scenario is like what occurred 25 years ago when gold rose sharply to $800 an ounce. A similar 6-fold rise today would push gold to $2,500 an ounce in a relatively short period of time. If this scenario occurred at any time in the next 10 years, the rate of return on gold would be a minimum of 17 percent a year at some point.
This scenario depends in part on a factor that is highly unpredictable, investor expectations. There is a herd mentality that can go to work in this case. If the general public gets the idea that major figures like Buffet are abandoning dollar assets and moving capital overseas, it will lose confidence. This then exacerbates the process of dollar flight.
Third. If there is a political-economic shock of some sort that creates flight from the dollar and dollar-denominated assets, then gold’s price in dollars will rise sharply although its price need not rise in terms of other currencies. For example, suppose that the U.S. Treasury had difficulty in refunding debt at some auction and domestic interest rates began to rise sharply. This might come about if foreign investors lost confidence in the revenue-raising ability of the U.S. government because of a tax revolt or a secession movement. Then people anxious to get rid of dollars might bid the price of gold up. Shocks are, by definition, unpredictable. However, coming events sometimes cast their shadows before. If gold should begin rising steadily for no apparent reason, or in spite of a friendly environment, it would be a sign that something is amiss.
Fourth. Hyperinflation is possible. This is a very rapid increase in money supply accompanied by increases in the velocity of money. These drive the prices of goods up dramatically. In this situation, gold in terms of the depreciating currency will rise to extremely high prices. Hyperinflation is unlikely in the U.S. for several reasons. Most of the U.S. debt is short-term. If the Fed prints too much money, interest rates will rise and the budget deficit will increase sharply. This factor restrains the Fed from letting M1 rise too fast. Also, hyperinflation is a government’s last resort way to finance expenditures when it has no other way. The tax system in the U.S. rakes in so much money that it makes hyperinflation less likely.
All of these inflation scenarios cause stock and bond prices to decline. Since bonds are contracts that pay off in dollars, they are negatively impacted as dollars are depreciated by inflation. The effects on stocks are more complex and vary across different types of stocks. Usually the markets are disrupted enough by inflation that the overall net effect is negative.
Suppose that a portfolio has $9 in stocks and bonds and $1 in gold. Under one of these extreme scenarios, suppose that the $9 investment falls to $1, while the $1 investment in gold rises to $9. Then the portfolio maintains its purchasing power, despite the enormous loss in the bond-stock component. That is, a 10 percent position in gold maintains the whole portfolio value if gold rises by a factor of 9 while other assets fall by 89 percent. There are many such possibilities. A 5 percent position maintains 70 percent of the portfolio value if gold rises 450 percent when stocks and bonds fall by 50 percent. The $0.95 worth of assets declines to $0.475 and the $0.05 of gold rises to $0.225, which add up to $0.70. A 20 percent position maintains total value if gold triples when the stocks and bonds lose 50 percent of their value. This scenario sounds fairly plausible. The $0.80 of stocks and bonds dips to $0.40 while the $0.20 of gold rises to $0.60.
Looked at this way, it is evident that gold in the portfolio is equivalent to insurance against some devastating contingencies. Complete or partial insurance are possible. The more gold you have, the more insurance you are buying. Over-insuring is costly because the overall rate of return of the portfolio goes down if nothing happens. That’s because gold historically provides no real return. Everyone has to decide for himself what the odds of these scenarios or ones like them are, how to insure against them as with gold or some other real assets, how high gold will go when other assets decline, and how much to insure against these events. There are no pat answers to these decisions, but they are within the realm of understanding and even sensible computation.
Michael S. Rozeff [send him mail] is the Louis M. Jacobs Professor of Finance at University at Buffalo.