For the past few weeks, I have spent a lot of time thinking about the valuation of gold. I have been struggling to figure out a metric for calculating how much demand the price of gold is discounting. Before I get into the details of this fairly simplistic approach, let me first explain this method in relation to equities.
When valuing an equity, it is fairly easy to see the relative demand for the security based on how much someone is willing to pay today for the security, relative to known earnings. Of course I am talking about the P/E ratio. While I see strategies of targeting purchases based on P/E ratios as being of little value, the underlying idea remains true. That is that a low P/E stock relative to comparable companies is cheap, while a high P/E is expensive. There is one discrepancy to clear up and that is that by using historical earnings, new information may distort P/Es. For example, a company reports earnings of $1 per share and then a few weeks later lands a deal that will double earnings for the foreseeable future. The price and the P/E ratio would advance significantly even though the security may still be cheap relative to future earnings.
Applying the P/E Concept to Gold
Valuing gold is very difficult because there are no underlying cash flows. In fact, the only cash flows that you can identify are negative (storage and security costs). So I have been looking for the underlying determinants of gold price. The problem - there are so many factors that determine the price of gold. After considerable thought, I decided to look at the money supply, a popular gauge used by speculators as to determine the future price of gold.
I also had to decide on what version of the money supply to use. Based on the data available I had to use M1, M2, MZM, or True Money Supply (TMS). I chose to use TMS, also known as Austrian Money Supply, after giving the matter a lot of thought (see Rothbard, Schotek). I also ran a study using M3 as I expected the bugs to give a lot of criticism had I ignored it. What I found was an R2 of 0.146 for TMS and 0.173 for M3. This shows M3 to be a slightly better predictor of gold price than TMS, but still not very significant. Either there is another, more relevant factor, or there are many factors and I would expect money supply to be one of the larger inputs. One could argue that the value of the dollar is a huge factor but for the purposes of this study it is irrelevant. The factors that affect dollar price would then have to be broken down to find the main determinant - which I would expect to be money supply in the long term. Immediately I observed the complexities of determining gold's price and how money supply is only part of the story.
click to enlarge images
Taking the Study Further
After deciding on TMS, I moved on. I wanted to see the premium that gold had historically traded at relative to TMS. I started by creating a TMS Index and a gold price Index - I did this by fixing the January 1971 value of both series to 100 and then building a new series using those derived values to the present day. I then subtracted the TMS index from the gold price index and plotted the new series. In theory, this should represent how gold has historically been priced relative the true money supply. What I found was very interesting.
I learned several things by doing this study. Firstly, the peak hit in 1979 was a bubble and using that inflation adjusted high as a target for gold price today is simply incorrect. Secondly, moves in gold price appear to come before moves in the money supply. In other words, gold is the leading indicator for money supply - not the other way around.
The implications for these observations are as follows. Firstly, the premium of gold price over TMS today shows that gold is discounting a future increase in the money supply - a pretty drastic increase. This leads to the second implication which is that the money supply can expand now without gold necessarily increasing in value. This is precisely what happened in the 1980's. True Money Supply grew 108.23% over the decade while gold price lost 39.64% of its value.
It's also interesting to note just how good a buy gold was in 2002-2003. Even though the money supply had been expanding for some time, the index spread was back to par. That shows just how much opportunity cost the market was pricing in terms of holding gold - Everyone wanted into equities and there was little demand for the precious metal.
Does This Mean That Gold Is Going Down?
Not exactly. All this tells us is that based on price, gold is already pricing in a sharp increase in the money supply. Should the money supply increase more than gold is already pricing in, gold can rise. It could also rise based purely on irrational exuberance as investors run to gold and the spread could increase further - but that is a bubble waiting to burst unless money supply grows rapidly.
If the index spread was zero today (if gold price tracked TMS perfectly), we would see a gold price of $385 per ounce. This is based on a 1018.0% increase in the TMS since 1971.
My advice is to find alternative ways to play the rising money supply. Identify assets/companies that will be affected and try to deduce if its current price is discounting a rapid increase in the money supply. Although both risky plays, treasuries and real estate are both on the opposite end of the spectrum. That being said, both have external risks outside of the money supply (i.e. quantitative easing and over supply in a slowing economy respectively).
Then there's the Nouriel Roubini opinion - that the Fed will be able to mop up a lot of the excess liquidity. After all, most of it is just sitting in excess reserves in the banking system. If that is the case then gold can move significantly lower as expectations of money supply increases diminish. What about the fiscal deficit? Ask Japan who is sitting with a debt-to-gdp ratio of 175% and isn't being forced to monetize their debt.
The 12 Trillion Pound Gorilla
The reason that things may be different in this scenario is that people may really start questioning the dollar as a reserve currency. I don't personally see the dollar losing its reserve status but I respect the argument. What I see as more likely is the dollar being one of several reserve currencies (likely with the Euro, and either the Yen, Yuan or an Asian currency basket). Such a move would really hit the dollar and send gold up in dollar terms. However that would alleviate a lot of the market's worries and gold globally could actually come down off such a development. That being said, we could also be moving to a new era where historical ratios are meaningless and the amount of inflation priced into gold is dwarfed by what actually comes. For this reason I am currently neutral on gold out of fear that the world as I know it is changing.
Disclosure: no positions