The continuing mystery of gold

Actually, as previous posts seem to indicate, there is no great mystery to gold: the executive summary is that it is not so much that gold is going up as that the dollar is going down. What is, however, mysterious, is that the gold miners are not doing very well. Indeed, this year, gold prices (as measured by GLD) have gone up some 8%, which is (as of today, June 8 2011) considerably better than any of the stock indices. Gold miners, however, are performing truly abysmally: the chart below shows the performance of gold vs S&P 500, vs NEM --  Newmont Mining (NEM), which is down almost twenty percent year to date. A picture is worth a thousand words:

 Newmont's friends and competitors which constitute the GDX ETF are doing abut the same. Now, if gold prices were very high compared to the cost of extraction, then gold miners should be doing very well. If, s previously suggested, we have no gold bubble, then the cost of extracting gold out of the ground should be growing roughly as the price of the metal itself, and if the margins were roughly constant in percentage terms, the prices of mining stocks should be growing roughly as fast  as the price of gold. It has been suggested that the wear and tear on the mines is adversely affecting the price of gold miners. Let's test this theory, and look at the Price/Earnings ratio of long-suffering NEM versus that of the S&P.

The sharp-eyed reader will see that until the end of 2009, gold miners were the toast of the town, but since the beginning of 2010, they suddenly became the black sheep of the investment community, and the current P/E of Newmont (around 11) is only slightly higher than that of the S&P 500 at the nadir of the stock market collapse (3/9/2009, when it was a little over 10). Indeed, while NEM's share price (adjusted for dividends) has risen by 25% over the last five years, its P/E ratio has gone from around 70(!) to the current 11.

The only at all plausible explanation  for this strangeness I have seen is that some hedge funds use gold mining stocks to hedge their gold positions. Well, actually, this does not really make sense to me -- the obvious trade at this point is go long miners (perhaps hedging with GLD), but nothing else does either, since it seems that GDX and friends appear undervalued by some 30% (and there have been massive selling in the last couple of days).

NOTE: Samsara is long gold mining stocks, and has been for quite a while (somewhat to its chagrin).

Trading strategies

I had fallen victim of Amazon's one-click Kindle purchasing one time too many, and, without thinking, bought this book, which, to my dismay, is another piece of chartist claptrap backing up its techniques with claims of years of impressive results -- I discovered this on a plane flight, and quickly diverted myself with a nap, but this experience turned my thoughts to that best of gambling (which is really the same as trading) systems: the martingale. To refresh the reader's memory, this system (already popular in 18th century France) allows one to win against any casino game, no matter how big the house edge. As an example, we will use Roulette, in its primitive red-black form: the gambler comes into the casino, and bets a dollar on red. If he wins, he walks away. If he loses, he bets two dollars. If he now wins, he has won two dollars after losing one, so walks away with a dollar net gain. If he loses, he now bets four dollars. It is easy to see that this sequence of double-downs allows our gambler to walk away with a dollar, with probability one. Somewhat unfortunately, this system works particularly well if you have unlimited bankroll, in which case winning a dollar might be viewed as rather small pickings.

The first (and most important) improvement to this strategy is to start a hedge fund. Let's assume, for the sake of argument, that the starting capital is $1000. Our intrepid money manager (based in Atlantic City, for convenience -- lets call him A) walks into Taj Mahal every trading day, and practices the martingale. Since his Sharpe ratio is infinite (every day he makes his dollar, like clockwork), his investors are happy. In the 250 trading days of the year, he makes \$250, which is a 25\% return, of which he collects a 20\% performance allocation of \$50. In addition, he makes \$20 in management fees, so in a little under four years our hero is around \$250 richer. A good thing, too, since this is roughly when A loses \$1000, to the chagrin of his investors.

The next, and also important step, in this process is leverage: Money manager B, having observed A, is horrified that the investors had lost money, while A has made out rather well. So, B borrows money on margin, so he now has \$2000 to play with, so he scales up his bets two times higher than A. At the end of four years, not only does he have \$500, but when the day of reckoning comes, a large chunk of the damage he suffers is allocated to the lender, so both B and B's investors are happy. Not so the bank's shareholders.

As a final(?) improvement, our manager C has not \$2000, but \$20000000000. All of the activity in the previous paragraph takes place, but now the bank's shareholders (and especially its bondholders) proclaim the bank too big to fail, and the bill goes to the taxpayers, of whom there are quite a few, and it takes only a few dollars of the savings of each of them to subsidize C, his investors, the bankers, and the vitally important luxury yacht industry, it seems that everyone is now happy.

Progress is a wonderful thing!

What is interesting is that the parable described above not only describes much of recent "investment" activity, but also encapsulates pretty much the entire content of The Black Swan, at greatly reduced investment of time, money, and pseudo-intellectual rambling.

More on inflation

This is a continuation of the last post. After looking at it, it seemed to me that the interpretation of the chart was a little oversimplified, and that was because the numbers had changed so much over the 18 years of the study that log scale would be more enlightening. No easier said than done (OK, a little easier said). Here is the same graph in log scale:

In this form, it seems clear that (other than some cyclical behavior) there was essentially no commodity price inflation from the beginning of the study (January 1993) until the end of 2001 --  beginning of 2002 (also known as the Dot Com crash). In those nine years, a conservative saver seems to have actually increased the purchasing power of his bank account quite considerably (by around 50%, in real terms). During that golden age (which we can call the Age of Clinton), there was a combination of the peace dividend, considerable technological innovation, a gridlocked government which actually managed to run a surplus, and fairly conservative fiscal policy. The chart below shows prevailing money market rates, which are closely tied to Fed Funds:

You will note that the "Greenspan put" coincides almost exactly with the end of the golden age (marked somewhat more memorably by the terrorist attacks of 9/11/2001). The massive money printing, did not wait to make itself be felt: tgold prices rise at a 1.5% a month clip -- the log scale graphs are amazingly linear (the F statistic, which shows the strength of the trend, is around 5000), commodity basket prices rise somewhat slower (1.1% a month clip), with more variability. During this period (the last ten years, approximately), our hapless saver's money market account gains around 20% in nominal dollars, but loses around 70% of its commodities purchasing power. Being no fool, our saver decided to invest her hard-earned dollars in real estate, but imagine her dismay when that investment did about as well (or poorly) as her neighbor's money market account (the graph below shows the change in the Case-Shiller index versus the Money Market account. You will see that nominally, Case-Shiller outperformed the money market account by around 15%, but all of that and more was eaten up by transaction costs, property taxes, and so on. I am only viewing housing as an investment, so am not counting the savings of rent or the mortgage tax deduction.

Since the move into housing was borne of desperation with the performance of the simpler ways of saving, it is not at all surprising that the returns are similar, though as we all know, the disruption caused by the fact that real estate investment is very far from risk free (as everyone now knows) has been rather considerable.

It is always harder to determine causality than correlation, so the graphs by no means prove that that monetary policy caused the end of the Golden Age of Clinton and the implosion of our savings, but they do provide some food for thought, I hope.

What is inflation really, or what's the use of gold?

Those who have read this previous post probably (and understandably) wonder how reasonable the Alternative CPI measure is. After all, it seems to indicate that our cost of living has increased almost four-fold since 1993, which seems rather steep (mostly since for only very few of us has our income increased four-fold in the same period). There is no question that the official numbers are seriously gamed (see this note, or this), but that, in and of itself mean that the alternative numbers are right. My personal view is that the Alternative CPI measure is more a measure of inflation (that is, the debasement of the dollar) than the actual CPI growth -- the latter tends to be smaller than the former, since the money-printing is offset by technological progress, which causes computers to drop from $6000 in 1981 to $150 at Walmart in 2011 (the latter computer also being several orders more powerful), and your car's fuel consumption to go from 8mpg in the 1960s to 40mpg today.

A good proxy for inflation is, however, provided by prices of commodities (still not perfect, since exploration and mining also have made considerable strides), and since a representative basket of commodities is rather cumbersome to hold, a good proxy for one is gold -- indeed, gold is almost miraculously convenient -- it is compact, it does not degrade, and it saves you from buying shares of oil tankers parked off Singapore.

The above seems counterintuitive (after all, we have all heard of the gold bubble, but [relatively]  few of us have heard of the the wool, rice, natural gas, or any one of the many other possible commodities bubbles. Well, luckily for us, the IMF maintains a commodities price index, so we can compare and contrast. Here is the requisite chart:

Some explanation might be in orer: The very smooth red line shows how many dollars you would have were you to invest $1 in a money market account in January of 1993. The almost-as-smooth green line shows how many of those dollars you would need according to the Alternative CPI computation to purchase a basket of goods worth $1 in Jan 1993. The jagged purple line shows how many of those dollars you would need to purchase however much gold you could buy for a dollar in January of 1993, while the really jagged light blue line shows the same for a dollars worth of a commodity basket. The conclusions, at least to me, are:

• The Alternative CPI seems, as advertised, to be a good measure of inflation (and is, therefore, a bit of an overestimate of the actual price inflation).
• There is no gold bubble (a conclusion also drawn in a previous post from other data).
• Gold (in addition to its compactness) is a better gauge of monetary inflation than the commodities basket (witness the huge volatility in the latter starting in late 2007  or, for that matter, just in this May.
• The red line (your risk free return) was added just to make the chart more depressing. Unfortunately, it has achieved its goal brilliantly.