systemic risk diagnostics, data, and policy tools

Systemic Risk Diagnostics, Data, and Policy Tools

Prepared for G20 Conference hosted by the central banks of Mexico and Turkey

Lasse H. PedersenNYU, CBS, FRIC, AQR, CEPR, and NBER

Systemic Risk Diagnostics, Data, and Policy Tools for Central Banks

Price stability:– Data required:– Policy tool:

Overall systemic risk:– Data required:– Policy tools:

Contribution to systemic risk:– Data required:– Policy tools:

Lasse H. Pedersen 2

Price stability:– Data required: inflation– Policy tool: interest rate

Overall systemic risk:– Data required: leverage and margin requirement– Policy tools: market-wide capital requirements, lending facilities

Contribution to systemic risk:– Data required: MES, SES based on returns, leverage– Policy tools: institution-specific capital required, systemic tax, mandatory systemic insurance

Main point:– Central banks have at least two monetary tools, not just one, and both are important:

• Interest rates• Managing leverage

Monitoring Overall Systemic Risk: Leverage

Source: Geanakoplos and Pedersen (2012), “Monotoring Leverage”

Monitoring leverage is “easy”: – Margin requirements (loan‐to‐value ratios) can be directly observed– Model‐free measure (in contrast to other measures of systemic risk that require complex estimation)

Monitoring leverage is monitoring systemic risk: – Monitoring leverage provides information about how risk builds up during booms as leverage rises– How crises start when leverage on new loans sharply declines

Liquidity crisis management: – Leverage data is a crucial input for crisis management and lending facilities– Helps ascertaining the state of an indebted economy in the aftermath of a leverage crisis

New vs. old leverage: – Leverage on new loans: timely measure of credit conditions– The average leverage signals the economy’s vulnerability. – The economy enters a crisis when leverage on new loans is low, and leverage on old loans is high

• A de‐leveraging event that starts a liquidity spiral• Early warning signal on the beginning of a systemic crisis

Lasse H. Pedersen 3

Monitoring Overall Systemic Risk: Early Warning from Leverage

Geanakoplos and Pedersen 4

Lasse H. Pedersen 5

Two Monetary Tools: Margin CAPM

PropositionThe equilibrium required return for any security s is:

where ψt is the leveraged agents’ Lagrange multiplier, measuring the tightness of funding constraints, xt is the fraction of constrained agents, mt

s is the margin requirement of security s, and λt is the risk premium,

1s f s s

t t t t t t tE r r x m

1M f

t t t tE r r

Sources:

- “Two Monetary Tools: Interest Rates and Haircuts,” - Ashcraft, Garleanu, and Pedersen (2010)

“- Margin-Based Asset Pricing and Deviations from the Law of One Price,” - Garleanu and Pedersen (2011)

Lasse H. Pedersen 6

Two Monetary Tools: Evidence on the Effect of TALF

Source: “Two Monetary Tools: Interest Rates and Haircuts,” Ashcraft, Garleanu, and Pedersen (2010)

Survey evidence from March 2009 on CMBS securities Demand sensitivity measured in terms of yields

– Improving funding conditions can lower required returns by several percentage points– Note that the Fed had lowered the short rate from 5% to zero and hit the zero lower bound

Lasse H. Pedersen 7

Two Monetary Tools: Interest-Rate Cuts Steepen Margin-Return Curve


Surprisingly, interest-rate cuts can increase required return for high-margin securities

Frictional Finance - Lasse H. Pedersen 8

Two Monetary Tools: Haircut Cuts can Ease Credit Frictions


Haircut cuts through central bank lending facilities alleviate funding liquidity frictions– by moving the affected securities down the haircut-return line– by flattening the whole haircut-return line as people’s funding conditions are improved

Lasse H. Pedersen 9

Two Monetary Tools: Margin-Constraint Accelerator


Funding constraints propagates business cyclesR

eal I

nves

tmen

t

Measuring the Contribution to Systemic Risk

Source: “Measuring Systemic Risk,” Acharya, Pedersen, Philippon, and Richardson (2010)

Conventional wisdom: Systemic risk =

– what would happen if bank X failed?

– or, what crucial infrastructure is operated by bank X? (triparty repo, payment system, etc.)

Our view: Systemic risk =

– Too little aggregate capital in the financial system

• Too little capital inhibits intermediation and credit provision

• A failed bank with crucial infrastructure can be taken over if there is enough capital in the system

– Example: Lehman vs. Barings

10Lasse H. Pedersen

Measuring the Contribution to Systemic Risk

Each financial institution’s contribution to systemic crisis can measured as its systemic expected shortfall (SES):

– SES = expected capital shortfall, conditional on a future crisis

A financial institution’s SES increases in:– its own leverage and risk– the system’s leverage and risk– the tail dependence between the institution and the system– the severity of the externality from a systemic crisis

Managing systemic risk: – Incentives can be aligned by imposing a tax or mandatory insurance based SES adjusted for the cost

of capital

Source: “Measuring Systemic Risk,” Acharya, Pedersen, Philippon, and Richardson (2010)

Lasse H. Pedersen 11

Measuring Systemic Risk: The Right Units for a Systemic Tax

How to regulate based on the systemic risk measure?We show that taxing based on SES implies that banks internalize externalities Taxing based on “crucial infrastructure” does not work since infrastructure crucial no matter how well capitalized

In case of tax, how to translate into right units? We show that SES is scaled in meaningful units

How to scale wrt. size of institution? Example, consider these three firms:Firm A = CitibankFirm B = 1 share of CitibankFirm C = 1 share of Citibank + $1 Trillion worth of Treasuries We show that SES taxes each correctly Other measure of systemic risk (e.g. based on “connections”) get this wrong

Same tax in dollars for A and B, or Way higher tax for C than B

How to handle if institutions merge or split up? We show that SES handles this immediately


Measuring Systemic Risk : Empirical Implementation

Empirical methodology:– We provide a very simple way of estimating SES

Institutions’ ex-ante SESs – predict their losses during the subprime crisis– with more explanatory power than measures of idiosyncratic risk

SES in the cross-section: – higher for securities dealers and brokers – every year 1963-2008– higher for larger institutions that tend to be more levered

SES in the time series:– higher during periods of macroeconomic stress, especially for securities dealers and brokers


Managing Risk Within and Across Banks

Standard measures of risk within banks:– Value at risk: Pr ( R ≤ - VaR ) = α– Expected shortfall: ES = - E( R | R ≤ - VaR )

Banks consists of several units i=1,…, I of size yi : – Return of bank is: R= ∑i yi ri

– Expected shortfall: ES = - ∑i yi E( ri | R ≤ - VaR )

Risk contribution of unit i: Marginal expected shortfall (MES)

We can re-interpret this as each bank’s contributions to the risk of overall banking system: The loss of bank i when overall banking is in trouble

We provide a more detailed economic theory for MES and SES

: |ii

i

ESMES E r R VaRy


15

Managing the Contribution to Systemic Risk

SES signals institutions likely to contribute to aggregate crises

Three approaches to limit systemic risk

1. Systemic capital requirement• Capital requirement proportional to estimated systemic risk

2. Systemic fee/tax (FDIC-style)• Fees proportional to estimated systemic risk

• Create systemic fund

3. Private/public systemic insurance

Lasse H. Pedersen

Empirical methodology

MES:– Very simple non-parametric estimation:

• find the 5% worst days for the market• compute each institution’s return on these days

– Parametric SES:

– Consider both MES and Leverage

Data: CRSP and COMPUSTAT


Stress tests: Predictive power of MES (equity)

AXPBBT BK

BAC

COF

CFITB

GSJPM

KEY

MET

MS

PNC

RF

STT

STI

USB

WFC

0.1

.2.3

.4.5

SCAP

/Tie

r1C

omm

4 5 6 7 8 9MES5 measured Oct06-Sep08

2007-08: Predictive power of MES (equity)

TROW

SOV

PBCT

BERBRK

SNVLUK

UB

CBSS

CINF

CMA

LTREV

FITBRF

MTB

WB

BEN

WFC

AT

HBAN

MMC

CNA

JPMHUM

UNP

LNC

BK

FNM

MI

MER

AGE

NCC

AFLNTRS

AXP

CB

BAC

SAF

TRV

PNC

AOC

TMK

CI

PGR

CFC

KEYLM

USB

SLM

AIG

STI

SEIC

BSC

MS

C

UNMBBT

STT

MBI

SCHW

FRE

CVHHNT

ABK

ALL

NYB

LEH

COF

WM

HIG

BRK

ETFC

ZION

AMTD

ACAS

CBH

GS

HCBK

BLK

MET

JNS

AET

FIS

WLP

PFGPRU

CG

CIT

CME

AIZ

GNW CBG

AMP

BOT

FNF

ICE

NYX

MA

WU NMX

UNH

-1-.5

0.5

Ret

urn

durin

g cr

isis

: Jul

y07

to D

ec08

0 .01 .02 .03 .04MES5 measured June06 to June07


NYU-Stern VLAB’S Risk Page

Directed by Rob Engle

We have introduced a page providing estimates of risk for the largest US and global financial firms

NYU Stern Systemic Risk Ranking: Risk is estimated both for the firm itself and for its contribution to risk in the system.

This is updated weekly/daily to allow regulators, practitioners and academics to see early warnings of system risks.

Extended to global firms: Collaboration with Universite de Lausanne and Australian Graduate School in Sydney


Conclusion

Central banks have at least two monetary tools, not just one, and both are important:– Interest rates– Managing leverage

• Overall systemic risk: – Good times: market-wide capital requirements– Crises: lending facilities

• Contribution to systemic risk: – SES, institution-specific capital required, systemic tax, mandatory systemic insurance

21

Systemic Insurance

Compulsory insurance against own losses during crisis– Payment goes to systemic fund, not the bank itself

– Insurance from government, prices from the market

• Say 5 cents from private; 95 cents from the government

• Analogy to terrorism reinsurance by the government (TRIA, 2002)

Advantages of private/public proposal– A market-based estimate of the contribution to crises and externalities

– Private sector has incentives to be forward looking

– Gives bank an incentive to be less systemic and more transparent:

• to lower their insurance payments


systemic risk diagnostics, data, and policy tools

Documents