For a while now, Paul Krugman has been tinkering with a model to simulate a Minsky moment. See, for example,
This sounds like a fun project, so I think I’ll do some brainstorming and see what happens. Comments welcome.
Finite Equity Market
One aspect I’d like to incorporate that is motivated by my experience at large asset management firms is to model the equity market as finite. Most models assume the market is infinite so that you can purchase and sell as much of any security as you like. In practice, we have size constraints. For example, when you own 10% of the outstanding market cap of any security, you cannot sell it without moving the market. Modeling a finite market will introduce liquidity. This will be important for modeling a Minsky moment.
Positions can be expressed as a fraction of the outstanding market cap.
I think we should model agents with continuous leverage, i.e. the amount of leverage is inversely proportional to market volatility.
This is motivated by both theory and recent experience. Clients expect stable or increasing returns. However, returns are commensurate with risk. Prior to the crisis, we were experiencing a period of decreasing market volatility with corresponding decreasing returns. To maintain high (and unsustainable) returns, investors began to increase leverage. This is also consistent with bank’s capital requirement being determined by value-at-risk (VaR).
To be continued…