Pricing algorithms can learn to collude with each other to raise prices
The headline here is in the context of pricing algorithms in a retail environment.
Its an interesting article with some eye-opening cocnlusions:
- two completely autonomous algorithms learned to respond to one another's behaviour
- quickly pulled the price of goods above where it would have been had either operated alone
- the algorithms leave no trace of concerted action
- They learn to collude purely by trial and error, with no prior knowledge of the environment in which they operate, without communicating with one another, and without being specifically designed or instructed to collude
Dunno what implications this might have for financial markets where its more about competition than cooperation … but, hey … dunno ….
Link to full (its only short) piece