04 financial crises - university of warwick€¦ · • default on their housing loans • february...

FINANCIAL CRISESStefania Paredes Fuentes

[email protected] Department of Economics, S2.121

University of Warwick

Money & Banking WESS 2016

THE FINANCIAL CRISIS

• Reinhart and Rogoff (2009) inflation crises, currency crises, sovereign crises and banking crises

• Banking crises ‘lead to the closure, merging or takeover by the public sector of one or more financial institutions’ Characteristics:(1) deep and prolonged asset price collapses(2) large and lasting adverse impacts on output and employment (3) government debt explodes

• IMF (2009) recessions associated with financial crises are more severe and long-lasting than those associated with other shocks

Stefania Paredes Fuentes Money and BankingWESS 2016

Crises are costly to the economy


• Great ModerationInflation targeting regime, calm macroeconomic environment, low unemployment levels

• levels of debt in households an banks increased dramatically- expansion of credit and debt in the financial and the household sectors - household debt/GDP from 44% in Spain and 69% in UK in 1999 to 88% and 105% in 2008 respectively

• booms in assets prices

• Central bank should only react to financial developments to the extent that they influenced forecast CPI inflation over the CB’s planning horizon


Crises are costly to the economy


• Banks are a special case for policy makers:- Economic dependence of core banking services - Contagion: Spillovers from one bank affect the whole system

• Bank bankruptcy is a special case vs other industries(?): → Expectation that govt. bails out failing banks to prevent crisis

• Thus, policy makers face a trade-off between: - Maintaining the continuity of core banking services - Avoiding moral hazard on the part of households and banks


Bank Behaviour and the Macro-Economy


• Banks do not take into account the negative externalities of its decisions (e.g. in excessive risk taking)

• There is also an information problem: The govt. may impose capital regulation (s.t. bank risks are at the socially optimal level), but this requires that risks are accurately observable.

• Banking Activities:We distinguish between 2 types of bank activity:- Retail banking: Deposit-taking, lending, mortgages etc.- Investment Banking (IB): Trading in financial products (securitised assets, derivatives etc.)Also assume: Assets are ‘marked-to-market’ , risk-neutral IBs follow a Value at Risk (VaR) business model




Excess risk taking by & excessive leverage of banks: Ignored externalities and relied on bail-outs by the govt.

Incentives of banks encouraged strategies that increased aggregate risk in the economy.

Regulators allowed banks to use their own models to calculate and report riskiness.

Widespread lack of concern for the financial system’s risk to the economy

Benign macroeconomic environment and inflation targeting encouraged the belief in low aggregate risk.


Origins - Risk


• Raghuram Rajan (2005) “Has financial development made the world riskier?,”- query the value of recent innovation in the financial markets

• Don Khon: “By allowing institutions to diversify risk, to choose their risk profiles more precisely, and to improve the management of the risks they take on, they have made institutions more robust … these development have also made the financial system more resilient and flexible - better able to absolve shocks without increasing the effects of such shocks on the real economy”

• Ben Bernanke (2006). “Banking organisations of all sizes have made substantial strides over the past two decades in their ability to measure and manage risks”


Risk

http://ideas.repec.org/a/fip/fedkpr/y2005iaugp313-369.html


• Is ‘risk’ another commodity?i.e. can standard economic tools be applied to the trading of risk?

• Greenspan’s doctrine: “The effects of trade is to allocate risk to those investors most able and willing to take risk”

• Behavioural finance: ‘irrational behaviour’: individuals chasing a dream, liking control, bias towards optimism‘rational’ people judge uncertain situations by attaching probabilities to the various outcomes and revise their probabilities in the light of new information

• Without the presumption that people act in rational self-interest the whole structure of risk evaluation fails (Markowitz model, Black Scholes model, EMH, Capital Asset Pricing Model, etc)


Risk



Risk

We are merely reminding ourselves that human decisions affecting the future, whether personal or political or economic, cannot depend on

strict mathematical expectation, since the basis for making such calculations does not exist; and that it is our innate urge to activity which makes the wheels go round, our rational selves choosing between the alternatives as best we are able, calculating where we can, but often

falling back for our motive on whim or sentiment or chance.

“The State of Long-Term Expectation” in Keynes (1936), Ch. 12


• Financial Cycle vs Business Cycle: Business cycles are based on GDP fluctuations; Financial cycles based on key financial variables (e.g. credit, house prices)



Key Features of the Financial Cycle:

1. Upswing: House prices ↑ → Household borrowing ↑ → Bank borrowing ↑ (positive feedback process)

2. Upswings often end with ↓ in house prices & a banking crisis

3. Downswing: Households & Banks deleverage (↓ indebtedness) → Banks ↑ int. rate spread, ↓ willingness to make loans

4. Housing boom reverses: Borrower households need to ↑ savings and recover from negative equity

5. 3. and 4. (balance sheet effects) imply a deeper recession

6. Public sector debt increases sharplyStefania Paredes Fuentes Money and BankingWESS 2016



Stylised Facts about the Financial Cycle:

1. Banks play a key role in the financial cycle through their lending and borrowing behaviour.

2. The housing sector is procyclical and house purchase is often financed by borrowing from banks

3. The inter-relationship between banks and housing is central to the fin. cycle; Peak of fin. cycle often followed by banking crisis.

• The BIS measure of financial cycles uses 3 variables: Private credit, Private credit-to-GDP ratio, Residential property prices

• Fluctuations in fin. cycle variables typically longer than business cycle fluctuations in output.


Financial Cycles and Business Cycles


1. Retail banks and govt-sponsored housing agencies

US Regulations: Hard to ↑ leverage to ↑ returns, so instead,↑ riskiness of asset pool to ↑ returns

∴ Made sub-prime loans and sold them (via MBSs) to eliminate own credit risks → ↓ incentive to check borrower’s credit status

2. Global banks, rating agencies & financial regulators

Europe: Lack of regulation on simple leverage; Lax treatment of ‘risk-weighted assets’

Basel II: Assets with a low risk weight do not need much or any equity to be held against it

∴ To purchase AAA-rated assets, banks can easily borrow to fund them, without equity (thus ↑ leverage)

Heavy reliance was placed on CRAs; AAA assets proliferated.


Origins - Actors


3. The shadow banking system

Bank-like entities outside the scope of banking regulation also traded heavily in securitised assets.

The rapidly growing shadow banking system further boosted aggregate risks in the regulated banking sector.

4. Bank concentration and interconnectedness

Consolidation of banks into ‘too-big-to-fail’ financial giants

Increased fin. dependence (mutual claims) between banks

Homogeneity in the financial system increases systemic risk

∴ Need for focus on Macro-prudential regulation (to manage system-wide risks), as well as traditional Micro-prudential regulation (of individual banks)


Origins - Actors


• US government promoting home ownership extended housing loans to previously credit constrained households rapid expansion of sub-prime lending

• source of aggregate demand(little attention to the potential risks of financial fragility from rising household indebtedness)

• rising house pricesincrease value of banks’ collateralincrease willingness of banks to lend without adequate checksFinancial Accelerator: More lending → House P ↑ → ↑ Collateral → ↑ Further lending

• expectations of rising house prices increased the incentive for households to borrow moretheir leverage increased


Origins - Subprime lending in the US


• Aron et al. (2012)- rise in house prices positively affects consumption in the UK and the US- the strength of this relationship increased as these economies became more financially liberalised during the Great Moderation

• a fall of house pricing can wipe out all the equity of the household- any greater fall in the house price means that they own an asset worth less than the amount they owe on it

• bubble in house prices- price of an asset rises beyond what is consistent with the fundamentals

• High cross-border banking flows: Financial boom from US sub-prime mortgages transmitted globally- Europeans banks borrowed from US money markets to buy securitised assets


Origins - Subprime lending in the US



Origins - UK banks’ leverage



Housing and Banking Booms and Busts

house mortgages - secured or collateralised loan

Household borrowing increases

Higher value of collateral

House price

booms

Purchases of housing

increase

Household borrowing

falls

Lower value of collateral

House price

declines

Purchases of housing

fall

ON THE WAY UP ON THE WAY DOWN



Housing and Banking Booms and Busts

house mortgages - secured or collateralised loan

bank borrowing increases

Stronger balance sheets

Asset pricesboom

Purchases of securitised

assets increase

bank borrowing

falls

Weaker balance sheets

Asset prices

decline

Purchases of securitise assets

fall

ON THE WAY UP ON THE WAY DOWN




Financial vs Business Cycles in the US: Fin. cycle upswings & downswings are more prolonged Peaks of the fin. cycles coincide with the onset of banking crises Preoccupation with stabilising the business cycle is problematic




Key indicators for measuring financial fluctuations: Note: real stock price growth is excluded (exhibits short-term volatility, less associated with crises)




• UK ≈ US

• Exception: Germany

Small fin. cycle amplitude:Different housing market(low home ownershiprates, no re-mortgagingbased on house prices)

� Institutions have animpact on both financialand business cycles.


• Pre-2000s factors- European regulatory regimes allowing for high leverage because of low risk weighting- incentives for risk taking provided by state guarantee for the banking system

• 2000s factors- financial innovations creating assets from US mortgages and consumer loans that could easily be traded around the world- growing dominance of ‘too big to fail’ banks- rise in importance of and trust in the ratings agencies for bank portfolio decisions- prevalence of incentives in banks for ‘search for yield’ behaviour, rewarded with high bonuses


Origins - The global upswing


• increase in oil and commodity prices

• Fed’s cycle of tightening monetary policy from late 2004

• pushed up the repayments of existing borrowers

• households began to go into negative equity

• default on their housing loans

• February 2007: index of the value of mortgage credit default swaps dropped sharply- US AIG sold insurance in the form of CDS on CDOs held by banks and pledged to repay them in full


Origins - The credit crunch: the collapse of the housing market


• Banks became increasingly dependent on short-term funding, especially the type that used mortgage-backed securities as collateral- MBS: financial product secured by a collection of mortgages and referred as securitised financial assets. They are subject to credit risk.

• Default of sub-prime borrowers and fall in house prices make these securities hard to value- lenders reluctant to accept them as collateral for loans

• differential of the interbank market rate and BoE official bank rate - risk premium attached to interbank lending- reduced supply of interbank loans- liquidity risk

• Northern Rock bank run in September 2007- BoE’s announcement of liquidity support for the bank


Origins - The seizing up of the money markets


• (1) Asset Price Bubbles

• (ii) Financial Accelerator Positive feedback process through which a change in the price of an asset affects the macroeconomy


Basic Mechanisms in Financial Crises:



Financial crisis based on house-price boom

Top: Changes inregulation or collateralrules lead to a houseprice boom.

Bottom: No direct linkbetween the rise inhouse prices and the3-equation model (ie.with AD and π );The CB just focusseson π –targeting.


• House-price based boom → Financial Crisis

• Not involving novel financial instruments

Summary:

1. Property bubble bursts, house P ↓ → Household net worth ↓ → Some households unable to service mortgages

2. Houses are repossessed by bank but sold at a loss (at a price below remaining mortgage value)

3. Losses on mortgage loans → Net worth of banks ↓

4. Sufficient exposure to falling prices → Bank asset value (mortgages) shrinks and wipes out its capital cushion → Banks become insolvent.




• IS Curve shifts rightwards from a boost in demand due to relaxed borrowing conditions

• Inflation increases and the CB tightens monetary policy

• The only link to the CB is due to changing inflation coming from the rise in loans and AD.

• No link between CB stabilisation and house price feedback process: CB’s response does not necessarily dampen the upswing of the financial cycle

• Higher int. rates for reducing inflation will dampen demand for mortgages but not necessarily cut off an asset price bubble and the financial accelerator mechanism




• Financial Accelerator

1. Credit-constrained households can borrow based on the value of their collateral, i.e. house value (= market price)

2. House prices ↑ → can borrow more (relaxed credit constraints)

3. Household borrowing ↑ (assume households are borrowed up to the limit set by credit constraints)

4. Borrowing is used for consumption and buying more housing → IS curve shifts rightwards

5. Increased demand for housing pushes up prices further ; the financial accelerator process begins again at step 1.




• Financial Accelerator

• Financial Accelerator (exists when there are credit constraints): Positive feedback process where P ↑ → Credit constraints ↓ → Demand ↑ → P ↑ → …

• Under no credit constraints: House P ↑ → Temporary shock to permanent income → Small effect on demand → no positive feedback process

• The Fin. Accelerator does not have to rest on bubble behaviour: It is driven by credit constraints and collateral effects.

• Fin. Accelerator : Demand shifts due to ∆P that increase wealth; Bubbles: Demand shifts on expectations of a rise in future price (self-fulfilling expectations)

• Both mechanisms can interact with each other: Bubble bursts → Fin. Accelerator amplifies and propagates the shock.




• The housing feedback process and the 3-equation model:

• We use insights from asset bubbles and the financial accelerator to describe a housing positive feedback process.

• The 3-equation model is used to show the relationship between: the house-price driven financial cycle, business cycle and the CB.

• Start of a house-price feedback process:

• Exogenous rise in house prices, or a change in banking regulation (increase in loan-to-value, LTV ratio; relaxing collateral rules)

• LTV ratio ↑ → Able to receive larger loans → Mortgage demand ↑

• Collateral rules loosened → Easier to borrow → Borrowing for consumption and housing ↑

• A bubble may develop: Mortgage demand ↑ → Prices ↑ → Future prices expected to rise further → …




No restriction to supply

• A rise in demand increases prices

• Supply is perfectly inelastic in the short-run, fixed at a given level

• Supply responds later to the rise in prices by increasing to a higher level

• Prices revert to long run equilibrium

• If steep rises in price is observed, a bubble process is likely (house prices should eventually fall, but do not)

Binding supply constraint

• A rise in demand increases prices

• There is no supply response, so no pressure for prices to fall

• House prices can rise persistently without a bubble, e.g. due to rising demand (fundamentals).

• Inelastic Supply Curve: Harder to distinguish a bubble from a price movement based on fundamentals




- Crises driven by bank leverage are newer vs. housing-based crises. - Dramatic increase in UK bank leverage prior to the financial crisis: (from 20 to 48 between 2000 and 2008)


The Bank Leverage-centred Process


1. Investment bank behaviour and leverage:• Risk neutral investment banks (IBs) and risk averse saver households both invest in risky

securities:• Upswing: Perceived risks ↓ → Demand for securities ↑ → Price ↑• Risk averse savers cut demand but IBs do not → Financial assets are transferred from

savers to IBs.• IBs are willing to hold risky assets up to the amt. that can be borrowed from savers.• Savers are willing to lend so long as loans to IBs are risk free• ∴ Fall in risk → IB leverage ↑

2. Second upswing: Asset price ↑ → IB capital gain → Larger capital cushion → Savers willing to lend more to IBs → IB leverage ↑

3. IBs now hold large volume of risky assets: If perceived risks ↑ → Asset price ↓ → Large capital loss → IB Solvency risk ↑






IB Leverage can increase dramatically from a fall in risk. Crisis: ‘Bad news’ that worsens risk beliefs → Asset prices ↓ →

Large capital losses wipes out equity → IB bankrupts.



The Bank Leverage-centred Process: The upswing

Starting from the bottom: (I) 3 factors reduce risk: Macro stabilisation Rising house prices Financial innovation (tranching;

pooling risk) (IIa) Lower risk → savers willing to lend more (IIb) Lower risk → asset demand ↑ → price ↑ → IB’s asset and equity value ↑ (IIa) + (IIb): IB can borrow more to buy more assets → IB leverage ↑



The Bank Leverage-centred Process: the downswingStarting from the mid panel:

(I) House prices start falling

(II) The feedback process is now in reverse.

(III) Retail banks, IBs, households deleverage (reduce debts) → AD ↓

The paradox of credibility:Benign period in the economy sows the seeds for the subsequent crisis.

Here, lower macro risk from inflation targeting etc. sets in place the conditions for the financial crisis.


Balance sheet recession & the Financial Accelerator :• Pre-crash: 2 household types: Savers (consume following PIH), Borrowers (credit-

constrained); Output is at 𝑦𝑒.• Crash: House and securitised asset prices fall, then:

• Retail banks reduce LTV ratio → AD ↓ • Value of housing collateral falls → Consumption loans ↓ → AD ↓• Balance sheet effect:

“Banks call in loans → Borrowers repay loans by cutting C → savers’ wealth rises from repayment → PIH: savers’ C increases by small amt.” ∴ AD ↓ as fall in borrowers’ C > rise in savers’ C.

• Rebuilding target wealth after fall in house & asset prices:Households save to rebuild wealth → AD ↓ (slows down recovery)




• The bursting of credit-fuelled asset bubbles → B/Sheet recession.• Negative impact on AD reduced if savers encouraged to spend.• But the Zero Lower Bound limits expansionary monetary policy• ∴ Threat of a deflationary trap (↓ 𝑝, ↓ 𝑦) in which the paradox of thrift applies

(↑ 𝑠𝑎𝑣𝑖𝑛𝑔𝑠 → ↓ 𝐴𝐷 → ↓ 𝑦)• Policy Implications of a balance sheet recession:

• Fiscal stimulus can be used to boost AD.• The increase in govt. debt can be repaid once the ‘deleveraging process’ is

over.• The balance sheet recession also dampens the multiplier, so the burden on

fiscal policy further increases.





2007-8



2007-8

GDP growth



2007-8 GDP growth



2007-8GDP growth



2007-8

• Bear Sterns was rescued in March 2008

• effects on the financial market bigger than expected

• policy interest rate had a weak effect on the lending rate• double spread in lending rates

- money markets required a risk premium on loans to banks - mortgages rates increased - breakdown of a key transmission mechanism of monetary policy

• conventional monetary policy loses its effectiveness



Policies after the 2008’s crisis



Post-crisis: Stabilisation Policy

• Fiscal stimulus coordination across countries in this crisis:• Stimulus leaks abroad through imports → Incentive to “free-ride” on other’s stimulus → Boost

𝑦 with no fiscal costs

• Under no coordination, countries do not internalise the positive externalities → Suboptimal stimulus amount

• FP also led to large deteriorations in fiscal balances → Constraints on govt. borrowing (e.g. Ireland)

• By economic theory FP is more effective when:• MP supports FP by preventing rising int. rates.• FP is a temporary rise in govt. spending; First round effect on AD is still not dampened by the

rise in taxes to fund 𝐺.

• Govt. borrows to fund spending; Same logic as above.• Tax cuts lead to higher C by credit-constrained households.




• Austerity in the post-crisis recession• Fiscal stimulus was followed by discretionary contractionary

FP in many countries before they had exited recession.• Aim: Reduce debt to GDP ratio, but there are issues• Domestic coordination failure: Deleveraging banks,

households and govt. → Balance sheet recession• International coordination failure: Countries save more,

contractionary policy spillovers abroad → AD ↓




• De Long and Summers (2012): Fiscal stimulus is self-financing (benefits from future growth > upfront fiscal costs)

• Possibility of hysteresis suggests caution towards premature fiscal consolidation in a depressed economy.

• Paradox of thrift: Attempts to save more under a negative output gap → ↓ AD & ↓ Output.

• The effect of austerity is ‘state and time’ dependent, so one can also learn about its effects from historical episodes.

• E.g. UK in the 1930s: High growth following fiscal austerity. Arguments:• Expansionary MP, exchange rate devaluation• ↑ Business profits and hence investment due to depressed real wages; Private

sector house building etc.




• Fixing banks first may mean less govt. debt later : • Quick return to well-functioning banks → shortened

recession → return to normalised economic activity & macro policy.

• Borio (2012): Increase in govt. debt is front-loaded, subsequently followed by smaller fiscal deficits


• Banks can take on more risk than socially optimal (externality)• The financial sector amplifies and propagates shocks in the economy via the

financial accelerator and asset bubbles. • The house price and financial asset price feedback processes cause IB leverage to

increase in a low risk environment.• Low perceived risks that preceded the financial crisis are due to: ↓ macro

volatility, ↑ house prices and financial innovation• Value-at-Risk behaviour leads to ↑ leverage when risks fall.• Rising house and asset prices do not dampen fin. cycle upswings due to the

above positive feedback processes.• Balance sheet recessions are slower to recover from due to the deleveraging

process by households, firms and banks.


Summary


• Pre-crisis fin. upswing due to the financial accelerator led to high bank leverage and household borrowing.

• The securitisation and trading of (sub-prime) mortgages as well as the involvement of ‘too-big-to-fail’ institutions greatly increased systemic risk.

• Downturn after the collapse in house prices and the Lehman Brothers was stabilised by swift FP and MP.

• However, austerity was implemented before deleveraging had ended. Lengthy recessions followed.

• Pre-crisis policy of π-targeting & light financial regulation may be inadequate


Summary

SOME READINGS

• Carlin & Soskice, Chapter 6, 7 & 13• Aron et al (2012). Credit, housing collateral, and consumption: Evidence from Japan,

the UK and the US. Review of Income and Wealth 58(3):397-423• Crafts N. & P. Fearon (2010). Lessons from the 1930s Great Depression. Oxford

Review of Economic Policy 26(3):285-317• Shin, H. S. (2012) Global banking glut and loan risk premium. IMF Economic

Review, 60(2)155-92

04 financial crises - university of warwick€¦ · • default on their housing loans • february...

Documents