> the time value of money also decreased over time, as the quickening of the pace of technological advances meant that lenders could expect more returns over a

> the time value of money also decreased over time, as the quickening of the pace of technological advances meant that lenders could expect more returns over a shorter amount of time. Isn't this backwards? In a world where money has no time-value, a world where a dollar today is the same as a dollar tomorrow, and the same as a dollar five years from now, then, if there's no default risk, you can pay me back the same dollar I gave you, without interest. Zero interest rates. In a world where everything changes rapidly and I might need money now, I'm going to demand a lot of interest before I'll lend you my cash. Because I'm incurring a lot of opportunity cost by parting with it. If this were a reinforcement learning problem, we might specify a discount factor. If that discount factor were near one, then we'd have a long time horizon. If it were closer to zero, then we'd have a short time horizon. I'd think that low interest rates would go together with a static, unchanging environment, in which money has very little time-value. I do agree, however, with your earlier sentence: > As the world became more stable over time, the default risk also lessened over time