Home > Commodities, Gold, Other Research Posts, Theory > Gold, and Dilemmas

Gold, and Dilemmas

At the start of another Employment week, the same refrain echoes: higher equity markets, soft commodities markets (because changes in China’s policies will hurt the demand for commodities…but I suppose that it will not hurt the profitability of U.S. shares?), and continued negative news from Europe that is mostly ignored during Employment week.

Actually, maybe the news from Italy is being mostly ignored here because it is hard for Americans to truly fathom what is going on. Remember that the basic issue is that a majority of Italians voted for one or another party that favored ending austerity measures and/or leaving the Euro, but left no single party controlling both houses of parliament. Until this morning, it appeared that no single party would be able to form a government, which meant that a new election would likely be called soon. But now it appears that the Five Star Movement (Beppe Grillo’s party) is offering to stage a walk-out from the senate. Now, that sounds negative, right? Well, actually it’s progress (and Grillo’s party would have to be given some policy concessions in exchange for walking out, which sounds like “lovely parting gifts” to me) since Five Star doesn’t have enough delegates to prevent a quorum from being established if they leave (with no quorum, the body cannot conduct business) but their absence would allow a majority to be established on a lower number.

In the U.S., the approach would be different: the Senators would reach a deal and then vote on the deal, with no one having to manipulate the process in an arcane Robert’s-Rules-of-Order fashion. On the other hand, they had a senate in Rome about 2,500 years before we had one, so who are we to question their parliamentary process?! And our institutions are no less clownish at times…such as right now, since despite so many dire threats the world apparently did not end over the weekend once the budgetary sequester went into effect.


Since the markets were quiet today (and likely will remain relatively quiet until the Employment report on Friday, if recent patterns hold true), I thought I’d take up a topic I’ve been meaning to discuss for a while: a look at the relative value of gold and a link to an interesting new paper on gold.

First, let me say that our systematic metals and mining strategy is currently approximately neutral-weight on gold itself, overweight on industrial metals, and deeply underweight on mining stocks. But that strategy relies on metrics I am not discussing here; nothing, moreover, that I discuss here should be taken as an indication of whether Enduring Investments would suggest an investor should add or subtract to his or her particular exposure.

Disclaimer completed, let’s look at the yellow metal relative to other assets, as I first did in this space back in August of 2010 when I concluded that gold did not look particularly overvalued. Gold subsequently rallied another 60%, then slid (in case you haven’t heard!). It is currently still 30% above where it was in August of 2010. So is it overvalued?

Some observers have noted that the ‘real price of gold’ (that is, gold deflated by the current price level) has recently risen to levels not seen since the peak of the gold market in the early 1980s (see chart, source Bloomberg, which shows gold in constant December 2012 dollars).


This is true, of course, but measuring the ‘real’ price of gold is a funny concept. The gold price relative to the cost of the consumption basket is a metric that has meaning, because it tells you how much consumption you displace to buy an ounce of gold, but unless you’re evaluating the consumption of gold I am not sure that’s a relevant metric.

On the other hand, it makes more sense to me to look at investments relative to gold, since that’s what is likely to be displaced by a purchase of gold. Some of these relationships are not particularly useful analytically, though, or at least appear at first blush not to be. For example, looking at gold versus the stock market (see chart, source Bloomberg) you can’t tell very much except that gold was rich or stocks were cheap (or both) in 1980 and gold was cheap or stocks were rich (or both) in 2000. Or, so I wrote in 2010.


However, I subsequently noticed another chart that looked somewhat similar. Below (source: Enduring Investments) I have put the data from the chart above alongside a measure of the volatility of inflation expectations, as taken from the Michigan Sentiment Survey. (As I’ve written previously, surveys of sentiment are not satisfying ways to measure true inflation expectations, but they’re all we’ve got and they might nevertheless be valuable in measuring the volatility of inflation expectations, which is what we’re trying to do here).


The notion is this: when inflation expectations are becoming both lower and more stable, then stocks become more valuable and gold less so as an investment item. But, when inflation expectations are rising and/or becoming less-stable, then stocks become less valuable and gold more so as an investment item. I haven’t worked very carefully to refine this relationship, but the Michigan series begins in 1978 so that’s the main limitation. Yet, without any lags nor tweaking of period lengths, the R-squared here (on levels, not changes) is 0.745, which is firmly in the “interesting” category.

Having said that, unless we’re able to forecast the volatility of inflation this isn’t particularly helpful in assessing whether gold is rich or cheap relative to stocks (although on the regression, not shown, the ratio of gold/S&P is 1.04 but ought to be more like 1.07, so gold looks slightly cheap to stocks). The main thing we can do with this is explain why gold prices have risen relative to stock prices over the last decade, and it makes sense. In this context, the recent slide in gold/rally in stocks can be attributed to a soothing, perhaps temporary, in consumers’ concerns about inflation.

The champion relationship, although less creative, is the ratio of gold to crude. Over a long period of time, an ounce of gold has bought between 15 and 20 barrels of crude oil (West Texas Intermediate), with occasional spikes wider and at least one lengthy period between 7 and 12. The chart below (source: Bloomberg) shows this classic relationship. It makes some sense that two hard commodities, both exchange traded and having no natural real return to them, ought to broadly parallel each other over time. Again, this isn’t a very good trading relationship but it is a decent sanity check.


By this measure, gold is approximately at fair value, although an argument could be made that WTI is no longer the fair price for crude. In terms of Brent Crude, Gold is only 14.3 barrels and so arguably slightly cheap.

None of this will delight the gold bulls, but it also won’t delight the gold bears. Gold, at least the way I look at it, seems to me to be somewhere between slightly cheap to roughly fair value versus a pair of comparables. Of course, it may be that stocks and crude oil are slightly expensive, on the other hand!

Gold bulls and bears also will both find things to like and things to dislike in a paper by Erb and Harvey called “The Golden Dilemma.”  Given that gold bulls tend to be more, er, passionate about the subject, they will likely be more strident in their disagreements but it is a capable attempt to tackle many of the well-known arguments for owning gold and put them to logical and empirical test.

These gentlemen (who have some serious chops in commodities research) conclude that as an inflation hedge, gold is (1) not an effective short-term hedge, (2) not an effective long-term hedge, (3) might be effective over the very, very, very long-term, and (4) probably effective in a hyperinflationary situation. Although this depends somewhat on your meaning of “hedge,” I concur that gold is not a hedge. It can, with some work, be made into a smarter hedge, which works better (especially in conjunction with other metals, and mining stocks). But they make a fairly powerful argument that if there’s even a teensy chance that hyperinflation happens, a high gold price can be rational since the tail of an option contributes quite a bit to its value.

Incidentally, a slide-show version of the paper is here and is pretty good even if you didn’t read the paper.

  1. Eric
    March 4, 2013 at 9:33 pm

    Can you say a little bit away why you are (“deeply”) underweight mining stocks? They seem ridiculously cheap relative the price of gold.

    • March 5, 2013 at 7:08 am

      Our model isn’t a relative-value model. It merely considers whether in general, the conditions favor stocks or direct metals investment. Another tilt determines whether industrial metals or precious metals are preferred.

      I have been reading for about five years how undervalued mining stocks are relative to metals. I had some of my personal account in GDXJ for a couple of years, waiting for “a pop.” Since these companies often have programs to hedge their output, it isn’t clear to me that they are as deeply undervalued as the popular wisdom seems to think they are. Because NO ONE says they are overvalued, why haven’t the stocks responded to pretty much everyone in the world saying they are undervalued in a rising market? It might merely be another case of market inefficiency, but I just know that I DON’T know how to properly value these stocks taking into account both the metal and the underlying market for equities.

      • Eric
        March 5, 2013 at 4:34 pm

        I’m kind of liking that everyone is pessimistic about the miners right now. Of course, sometimes everyone is just correct. But NEM pays a dividend of nearly 5% that is not likely to go down unless the price of gold does. So, I think I would rather own that than something that just sits around looking shiny. I don’t have the faintest idea how to value the juniors, so I stay away from GDXJ. Also, there was a recent piece on SA somewhere that evaluated the production-value-weighted country/political risk of various majors, and NEM did very well.

        NEM also produces quite a bit of copper, I believe, which should be a better inflation hedge than a pure gold play.

      • March 5, 2013 at 4:44 pm

        I don’t disagree with any of that. On a security-by-security basis, and compared to OTHER equities, I see value there. It depends on what you’re comparing to, as always. But again, this is a systematic model, which doesn’t look at micro data like that.

  2. Jim H.
    March 5, 2013 at 2:31 pm

    Valuing gold is a problem that has perplexed me for years, and still does.

    Your chart of the Gold/SPX ratio vs. infl. exp. vol. is a good shot at it. But still … at 1/1/1988 (eyeballing the chart) the ratio was twice the inflation measure. Ten years later, gold/SPX was less than half the inflation volatility — a difference wider than the side of a tobacco barn.

    For now, I’ve given up on valuing gold. If gold were used as money, as nature intended, we wouldn’t be concerned with valuing it versus a monetaire consisting of cockamamie Bernanke Bux. Rather, I just use a plain old, brute-simple trend-following method:

    “Each month, gold’s percentage change over the past ten months is compared to the yield on three-month Treasury bills. If gold’s ten-month change exceeds the T-bill yield by ten or more percentage points (e.g., with T-bills yielding 2%, gold increased 12% or more in the past ten months), then gold is bought. If gold’s ten-month change is ten or more percentage points below the T-bill yield, then gold is sold.

    “When gold’s change is within a ten percentage point range of the T-bill yield, then gold’s weighting remains the same as last month.”

    This simplistic little model booted gold last June, after holding it for all but 8 months of the past decade. And it’s nowhere near buying it back.

    When the time comes, I’ll buy back my precious-s-s. But for now, gold is dead to me. Other fish to fry!

  1. January 13, 2015 at 6:58 pm

Leave a Reply

%d bloggers like this: