The 2004 Diamond Pipeline

Downstream, on the other end of the pipeline, diamond jewelry retail consumption in 2004 has risen by some 8.5-8.7% in dollar terms and 6.8% when measured in local currencies. This makes 2004 the best year, in terms of sales growth, that the industry has seen for decades. What is more significant is that this growth doesn’t present a statistical aberration, but rather reflects a fairly persistent trend: 2002 recorded a growth of 2.6%; in 2003 it went up to 5.9% and 2004 saw a further increase of 8.7%. According to DTC director Gareth Penny: “In the five year 1999-2003 period, we saw an average annual worldwide retail growth (expressed in dollars) of 3.6%. In the five preceding years, from 1994 to 1998, sales showed an average negative result of -0.2%.” Last year’s performance should make the publication of the annual “2004 Diamond Pipeline” a joyful event, so why do we have the sense of trepidation that we actually feel by presenting this information? It is because, more than ever, the figures don’t add up - and larger volumes don’t mean “more money”. 

At the retail level we see that the 2004 growth is achieved at significantly lower margins. We see that “discounting” (already well before Thanksgiving) has become an integral part of the selling strategies. Moreover, we see that the jewelry retail and wholesale sales levels are below the level of new supplies - so the overhang at these downstream levels is increasing, notwithstanding the recorded growth in consumer offtake. De Beers has sold fewer carats at higher unit prices. The managing director of De Beers, Gary Ralfe, notes that the DTC’s sales of $5.7 billion were, at an average, at 14% higher prices than the $5.5 billion of the previous year. So “in real terms” (or in “carats”), the DTC sales actually declined - it supplied less volume to the markets. Cutters process carats, not money. The lower carat sales helped in creating a feeling of shortages (“we cannot find the goods we need”) while, in fact, the new polished production was not readily sold.

De Beers knows this. Gary Ralfe has said that “exports were robust from the cutting centers through to the retail market. It could have been anything between 15% and 20% growth there”. This growth rate is more than double, almost triple, the retail growth. So while inventories in the cutting centers may have come down, as De Beers believes, there has been a shift in inventory holding downstream. A stone sold (or that has moved) from the cutting centers to the jewelry industry is not necessarily a stone that has been sold to consumers.

The managing director of De Beers sees a causal relationship where there isn’t one. Says Ralfe: “...the debt level [in the cutting centers] has continued to rise, it is standing at $9 billion now. This is clearly the result of the cheap US interest rates that have been available. The results of that have been that there are longer credit terms that have been available to retail customers...” Actually, the debt levels are higher - and closer to $9.8 billion plus $0.6 debt securitization - and bankers find it more realistic to refer to $10.4 billion.

The growth in the debt is almost fully attributed to the increase in stocks on the jewelry wholesale and retail levels. We think that the polished overhang has created a “buyers’ market” and that the fierce competition among the polished manufacturers has forced them to agree to irrationally long credit periods. Low interest rates may have facilitated these terms of trade; they are not the cause.

Through concerted efforts, polished prices may have gone up by some 10% during the year, depending on the type of goods, but retailer margins on the sale of loose diamonds and of diamond jewelry have declined by some 15%-20% Given the cumulative growth of the diamond jewelry retail market of the past five years, it doesn’t appear that the end consumer is paying anything more for the diamond he/she is buying, (except, maybe, for branded goods.) Thus the 2004 record rate of growth story cannot be viewed as an unmitigated achievement in each of the pipeline phases.

Looking Behind the Figures

With all the immense progress in industry transparency, accountability and data availability, it is precisely the sheer abundance of data that reveals and accentuates the inconsistencies, the distortions and the flaws in the flow of goods moving through the diamond pipeline. If in 2004 the free-trade zone of Dubai registered $883.9 million of imports of 28.2 million carats of rough diamonds at an average of $31.31 per carat and when the same volume is exported for the double value of $1.7 billion of exports at $58.25 per carat - one knows that something doesn’t add up. When 12  million carats go into Swiss free trade zones at $69.55 per carat only to be re-exported at $132.42 per carat one knows this isn’t correct. And when these (partial) re-exports of rough from Dubai and Switzerland equal some 30% of new world production, these figures are bound to impact the pipeline.

But free trade zones present only a minor dilemma as these distorted figures can be neutralized in a pipeline analysis. (There are, however, some less comfortable implications: either the values of the relevant exports from the producing countries are understated or the subsequent re-export to cutting centers are at grossly inflated values. It is probably a combination of both) And it may be unfair to single out Dubai and/or Switzerland. Elsewhere too one can find amazing added values for a rough product that only moves from one parcel into another.

In the cutting centers (and especially India) a misstatement on the rough side is matched by similar adjustments on the polished side, to avoid the creation of fiscal liabilities. It is for sound reasons that the figures for India in our pipeline are not the same as the officially published figures.

No, some of the reluctance comes from the retail market diamond jewelry data. No organization spends more resources on the collection of consumer market data than De Beers through its continuous global tracking surveys. And though we don’t question the validity of this data - as there is probably no more accurate data on the market to make an intelligent challenge - we are concerned that the DTC uses the retail sales data to vindicate and to demonstrate the enormous success of its Supplier of Choice marketing strategy. We tend to wonder whether the “cause and effect” relationship between SoC and market growth has not been overstated - there are too many other significant macro-economic drivers which impact diamond jewelry sales. But that isn’t the issue. We are slightly concerned that the use of data to justify a policy, rather than to inform, may unwittingly bias the presentations, especially since the release of this data is selective and limited.

A DTC inside source pointed out that under Supplier of Choice the “market growth” has become part of the KPI for De Beers marketing and sales people. (For those uninformed of De Beers jargon, KPI stands for Key Performance Indicators - and these are used to set the bonuses of some employees.) So many people have a stake in high growth figures... Nevertheless, they still seem the best we have - at least on the retail side. And as these figures are based on consumer surveys, rather than statistics, the fact that huge amounts of polished are smuggled into markets such as Canada, China, Italy, Korea, Taiwan, India, etc. are not impeding the accuracy of the data.

Dollar versus Local Currencies

The dollar dramatically fell last year against other major currencies - a fall which was steepest in the last quarter of the year, the period in which some 40% of all diamond jewelry sales takes place. A depreciation of, let’s say, 10% against a local currency, would also make diamonds more attractive by that percentage expressed in local currencies. If the increase of diamond jewelry retail value is more moderate than the local currency depreciation against the dollar, there was no growth in diamond sales in real terms.

Another factor of concern to us is the fact that the retailer margins on loose diamonds and diamond jewelry is continuing to reduce. This leads to a realignment of the ratio between the diamond jewelry piece’s retail price and its diamond content measured in polished wholesale prices (pwp). The diamond content increases and the consumer simply gets “more diamond” for his money. It also means that more diamonds (carats) are sold for less money. In Turkey, for example, the ratio of pwp diamond content to the diamond jewelry retail values exceeds 50% and it is not surprising that diamond jewelry sales in that country registered a 25% growth in local currency in 2004 - the highest growth recorded anywhere in the world in 20041. But, let’s go through the pipeline systematically.

World Production

Botswana is the world’s largest diamond producer (by value). Through its joint venture with De Beers, its Debswana mining company produced in 2004 some 31.1 million carats, an increase of 2% over the preceding year. Quite a performance given the fact that Debswana faced technical problems and a two-week strike in 2004. Its 2004 production has a value of $2.32 billion. However, our pipeline depicts the actual flow of goods during the year. In the calendar year 2004, the DTC imported 38.6 million carats from Debswana at an average value of $74.58 per carat, making for a total of $2.879 billion in actual sales and deliveries to the DTC. (This explains why in the pipeline mine sales are higher than production figures.)  

In Namibia the largest mining producer is Namdeb, a joint venture with De Beers. Namdeb produced 1.9 million carats of high value diamonds (mostly offshore), which, at a realized sales value of $346.05 per carat, comes to a total value of $657.5 million. Namdeb was undoubtedly the success story of De Beers in 2004. Its marine operations mined 800,000 carats, which is 31% more than in the previous year. Total carats recovered were 28% higher than in 2003. The other producer is Samicor, which had a production target of 150,000 carats in 2004, but reportedly remained below this objective. In our pipeline, we estimate Namibia’s total production at $700 million.

Russia’s Alrosa mining conglomerate is the world’s second largest diamond mining company by diamond production. For many years, the exact figures on diamond production were “guestimates”, as carat production figures were considered a state secret. Some of these figures have been declassified and we now know that in 2003, Alrosa produced 33.02 million carats of rough diamonds which were sold for $1.7 billion - an average of $51 per carat. In the first half of 2004, the corresponding data were 17.8 million carats produced, worth $944 million, for an average of $53 per carat. The problem with setting a precise value is that there are great discrepancies in sales prices, discrepancies which are currently under investigation by the Federal Antimonopoly Services. Factories affiliated to Alrosa and the government of Yakutia are said to secure a far more favorable price than other local factories. De Beers is believed to get its diamonds at a significant (18% in 2004) discount to the price paid by domestic manufacturers. As Alrosa (and Russia) sells also from inventory, its rough sales tend to be higher than its actual production figure.

Our own research estimates the 2004 Russian production at 31.71 million carats, valued at $62.71 per carat (2004 world prices), making for a total of $1,988.8 million. Of this production some $700 million was marketed through the DTC.

In Canada the two diamond mines registered record output, as Rio Tinto and Aber’s Diavik mine operated at full production. In terms of carats, Diavik produced 7.575 million carats, an increase of 98% over 2003. BHP Billiton’s Ekati mine produced some 5 million carats. As the production mix (different pipes) changed throughout the years and as average prices changed significantly throughout the years, we have, for pipeline purposes analyzed the rough imports from Canada into the United Kingdom. Both Ekati and Rio Tinto’s 60% share are first shipped to the U.K. before being re-exported to Belgium (for fiscal reasons and, it is said, to preserve the integrity of the data). The 2004 Canadian production - based on export data - of some 12.6 million carats is valued at $1.6 billion.

In South Africa, De Beers Consolidated Mines produces some 95% of the nation’s output. Statistics are difficult to interpret as many alluvial diamonds from Angola, DRC and other countries are smuggled to S. Africa where, for a fee, diggers are willing to declare these goods as having been recovered locally. De Beers produced some 13.7 million carats, an increase of 15% over the previous year. Even though five out of De Beers’ seven S. African mines are reportedly unprofitable because of the appreciation of the rand, 2004 was a record year. The Kimberley mine produced 2.05 million carats, the highest annual production of this mine since 1914. The “jewel in the crown”, from a revenue perspective, is Venetia which produced 7.2 million carats, an increase of 9% over 2003.

The future of many of the South African mines depends on production. DBCM managing director, Jonathan Oppenheimer, has introduced a “thrive at 5” strategy, aimed at being profitable also when five rands are one dollar. We estimate the total South African production at 14.6 million carats at an average of $72 per carat, making a total of $1.05 billion.

Argyle’s production in Australia saw a steep decline from 30.9 million carats in 2003 to 20.6 million in 2004. At $14 per carat, Australia’s Argyle production is estimated at $290 million.  The Williamson mine in Tanzania outperformed itself by going from below 200,000 carats to well over 300,000 - a growth of 38%. At $87 per carat, this mine contributes merely $25 million to the world totals. The less certain data comes from the mostly alluvial countries such as Angola and the DRC which we conservatively put at $0.9 billion each.

Rough Sales to the Market

De Beers, as a private company, doesn’t provide as much corporate data as it used to - and, of course, it doesn’t have to. Though in the past few years it had also drawn from stocks, at the end of 2004, its diamond stocks were about $100 million higher than the year before, at $1.856 billion. The company’s finance director, Paddy Kell, did, however, note the “sharp drop in the volume of the diamonds which were sold” in 2004. The DTC is believed to be having now only “working stock”, which would be sufficient for 2-3 sight cycles.

Debswana did, however, sell more to the DTC in 2004 than it produced. Also the Russians have made further withdrawals from stocks. Other mining companies are selling their current output without unnecessary delays. Thus the “intake” of the DTC, from the De Beers producers ($4.0 billion production), from inventories of associates (Debswana approx. $400 million) and Russia (slightly below $700 million) gets to a DTC intake of around $5.1-$5.2 billion. Because of the 14% average price increases, the nominally lower intake didn’t affect the overall inventory levels.

The total mining sales to the DTC and the market were about $11.388 billion. Assuming that the DTC, as an independent profit center also earns some money itself, the total rough supply to the markets in 2004 may have been closer to $11.9 billion. Of this, the share of the DTC was $5.695 billion and of the others $6.288. The sharp increases in Canadian and Russian production may have triggered, for the first time, a fairly balanced position in which the DTC has only some 50% of the market. We say this with caution, because De Beers produces some 60% of all rough stones higher than 2 carats - and these are the ranges in which the price increases were the steepest. This is also why the DTC was able to increase its earnings in the diamond account by a hefty 27% to slightly over $1 billion. As these increases have not (yet) been fully reflected in the production figures, we assume that in 2005 the DTC will again hold 55%-60% of the market.

Manufacturers and Retail Levels

Whenever the prices of rough grow faster than those of polished, and are - as they say - “out of sync”, the manufacturers and traders find themselves squeezed and complaints about decreased profitability tend to fill the (media) air. Nevertheless, anecdotal evidence suggests that 2004 was a better year than 2003, also from a profitability perspective.

The trend towards industry consolidation (and greater concentration of more business in fewer hands) continues unabated. An interesting yardstick to measure this phenomenon is viewing the composition of the banking debt. In Israel, for example, 1% of all diamond industry banking clients (i.e. six clients) account for 33% of the $2 billion Israeli diamond sector’s banking debt. Ten percent of clients hold 70% of the debt. The 2004 growth in concentration is well over 10%. As the industry still remains rather fragmented, this trend is expected to intensify this year - and may accelerate into higher gear, depending on the number of Sightholders which the DTC will drop when it announces the names of those to whom new (2.5 years) contracts will be offered.

During the year there were many complaints about “shortages” of rough and the rough market was “hot”. One must be careful not to draw the wrong conclusions. To a large extent these shortages reflected over capacity in the manufacturing centers - especially India. (India was affected by production declines of the Australian Argyle mine.) The manufacturing sector is still largely production-driven, rather than demand-driven. This enabled the rough producers to comfortably increase selling prices - the manufacturing sector absorbed the goods eagerly. Then they got stuck at the retail level.

This is reflected in the Diamond Pipeline. The worldwide polished production is estimated at $16.74 billion, while the polished diamond content in diamond jewelry retail sales (measured in polished wholesale prices) is some $750 million lower at $16.04 billion. The retail sector is facing fierce competition from other luxury items and diamond jewelry (and loose diamond) internet sales.

There were some particularly good diamond jewelry consumer markets in 2004. In India a 19% growth was recorded (in local currency), Turkey 25%, the Gulf countries 14% and China 11%. The major markets showed smaller gains. The United States grew by 8%, Japan by 3%, the United Kingdom 4% and Canada 5%.

The declining retail margins and the higher retail stocks (many of which are “unpaid” goods received on consignment basis) may make it difficult to repeat in 2005 the 8.7% worldwide diamond jewelry retail growth performance. Meeting that level will be quite an achievement; beating it will simply be a miracle.