Working Paper: CEPR ID: DP14627
Authors: Ricardo Caballero; Alp Simsek
Abstract: In this paper we: (i) provide a model of the endogenous risk intolerance and severe asset price and aggregate demand contractions following an adverse real (non-financial) shock; and (ii) demonstrate the effectiveness of Large Scale Asset Purchases (LSAPs) in addressing these contractions. The key mechanism stems from heterogeneous risk tolerance: as a recessionary shock hits the economy and brings down asset prices, risk-tolerant agents' wealth share declines and their leverage rises endogenously. This reduces the market's risk tolerance and generates downward pressure on asset prices and aggregate demand. When monetary policy is unconstrained, it can offset the decline in risk tolerance with an interest rate cut that boosts the market's Sharpe ratio. However, if the interest rate policy is constrained, new contractionary feedbacks arise: recessionary shocks lead to further asset price and output drops, which feed the risk-off episode and trigger a downward loop. In this context, LSAPs improve asset prices and aggregate demand by transferring risk to the government's balance sheet, which reduces the market's required Sharpe ratio and reverses the contractionary feedbacks. Quantitatively, we show that aggregate shocks and LSAPs have large impacts on asset prices when the model is calibrated to fit the inelastic demand for aggregate assets uncovered in recent literature. We also show that heterogeneity in risk tolerance explains part of the demand inelasticity in normal times, and further reduces the elasticity after a recessionary shock. The Covid-19 shock and the large response by all major central banks provide a vivid illustration of the environment we seek to capture.
Keywords: risk intolerance; leverage; asset price spirals; aggregate supply and demand; recessions; conventional and unconventional monetary policy; multiple equilibria; LSAPs; asset demand inelasticity; COVID-19
JEL Codes: E30; E40; E50; G01; G10; G20
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| recessionary shock (E32) | decline in asset prices (G19) |
| decline in asset prices (G19) | decrease in wealth share of risk-tolerant agents (banks) (G21) |
| decrease in wealth share of risk-tolerant agents (banks) (G21) | increase in leverage (G32) |
| increase in leverage (G32) | reduce overall market risk tolerance (G40) |
| reduce overall market risk tolerance (G40) | increase downward pressure on asset prices (E44) |
| reduce overall market risk tolerance (G40) | increase downward pressure on aggregate demand (E00) |
| interest rate cuts (when monetary policy is unconstrained) (E52) | mitigate effects of recessionary shock (E65) |
| interest rate cuts (when monetary policy is unconstrained) (E52) | improve market's Sharpe ratio (G19) |
| constrained interest rate policy (E43) | new contractionary feedback mechanisms (E61) |
| new contractionary feedback mechanisms (E61) | additional declines in asset prices (G19) |
| new contractionary feedback mechanisms (E61) | additional declines in output (E23) |
| Large Scale Asset Purchases (LSAPs) (E44) | improve asset prices (G19) |
| Large Scale Asset Purchases (LSAPs) (E44) | improve aggregate demand (E00) |
| transfer of risk from market to government's balance sheet (via LSAPs) (H81) | reduce required Sharpe ratio (G40) |
| reduce required Sharpe ratio (G40) | reverse contractionary feedbacks (E62) |