Here’s one of several economic indicators: Based on official US government figures, between October 2014 and March 2015, retailers lost $550 million in fine jewelry and watch sales.
During the equivalent period in 2013-2014, overall fine jewelry and watch sales in the US totaled $43.58 billion, falling to $43.03 billion during the six months of October 2014 to March 2015.
Industry Indicators Say
In the 2013-2014 period, specialty jewelers’ sales totaled $17.31 billion, declining to $16.49 billion in the 2014-2015 period, a loss of $816 million. You may look at the figures and ask yourself how it is possible that the total decline is smaller than the decline among a segment. The answer is that sales at “other” retailers, multi-item retailers, actually increased in the period. Specialty jewelry retailers, on the other hand, lost business and market share.
Consumer Prices Declining
The shrinking sales figures indicate a decline in demand. Retailers are reporting less traffic into their stores and that sales are shifting to smaller, less costly items, pointing to a rise in prices of fine jewelry that has led to further decline in interest.
Prices of fine jewelry have been declining since early 2012, so is this really the issue? Following is an index of consumer prices (JCPI – Jewelry Consumer Price Index by BLS), a measure of the average change over time in the prices of fine jewelry as paid by consumers.
With the exception of a surge in mid-2013, prices have been heading south for many long months. So where are they today? Imagine a fine jewelry item that is representative of the changes in average prices in the US market. Let’s say that it cost $1,000 in December 2010, the month before the above graph starts, and let’s follow the month-over-month changes in its price since that time.
The small, yet constant price increases throughout 2011 brought the price of this sample item to $1,078. The compounded monthly price increase hiked up its price 7.8% by December 2011. By December 2012, it softened to $1,047. It hit a record $1,120 in August 2013, and despite some price increases, by December 2014, the price of this item fell to just $987, a 1.3% decline from December 2010. According to the latest available data, prices increased and brought the item’s price full circle back to $1,000 in April 2015.
Economic Indicators Say
We know diamond prices peaked in mid-2011 and have been slowly softening ever since. The price of gold peaked in September 2012 and has also been heading down since then. With the cost of the main components decreasing for so long, and with a strong dollar compared to the Indian rupee, for example, the cost of labor is decreasing too. So is cost of jewelry really the issue?
Another piece of data tracked by the US government is producer prices. Its Jewelry Producer Price Index (JPPI) is an index that tracks average change over time of real selling prices by US jewelry makers. Here, too, we see a decline in the price producers are charging retailers.
The decline in sales is not necessarily an issue of high prices or a bad price reputation; after all, diamonds and diamond jewelry are luxury items, not a necessity. The half-a- billion dollars in lost sales is a figure significant enough not to be dismissed as a phase. Instead, it warrants a closer examination. In the past, a decline in demand in diamond jewelry preceded wider economic declines. Closer to home, it may simply be a fading awareness of this category among consumers.
First I think you meant to explain that the first chart shows year-over-year change in total sales of jewelry and watches in GREEN while change in sales by jewelry specialty stores is shown in BLUE.
First, a nit of a comment.
You say: “… look at the figures and ask yourself how it is possible that the total decline is smaller than the decline among a segment.” The way that happens, you claim “is that sales at ‘other’ retailers, multi-item retailers, actually increased in the period.”
Your answer doesn’t necessarily follow: For example, let’s use the March numbers of -1.7% and -5.8% . Let’s say the subset of jewelry specialty stores represented 20% of total sales, sales of that subset declines by 5.8% while sales by the whole set declines by 1.7%. Then sales by the complement of the subset (the other 80%) would in fact be in DECLINE (by .7%.)
You need another piece of information to reach your conclusion: In reality the subset of specialty jewelers represents roughly 40% of total jewelry sales. Using that and the same March numbers we get that the sales of the complement (the general retailers) would then, indeed, be INCREASING (by 1%.)
But there’s a deeper issue that’s not a nit: the source of this data may simply not allow such close examination. First, taking a Stats 101 look: The standard errors for year-over-year change, which are published by the Census Bureau, right along with monthly retail sales estimates, are quite large. In fact the 95% confidence interval for 2014-to-2015 March jewelry specialty stores sales (which you assume to be -5.8%) is in fact -12% to 1%. This is not surprising as it comes from a small sample. Even if we were sure that the change in total sales was -1.7% for that period, we could NOT CONCLUDE that there is any difference between the year-over-year changes in sales by jewelry specialty stores vs total jewelry sales.
The situation is actually worse still. The Bureau of Economic Analysis, while producing the total Jewelry Sales estimates, (the one with the 1.7% decline) uses the Census Bureau’s estimates of the subset of sales (the one with the 5.8% decline) as its principal starting point. In fact, as far as I can see, they have no other direct source of information on jewelry sales. I’m not arguing with the BEA’s models – the econometrists and statisticians at the BEA are pros and are doing what they can to get the best monthly GDP growth estimates out of the door. But they themselves warn against using the data in the ways this industry, not just you, is using it.
Any comparison of these two data points may, in great part, be revealing artifacts of the BEA’s models rather than any market reality.
Where does that leave us? We should probably not depend so much on these two, decidedly non-independent, estimates. Instead, this industry needs to develop more data sources and apply more rigorous methods – and focus on understanding the systemic changes and long term trends of this industry.