Dualing Auctions
We have been discussing various types of auctions for a while now in class, but in all of the examples we have studied, it has always been one seller and multiple buyers. Sometimes the seller is selling multiple items, sometimes only one, but it’s always been one seller. In this short blog article by Jensen Loke, they bring up an example where they compare the market clearing strategies of Uber and Lyft to that of a Chinese company Didi Dache.
Uber and Lyft algorithmically generate a price based on demand and gives that to both the rider and the driver, whereas Didi Dache gives riders the ability to tip varying amounts depending on their own value for the ride based on their probability of getting the ride for that price. This means that they are competing with other riders (who’s value they don’t know) for a ride. Additionally, drivers are now offered different prices depending on how long they wait. The longer they wait, the more likely riders will continue to outbid each other, but the probability also increases that another driver will agree to the current price. Thus drivers are also competing in an auction for the riders.
In this type of situations all parties are on both sides of an auction: they are all competing both to receive an object, a ride & payment, and to supply an objective, payment & a ride, respectively. In the article they presented game theory tables to construct Nash equilibria for each auction, and they mostly analyze the two situations separately. I am still curious, however, if you can gain any additional information by analyzing it as one system, rather than two separate systems.
Blog Article:
https://link.medium.com/Uo4qNNMPMkb