Tuesday, May 11, 2010
Today gold bugs are celebrating as the price of gold rises by $30 dollars and €31 euros. Gold has now reached a new all-time against both currencies. In yen, however, gold is still almost 30% below its 1980 high (1980 figures in this chart are month-end, not daily, so the actual gold high was somewhat more than what is reflected in the chart). Thus we see how dramatically the yen has appreciated against the dollar and the euro in the past 30 years. Not surprisingly (since inflation and the value of a currency are intimately related), Japanese inflation has been much lower over this same period than inflation in the U.S. and the Eurozone: since early 1980, Japanese inflation has totaled a mere 35%, while U.S. inflation has been 150% and German inflation 90%.
Supply-siders like me believe that gold is a good common denominator against which to measure currencies over long periods. Currencies that hold their value better against gold invariably have lower inflation than currencies that don't. A corollary to this is that when all currencies decline meaningfully against gold, as they have been doing in the past 4-5 years, then global inflation is quite likely to rise. Given the substantial degree to which currencies have fallen relative to gold since 2005 (a few years after most central banks adopted accommodative monetary policies), we should therefore expect a substantial pickup in inflation around the world in coming years, and the cause will be easy money. Since the end of 2005, gold is up 80% vs. the yen, 120% vs. the euro, and 140% vs. the dollar.
Unfortunately, I am unaware of any formula that relates changes in gold prices to changes in future inflation. The linkage is loose, the lags are long, and there are other things which get into the mix—such as geopolitical risks—that muddy the waters. But if this theory of gold and currencies holds any water at all, we should see rising inflation in the future, and that should be quite a surprise to most global bond markets, since they are currently priced to inflation remaining very low and stable.
By way of illustrating how this process works, I offer the following simple rule of thumb for any central bank desiring to keep its currency stable against gold (and thus replicating a gold standard): add or subtract whatever liquidity is necessary to keep the price of gold within a relatively narrow band. In practice, that means trying to find the short-term interest rate that makes the public indifferent between owning a short-term deposit or owning gold.
If interest rates are too low, gold becomes more attractive and rises; people prefer to hold less money and more gold, and the unwanted money tends also to get spent on things, pushing up their prices in the process (this is similar to the velocity story I have been highlighting in recent months). If interest rates are too high, the public prefers to own bonds rather than gold, and gold prices fall. When gold prices are stable, a currency is "as good as gold;" demand for the currency exactly matches the supply of the currency, and inflation is negligible. The history of gold standards tells us that when implemented correctly, a gold standard is virtually guaranteed to deliver very low inflation.
Is the rise in gold prices a signal to buy gold? Not necessarily. Indeed, I would argue that gold prices at today's levels already reflect a substantial amount of monetary inflation. Inflation almost has to go up to validate current gold prices, and central banks almost have to continue making the mistake of keeping interest rates too low. Gold has had a substantial run, and it can't keep rising at this rate forever. If the past is any guide, gold might rise about 1-2% a year on average over the next several decades. Buying gold today for the long haul is almost certain to be a poor investment, although gold could certainly rise further over the next several months. Gold is going to be the asset most vulnerable to the first indications of central bank tightening, whenever that happens to occur.
Remember when gold used to be worth about $35/oz. back in the1930s, 40s, 50s, 60s, and 70s? Taking early 1940 for purposes of comparison (inflation was about zero that year, after having been negative for most of the 1930s), the U.S. CPI today has increased by a factor of 15.6, while gold has increased by a factor of just over 35 (a strange coincidence). That sounds impressive from gold's perspective, but on an annualized basis, gold has appreciated only 1.2% more than the rate of inflation over the past 70 years, and that's not very impressive. You can't expect that buying and holding gold for the long haul will do better than most alternative investments.
Finally, as these next charts show, the price of gold in constant dollars tends to oscillate around some value (e.g., the real price of gold is mean-reverting) over time. By any historical standard, the price of gold is quite high in real terms relative to its long-term average. (Note that the first chart uses year-end values, while the second chart uses month-end values.)
Posted by Scott Grannis at 7:09 PM