I was reading the thread over on NK about MTH All-Door Boxcars. A poster claims that
Brad's Trains (bmartz-1991) is buying stuff from Trainz dot com and then relisting it (& letting it sit) on ebay for twice what he paid Trainz. I think the NK poster is correct -- all the 2-rail locomotives bmartz-1991 currently has on ebay are models I think I recall recently seeing on trainz's site.
This strikes me as a really weird business model -- bmartz first buys stuff from Trainz, then has to create listings on ebay for the same items, then waits for something to sell, and then has to pack it up and ship it. Bmartz then gives ~25% of the selling price to ebay. Roughly ~50% to ~55% of the selling price was spent buying the item from Trainz, so bmartz is doing a lot of work just to earn a maximum of 20-25% on low-dollar items.... and that's not even considering the cost of the capital bmartz has tied up in bmartz's inventory.
Meanwhile the S&P is up 20.54% year-to-date.... bmartz would be ahead if they had just deposited their capital in a low-cost index fun and they would not have had to do any work to have earned that return.
I also wonder if there might be an algorithm feedback problem here for bmartz. TrainZ seems to price a lot of their stuff automatically using a database of past sales; I assume their pricing database uses data from their own website and data scraped from ebay. I'm thinking that each time bmartz resells a Trainz item on ebay, bmartz is causing the prices in Trainz's database to artificially inflate.
For example, assume Trainz's database has the prices of 9 individual sales of a particular item, let's say a Weaver Lehigh Valley RS3. The first 8 sales have an average selling price of $100. The 9th sale is when TrainZ uses their price database to sel one to Bmartz for $100.... Bmartz then (eventually) sells it on ebay for twice as much ($200) and Trainz adds that sale to their database. Trainz's software automatically crunches the numbers, and the next LV RS3 they sell will be offered for $110.