On the Macroeconomics of Risk Intolerance · On the Macroeconomics of Risk Intolerance Author:...
Transcript of On the Macroeconomics of Risk Intolerance · On the Macroeconomics of Risk Intolerance Author:...
On the Macroeconomics of Risk Intolerance
Ricardo Caballero
MIT and NBER
BIS, Summer 2017
Caballero (MIT and NBER) Risk Intolerance BIS, Summer 2017 1 / 25
Background papers
1 Caballero, R.J., E. Farhi, and P.O. Gourinchas, “Rents, TechnicalChange, and Risk Premia. Accounting for Secular Trends in InterestRates, Returns on Capital, Earning Yields, and Factor Shares,” inAmerican Economic Review, Papers and Proceedings, 2017, 107(5):614-620
2 Caballero, R.J., E. Farhi, and P.O. Gourinchas, “The Safe AssetShortage Conundrum,” forthcoming in Journal of EconomicPerspectives, Vol 31(3), Summer 2017
3 Caballero, R.J., and A. Simsek, “A Risk-centric Model of DemandRecessions and Macroprudential Regulation,”MIT Mimeo, May 2017
Caballero (MIT and NBER) Risk Intolerance BIS, Summer 2017 2 / 25
Real (Riskfree) Interest Rates (U.S.)
1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 2010 2013 20166
4
2
0
2
4
6
8
10
12
Caballero (MIT and NBER) Risk Intolerance BIS, Summer 2017 3 / 25
Real Return on Capital (U.S.)
1981 1984 1987 1990 1993 1996 1999 2002 2005 2008 2011 20143.5
4
4.5
5
5.5
6
6.5
7
7.5
8
8.5
Caballero (MIT and NBER) Risk Intolerance BIS, Summer 2017 4 / 25
Labor Share (U.S.)
1981 1984 1987 1990 1993 1996 1999 2002 2005 2008 2011 20140.6
0.62
0.64
0.66
0.68
0.7
0.72
Caballero (MIT and NBER) Risk Intolerance BIS, Summer 2017 5 / 25
Risk Intolerance
Note: unlevered premium
Caballero (MIT and NBER) Risk Intolerance BIS, Summer 2017 6 / 25
Earnings Yield (S&P 500)
1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 2010 2013 20160.02
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
Caballero (MIT and NBER) Risk Intolerance BIS, Summer 2017 7 / 25
Equity Risk Premium (U.S.)
1980 1985 1990 1995 2000 2005 2010 20150
0.05
0.1
0.15
0.2
0.25
Caballero (MIT and NBER) Risk Intolerance BIS, Summer 2017 8 / 25
Equity Risk Premium (Global)
Caballero (MIT and NBER) Risk Intolerance BIS, Summer 2017 9 / 25
Risk Intolerance
1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 20170
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
Important caveat: Evidence of reach-for-yield within asset classes(heterogeneity)
Caballero (MIT and NBER) Risk Intolerance BIS, Summer 2017 10 / 25
The Safe Assets Shortage Perspective
Steady increase in demand for store of value and “safe”assets
Demographics, international reserves accumulation, regulation...
What is a safe asset?
Depends on time and context: Individual and collective fears andconcerns, on coordination and liquidity, and so on...Operational definition: A simple debt instrument that is expected topreserve its value during adverse systemic events
The supply of these assets is not keeping up with demand
The financial system reacted to that gap (financially engineered) andmasked the underlying trend for a while, but it didn’t end up well(micro vs macro safe assets)
Caballero (MIT and NBER) Risk Intolerance BIS, Summer 2017 11 / 25
The Safe Assets Shortage Perspective
Caballero (MIT and NBER) Risk Intolerance BIS, Summer 2017 12 / 25
The Safe Assets Shortage Perspective
1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 20170
0.02
0.04
0.06
0.08
0.1
0.12
Caballero (MIT and NBER) Risk Intolerance BIS, Summer 2017 13 / 25
Consequences
Early phases: Mostly “benign”
Overtly benign: Lower cost of capitalOvertly concerning: Global imbalances and the Greenspan conundrumCovertly dangerous: Financial engineering to manufacture safe tranchesfrom subprime inputs (... and partially held by levered institutions)
Post 2008: The Safety Trap
With downward rigidities in rates adjustment (ZLB, reversal rate, etc.),goods markets takes the hitA stubborn form of liquidity trap, less responsive to standard (generic)wealth boosting macro policies
Caballero (MIT and NBER) Risk Intolerance BIS, Summer 2017 14 / 25
The Medium Term
Obviously there are cycles, waves of euphoria won’t go away, and soon.... but the secular forces seem robust
Moreover, this is a global, not a U.S. phenomenon
A country with an acute shortage of safe assets spreads its downwardpressure on rates all around
Incentives aside, this is mostly good news for the recipient country ifinterest rates are flexibleNot so, if against “ZLB”
We may well be in a recurrent global safety traps environment
Caballero (MIT and NBER) Risk Intolerance BIS, Summer 2017 15 / 25
Scape Valves
The main market mechanism to restore equilibrium in a safety trap isan increase in the valuation of safe assets
Interest rate drop is the first response, but not much space thereNext is an appreciation of the currency in which these assets aredenominated, primarily USD – paradox of the reserve currency
Issuance of public debt
If asset producer economies grow less than asset demander economies,we get a modern “Triffl in dilemma"
Risk of fiscal fragility. In this environment the main concern is theappearance of credible substitutes (not too worrisome for the U.S.... atthe moment)
The value of the new safe debt created is distinct from the use of thesefunds (cheaply funded fiscal expansion). QE swapping risky for safeassets does it too (while Operation Twist doesn’t)Public good dimension of safe asset issuance
Caballero (MIT and NBER) Risk Intolerance BIS, Summer 2017 16 / 25
Scape Valves
Private substitutes.
Endogenous risk reduction via technology/sector selection and liabilitymanagementFinancial system (to the extent that it doesn’t consume too many safeassets in the process). Needs public insurance overlay (micro vs macroinsurance)
Reducing the (net) demand for safe assets
International reserves
EM: Self-insurance. Partially replace with multilateral insurancearrangementsDM: QE makes sense when absorbing risk, otherwise it can becounterproductiveRegulatory framework
Caballero (MIT and NBER) Risk Intolerance BIS, Summer 2017 17 / 25
A Macroeconomic Framework for the Times (Risk-centric)
A productive capacity (and its expansion) generates output and risks,both of which need to be absorbed by economic agentsIn principle, we could have output- and risk-gapsIn most macro (unlike finance) modeling the emphasis is on theformer
Often the risk side is ignored (log-linearizations)Or it remains in the background (discount factor)
Main “big picture” goal of paper with Alp Simsek: To provide amacro model to study both gaps jointly, but inverting the hierarchy
Caballero (MIT and NBER) Risk Intolerance BIS, Summer 2017 18 / 25
What do we do?
AD/AS:
Aggregate supply (potential output): Stochastic endogenous growthmodel with investment costs and (implicit) NK-structureAggregate demand: It may become an additional constraint toproducersInterest rate has downward rigidity (ZLB or others)
Two shocks: productivity (diffusion) and ERP (Poisson)Heterogenous beliefs (about ERP shocks) and speculationPolicy: r-policy, FG, macroprudential
Caballero (MIT and NBER) Risk Intolerance BIS, Summer 2017 19 / 25
What do we find?
Cyclical:
In the absence of r -friction: Productivity shocks only affect growth,while ERP shocks are absorbed one-for-one by r−adjustmentWith r ZLB:
when (implied) vol spikes, asset prices drop to restore equilibrium inrisk markets, and drag down aggregate demand,...which drags downasset prices, and so onoptimism (about exit from high vol state) is key to limit the fall in ADwith heterogeneous beliefs, share of wealth owned by optimists is a keystate variablespeculation is destabilizing (economy is effectively extrapolative)volatility rises endogenously and feeds back into asset prices andaggregate demand
Caballero (MIT and NBER) Risk Intolerance BIS, Summer 2017 20 / 25
Policy:
Interest rate policy is highly effective while it lasts
it controls the Sharpe ratio, leaving asset prices to equilibrate thegoods market
Forward Guidance is also effective, but its robustness to deviationsfrom "RE" drops with pessimismMacroprudential policy makes everyone better off (evaluated undertheir own beliefs, or belief-neutral criterion)
Goal is to provide additional insurance to high valuation investors tointernalize aggregate demand externalitiesIt is naturally procyclical as the negative effect on AD can be offsetwith r -policy during the boom but not during deep recessions
Volatility stabilization measures are also powerful
Caballero (MIT and NBER) Risk Intolerance BIS, Summer 2017 21 / 25
The Basic Mechanism
Goods Markets Eq
(Ak) η = ρQk + ι(Q)k
Monetary Policy
Q ≤ Q∗, r ft ,s ≥ 0, with complementary slackness
Risk Markets Eq.
σs =ρ+ g(Qs ) + λs
(1− Qs
Q∗
)− r fs
σs
Speculation leads to sharper drops in RM; Interest rate policy, FG,macropru
Caballero (MIT and NBER) Risk Intolerance BIS, Summer 2017 22 / 25
Caballero (MIT and NBER) Risk Intolerance BIS, Summer 2017 23 / 25
Final Remarks
The world economy experienced a prolong period of“risk-intolerance,”which is likely to remain around the(macroeconomic) corner
Risk-intolerance has strained the safe-assets markets.
The natural market solutions to the problem are not encouraging(appreciations, financial instability,...)There is a need for policy support in the creation and use of safe assets
Caballero (MIT and NBER) Risk Intolerance BIS, Summer 2017 24 / 25
Final Remarks
High frequency macroeconomics is also affected by this environment
Strong feedbacks between risk-markets, volatility, and the output-gap
The key tension is that asset prices have the dual role of equilibratingfinancial markets and supporting aggregate demand
Interest rate policy works by taking over the role of equilibratingfinancial markets, which then leaves asset prices free to balance thegoods marketsAt the ZLB the dual role problem reemerges and asset prices are drivenprimarily by financial markets considerations...
triggering a perverse feedback between asset prices and ADwhich can only be stopped by hope (optimism about recovery)
(Procyclical) Macroprudential regulation reduces the gap between theasset prices that equilibrate the financial and goods markets at theZLB... by keeping high valuation investors well capitalized
Caballero (MIT and NBER) Risk Intolerance BIS, Summer 2017 25 / 25