Finding A True Value For Gold And Silver: Part I

by: Retired Aviator

One of the greatest difficulties with deciding whether to invest in gold or silver is to come up with a fair valuation for the precious metals. As investors, we invariably have to make a buy decision and a sell decision. The hard question is, at what price do we buy or sell? With income-producing investments such as rental real estate or businesses (stocks), our buy and sell prices can be calculated based on projections of the income produced (plus maybe something thrown in for the net assets). That is standard security analysis; the "discounted cash flow" method, for example.

With commodities such as precious metals, however, we know there is no income produced or business growth, so it becomes very difficult and highly subjective to judge just how much gold and silver are worth. There is little that is both objective and substantive to go on in coming up with a number, at least based in the traditional sense of investment analysis. So investors' estimates of the 'true' value of gold and silver range widely-- all over the map even-- based on other things.


These other things include benchmarks. Frequently the ratio of silver to gold in the earth's crust (roughly 15 to 1) is used as a benchmark to justify the idea that the prices of silver and gold will tend toward an equilibrium of about the same 15 to 1 ratio. I'm highly skeptical. Their values were artificially pegged at approximately that ratio for hundreds of years by governments' decree (in the relative values assigned to gold and silver coinage). However, since gold and silver were de-monetized and were floated freely on commodities exchanges, they have spent much more time wandering far from the 15 to 1 ratio than they have been anywhere near it.

In fact, since the 1970's when the metals first floated freely to find their values in the marketplace, they have never traded anywhere near the 15:1 ratio, with the lone exception of a very brief period in 1980 when the Hunt brothers famously tried to corner the silver market and briefly caused silver to spike way up to $50 an ounce. So hoping for a particular ratio that has virtually never been a reality in forty odd years of unfettered free markets strikes me as a pretty long shot and not a very good bet. (I personally know investors who are buying silver but not gold specifically because they expect the ratio to move toward the historical 15:1. I have wished them luck; they will need it.) Currently, with gold at $1650 an ounce and silver at $31.50, the ratio is 52 to 1.

A second benchmark that is often cited as a 'proper' valuation for gold is that one ounce of gold should equal the price of "a mens suit of clothes." This however poses some problems. First, since floating freely beginning in the early 1970's, the price of gold underwent a tremendous boom. Gold rose over 20 fold in less than 10 years to peak in 1980, followed by an excruciating, 22 year bust that wiped out a staggering 75% of its peak value (significantly more than 75% when adjusted for 22 years of inflation). This was followed by the 2002-2012 current boom, when the price of gold has again multiplied (six fold to date for the decade). The price of mens clothing has risen, well, more gradually, to put it mildly. Clothing certainly did not predict that kind of volatility! Second, "a mens suit of clothes" or even "a mens fine suit of clothes" as is sometimes cited, is vague enough to cover a very wide range. Basically, an ounce of gold during its forty year free floating period has at certain times bought you the finest of custom tailored suits, but at other times only bought you the cheapest off-the-rack suit from a discount men's clothing chain. The idea that an ounce of gold has thus 'bought you a suit of clothes' throughout this period might be somewhat technically accurate, but only in the sense of accurate enough to hit the broad side of a barn. Hardly useful in coming up with a buy or sell target price.

A Better Benchmark for Gold Needed

Because of the aforementioned problems, I herewith propose a new benchmark for estimating a fair value range for gold, rather than mens clothing or 15 times the value of silver. Silver itself is a highly volatile moving target.

Nearly any gold investor would agree that gold is an inflation hedge over the very long-term (decades or centuries) that should rise in price (at least approximately) by a factor somewhat similar to that which goods and services in general rise in price. I propose a new, low volatility gold benchmark; one that is much more precise than mens suits of wide-ranging quality. This benchmark rises steadily along with general inflation but smoothly over time rather than with jagged ups and downs.

Such a benchmark, I argue, should be non-volatile. For example, using one ounce of gold as equal to some number of barrels of oil would be a terrible idea as the price of oil can easily move 75% in one year and then retrace the entire move in the next. It would be absurd to think that the dollar value of gold should peg to oil during such drastic, temporary moves. In fact, using any one commodity price should thus be ruled out because supply or demand shocks in one particular commodity can move its price dramatically and rapidly even while prices in general remain unchanged. We don't want a target price for gold set at x bushels of wheat when the price of wheat might move only because of widespread drought or blight or a demand shock from emerging markets.

A Composite Good is Best

That leaves for consideration as our benchmark, 'composite' type goods which, unlike commodities, are consumer goods that are made up of many diverse cost sources. That diversity of cost types ensures that no single commodity price weighs too heavily in the price of the benchmark good but that many commodity prices, as well as labor costs, do get factored in over time.

Also, our benchmark good should be as ubiquitous in demand as possible, and fairly elastic in supply over time. We don't want to peg gold to any particular good that can too easily experience either a pinched supply or a glut affecting its price. The benchmark should also be priced in very price-competitive markets (with transparent pricing). No sense benchmarking to attorneys' or doctors' fees when many people don't even know what fees their attorney or doctor is charging until they receive the bill. Many cannot or do not comparison shop these services. Such opaquely priced markets are inefficient and their prices can thus be capricious and wide ranging.

Proposed: A New Gold Price Benchmark

Then what would be a good benchmark for pricing gold? I suggest housing. Specifically, median-cost rental housing in a U.S. city of substantial population, with per capita net worth and income in the middle quintile range for the nation. In other words, I would use a proxy estimate of the 'fair value' of gold as indexed to the rising cost of rental housing in a typical, Midwestern capitol type city that is neither rich nor poor, neither booming nor busting.

With rental housing you have a built in cross section of cost factors: Fixtures and building materials, labor for maintenance, repairs, remodeling and property management, property taxes (includes a variety of government services), insurance, capital costs (mortgage), and utilities; all being factored into rents over time. Yet the landlord who overprices will learn the lesson quickly that uncompetitive pricing leads to high turnover and high vacancies. Housing is a competitive and price-transparent market with diverse cost inputs. It is also a major expense for most people and thus a very important component of inflation as it directly affects consumers. Further, since those who own their own home can freeze much of their housing expense for as long as 30 years in the form of a fixed mortgage payment, it makes sense to index to rental housing, leases of which typically can be re-priced annually.

But how much rental housing should equal an ounce of gold? Let's look at history to answer that question. If gold were truly a 'constant value' form of money over time, an ounce of it should generally buy more or less the same amount of housing in any given time period. But first let's smooth out the historical price of gold into a broad, long-term trend line so we can compare a smoothed out average gold price trend line to historical housing prices.

The 20th Century Gold Price History Trend Line

Gold was pegged very near $20 per ounce from the time of the Coinage Act of 1792 until 142 years later in 1934. Then President Roosevelt signed the Gold Reserve Act, which revalued gold overnight from $20.66 to $35 an ounce. This was essentially an inflation adjustment to help alleviate federal debt obligations payable in gold. Though this $35 peg remained constant from 1934 until 1971, I will extrapolate a gradually rising gold price during that 37 year period, since there was indeed inflation in general and in housing, our benchmark.

To extrapolate the trend, I will need to assume two things: That the gold re-peg to $35 in 1934 was commensurate with a reasonable inflation adjustment at the time, and further that the gold market's 15 year period from 1982-1997--which settled on a relatively stable value for gold averaging around $375 an ounce--was also a 'fair value' for gold for that particular period. (It's hard to argue that free markets could be too far out of whack from 'fair value' for fully 15 years.) I'll use 1990 as the midpoint of this range-bound 15 year period and thus extrapolate a 'smoother' long-term gold price inflation as follows. The idea is to use an earlier date (1934) when the price of gold was 'reset' after a long period of artificial suppression (the $20 peg), and to use a much later 'free-floating' price when gold exhibited long-term relative price stability in free markets. That is to say, when gold was neither 'hot' nor 'cold' during the long 15 year stretch of relatively stable gold price. This methodology thus yields a ball park estimate of gold's 'fair' value of $35 in 1934 and $375 in 1990. Though no estimates can ever claim title to being the precise fair values, the fact that these figures are separated by a long 56 year period helps them to be reasonably accurate, much as a long barreled rifle is much more accurate than a short barreled handgun.

Gold's Rise Smoothed Out

It turns out that the annual inflation rate that turns $35 in 1934 into $375 in 1990, is 4.326%. The extrapolated hypothetical fair values for gold then at the turn of each decade, simply using 4.326% as an annual inflation rate, are as follows:

Gold Hypothetical Trend Line Assuming 4.326% Annual Rise

























These are essentially 'fair value' estimates of the value of an ounce of gold, assuming it should have constant purchasing power. Also assumed are: (1) the 1934 gold re-peg was set at a reasonable price, (2) the 1982-1997 period of relative stability of the price of gold was at a fair price, on average, and (3) consumer price inflation in general after 1990 has been at a rate quite similar to the average for the 1934-1990 period. I could use government inflation statistics such as CPI to test the third assumption, but I'm quite sure that many gold investors would howl about those supposedly 'rigged' statistics. (I have learned that many, perhaps most, gold investors cannot be swayed by any argument that calls upon CPI--gold affinity and deep distrust of government often go hand in hand.)

Actually, hard money crusader Congressman Ron Paul has implicitly accepted an average inflation rate of just 3.2% annually for the whole 20th century in his writing, so 4.326% looks quite generous if anything. I believe that 3.2% is much closer to the truth for a 20th century overall average consumer price inflation rate. It's the rate that turns $1.00 in 1913 into $.05 today. If so, then an ounce of gold, rising 4.326% annually, had in fact appreciated substantially by 1990-- in terms of buying more goods-- compared to what an ounce bought in 1934. However, it may simply be that FDR's administration did not re-peg gold high enough, so that the 4.326% imputed rate is thus too high because of coming off of a slightly depressed $35 base. But even at the more generous 4.326% rate trajectory, gold still would only have a constant-purchasing-power fair value of about $900 today in 2012 and $1081 by 2015, as shown in the table.

This suggests that the current $1650 an ounce price is way overinflated. My gut has told me as much for several years now and this analysis showing that gold's real purchasing power has in fact risen a great deal confirms what many of us have felt intuitively. So much for the oft-quoted idea that "the value of gold doesn't change, it's the currency that's moving."

To double check we still should perform an analysis of the rate of consumer price inflation post-1990, using actual residential rent statistics instead of CPI, to determine whether the 1934-1990 average inflation rate should be adjusted post-1990. I will use this analysis to recalculate the above table of fair value estimates for gold based on the actual inflation experienced by the rental housing market post 1990. This will come later in Part II of this article. Also in Part II, I will use historical data to determine just how much rental housing an ounce of gold should be expected to purchase. With that benchmark determined, we will gain another and I believe a better yardstick for gauging gold's valuation now and in the future.


In conclusion, the two most commonly cited benchmarks for gold and silver valuation have serious flaws and miserable track records. They should be abandoned and replaced with something better. Using the 'constant purchasing power' philosophy of valuing gold, the yellow metal appears way overpriced today.

Silver also appears overpriced as its current decade long boom has coincided exactly with gold's, albeit with greater volatility. Silver's ten year chart looks like gold's hyperactive junior sibling that has come along for the ride. It's wise to remember that circa 1980, many people who feared fiat money inflation felt that as 'real money' silver was still cheap between $20 and $50 an ounce. They were buying heartily, only to discover that it could get a whole lot cheaper. It traded well under $5 a decade later. Most anything can happen to precious metals prices as they are heavily influenced by powerful emotions.

It is widely accepted among precious metals experts that the flood of investment demand (from ETFs, Central Banks, & individuals) is responsible for the current boom in the metals' prices, not industrial or jewelry 'fabrication' demand. I would recommend zero allocation to gold or silver until their bubbles have burst several years or more from now, and even then to tread lightly. There simply are much safer alternative inflation hedges that are not bubble priced. I would avoid gold and silver ETFs and silver or gold mining stocks until the metals' prices have come back to earth: GLD, IAU, GDX, GDXJ, SGOL, PHYS, ABX, NEM, AEM, AU, GG, GOLD, DGL, EGO, DGP, UBG, UGL, RGLD, GBG, GFI, GRZ, GSS, IAG, HMY, CDE, HL, RBY, RIC, NGD, KGC, AUY, NG, SLV, PSLV.

Disclosure: I am short GLD.

Additional disclosure: I currently have a small short position in GLD but will short more if gold rises much over $2000/oz. I have almost completely unwound my long position in SLV (over a period of years). I have no positions in any miners nor plans to initiate any.