The monthly model (R-squared = 0.95) going back to 2001, shows some major froth on the upside. The intra-month highs in August and September approached 1900 and reached more than 4 sigma above the linear regression. The closing price on 9/30/2011 was $1,620. Here are the prices predicted by the model:
Two sigma below: $1,335.27
One sigma below: $1,417.52
Best fit: $1,498.20
One sigma above: $1,578.87
Two sigma above: $1,661.13
Expanding the model to yearly prices back to 1971 (R-squared = 0.64), prices are well above trend (predicted best fit price = $1,126.52). The long bull market periods are marked by extreme stress in the financial system. I could be wrong, but the nature of the stress in banks and sovereigns seems much more severe than it did in the 1970s and 1980s. European countries were not going bankrupt in the 1970s while tied to a Euro wide currency. The private debt levels, financial sector debt levels, and sovereign debt levels were not close to what we have now, and no resolution seems to be near. Instead, every effort is being made to sustain the unsustainable. A lot of paper promises are going to be broken. That doesn't mean the gold price hasn't already discounted all of the turmoil to come.
Whether the big gold sell off in September was the start of a bubble collapse or just a correction in an ongoing bull market, I don't know. I think we will know within six months. While I made some good trades during the August and September froth, I also gave some back. Until I have a better understanding of what story gold is telling, I am going to back to the discipline of my statistical models.
Here's something else you might want to factor into your model.
ReplyDeleteYou are not adjusting for increase in the gold supply.
As seen here, world gold production increased 65% from 1970 to 2009. Production did not add 65% to the gold supply but it certainly added a non-trivial amount.
Further, the high price of gold should lead to more production in theory. There was a 7.0% increase from 2008 to 2009. Not sure if it is sustainable or not but there is a risk clearly.
Put another way, from 1980 to 2000 there was no real reason to increase gold production much. Prices were depressed. It was not a desired asset. There are lots of reasons to increase production now though. How much it can be increased is open for debate. I am mostly agnostic on the topic, due mostly to general skepticism.
Stagflationary Mark,
ReplyDeleteGold supply went up, but I think dollar supply went up more. It is true that gold tends to accumulate. Very little is lost to consumption unless it is actually lost.
In my opinion, the risk you are running by ignoring the supply of gold is somewhat similar to the risk investors ran by ignoring the supply of housing.
ReplyDelete2450 metric tons of gold were produced in 2009. At $1600 gold, that's $126 billion.
That might not seem like a lot compared to your dollar supply chart, but consider that gold is just ONE thing out of many things that people can spend their money on.
In other words, if the dollar supply goes up $1 then that does not mean every asset can go up a dollar. It has to be spread around.
Stagflationary Mark,
ReplyDeleteIn my opinion, the risk you are running by ignoring the supply of gold is somewhat similar to the risk investors ran by ignoring the supply of housing.
A gold bubble could, and likely, will happen, but I don't think it will unfold like the housing bubble. I doubt banks will make half million dollar fog-a-mirror gold loans. The credit in the gold markets comes from speculative futures traders, and margin hikes have already been reduced it 3 times this year. I am not ignoring supply, but price action should already reflect supply/demand pressures.
In other words, if the dollar supply goes up $1 then that does not mean every asset can go up a dollar. It has to be spread around.
Very true. But gold is not just purchased with dollars. The euro supply, pound supply, yuan supply, and yen supply also went up a lot from fiscal stimulus and QE programs. Gold trades more like a currency than a commodity. My risk has also been reduced by a reduction in my position (on both sides of the recent spike).
"Gold trades more like a currency than a commodity."
ReplyDeleteInvestors were pretty sure gold traded like a currency in 1980. It traded like a bubble asset from 1980 to 2000 though.
To me, gold trades like an extremely leveraged hard asset/commodity (and therefore nothing at all like a currency).
The following was one of the first charts on my blog.
Fed Funds Rate vs. Commodities
"I bought gold and silver for the first time in my life back in 2004. I know exactly why I did it. It is still fresh in my mind. My gut said that earning 1% in three month treasury bills when inflation was clearly higher than that was not a good thing. I didn't know how long it would go on. The Fed pretty much forced me to own something of substance. Since I'm a saver by nature, I chose rocks. I already had a house. I can understand why renters might choose to buy. It surprised me that the rocks did so well. It surprised homeowners that real estate did so well. I just wanted a hedge against inflation but I got so much more than that. If you can believe in that chart above, you'll see that I actually bought a piece of heavily leveraged inflation. It seems a very small change in the Fed Funds rate can (not always) have a dramatic effect on commodity prices. Why? We live in an overleveraged society. Everything is overleveraged. Just look at how many hedge funds we have. That's my opinion and I'm sticking to it."
Investors were pretty sure gold traded like a currency in 1980. It traded like a bubble asset from 1980 to 2000 though.
ReplyDeleteBetween October, 1979 and January, 1980, the Fed raised short term interest rates from 12% to 20%. It killed inflation (and the economy) and rates have been dropping since then, currently 0-0.25%. If I could get 20% on T-bills, I would drop gold instantly, but I think the chances of that happening are 0-0.25%.
That doesn't mean gold will perform well in the future, but I feel very comfortable holding some percentage of my investments in gold in the current environment.
I hear where you are coming from but this isn't the 1970s.
ReplyDeleteSince 2000 inflation has averaged 2.5% per year. The Fed Funds Rate has averaged 2.5% per year.
As seen in the chart of my last link, a 0% real fed funds rate should generate about 50% annual returns for gold.
There's just one problem. It can't do it year after year until the end of time. Eventually the high price of gold (bubble) limits further increases.
Put another way, I do not believe even 2% real losses in short-term savings per year can permanently justify 20% increases in the price of gold per year.
That's a 10-1 leveraged ratio.
Eventually the prosperity creating gold bubble (for investors) dies even if the Fed does nothing (just like housing did even as the Fed made every attempt to prop it up).
I see toilet paper as one of the few things acting like a true currency in this environment. Its price is VERY stable compared to much riskier assets (gold going up, real estate going down, stocks going up and down, and up and down ;)).
Also consider that hindsight may show that even a 0.0% fed funds rate offers a reasonably high real yield if the price of oil continues to struggle. That said, I do have concerns that it will once again be temporary (and hence why I continue to favor TIPS for the long-term).
Hey, just opinions. What's new? :)