Shadow Finance, Order flows and Cream Skimming Patrick Bolton Tano - - PowerPoint PPT Presentation

Shadow Finance, Order−flows and Cream Skimming

Patrick Bolton Tano Santos Jose Scheinkman Columbia University

Gerzensee − July, 2019

An overview of the research agenda

• In the period preceding the crisis

– The US financial sector accounted for an increasing fraction of corporate profits – Strong growth in compensation in the financial services industry – Finance as a favorite occupation – Deterioration of origination incentives – Increase in balance sheet size and leverage amongst some financial intermedi- aries (IBs and broker-dealers) – Growth of the “originate and distribute” model: Shadow finance

• This research agenda: Models that are aimed to shed light on these facts

The papers

• Theory

– “Savings Gluts and Financial Fragility” (with P. Bolton and J. Scheinkman) – “Outside and Inside Liquidity” (with P. Bolton and J. Scheinkman) – “Cream Skimming in Financial Markets” (with P. Bolton and J. Scheinkman) – “Shadow Finance” (with P. Bolton and J. Scheinkman)

• The political economy of liquidity cycles

– “Antes del Diluvio: The spanish banking system in the first decade of the Euro” – “El Diluvio: The Spanish Banking Crisis, 2008−2012” – “Political Credit Cycles: The Case of the Eurozone” (with J. Fdez-Villaverde and L. Garicano) – “Credit Booms: Implications for the Public and the Private Sector,”

The deeper questions

• Normative
• 1. What does the financial industry add to the real economy?
• 2. What is the optimal organization of financial markets and what do we mean

by optimal?

• Positive
• 1. How does the financial industry originate and distribute risk across different

sectors of the economy and how much capital is needed to perform these activities?

• 2. How does the information get produced and used in financial markets? Is there

“too much” information produced in financial markets?

• 3. How much talent is required in the financial industry?

Three ideas

• 1. Cream skimming
• a. Interaction between “private” and “public” markets
• b. Inefficient information production and “excessive rents”
• c. Effect of IT improvements on financial markets
• d. Private markets and inequality
• 2. Liquidity
• a. Origination incentives
• b. Balance sheet size and leverage
• 3. The political economy of credit booms
• Liquidity and structural reforms

.

Cream Skimming

Originators: 1 Incentives: e Uninformed investors M

Originators: 1 exh + (1 − e) xl Fraction of good projects e Uninformed investors M p = exh + (1 − e) xl

Originators: 1 Fraction of good projects e Informed investors: N

Assets Liabilities

Uninformed investors M

Originators: 1 Fraction of good projects e Informed investors: N

Assets Liabilities

Uninformed investors M

Originators: 1

e(1−m)xh 1−m(e+α(1−e))

Fraction of good projects e gxh pd Informed investors: N

Assets Liabilities

qpd K Uninformed investors M p =

e(1−m)xh 1−m(e+α(1−e))

m =

Nq e+(1−e)α

g ≡ prob (xh|σh) =

e e+(1−e)α

Originators: 1

e(1−m)xh 1−m(e+α(1−e))

Fraction of good projects e gxh pd Informed investors: N

Assets Liabilities

qpd K Uninformed investors M p =

e(1−m)xh 1−m(e+α(1−e))

m =

Nq e+(1−e)α

g ≡ prob (xh|σh) =

e e+(1−e)α

ϕ (i, c) for i ∈ M

Cream skimming - The main mechanism

• Bargaining:

pd ≡ arg maxs

• (gxh − s)1−κ (s − p)κ
• ⇒

pd = κgxh + (1 − κ) p

• Equilibrium payoff of the originator (α = 0)

U (N) = em (N) pd (N) + (1 − em (N)) p (N)

• Equilibrium payoff of the informed investor i

V (i, θ|N) = −ϕ (i, c) + (1 − κ) p (N) ⇒ VN > 0 and VNN > 0

• Measure of informed dealers:
• 1. Optimal measure of informed investors Nopt. = 0.
• 2. Equilibrium measure of informed investors: If ϕ (0, c) = ϕ′ = 0 then N ∗ > 0

N U, V V (i, c|N) U (N)

.

Cream skimming - Implications

• 1. The size of of the “informed segment” of the market is too large
• In this version of the model there is no social value to information: Information

is pure rent extraction from originators

• The marginal incentive to become informed increases with the measure of in-

formed agents N ∗ – Assumptions on ϕ (·, c)

• 2. Drop in IT costs c, shifts the curve ϕ(·, c) inwards and thus results in
• a. More misallocation of talent
• b. More rents flow from originators to informed investors

Cream skimming - Implications: Endogenous origination

• Consider now the case of endogenous quality origination

U (N) = −ψ (e, θ) + em (N) pd (N) + (1 − em (N)) p (N)

• Then

eθ < 0 and Nθ > 0

• Efficiency

– The size of the informed sector, N ∗, is too large precisely when the social value

• f information is at its lowest (when θ is high)

– The size of the informed sector, N ∗, is too small precisely when the social value of information is at its highest (when θ is low)

. Liquidity

Figure 1: USA: Current account as a percentage of GDP (right axis) and total US household mortgage liabilities per household (deviations from trend; left axis). Sample: 1960 − 2018. Source: Federal Reserve Bank of St. Louis

• 7.0
• 6.0
• 5.0
• 4.0
• 3.0
• 2.0
• 1.0

0.0 1.0 2.0

• 3000.0
• 2000.0
• 1000.0

0.0 1000.0 2000.0 3000.0 4000.0 5000.0 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 Mortgage per HH (deviation from trend; left) Current account as % GDP (right)

0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00 20.00

• 200.0

0.0 200.0 400.0 600.0 800.0 1000.0 1997Q1 1997Q4 1998Q3 1999Q2 2000Q1 2000Q4 2001Q3 2002Q2 2003Q1 2003Q4 2004Q3 2005Q2 2006Q1 2006Q4 2007Q3 2008Q2 2009Q1 2009Q4 2010Q3 2011Q2 2012Q1 2012Q4 2013Q3 2014Q2 2015Q1 2015Q4 2016Q3 Corporate savings (left) Yield (right)

Figure 2: Corporate savings in billions of US$ (left axis) and BofA Merrill Lynch US High Yield Option-Adjusted Spread in % (right axis). Quarterly:1997Q1-

• 2017Q1. Not seasonally adjusted. Data Source: Federal Reserve Bank of St. Louis

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007

Trillion pounds Total Assets Risk-Weighted Assets

Barclays (1992 – 2007) Société Générale (1999 – 2007)

0.0 0.2 0.4 0.6 0.8 1.0 1.2

1999 2000 2001 2002 2003 2004 2005 2006 2007

Trillion Euros Total Assets Risk- Weighted Assets

Figure 4. Total assets and risk-weighted assets of Barclays and Soci´ et´ e G´ en´ erale (Source: Bankscope)

Figure 3: Securities broker dealer: Leverage (Total assets over equity) and Total Assets. Quarterly data: 1980Q1-2012Q4

5 10 15 20 25 500000 1000000 1500000 2000000 2500000 3000000 3500000 1980Q1 1980Q4 1981Q3 1982Q2 1983Q1 1983Q4 1984Q3 1985Q2 1986Q1 1986Q4 1987Q3 1988Q2 1989Q1 1989Q4 1990Q3 1991Q2 1992Q1 1992Q4 1993Q3 1994Q2 1995Q1 1995Q4 1996Q3 1997Q2 1998Q1 1998Q4 1999Q3 2000Q2 2001Q1 2001Q4 2002Q3 2003Q2 2004Q1 2004Q4 2005Q3 2006Q2 2007Q1 2007Q4 2008Q3 2009Q2 2010Q1 2010Q4 2011Q3 2012Q2 Total Assets Leverage

.

Assets Liabilities qpd D K Assets Liabilities qpd D K Assets Liabilities qpd D K Assets Liabilities qpd D K Model 1 Model 2

Panel A Panel B

0.05 0.1 0.15 0.2 0.25 1 2 3 4 5 6 7 8 Quarter Since Origination

2007 2006 2005 2000 2004 2001 2002 2003

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 2000Q1 2000Q3 2001Q1 2001Q3 2002Q1 2002Q3 2003Q1 2003Q3 2004Q1 2004Q3 2005Q1 2005Q3 2006Q1 2006Q3 2007Q1

Figure 4: Panel A: Fraction of loans that were originated with no/low/reduced documentation among non-agency mortgages with known documentation

status in the Loan Performance (LP) database. Panel B: Average cumulative default paths for the non-agency securitized loans by year of their origination. The cumulative delinquency rates are of all privately securitized residential mortgages calculated from loan level, monthly Corelogic-Blackbox data. A loan is defined as delinquent if it ever becomes 60 days past due and is considered delinquent thereafter. The graph is separated by year of mortgage origination, the y-axis plots cumulative delinquency rates in each quarter following origination which is depicted on the x-axis. Data source: Corelogic

Originators: 1

e(1−m)xh 1−m(e+α(1−e))

Fraction of good projects e gxh pd Informed investors: N

Assets Liabilities

qpd K Uninformed investors M p =

e(1−m)xh 1−m(e+α(1−e))

m =

Nq e+(1−e)α

g ≡ prob (xh|σh) =

e e+(1−e)α

Originators: 1

e(1−m)xh 1−m(e+α(1−e))

Incentives: e gxh pd Informed investors: N

Assets Liabilities

Dr D qpd D K Uninformed investors MK − ND p = min

• MK−ND

1−m(e+α(1−e)), e(1−m)xh 1−m(e+α(1−e))

Liquidity, origination incentives and leverage - The main mechanism

• 1. Liquidity and origination incentives
• The first order condition

ψe (e − e) = (1 − α) m (K) κ [g (K) xh − p (K)]

• Non-linear impact of liquidity K on origination incentives
• a. Capital in the hands of informed investors: mK > 0
• b. Cost of distributing in the uninformed exchange: [gK (K) xh − pK] < 0
• When K is low, m is low and the spread g (K) xh − p (K) is wide, an increase

in K improves origination incentives

• When K is high, m is high and the spread g (K) xh − p (K) is narrow, an

increase in K worsens origination incentives

Figure 5: Panel A: Origination effort e, as a function of capital, K, for the cases α = 0 (dashed line, in green) and α = .2 (continuous line, in blue). Panel B: price in the exchange, p, as a function of capital K (continuous line, in blue, left axis) and matching probability m as a function of capital, K, (dashed line, in green, right axis). In this example, ψ(e) = θ e2

2 with θ = .25. In addition, κ = .15,

η = .5, M = .75 and xh = 5.

1 1.5 2 0.6 0.65 0.7 0.75 0.8

e(K)

K α = 0 α = .2

Panel A 1 1.5 2 2 2.5 3 3.5 4

g(K )xh − p(K ) K (gxh − p) m

Panel B 1 1.5 2 0.3 0.4 0.5

m(K )

• 2. Liquidity and leverage
• Definitions

– Size of the balance sheet: qpd = K + D – Leverage: L := K+D

K

= 1 + ℓ with ℓ = D

K

• Two results
• 1. Debt D is an increasing function of savings K.

– Demand: ∗ An increase in p increases debt capacity: (1 − η) qp ≥ D with η ∈ (0, 1) ∗ In addition, pd goes up by less than p: Less equity needed to buy assets – Supply: An increase in p, lowers rx and investors reallocate part of the funds to the repo market

• 2. Leverage L is an increasing fraction of K.

Figure 6: Fragility. Panel A: Fraction of high payoff assets in the balance sheet of financial intermediaries, g, as a function of capital,

• K. Panel B: Leverage, ℓ := D/K, as a function of capital K for the cases α = 0 (dashed line, in green) and α = .2 (continuous line,

in blue). In this example, ψ(e) = θ e2

2 with θ = .25. In addition, κ = .15, η = .5, M = .75 and xh = 5.

1 1.5 2 0.89 0.895 0.9 0.905 0.91 0.915 0.92

g(K)

K

Panel A 1 1.5 2 0.35 0.4 0.45 0.5 0.55 0.6 0.65 0.7 0.75 0.8

ℓ(K)

K α = 0 α = .2

Panel B

. The different phases of a liquidity cycle

t p (t) Cash-in-the-market region Discounted cash-flow region