Horolonomics

There's a quip in the world of finance: "nobody rings a bell at the top or bottom of the market."

If you're on a trading floor, you know it is the start of the day because of the "opening bell." But, if you reach the low point for prices, there is no bell. Everyone just has to do their best to figure out if prices are finally going to start increasing again.

This is certainly true of watches. According to Chrono24's Chronopulse index of watch prices (full disclosure, I've contributed to this index), we've just begun the third year of a "bear market" for preowned watches.

The rate of decrease in prices has certainly slowed, but there hasn't been an overall turnaround in prices just yet. It is hard to say if we're at the bottom or if we're just on a gentle downslope that will continue for the near term.

Right or wrong, I sometimes use emails about discounts on preowned watches as a bellweather for market conditions. My thinking goes like this: at some point, a shop or preowned dealer spends time determining a price that will: a) earn an adequate markup while b) also ensuring that a watch doesn't occupy inventory for an intolerable length of time. A subsequent discount suggests that the watch has remained unsold for too long which, in turn, means the market is softer than the seller believed at the time they set the price.

A few days ago, I received a "sale" email from a very well-known preowned shop. I started looking through the sale prices and they seemed far more steeply discounted than I was used to seeing, particularly from this seller. But, I prefer actual calculations (rather than informal impressions) so I scraped the original and sale prices from the seller's web page. The discounts were, in fact, fairly steep.

I pulled the prices and discounts for 33 watches that were still available for purchase when I opened the "sale" email.

Each timepiece was from a well-known luxury brand (Patek, Rolex, Panerai, Vacheron, Cartier, Zenith, etc). The minimum undiscounted price was $4,000 and the maximum was $72,500 with an average price of $12,800. The distribution of prices was skewed, though (see graph). In particular, the highest priced watch was far from the rest of the sample and seemed to be an outlier. In the subsequent analysis, I leave out the outlier watch.

The average discount on the remaining watches was 24.9%.

This was, honestly, much higher than I expected. The minimum discount was 6.3% while the maximum discount was a whopping 40%. The discounts were, more or less, symmetrically distributed around the average of 25% (see chart). Dollarwise, the average discount was $2,284 with a minimum discount of $400 and a maximum discount of $4,500. Interestingly, the dollar discounts look bimodal.

I was curious to see if the level of discount varied with the original price. My casual observation has been that, at certain times certain watches at the extreme upper end of the price distribution appear on the grey market with a noticable discount. I even observed this outcome during the height of watch mania when numerous, more modestly priced, timepieces were selling for a premium on the secondary market.

Because the discount percentages fall within a specific range (between zero and one), I couldn't really use standard techniques to estimate how the discount percent varies with price. After figuring out a decent approach (for those stats obsessed out there, I used a generalized linear model with a logit link function and binomial family), the numbers suggest that this particular retailer was actually offering larger discount percentages at lower prices.

I graphed the relationship between the discount percent and the price (see figure) and also calculated the predicted discount for prices between $4000 and $12000 in increments of $2000. The discount begins at 28.9% for the lowest priced watched ($4,000) and declines to 23.5% for the highest priced watch ($12,000). In essence, the discount decreased by 1.3% per $2000 of watch value (although it decreased at a decreasing rate).

In summary, based upon the sale I observed, it does not appear that the watch market has reached bottom yet. A 24.9% average discount across a very wide price range suggests that, at least for this retailer, the market is still much softer than the retailer expected when they originally set prices. It is also interesting to note that the empirical estimates suggest the lowest priced watches are most heavily discounted. The longer run trend in watch sales seems to suggest that the more affordable segment of the watch market is "drying up" a bit, particularly as demand is diverted to wearables and smart watches. Perhaps that is part of what is happening here. The good news is that we are heading into the graduation season, which is one of the brighter spots in the annual cycle of watch sales. As a result, there is a possibility that we will see a reprieve from a softer market in the short run.

---

My book on the history of Rolex marketing is now available on Amazon! It debuted as the #1 New Release in its category. You can find it here.

You can subscribe to Horolonomics updates here.