The short answer is no, if you are not willing to speculate. In other words, you will generally not be buying in large volumes at an acceptable discount to spot prices.

And if you asked me this question, off the cuff I would look at you like you were crazy and probably say, “You want to buy gold? Why don’t you just go downtown? Who would buy gold piecemeal with all the hassles? And public markets don’t tolerate blatant arbitrage!”

But there are some interesting lessons from such an exercise, especially when compared to markets for art, watches, fine (signed) jewelry, and other collectibles.

In the month of October 2024, gold averaged approximately $2,700/t oz.

I compiled a list of no- or low- reserve auction results of 23 (22 unsigned and 1 signed) lots of 18K+ YG jewelry averaging 70g, all from LiveAuctioneers.com and hammering within the latter half of the month of October 2024. Most were in the US, but 2 lots were from the UK. For simplicity, I assumed a 0.5% melt fee and did not factor in shipping costs. I included one signed piece from an obscure brand for comparison. Below is a screenshot1:

Here are the takeaways:

• Firstly, a sample of 23 listings2 seems too few, but all other unsigned 18k+ gold listings either had reserves3 or starting prices at or higher than gold melt value. So this is it in terms of volume for the latter half of the month of October 2024 - a paltry 1,627 grams. If I expanded my target to auctions with starting all-in prices at melt, there would have been much more, but I did not have time to calculate this. Just eyeballing, I estimate 8-10x more reserved auctions vs no or low reserve.

• The 22 unsigned lots hammered at a weighted average of a 4% premium to melt value on the day of auction (including buyer’s premium but excluding shipping costs). This is similar to what some gold dealers will ask as commission.

• As expected, signed lots (Lot 1) or lots with intricate handiwork (Lot 9, the San Marco bracelet) trade at a premium to unsigned, simpler lots. Cuban bracelets (Lot 13) also seem to command premiums.

• If lots 1,9 and 13 are excluded, the remaining 20 lots (1,396.62 grams) would have hammered at a weighted average of 0.14% premium to melt value.

• Although the highest variations to melt value were a 13% premium and a 19% discount, these were outliers and most variations were well below +/-10%. Volatility is relatively low.

• 6 out of the 23 lots hammered below gold value. Of those, only 3 had a spread of more than $500. Heavier does not necessarily equal larger absolute profit.

• There is a significant difference between auction houses. Gray’s Auctioneers, a small auction house in Ohio (lots #14-17) was not able to get a premium to gold value on any of their no reserve lots.

• Shipping costs will eat into margins, but registering directly with the auction house would offset this somewhat.

• If you speculate that gold will rise, buying unsigned jewelry at auction may be a low-overhead, low-labor albeit risky4 and low volume way to source. See below sensitivities, excluding lots 1, 9 and 13 which were outliers.

As of this writing, gold is at $3,008/t oz, so if someone had bid on all the lots in the list excluding lots 1, 9, and 13, total investment would have been $91,192 for a total 1,396.62g and the total discount to melt would be $10,685, or 11.7% implied gross profit over an approximately 5 month period. Whether that is a good risk-adjusted return or not is in the eye of the beholder.5 However there are some important lessons here:

1. All-in auction prices for secondary finished goods such as art, watches, fine jewelry etc tend to be “fuzzy”. As I’ve written before, auction prices tend to over- and under-shoot and can be unreliable. This is because of A. the heterogeneity of the items themselves, plus B. differences in intermediation value and price discrepancies due to market fragmentation, of which auctions are one island. Price volatility is high. Hence many bidders tend to “feel their way up” or “snipe” the hammer price, which usually climbs abruptly in the final seconds.

2. Gold on the other hand is a commodity that is traded on a large, centralized public exchange with broad global participation - unless it is a signed (branded) piece and/or has intricate/popular handiwork, the market price is clear, and any price discrepancy depends on the bidders’ future price expectations, and to a lesser extent transaction/shipping costs. Price volatility is low. When bidding for gold, it is better to place your bid faster to secure your maximum increment - one higher increment could put you in uncomfortable speculation territory.

3. I’ve argued before about the necessity of a global, centralized b2b trading platform for finished goods such as art, watches, fine jewelry etc. While we would not be able to eliminate all the “fuzziness” of prices due to the heterogenous nature of finished secondary goods (they are not commodities), we should see a reduction in the degree of fuzziness if there is a price index (like gold) based on pre-intermediation, wholesale/b2b transaction values with broad, global participation. This will benefit end-consumers, as retail prices will better reflect intermediation value as opposed to a perceived or false supply constraint6. It should also lead to lower retail price volatility. Implementing a centralized b2b platform is the most efficient method of developing such a price index.

Comparing auctions for commodities such as gold with heterogenous finished goods such as art, watches and fine jewelry (which many readers of this Substack are familiar with) makes the differences more palpable and underscores the importance of a centralized b2b trading platform. That is the real purpose of this post.

6

I wager that the word “rare” will appear less in listing description and titles.

Share

Leave a comment

Thanks for reading Infinite Breguet Substack! Subscribe for free to receive new posts and support my work.

Subscribe