Financial Disintermediation and Financial Fragility Aoki and Nikolov U of Tokyo and ECB 21th October 2013 Aoki and Nikolov (U of Tokyo and ECB) Financial Disintermediation and Financial Fragility 21th October 2013 1 / 31
Motivation Markets for debt securities have grown rapidly in the past 20 years - corporate bond markets - securitisations Aim of the paper: build a limited commitment economy in which banks and markets co-exist Main question: what is the e¤ect of larger debt markets on the likelihood and real economy impact of banking crises in our model economy? Aoki and Nikolov (U of Tokyo and ECB) Financial Disintermediation and Financial Fragility 21th October 2013 2 / 31
Corporate bond markets have grown Corporate bond stock as a fraction of commercial bank loans in the US (Source: FDIC, Flow of Funds) 0.5 0.45 0.4 0.35 0.3 0.25 0.2 0.15 0.1 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 Aoki and Nikolov (U of Tokyo and ECB) Financial Disintermediation and Financial Fragility 21th October 2013 3 / 31
...banks moved to housing …nance Commercial & Industrial and Real Estate loans as a fraction of total loans in the US (Source: FDIC) 0.7 0.6 0.5 C+L loans as a % of total loans and leases 0.4 0.3 0.2 Loans backed by real estate as a % of total loans and leases 0.1 0 1938 1943 1948 1953 1958 1963 1968 1973 1978 1983 1988 1993 1998 2003 2008 Aoki and Nikolov (U of Tokyo and ECB) Financial Disintermediation and Financial Fragility 21th October 2013 4 / 31
...but faced competition from ABS markets there too Assets of ABS issuers and Broker-Dealers as a fraction of commercial bank loans in the US (Source: FDIC, Flow of Funds) 0.5 ABS issuers 0.45 0.4 0.35 0.3 0.25 0.2 Broker-dealers 0.15 0.1 0.05 0 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 Aoki and Nikolov (U of Tokyo and ECB) Financial Disintermediation and Financial Fragility 21th October 2013 5 / 31
Question and Approach Main question of the paper: - What e¤ect has the growth of markets had on the likelihood and severity of banking crises? Paper is part of our work on understanding the role of asset price bubbles in …nancial crises. Aoki and Nikolov (2012): - bubble collapses have much larger e¤ects if banks exposed (’Dot Com’ vs ’Subprime’) - But what explains banks’ high exposure in 2007? This paper: the type of …nancial innovation that occurred in the 2000s Aoki and Nikolov (U of Tokyo and ECB) Financial Disintermediation and Financial Fragility 21th October 2013 6 / 31
Preview of Results We build a limited commitment economy with two types of markets - Ones that operate independently of banks: corporate bond markets (1980s and 1990s) - Ones that need bank guarantees to work: ’Shadow banking’ (1990s and 2000s) The growth of both types of …nancial markets stimulates bank risk-taking in the model. - Markets compete with banks and errode their pro…tability. - They increase their exposure to risky assets = ) risk of banking losses increases But: di¤erent markets have a di¤erent impact on systemic risk - Dot Com vs Subprime Aoki and Nikolov (U of Tokyo and ECB) Financial Disintermediation and Financial Fragility 21th October 2013 7 / 31
Preview of Results (cont’d) Corporate bond market growth more benign - corporate bond market expands at the expense of banks - lending-deposit rate spreads fall and risky investments grow - BUT: consolidated bank leverage falls = ) small banking losses ’Shadow banking’ more likely to cause a crisis - allows commercial banks to evade regulation (e.g. ABCP) = ) aggregate …nancial system leverage expands - higher leverage leads to an even larger fall in lending-deposit spreads = ) risky invesments grow even more = ) large banking losses Aoki and Nikolov (U of Tokyo and ECB) Financial Disintermediation and Financial Fragility 21th October 2013 8 / 31
Related literature Bank and bond …nance Holmstrom and Tirole (1997), De Fiore and Uhlig (2011), (2012), Adrian, Colla and Shin (2012), Gertler and Karadi (2012) Shadow banking Gennaioli, Shleifer and Vishny (2011, 2012), Acharya and Schnabl (2011), Goodhart et. al. (2012), Alessandri, Meeks and Nelson (2012) Franchise values and bank risk-taking in macro models Martinez-Miera and Suarez (2012), Corbae and D’Erasmo (2012) Asset price bubbles in macro models Farhi and Tirole (2011), Martin and Ventura (2012), Aoki and Nikolov (2012), Hirano and Yanagawa (2012), Miao and Wang (2012) Aoki and Nikolov (U of Tokyo and ECB) Financial Disintermediation and Financial Fragility 21th October 2013 9 / 31
Model Overview Limited commitment economy with 2 main types of agents: entrepreneurs and bankers Only source of aggregate risk: stochastic rational bubble - bubble modelled as an intrinsically worthless asset - bubbly equilibria exist due to credit frictions Bubble impact: - large when bubble held by banks - small when held by savers This paper: does the growth of …nancial markets increase banks’ incentives to hold the risky bubble asset? Aoki and Nikolov (U of Tokyo and ECB) Financial Disintermediation and Financial Fragility 21th October 2013 10 / 31
Model Overview (cont’d) Compare the e¤ects of the introduction of two types of …nancial markets into a hitherto bank-centred …nancial system ’Corporate bond market’ - operates without bank involvement (genuine disintermediation) ’Shadow banking’ - operates with the help of bank guarantees (motivated by Acharya and Schnabl (2011) evidence on the ABCP market) Aoki and Nikolov (U of Tokyo and ECB) Financial Disintermediation and Financial Fragility 21th October 2013 11 / 31
Model of …nancial disintermediation 1: corporate bond Aoki and Nikolov (U of Tokyo and ECB) Financial Disintermediation and Financial Fragility 21th October 2013 12 / 31
Model of …nancial disintermediation 2: shadow banking Aoki and Nikolov (U of Tokyo and ECB) Financial Disintermediation and Financial Fragility 21th October 2013 13 / 31
Financial markets and bank risk-taking Baseline economy: - only banks intermediate - economy is in a bubbly equilibrium - bank bubble holdings are zero due to high lending spread Contrast the impact on this economy of two …nancial innovations - corporate bond market expansion - shadow banking expansion Crisis: bubble collapses in period 5 of the simulation Shadow banking system is closed by regulators Aoki and Nikolov (U of Tokyo and ECB) Financial Disintermediation and Financial Fragility 21th October 2013 14 / 31
Financial markets and bank risk-taking (cont’d) Aoki and Nikolov (U of Tokyo and ECB) Financial Disintermediation and Financial Fragility 21th October 2013 15 / 31
Financial markets and bank risk-taking (cont’d) Aoki and Nikolov (U of Tokyo and ECB) Financial Disintermediation and Financial Fragility 21th October 2013 16 / 31
Why does leverage depress spreads? Rate of return on bank equity: � � φ t R l R l t � R d t + t ( 1 � λ ) φ t ( 1 � λ ) is bank leverage. Consists of: - φ t > 1: the ’charter value’ of a unit of net worth inside the bank: " when " R l t � R d t - 1 � λ : depositors/regulators’ perceived bene…t from diverting funds: # under shadow banking )# R l t � R d Holding equity returns …xed " λ = t Aoki and Nikolov (U of Tokyo and ECB) Financial Disintermediation and Financial Fragility 21th October 2013 17 / 31
Conclusions We develop a model of the corporate bond market and shadow banking Bank risk taking impact - both decrease bank pro…tability and stimulate bank risk-taking - shadow banking increases …nancial sector leverage = ) " risk-taking compared to the corporate bond market Aoki and Nikolov (U of Tokyo and ECB) Financial Disintermediation and Financial Fragility 21th October 2013 18 / 31
Entrepreneurs Utility t = 0 β t ln c t Σ ∞ Production function a H > a L a = a H , a L , y t + 1 = ah t , Budget constraint c t + w t h t + m e t µ t � b m t � b l t t � 1 � e a i h t � 1 � R i t � 1 b l R i t � 1 b m t � 1 + m e = t � 1 µ t We consider two alternative credit market arrangements Aoki and Nikolov (U of Tokyo and ECB) Financial Disintermediation and Financial Fragility 21th October 2013 19 / 31
The corporate bond market e t = e R l R d t = R m t Banks and savers have di¤erent enforcement ability - Can pledge up to θ to banks = ) total (bank + bond) debt is R l t b l t + R m t b m t � θ y t + 1 , 0 < θ < 1 - Can pledge up to θ ( 1 � χ ) to savers = ) total bond debt R m t b m t � θ ( 1 � χ ) y t + 1 , 0 < θ < 1 , 0 < χ < 1 Consistent with several di¤erent micro-foundations - Di¤erent enforcement ability of banks and savers (our assumption) - ’Skin-in-the-game’ constraint in ’plain vanilla’ securitisations (Shleifer and Vishny (2010)) Aoki and Nikolov (U of Tokyo and ECB) Financial Disintermediation and Financial Fragility 21th October 2013 20 / 31
Recommend
More recommend
