Your hypothesis could very well explain the runup in prices for "high-end" coins & collectables, however, the "top 20% of income earners" aren't the only force driving the ancient coin market. The largest group of collectors who form the ancient coin market are people with a modest income, collectors willing to spend $50-$1,000 for the coins they like. As I've seen first hand after paring down my collection the last couple of years, coins in the $50-$1,000 range have risen in value from 25 to 200%. Many people of modest incomes have not gone on vacations & spent money on recreation because of the Covid-19 pandemic, instead they have added to their collections, as I have. Don't discount the "small fry" when it comes to hobbies, we are the real driving force, not the "high-end" earners .
Just out of curiosity is there a similar chart for auction values of US coins over the same time period? Is this just an ancient coin thing or is it for other numismatic collectibles?
Certainly my observation was focused on high-end collectibles, not just coins, and I noted only the correlation. Proving causality would require a lot more research and analysis that I'm, well, too lazy to attempt! At this time I could provide only anecdotal data, not statistically valid data. It would be interesting if you have any data and analysis that supports your assertion about the rise in coins whose prices are in the $50 - $1,000 range.
I based what I posted in this thread not only on my own experiences & observations of market activity, especially auction results, but also on written assessments from key figures from a number of different auction houses like CNG, Heritage, & Stack's & Bowers. If you need hard core data I'm not a statistician, however, if you approached the major auction houses world-wide you may get the data & answers you're looking for.
If cost of goods and interest rates return to pre-Covid levels is that really inflation? Or just a pot hole in the road?
My main anecdote to lower value coin appreciation doesn't even come from prices. I have noticed a lot more coins that normally wouldn't be offered individually being sold that way. Stuff like common Mithradates IV bronzes with scruffy surfaces being sold per coin. That, and the size of the group lots are going down, (except for some outliers like Leu). So, either the material is getting rarer, or the prices are going up so more coins are worthy of being sold individually, and lots are smaller size because more are being picked out of them for individual sale. Either way making me happy I bought stuff when I did. Sometimes I will just pull out a lot I bought a few years ago to get my group lot fix, (well, that plus I am shipping more stuff out of state and haven't been able to go get them yet ).
So dividing the realized amount and the number of lots the "average" price per coin was around 450 CHF. There also was a large number of gold coins for sale, particularly Byzantine.
I would argue that the price has gone up more on mid If you consider that estimate was 2x starting bid, which in most cases it was, you did well with 2x of estimate. So what is the correct description of the situation here: 1. Our money is worth less, and estimates haven’t been adjusted to this yet. 2. Estimates are correct, but the market is full of people with no sense of price. A bubble. 3. There are more collectors, creating higher demand, and prices will stay high in the foreseeable future.
Yup, one or two of those three. Now, which one they are will determine future price directions. I myself am torn. I think its a combination of money worth less and Covid bubble. Unfortunately those two work in opposite directions. If money stays the same I expect less money going to collectibles when vacations, (especially international), open up. However, that is a big assumption. Personally I am not thinking more collectors, simply more money from existing ones. Sure, at the top end one new billionaire being interested can change things, but down here in average coin territory I am not seeing a blossoming of new collectors any more than usual. New collectors are always born, but unfortunately we also lose some every year.
I took the liberty to redo the calculations, taking into account: 1) inflation, using data from https://www.in2013dollars.com/switzerland/inflation/ 2) standardizing to number of coins per auction 3) bids per coin Overall, Leu held 15 Web Auctions, starting June 2017: 2017 - 2 auctions 2018 - 4 auctions 2019 - 4 auctions 2020 - 4 auctions 2021 - 1 auction In these auctions, a total of 29.250 coins were auctioned for a total of 10.045.252 CHF, on an estimate of 3.049.685 CHF (average ratio hammer:estimate 3.29, meaning that, on average, every coin hammers at 3.29 times the estimate). These 29.250 coins attracted a total of 208.322 bids, which resulted in an overall average percentage sold coins of 96.5%. First, the correct graph showing the data corrected for inflation: This graph is obviously very similar to the graph posted here before. The cumulative price change of 2017 to 2021 is 2.04% (meaning that every CHF of 2017 is now worth 1.0204 CHF). In other words, the increase is flattened just a little bit compared to Leu's graph. Second, the data standardized for number of coins per auction. Obviously: including more coins in an auction will result in a higher gross hammer price. For this graph, I divided the hammer price (corrected for inflation) by the number of coins in the auction, and the estimated price (corrected for inflation) by the same number of coins included in auction: Now this is somewhat different! First, the average hammer price per coin stays more or les stable over time (between 256 CHF and 437 CHF), but the average estimated price per coin decreases over time (from 170 CHF in the first auction to 82 CHF in the final auction). Next, lets look at the number of bids per coin Leu might show a graph like this, illustrating the explosive interest in their coins! ... however, if we again standardize to the number of coins included in the auction: ... then, this effect is less exaggerated, with an average of 6.88 bids/coin. If we exclude the first two auctions (when Leu was the new kid on the block), the number of bids per coin was 7.19, which is not that much less than the 8.46 average bids per coin during the COVID pandemic. Finally: what would be the correct graph? In this final graph, I combine above graphs: 1) prices corrected for inflation 2) standardized for the number of coins included in the auction Thus, we get a ratio of the estimated and hammered price, corrected for inflation per coin included in the auction: What does this learns us? 1) Estimated prices remain low and even decrease over time, regardless of 'the high demand for antique coins during the pandemic'. High demand would result in higher estimated prices, yet estimates have only dropped since the first Leu auction (except for one auction, auction 4). 2) Hammer prices indeed increased, but less dramatically so than Leu wants you to believe 3) The number of bids have increased as well, but only marginally so This little exercise demonstrates how easy it is to play with data. None of these graphs are faulty, but they all present part of reality. Still, even a graph presenting data corrected for inflation and standardized for the number of coins auctioned, is incomplete: the higher demand may well be explained by higher quality collections auctioned. Ask yourself: would you auction your top-dollar collection to a company that only just started selling ancients? (As an epidemiologist, I would be hesitant to sell my coins to a company that incorrectly displays their data, though). Next, it may well be that the composition of the auction differed: e.g. Greek coins usually attract less bidding than Roman coins. A better analysis would be to standardize on quality of the coins, the type (ie. Roman, Greek, Celtic, etc.) and known provenance, amongst others. In conclusion: prices of Leu may have increased, but less dramatically than Leu wants us to believe. We really don't know what caused it.
Thanks for taking the time to play with these numbers ! I'm not sure if anything was gained from your exercise, but your diligence is commendable .
A superb analysis, thanks for the effort! It seems that auction houses should start estimating coins at $1 and bragging about their ratio over the estimate. ;-)
A parameter that is not captured in your figures, is that the quality of coins offered in their esales - and I speak for the Greek only- has been steadily decreasing over time. What I can keep out of your analysis, is that the auction house made a gross profit of 2-4mE out of the buyers' and sellers' fees in less than 4 years by mainly selling low quality coins. I think that's a colossal achievement. I refrain from commenting on how clever we (collectors) are, because no one will like it.
True, especially for the high-end "glamor" coins. Still, the base of the hobby, actually I would argue the foundation, of the hobby has always been the collectors of low cost coins. This is especially true for the growing middle class, a result of the Industrial Revolution. The growing income gave families to purse activities outside the necessities of life, including coin collecting. I think you are right pegging the date 1945 forward (the post-war boom) for the expansion of coin collecting, but I think the collecting of common coins at the time, such as Indian head pennies, buffalo nickels, often in highly circulated condition was widespread prior to 1945 as well. Getting back to the present, yes, prices seem to have increased for certain coins, again mostly for super rare or high grade coins, and there have been more modest gains for mid-range coins as well. Other coins, especially those that have flooded the market over the past few years, may offer good buying opportunities, such as the Athenian tetradrachms that seem to be washing onto the shores of auction houses and retail seller websites in considerable numbers. While the rare Starr groups and super high-grade/desirable coins have seen major hammer price jumps, the mid-range (fine to very fine) coins may pose buying opportunities for astute collectors seeking one of these coins for type. This is a classic situation of a coin saturating the market, and the market absorbing the supply over time. This has happened time and time again.
These coins come from disparate sources, but primarily from a gigantic hoard from Turkey, but also from innumerable hoards and individual finds that are on-going and never documented, for various reasons. Other sources are coins coming out of private collections. Currently, Jordan and Syria appear to be prime sources for Athenian and imitative owls, based on what little information I can glean from various sellers. Here are some coins from Syria about two years ago. They were being offered as a group.
What I'm getting out of this is that the auction house has neglected to adjust their starting prices to align more closely with market prices.
Thanks @AncientJoe. I think that's basically what auction houses are doing (that is, at least Leu). You are of course right, and I do stress this point in my final paragraph: I would be happy to perform such an analysis (I have made some attempts in the past using automated text mining algorithms), however the main problem is the broadness of the data. I was able to 'explain' about 70% of the variance in the hammer price for some coin types, including variables such as assigned grade, comments ('broken, holed, ..."), provenance, etc. Also, it's difficult to extract the data: I used data from Sixbid, but this extraction of up to ~700 coin lots.