To 321gold home page

Home   Links   Editorials

Where's the Silver?
A Review of the Silver Institute's
World Silver Survey 2005

By Franklin Sanders
Editor, The Moneychanger
Jul 07, 2005

Every industry boasts its yearly surveys of global supply and demand, and one always stands out as the standard. In silver's world that distinction belongs to the Silver Institute's World Silver Survey.

Having read the 2005 edition, I have to wonder why.

When will the Silver Institute wake up and figure out that Gold Fields Mineral Services is short-changing it? Poor writing, analysis unsupported by data, conclusions that leave the reader clueless, and poor writing all drain away the Survey's usefulness. It gets harder and harder to swallow GFMS analysis and conclusions, or even respect them. This year's Survey contains puerile errors and in some points is almost incomprehensible, with half explanations and hints at explanations which are in themselves incomprehensible.


Did I say "poor writing" twice? I meant to mention it three times. Who did they hand this off to? Some apprentice analyst? A speaker of English as a second language? Somebody please tell them that English has more verbs than "to be," and that 'is" plus a verbal adjective and "of" does not satisfactorily substitute for a real verb. "Is supportive of" will never replace "supports" in the hearts of any true English speakers except hopeless bureaucrats.

Here's a howler from page 18, col. 2: "Demand for US$1,000 face value 90% and 40% coin bags is more clearly investment related as these typically contain 710-800 oz. silver." Can it be that GFMS, world leader in gold and silver market intelligence, does not know that US 90% silver coins were minted at exactly 723.4 fine ounces to the dollar face value? Could they possibly fail to know that the industry counts this most popular form of physical silver investment as 710 - 715 fine ounce? Or that 40% silver half bags contain only 295 fine ounces? This statement leaves the informed reader gasping like a bass out of water, first for its ignorance, and second for its puerile truism. Of course demand for US silver coin is "investment related." Why else would anyone buy a 55 pound bag of common silver coins? To drill holes in them and make washers?

Here's another sample of the Survey's somewhat less than pellucid style: "The addition left the total outstanding delta-adjusted book at end-2004 at a provisional 53.5 Moz (1,665 t), representing roughly 8% of global mine production. A relevant comparison here is of one against the gold book, which at end-2004 stood at 57.2 Moz (1,779t) or 72% of annual gold output." (p. 28) Now, those look like English sentences, but they aren't. I only think I know what that second one means, but I'm not wholly sure.

Here's another one from p. 31: "As such, this does not include the large amount of bullion held by many private individuals or in non-recognised depositories, which by its nature is impossible to quantify on a fully scientific basis." I am lost. What foregoing word does "its" refer to? What makes it "impossible to quantify"? What is a "fully scientific basis"? Does this mean the whole Survey was written on a "partially scientific basis"? Where am I? What's my name?

I like this one, too, on p. 32: "In the context of increased investment activity throughout the year -- discussed extensively in Chapter 3 -- such a possibility would not seem illogical." Go get 'em, Tiger. There's a bold, clear cut statement if I've ever seen one. No doubt where the writer stands - I think.

The GFMS style rule seems to be "never use one plain word where two imprecise ones can be substituted." So on page 34 I find "[I]ndeed the country operated a silver monetary standard through to 1935." Would this mean "until"?


Apparently the GFMS style book rations the number of sentences which may be used per Survey, so the poor writers have to pack two or three sentences into one to save periods. On page 34 I read, "As we have pointed out in previous World Silver Surveys, silver mining in China only really expanded in the late 1970s, when it was realised by the governing authority that, in order to support the nation's push to industrialise and become self sufficient, it would need to actively promote industry sectors such as silver mining." I'll bet that "it was realised by the governing authority" means "the government realised," but I'm not certain.

Here's one last mysterious paragraph-sentence from page 37. I begin with the preceding sentence to give you at least some hope of understanding. "The surging silver price in the early part of 2004 led to a rise in scrap supply (mainly in the form of Maria Theresa Taler coins) in Saudi Arabia, Yemen and north Africa. While the trade spoke of significant volumes, these paled against those seen in 2002 and it was perhaps the concentrated nature of the flurry of activity that led to some exaggeration regarding annual quantities supplied." Supplied? What? Scrap? What? What? I'm dying of preposition poisoning!

Naww, let's not quit yet. I'm having fun. On page 38 my bleeding eyes read, "One important feature (and a differentiator from gold) is that imports of scrap from price sensitive regions are rarely substantial." Does "rarely substantial" mean "usually small." And what, pray tell, is a "differentiator"? Is that a Swarzeneggerian Terminator bred to differentiate instead of terminate?

I really like the "flying references" that throw you back to another chapter or Survey without mentioning exactly where you should look. "In addition, 2003 had seen a sharp rise in exports to the United States, in response to an arbitrage opportunity (see last year's Silver Survey for more on this), of which there was no repeat last year." The yearly Survey generally contains about 90 pages with 7 chapters, numerous appendices, and focus boxes. It is a rich and tiresome field for research when given only the vaguest notion what you're looking for.


Unhappily the GFMS folks haven't discovered that neither a computer spell checker nor your ear can be trusted. Page 43 makes this plain, speaking of South Korea. "The large increase in exports followed a commensurate rise in silver recovery from the imported concentrates processed by the two main smelters based on the peninsular." Granted, there is such an English word as "peninsular," an adjective, and the spell checker would pass it. Unfortunately, it's not the word wanted in this place. Rather, the writer wants the noun. He means "peninsula", but in many English (and Southern) dialects, final a is pronounced er. So my Mississippi-born father-in-law, God bless his memory, would have pronounced it "peninsuler", but spelled it "peninsula." This might seem a picayune and picky objection, but remember that the Survey is not a high school term paper. It claims to be a professional statement produced by an organisation standing at the top of its class in the world. Can they not afford a competent proof reader?


I will grant that in the past few years GFMS has improved in supplying supporting data. For a few years there they offered mystery conclusions without any data, or two by two and a half inch "charts" without data tables. (Does that thing read "500" million ounces, or 550 million?) And Oh, how painful is their unapologetic mixing of metric tonnes on some tables and millions of ounces (Moz) on others. That makes analysis almost as much fun as comparing Greek and Old Slavonic translations of the prophet Malachi. (One metric ton equals 32,150.8 troy ounces.)

I concede that GFMS must transmute an ocean of data into a short, comprehensive report, but too often past Surveys left the reader guessing what on earth might be the factual basis of their conclusions. The Surveys often betray a foundational editorial bewilderment. Which data should we include and which excise? How many details -- if any -- should we give about this or that controversy? Too often the Survey merely alludes to some recondite controversy without data or explanation, offering a conclusion only, with its feet firmly planted in the air.

For example, over the past 16 years the silver market has faced a single giant riddle: how is the ongoing deficit of supply over demand being filled? Nearly one and a half billion ounces of silver doesn't just pop out of the ground. For the past couple of years the Silver Survey has concluded that China supplied the deficit silver. Well, that conclusion may be correct, and it may not, but the Survey doled out facts and arguments so niggardly that the reader could not rationally evaluate that conclusion. Therefore GFMS's conclusion sounded like "magic Chinese silver" that appeared and vanished according to their need to balance the Survey, even though it may very well be correct.

This year's Survey offers an example of how they should address controversies and mysteries. Page 41 discusses certain tax changes by Indian states that altered long-established flows of silver into India. This is very important because India ranks as the world's hungriest silver consumer. The world's second largest silver consumer (after the US), the Indian silver market is also the world's most volatile. A change in the monsoon there, for example, can exert profound influence on silver, raising or lowering farmers' income and with it their appetite for silver. Details such as these serve a useful purpose, if they are in context. Otherwise they become surplusage -- packing to camouflage lack of substance.


This year's Survey also offers an example how not to address mysteries. There's nothing unusual or sinister about revisiting previously published statistics and correcting them when better data become available. In fact, integrity obligates it. But this years' corrections were huge, and offered without any explanation I could find.

This year's revised mine production figures were astonishing, raising mine output over the last ten years by 54.9 million ounces (1,707.6 tonnes). Combined with other changes, notably in fabrication and photographic demand, GFMS materially lowered the cumulative silver supply shortfall. Using the 2004 Silver Survey, the cumulative deficit from 1989 through 2003 was 1,521.6 Moz (47,327 tonnes). Using the 2004 Survey, however, with all its revised figures, the same shortfall becomes 28.2 Moz (877 tonnes) less. Hey, what's 877 tonnes among friends? Perhaps GFMS offered some explanation of these large revisions in the 2004 Survey, but if they did I overlooked them.


Here are the highlights of the 2005 Survey:

  • Mine production increased from 611.2 Moz to 634.4 Moz (by 4%), an all time high.
  • Silver production from primary mines (as opposed to by-product mine) grew to 30% of all silver mined, up from 28%.
  • Scrap supply fell to 181.1 Moz, a four year low and 2.5 Moz drop from 2003.
  • For the seventeenth straight year the marked showed a "structural deficit." That is, production and old scrap fell short of meeting fabrication demand.
  • Government silver sales fell by 30%, furnishing only 7% of all supply. Chinese government sales dropped by nearly half.
  • Fabrication demand fell 2%, but this was a strong showing in the face of a 36% higher average silver price.
  • Photographic usage dropped 6%, with some sectors falling and some rising. Chinese photo demand grew 6%.
  • For their third year running industrial fabrication demand grew, by 5% to 367.1 Moz.
  • Jewellery and silverware slumped to 247.5 Moz or 9-3/4%, thanks to a huge drop in Indian demand. (Silver rose 30% in rupee terms and the monsoon was poor, pinching Indian farmers' incomes.)
  • Implied net investment jumped from 2003's 8.7 Moz to 42.5 Moz.


Touching silver demand, the loudest argument for two and a half decades has raged around digital replacing silver in photography. Those who understand silver photographic demand have always known that this transition would not take place overnight. Callow analysts and green reporters new to silver, on the other hand, periodically "discover" that digital is about to replace silver in photography. Amazed by their "discovery," they publish articles shouting that silver is about to plunge because of digital competition A suspicious mind might conjecture that scared people longing to escape their short silver positions must hire these analysts to force silver to jiggle downward a couple of days, allowing them to flee. A serious mind must laugh at this tired old red herring.

In our 1993 book, Silver Bonanza (pp. 69-77), Jim Blanchard and I concluded that rising economic development and incomes in China and elsewhere would keep silver photographic consumption high. Shifting photography from silver halide to digital would take a long, long time for several reasons. First, technology never changes overnight. The new must wait until the old needs to be replaced. Next, the new technology costs many times as much as the old, and offers much lower resolution. Third, enormous pent up Chinese demand would run into cheaper silver halide photography first. Fourth, digital photography would actually increase demand for some silver halide products, such as photographic paper. The bottom line was that digital would not replace silver at any time material to our case for a bull market in silver. Eventually, maybe, but not soon enough to hurt silver.

It's comforting to see that the 2005 Survey's photography section (pp. 57 -61) confirms our 1993 forecast. "Chinese photographic demand rose 6% year-on-year . . . higher than the 2% growth rate seen in 2003. Despite the penetration of digital cameras, the consumption of traditional film in China has actually grown because of an expansion of the total photographic market. Although the prices of digital cameras are falling, they are still expensive for low income families. Personal incomes in rural areas have increased enough to encourage the purchasing of discretionary items, putting low cost film cameras within the budgets of many consumers, but digital cameras by and large remain out of reach. [T]he preference for film cameras is also driven by the fact that they are easier to use in terms of obtaining and sharing the . . . photograph." (pp. 60, 61)

Relief! Silver is safe in photography for yet another year.

Another interesting twist remains to the photographic silver controversy. Silver photographic use has always ranked as one of the most active re-cyclers of scrap. As silver photographic use declines, so does the silver supply from photographic scrap.


Here's another tired old argument against silver that often raises its balding head: "gold is money but silver is an industrial metal." From that descends the objection that a world economic downturn would send gold up but silver down. As a slower world economy reduces industrial demand for all commodities, silver demand would drop, too, regardless how gold moved.

What this argument overlooks is that over half of silver production (58%, in 2004) comes as a by-product of lead/zinc or copper mining. (Primary silver mines produce 30% of all silver mined, and gold mines 12%.)

My German friend Dietmar Siebholz points out that this production profile at least partially contradicts the whole "industrial metal" argument. "Compared to last year, silver mine production's dependence on base metal demand has relaxed somewhat. Besides, the assertion still remains valid that any cyclical economic decline that curtails zinc, lead, and copper usage curtails silver production at the same time. On the other hand silver usage, which resists cyclical economic declines much more readily because of its unique strategic qualities, will suffer a much lower percentage drop in demand." ("Meine Meinung zum", 27 May 2005, my translation).

In other words, as demand for base metals decreases, so does their silver production. That doesn't completely correct silver's problem, but it offsets part of it. Beyond that, I don't buy the "silver is an industrial metal and only gold is money" argument. At the margin investment or monetary demand primarily drives both silver and gold prices. The fundamental demand pictured by supply/demand analysis changes only slowly, while monetary demand hits and grows suddenly. Is silver money or not? It's beside the point. Since 1960, silver has moved with gold in every gold bull market, spurred by the same investment demand. Before 1873, silver moved with gold, spurred by the same monetary demand. Overwhelming odds are, silver will move with gold in the future, although certainly the past doesn't guarantee that.


Some silver stocks you can see, and some you can't see. Some, like the reported stocks in the New York or Tokyo commodity exchanges, or in government hands, are plainly visible. Others, like European dealers' stocks, are invisible, and must be estimated. GFMS estimates these at 332 Moz at end-2004, about three times Comex stocks of 104 Moz.

That leaves me nervous.

As far as I know there is no government mandated or even official account reporting required from these dealers. The dealers are notoriously tight-lipped about their business (witness GFMS remark on page 32 about "obvious confidentiality issues" with individual dealers.) Since not every dealer will report stocks to GFMS, I reckon that they must infer the total. How? By a statistical abstraction. That is, they get samples from 30 or more dealers, compare those to last year's statistics from those same dealers, and from that and other factors infer the current level of all European silver stocks.

There's nothing wrong with that, it's a standard statistical procedure, except when dealers aren't reporting their exact physical stock levels. Not long ago Dietmar Siebholz reported a conversation with a Swiss banker that hints that most of the silver European banks hold on deposit has been loaned out.

Whether that's true or not, I can't verify. However, since I don't know what statistical procedure GFMS used to infer stocks, or how accurate dealer reports might be, or how many dealers report out of the total, a large shortfall in European stocks is certainly plausible. Add to that seventeen years' supply deficits, and the mysterious "appearance" of enough silver to fill that void, say, 1.5 billion ounces.

Now consider the silver leasing market, where rates at the end of 2004 were a minuscule 1/10 of one percent. With an interest rate that low, either demand is awfully low, or supply is awfully plentiful. But GFMS reports that the silver leasing market is growing, reaching at year end 275 Moz or 8,560 tonnes (p. 33). In fact, GFMS reports, "[I]n recent years intermediaries have aggressively targeted such manufacturing companies that use high volumes of silver. A number of these users were arguably 'under-borrowed' and their silver leases have seen an important growth over recent years. . . That the pick-up in borrowing demand has had little or no effect on leasing rates owes much to the surge in available liquidity from investors' long positions in silver."

Hmmmm . . . Why exactly would investors' long positions in silver cause a "surge in available liquidity"? First guess is that GFMS believes the bankers holding silver deposits for those investors are loaning it out. In fact, the bankers are "aggressively" beating the bushes for borrowers, to make a measly 1/10 of one percent. Somebody is mobilising lendable silver stocks and actively promoting borrowing, and only the banks possess the silver.

Now why would they take all that risk for a lousy 1/10 of one percent? Because they plan to invest the loan proceeds into something else which, they expect, will pay a higher return than silver before they have to buy the silver back. In plain English, they have sold massive amounts of silver they owe to depositors.

They are short physical silver.

And that may furnish us yet another reason to buy and take delivery of physical, not paper, silver, before the investors make a run on the banks to retrieve their missing silver.


The way GFMS reports world silver supply and demand confuses me somewhat -- and others, too, I don't doubt. They don't report supply and then subtract fabrication usage from that to arrive at a surplus or deficit. Rather, they treat supply and demand like a problem in double entry bookkeeping. Total silver supply must meet total demand, by definition. If they don't match perfectly -- and they never do -- then GFMS infers balance figures such as "Producer Hedging," "Implied Net Disinvestment," "Producer De-Hedging," and "Implied Net Investment." . Yes, "infers" is a nice word for "makes up to balance." (You know somebody who "infers" balance figures on his income tax return, I'll bet.)

So if you want to know how silver consumption compared to silver supply, you have to throw out those balance figures and calculate the "structural deficit or surplus." For the past 17 years that calculation has produced a shortfall (deficit). As a result, the world has burned a cumulative 1.5 billion ounces of silver since 1989. Nobody seems to know where all that silver came from, but they infer, pretty soundly, that it came onto the market following the 1980 price surge.

Since most everyone believes that price eventually balances supply and demand, that cumulative deficit is the first and best argument for higher silver prices.

But a related question must be asked: Is the deficit increasing or decreasing? For this reason I dislike very much GFMS's unexplained diddling with past figures in the 2005 survey. Here's the change in the structural deficits, according to GFMS changing figures:

   2004SS  2005SS:
 2003  72.0 Moz  58.6 Moz
 2002  62.6 Moz  44.2 Moz
 2001  89.5 Moz  72.8 Moz
 2000  147.6 Moz  136.3 Moz

These changes affect the shortfall's chart even more dramatically. From a chart that appears to have bottomed and begun to climb back toward even, the structural deficit chart now ascends very steeply. This implies the deficit's days are numbered.

I'm not accusing GFMS of fudging the numbers to make the deficit look smaller, but the change did accomplish that. I'm curious why they made those large changes, but GFMS remains silent.


Another good work from GFMS is their attempt to establish cash production costs for silver mines, a statistic that remained a mystery until recently. While these cannot comprehend all silver mines, because 70% of silver comes as a by-product of mining some other metal, at least GFMS' estimate gives us an annual number to compare. Wonderful!

In 2004 simple cash margins (average annual silver price less cash costs) were $4.29/oz, which leaves healthy room for profit in primary silver mines. Total production costs (including depreciation, depletion, and amortisation) were estimated at an average $3.40/ounce, and ranged from $2.11 to $8.12. (pp. 27 & 28).


Nothing I found in the 2005 World Silver Survey changes my outlook for silver. All the following factors remain, all working for higher silver prices:

  • Long term shortfall of supply against demand (17 years)
  • Long term undervaluation (much of the 1990s spent with silver's price under production cost)
  • Wearing out old bulls. No doubt 1970s investors (and their heirs), tired of holding silver so long, supplied much of the 1.5 billion ounce cumulative deficit. The last 17 years has burned up their silver.
  • No overhanging stocks. Hoards such as the US government's or India's have been dissipated. By the 1980 peak all the U.S. "attic silver" that could be bought under $20 (equal to $50.79 in 2004 dollars) was bought. The rest of the world doesn't have any attic silver.
  • Demand inelastic to price. In most industrial applications only a tiny amount of silver is used in relation to the end product's price, so usage doesn't shrink much when silver's price climbs. There's more: in most applications no good substitute exists for silver at anything near silver's efficiency or price.
  • Technological reductions in silver use. Stung badly by the 1980 silver peak, for more than two decades silver users have been applying every technological innovation to squeeze down silver use, so further reductions will be tiny.
  • Primary trend of falling paper asset, rising hard asset cycle. Paper assets like fiat currencies, stocks, and bonds move inversely to hard assets like metals. Stocks peaked in 2000, the US dollar in 2002, and bonds in 2003. These trends will last at least a decade. Gold began rising in 2000 and silver in 2001 after more than 20 year bear markets.
  • Investment or monetary demand moves the silver market at the margin. Only that can take silver to new price heights. Falling confidence in dollars and other fiat currencies stimulates monetary for both silver and gold as alternative monies.

Since only investment demand for silver really drives the price wild, fundamental demand doesn't matter much for our investment case. Certainly fundamentals must be on our side, since large surpluses or supply overhangs would create a strong headwind. However, since silver has already racked up big deficits, we only want to check fundamentals to make sure the trends that favoured silver in the first place still exist.

They do, and I still cannot see a single investment on the horizon that will outperform silver. I remain convinced that silver will outperform gold by 400% and will reach US$78.80 or more in the next ten years.

World Silver Survey 2005, US$195 from the Silver Institute, 1200 G. Street NW, Suite 800, Washington, DC 2005. (202) 835-0185; fax (202) 835-0155 or Outside North America, contact Gold Fields Mineral Services at +44 (0) 20 7478-1777 or, who produce the report for the Silver Institute.

Copyright ©2005 The Moneychanger

July 2005
-- F. Sanders


Franklin Sanders has edited and published The Moneychanger newsletter and brokered physical gold and silver since 1980.

You can find more of his articles at at

321gold Inc