February 06, 2006
Gold and inflation
What's behind the ongoing run-up in gold prices? One popular interpretation is that investors fear a resurgence of U.S. inflation. But that story just doesn't square with the facts.
The above graph shows the spread between the yield on standard 10-year U.S. Treasury bonds and the yield on 10-year U.S. Treasury inflation-protected securities (TIPS), whose coupon and principal both increase each year with the consumer price index (CPI). This spread is basically a measure of the expectations of inflation that are built into the TIPS yields. This expected inflation rate rose 100 basis points between 2003 and 2004, but has been virtually constant for the last year and a half at a value around 2-1/2 percent. Fears about a resurgence of inflation? Not in these bond yields.
I often hear stories along the lines of, "the bond market doesn't expect inflation but the gold market does." But this makes no sense. Markets are fully integrated, and there's no reason why somebody with a particular world view wouldn't try to grab spectacular returns wherever they might be found. If you think that U.S. inflation is going to be substantially above 2.5%, there's no question that you want to go long TIPS and short the nominals. If gold and bonds are sending conflicting signals, it has to be due to some fundamental differences in the conditions under which the two assets will perform.
For example, if you were worried about a surge in prices that was not that strongly correlated with the CPI-- perhaps a rise in the price of traded goods relative to nontraded goods-- that might do it. But the story then is not a fear that the Fed wouldn't do its job. On the contrary, bond yields are reflecting confidence that the Fed will keep the CPI proper under check. Or the gold market could be responding to concerns not about inflation per se but rather broader financial or political instability. Certainly there are concrete developments that could warrant heightened concerns about the latter over the last few months.
Institutional Economics notes that the gold moves should also be seen as part of a broader trend. It seems a mistake to come up with a completely different story for gold, oil, copper, aluminum, zinc, and so on, when they've all been exhibiting similar behavior. To be sure, there are separate important factors in supply and demand that are unique to each one. But there is also something to be said for focusing on the factors that are common to them all.
We've seen a pretty impressive surge in commodity prices relative to other prices. But that is not quite the same thing as inflation, in which the price of a broader basket of goods and services (such as the CPI) goes up at a faster rate. If you want to claim that the current gold price reflects a fear about the latter, you have to reconcile it with the way that nominal and inflation-indexed bonds are currently priced.
Posted by James Hamilton at February 6, 2006 11:51 AMdigg this | reddit
Listed below are links to weblogs that reference Gold and inflation:
» The Myth of Diverging Markets from The Stalwart
Economist James Hamilton makes a good, and what should be obvious, point:I often hear stories along the lines of, the bond market doesn't expect inflation but the gold market does. But this makes no sense. Markets are fully integrated, and [Read More]
Tracked on February 7, 2006 06:30 AM
Tracked on February 10, 2006 01:59 PM
Excellent. I was hoping someone would take the time to produce this graph in blogspace and write the piece you just did. For an example of the insanity of the commodity price rule aka gold bugs, check out Jerry Bowyer's grading of the Greenspan FED over at National Review Online.
Posted by: pgl at February 6, 2006 02:23 PM
It seems a mistake to come up with a completely different story for gold, oil, copper, aluminum, zinc, and so on, when they've all been exhibiting similar behavior.
No kidding, aluminum has gone from around 80 cents a pound a year ago to about $1.20 recently... zinc from less than 60 cents last summer to over $1.10.
I just got a newsletter about special steel alloys telling me to expect further hikes through 2006... on top of past ones!
Its not just gold.
Posted by: dryfly at February 6, 2006 03:35 PM
So you see the base metals are going up 1:1 with gold and no reaction in bonds. What is your explanation? Are you sure the CPI numbers are correct, after all you see prices going up everywhere? Could it be that the gold bugs are right that gold leasibg is part of the carry trade. That would keep bond interest rates down and keep gold from registering a warning about inflation.
Posted by: Mike Landfair at February 6, 2006 03:56 PM
"Are you sure the CPI numbers are correct, after all you see prices going up everywhere? "
My thoughts exactly Mike.
Posted by: me at February 6, 2006 04:02 PM
Does the TIPS CPI exclude food and energy? If so then this might reflect recent higher-than-usual increases in those commodities, particularly energy.
Posted by: Hal at February 6, 2006 04:28 PM
What if someone were "gaming" the bond market?
That is, buying bonds at any yield, thus driving yields down for everyone.
What if, say, a large nation, with a powerful central bank, which was suddenly flush with dollars, had nothing to do with them except buy bonds like this?
What if this nation didn't care about inflation in the US dollar, or trade imbalance? What if they only cared about selling their goods abroad and industrializing at the most rapid pace the word has ever seen?
What if they didn't care about the long-term consequences of these actions on the US economy?
Seems to me like this would establish a bond market with degenerate behavior; one that would no longer act as a useful predictor of inflation.
But, nah, not gonna happen. Therefore I deem the gold market insane.
Posted by: Aaron Krowne at February 6, 2006 04:55 PM
One last addendum.
The situation would really be abysmal if the country wasn't democratic, and people couldn't respond to the negative effects of such a policy by voting out their government.
But where's there a country like that in the world?
Posted by: Aaron Krowne at February 6, 2006 04:58 PM
So, the inflation story doesn't add up given the disconnect between commodity prices and bond yields. Perhaps we could look at the demand and supply dynamics in the bond market to help explain. For example, in the UK long real yields have been squeezed right down not due to a view on inflation, but due to restricted supply and super strong institutional demand. Given that the pension crisis is also an issue in the US I wonder if long yields are also being held down by institutional demand and are a false representation of the inflation outlook?
If we assume a world of surplus liquidity, it is easier to explain high commodity prices. Bonds, equities, housing markets around the world, and bonds have all faced very strong demand. I wonder if the commodity market is just late to the party? We know for sure, that a lot of funds are moving in to this space, many with the preconceived notion of a new world of permanently high commodity prices.
The other thing with gold is that the traditional correlation with the USD has inverted. Maybe this relationship can help square the circle? I'm not sure myself.
Posted by: Abobtrader at February 6, 2006 05:10 PM
Hal, the TIPS are indexed directly to the regular "headline" CPI, not to a CPI excluding food or energy.
Posted by: JDH at February 6, 2006 05:13 PM
How is the supply-demand equation of gold and other metals determined?
JDH’s caveat that “there are separate important factors in supply and demand that are unique to each one” should be pursued before we jump to conclusions.
As dryfly noted, the cost of aluminum is rising sharply and is expected to continue to rise. World demand is not keeping up with supply. Real market forces are at work here. Such may not exactly be the case with gold. Supply may well be in the hands of Central Banks.
Consider the following two articles, one from Reuters on aluminum; the other from Huynh Trung Khanh, chief representative of the World Gold Council in Vietnam.
“For the wider market, Alcan forecast a 4.9 percent rise in world primary aluminum consumption in 2006, up from a 4.5 percent increase in 2005. It expects production from new capacity and restarts to boost world supply by about 4 percent during the year, compared with 6.9 percent last year.
“That would produce a market deficit of about 300,000 tonnes of aluminum in 2006, compared with a balanced situation in 2005, it said.
“Engen said Alcan's forecast for demand growth to outstrip supply growth bodes well for aluminum prices , which earlier on Tuesday touched a 17-1/2 year high of $2,678 a tonne.”
“Dao Hong Chau, deputy CEO of Export Import Commercial Bank said mounting Middle Eastern political term oil was responsible for the jump, adding the stronger US dollar was in its way though.
“He added gold supply was far outstripping global demand, worsened by intense buying from international banks.
“Huynh Trung Khanh, chief representative of the World Gold Council in Vietnam, said Europe’s central banks would actually offload over 1,440 tons of gold from 2005-10 though they had agreed to sell 2,500 tons at most during the period. “
There seems to be more going on than meets the eye or than can be seen in one graph.
Posted by: Stormy at February 7, 2006 09:04 AM
All this talk of bond yields seems to miss somewhat the point that Jim is making: the SPREAD between TIPS and nominals hasn't trended wider in a year and a half. As further support, I would note that during calendar year 2004, when the central banks were purportedly engaged in their largest purchases of Treasuries, the spread was also pretty stable. Since then, the TIC data show official purchases as having moderated last year relative to '04. I guess the Bank of China must be buying TIPS and nominals in just the right proportions to manage the spread. Those cunning central bankers - what will they think of next!
Posted by: Kevin T at February 7, 2006 09:19 AM
That is a good point. I'm looking at the TIC data, and that's some interesting stuff. It looks like transactions with private individuals are actually the biggest factor.
But otherwise, the dynamic seems the same.
If I was Chinese (either the central bank, a private individual, or a corporation) and I needed to park massive quantities of dollars in US bonds, I would probably do so in as "diversified" a manner as possible. I.e., various maturities and classes (TIPS vs nominal).
Presumably, there should still be some bias towards the TIPS, if these buyers believe the CPI stands to rise.
But what if there isn't, for some reason? Say, for example, these investors being really immature, only entering into bonds and treasury notes to make hasty use of their dollars.
It does seem like the rank-and-file Chinese should have some intuitive feeling that the dollar stands to see some inflation, however. I mean, they are buried up to their necks in the damn things.
Posted by: Aaron Krowne at February 7, 2006 10:49 AM
Presumably we could have, as we have had, considerable increase in commodity prices that affect consumer prices very limitedly. Given commodities provide no income, their should experience much more significant fluctuations. Are we headed for another 87?
Posted by: Lord at February 7, 2006 01:05 PM
"If gold and bonds are sending conflicting signals, it has to be due to some fundamental differences in the conditions under which the two assets will perform."
Absolutely, but that doesn't mean to say there's not a rational explanation for the apparent anomaly.
As the FT's investment editor Philip Coggan pointed out a couple of weeks ago, the bears are currently in one mind as to the basic problem we're facing today, but in two minds as to the outcome.
They see a world awash with liquidity engineered by overly loose central bank monetary policy, with credit expansion out of control. As far as they are concerned, a day of reckoning *must* come, but what will be the result?
On the one hand we have the 1930's debt deflation, where cash is king. On the other, we have 'helicopter commander Bernanke' (as the gold bugs term him) ramping up the printing presses to prevent deflation at all costs.
Since either outcome is possible (if you accept the premise), the prudent investor hedges both outcomes, allocating the majority of his portfolio to cash and AAA bonds, with a sizeable portfolio alloaction also going into gold and other inflation hedges.
Do you include TIPS in the inflation basket? Doubtful. If the presses roll, what will happen to promises on such instuments? Can the indexing be trusted? And if TIPS are anything like their UK counterparts - Index-Linked Gilts - they would be next to useless in a period of very high inflation due to the lag in indexing.
Too conspiratorial? Ah, but you're not thinking like a gold bug...
Posted by: FTX at February 7, 2006 02:47 PM
JDH misses the bigger picture imo. TIPS are dollar-denominated. Gold seems to be rising due to international demand and the beginnings of, and expectations of, a move away from the dollar by private investors and some central banks, and has little to do with US inflation expectations. Why hold TIPS if you expect a falling dollar will counteract much of your inflation-protection? So not necessarily related issues.
Posted by: RN at February 7, 2006 04:05 PM
RN, Shouldn't TIPS (and treasury bonds) be falling for the same reason that you think gold is rising? If foreign investors are beginning to lose confidence in the dollar, bond prices should be falling as gold rises. That's not happening. Gold and bond prices are both high.
Posted by: sestina at February 7, 2006 04:56 PM
Someone please explain how the 10 year bond is no longer useful in predicting recessions (via the slope of the yield curve) but still very useful in predicting inflation.
And while you're at it, since when do central banks care about the return on their investments? Surely if they cared about things such as this, they wouldn't have dumped so much gold at such ridiculously low prices in the last ten years.
Posted by: Tim at February 7, 2006 09:44 PM
Warren Buffett famously said that in the long run the stock market was a weighing machine but in the short run it is a voting machine.
Bond markets are no different. UK index linked gilts are at 0.5% real yields. This is not because this is a reliable forecast of future UK inflation, but rather because UK pension funds, in the face of new legislation, are increasing their weightings in very long duration inflation linked instruments (to reduce liability risk).
There is a bubble out there in government debt. It reflects the weight of money globally (liquidity) going into asset prices.
So far, due to supply side effects (chiefly the movement of production of industrial goods from the West to China, and also service outsourcing to the likes of India), this has not translated into price inflation.
The flows of money are absolutely enormous. Since the US index linked treasury is the safest debt instrument in the world, there is going to be excess demand, given that the supply is anything but elastic (US isn't even issuing 30 year TIPS, and those that have been issued are seldom traded).
Luiquidity has however translated into asset price inflation. One symptom of which is low bond yields, another is a higher gold price (more money chasing essentially a fixed supply of gold).
Hence the divergence.
Using a very thin and illiquid market like that for inflation linked bonds (TIPS, indexed linked gilts, Canadian RRBs etc.) to predict inflation is the equivalent of predicting the weather using a very bad an inaccurate thermometre.
Posted by: John at February 8, 2006 03:27 AM
Professor Hamilton- Reading your chart, it appears that gold and base metals peaked in 1987-88 after which their prices declined until 2001-02. I assume that your chart uses nominal dollars, so the decline in real terms is even greater. Is not the run up in prices simply a recovery to "normal" levels, reflecting increased global economic activity?
Posted by: Rich Berger at February 8, 2006 04:13 AM
"It seems a mistake to come up with a completely different story for gold, oil, copper, aluminum, zinc, and so on, when they've all been exhibiting similar behavior."
Is the Gold vs Base Metals chart using different scaling? Eyeballing it, since mid 2002 Gold appears to be up 50% vs Base Metals of over 100%. Is that right? Is that "Similar Behavior"?
Posted by: Press at February 8, 2006 08:34 AM
I suspect Prof Hamilton is having a little fun here with the gold bugs and the peak oilers. Maybe he has a bet with the other contributers to see how many posts he can get.
More seriously - there are two possible consistent theories for the high prices in the commodity markets and the low inflation expectations of the bond markets. The first one is that they are both being driven by rapid industrialisation of the BRICS - accelerated demand by BRICS increases prices of commodities and accelerated supply by BRICS of manufactured items reduces inflation in the consumer market place. The second theory is that this is all to do with mysterious in-balances in the world economy possibly due to some clever manipulation by the Chinese in cahoots with Greenspan, which has resulted in a multi-trillion dollar markets being moved in opposing directions. I know which one I believe.
Posted by: ChrisA at February 8, 2006 11:14 AM
I think in theory you're right, but in reality now there are two different sets of parties at work. Those beginning the hedge against the dollar and driving up the gold price and those who remain dollar buyers at any price (like the Chinese who must maintain the RMB/$ peg).
I haven't done the hard research on this yet, mind you, so I could be way off, but it's difficult to see other reasons for such a broad move up in gold while the dollar is still as strong as it is.
Posted by: RN at February 8, 2006 03:58 PM
In particular, gold prices have been affected mainly by Japanese investors (and others) trying to protect themselves from a falling Yen, which is the Japanese government's policy as of late.
The others are playing the carry trade; borrow Yen to buy gold.
As a matter of fact, BoJ recent appeals to tighten their currency provoked sharp falls in gold prices; the reassurances that the BoJ would sustain their low rate sent gold prices skyrocketing (both moves over 3.5% for the day, IMSR).
But, in general, John is on the money; there are too many USDs out there searching for a home, a bubble in commodities, why not?
Also, China is recognized to be actively pursuing commodities and resource extraction facilities for two very good reasons:
(1) Strategically it makes sense for the long run, they donï¿½t want to be left out of scarce resources.
(2) Buying commodities is a very smart way to diversify out of the USD without provoking a drop in its price, which would hurt their enormous USD holdings, and potentially provoke a rise in the Yuan, which in turn would hurt their pricing competitiveness in the world.
Posted by: Joe Rotger at February 11, 2006 05:59 PM
Most posts here are completely off-topic. The original post is about gold, bonds and inflation, not china. I will offer my 2c on the off-topic first. The outstanding value of US treasuries is over 8 Trillion, China holds about 250 Billion (http://www.ustreas.gov/tic/mfh.txt) , or about 3%. Moreover, the Chinese government does not print US dollars, it extracts them from trade and investment flows. This means that the same dollars have been previously held by someone else, likely in treasuries as well. The entire story of china "manipulating" and "threatening to dump" US treasuries has been invented by people who do not own their own car much less billions in US bonds.
As for gold and TIPS/bond spread, I personally do not know anyone who, while worried about inflation bought both gold and TIPS. These two "investments" are held by two entirely different groups of investors, and both are suckers bets. Neither of them protects you in the long run from inflation, and assuming you buy gold futures and not physical gold, should produce negative real returns to the tune of -3%pa. TIPS are also likely a suckers bet, because Tbills are likely to outperform TIPS over almost any investment horizon whether in real or nominal terms.
The only good protection against losing savings to high inflation, my father taught me in the dying days of the Soviet Union, is to spend your money now. And if and when the hyperinflation comes, you should spend your remaining money on an airplane ticket out of the country. His advice is still right.
Posted by: Kotika at February 16, 2006 05:20 AM
Gold & inflation, gold & dollar, stocks & bonds, oil & equities markets, all these formerly meaningful relationships have been decoupled in recent years.
All asset classes are rising due to overliquification caused by lax monetary policies by the central banks.
In an age where ruling parties distort data to serve their own interests, it is hardly surprising that financial ministers present price indices and inflation data in a manner best suited to their immediate needs.
Likewise, central banks and vested economic interests will manipulate forex exchange and bond markets to suit their immediate & near term financing needs.
Traders should always trust the money flowing through the commodity markets for real time indications of future inflation, and not government statistics or any instruments based on those misleading statistics.
Posted by: The Nattering Naybob at February 20, 2006 06:20 AM