Saturday, December 10, 2016

RM Sotheby’s Duemila Ruote Auction

Bidding Wars for Basket-Cases

RM Sotheby’s Duemila Ruote Auction in late November was unusual in many respects. The Italian Government needed to sell a seized collection of cars, boats, motorcycles, bicycles, and automobilia - and hired RM to do it. Held at the Fiera convention center outside Milan, close to 900 lots were sold over three days, and total sales exceeded €51 million. RM stated that 83 percent of the bidders were new to their registry, and over 1,000 bidders participated online.

I evaluated approximately 400 of the cars sold to understand why so many lots sold for multiples of their pre-auction estimates. I could not quantify if amateur hour bidding from the 83 percent was a factor; instead, I investigated if the condition of a car was a factor in it selling for over its pre-auction estimate. Classifying the cars into five categories according to the definitions in the Hagerty Price Guide - with an added condition (#5) for non-running basket-case cars - over 12 percent of the cars were basket-cases, none were best-in-the-world (#1), over 25 percent had obvious flaws (#4), almost 48 percent were good (#3), and close to 15 percent were local show winners (#2). While cars in the poorest condition were more likely to sell for many times their pre-auction estimate, cars in show condition often sold for prices close to their pre-auction estimate.

The scatterplot below shows the cars’ sale price on a log-scale compared to the pre-auction mid-estimate (an average of the low estimate and the high estimate), also on a log-scale. Pre-auction estimates for cars in condition category #5 were least accurate in relation to the actual sale price, which resulted in a flatter slope in the scatterplot (a regression coefficient of 0.70). Cars in excellent condition #2 had pre-auction estimates that were much more accurate, and the slope of the regression model for that group produced a line that is much closer to one (0.84). A slope of zero would mean the pre-auction estimates had no relation to the final sale price, while a slope of one would mean that the pre-auction estimates closely predicted the final sale prices.




The scatterplot will show the lot behind each point when hovering the cursor over a given point, and several outliers can be picked out in this way. Cars in poor (or worse) condition (#4 or #5) that sold for way more than their estimates include lot 567, the 1952 Aston Martin DB2 DHC Volante, which sold for €504,000 or 3.6 times its mid-estimate of €140,000. Another example is lot 265, a 1971 Maserati Ghibli 4.7 basket-case which sold for €179,200, or almost five times its mid-estimate. Cars that are potential show winners such as lot 602, the 1966 Ferrari 275 GTB 6 carb with an alloy body, sold for the auction high of €3,416,000, but that was just 30 percent more than the pre-auction mid-estimate of €2,600,000.


While the RM Duemila Ruote (two-thousand wheels) auction was unique in many ways, the strong sale prices achieved by some of the cars could be repeated at other auctions. The large percentage of new bidders may have played a role, but low pre-auction estimates on cars in poor (or worse) condition, and the subsequent bidding wars are the most likely cause.

RM Sotheby's Duemila Ruote auction catalog PDF (28 MB)


Disclaimer

JohnWileyCars provides no warranty about the content or accuracy of the content published here. JohnWileyCars shall not be liable for any loss of profit or any other damages resulting from the use of the information published here.


No comments:

Post a Comment