Hyperbolic discount curves: a reply to Ainslie