Too Big to Fail Causes, Consequences and Policy Responses. Philip E. Strahan. Annual Review of Financial Economics Conference.

Transcription

1 Too Big to Fail Causes, Consequences and Policy Responses Philip E. Strahan Annual Review of Financial Economics Conference October, 13

2 Too Big to Fail is a credibility problem Markets expect creditors to be bailed out when Borrower is large Borrower is financial and relies on short term debt Borrower is inter connected Promises to eschew bailouts are not fully credible Credit market distortions Bond yields don t fully reflect risk Weakens incentives to monitoring, enforce margin, impose haircuts, etc.

3 Why do markets expect bailouts? Short run spillovers may outweigh long run costs of bailouts Liquidation of financial institutions is costly Large failure may cause contagious panics ( liquidity spirals ) Declines in money & credit Large costs to non financial sector Government agents may prefer bailouts Political pressure correlated with size High short run costs of financial disruptions (whereas bankrupt airline still flies jets)

4 Four distortions for TBTF borrowers Too much leverage (since debt is cheap) Too much risk (since risk not fully priced & leverage increases risk taking incentives) Distortion propagates to non financial firms Financial firms grow inefficiently large To capture subsidy Small firms competitively disadvantaged

5 Fannie & Freddie: Paradigms of TBTF Distortions Cheap debt Lack of risk pricing (near Treasury Rates) Highest leverage among all US financial firms Increasingly aggressive investment in risky mortgages during the s Rapid (and unsustainable) growth in 199s and s Other institutions unable to compete in mortgage guarantee business

6 What Do We know from the Empirical Literature? Borrowing costs are lower and less risk sensitive for large banks But also true for large non financials Large banks have more leverage & take more risk than smaller ones But large banks are also better diversified than smaller ones Are large banks less efficient than small ones? Mixed evidence from cost studies of banks Some evidence that bank M&As motivated by desire to become very large

7 TBTF Problem has varied over time Continental Illinois (198s) Comptroller admits some banks are too big to fail Wealth increases at large banks, less risk pricing in debt markets FDIC Improvement Act (199s) Least cost resolution More risk pricing in bank debt LTCM (s) Large hedge fund drew Fed attention No follow up on regulating large institutions 8 Financial Crisis (1 and beyond)

8 Did TBTF cause the Financial Crisis? Proximate cause: A run on the shadow banking system But TBTF likely helped sow seeds of the run Increased leverage in s Off balance sheet leverage (OTC derivatives) Increase in securitization without risk transfer Increase in origination of risky loans Subprime, leveraged loans, covenant lite, etc. And little to no risk pricing at largest banks Graph of CDS

9 CDS Spreads for Top 1 Banks, 4 to 6 1.%.8%.6%.4%.%.% 1 Aug 6 1 Jan 4 1 Feb 4 1 Mar 4 1 Apr 4 1 May 4 1 Jun 4 1 Jul 4 1 Aug 4 1 Sep 4 1 Oct 4 1 Nov 4 1 Dec 4 1 Jan 5 1 Feb 5 1 Mar 5 1 Apr 5 1 May 5 1 Jun 5 1 Jul 5 1 Aug 5 1 Sep 5 1 Oct 5 1 Nov 5 1 Dec 5 1 Jan 6 1 Feb 6 1 Mar 6 1 Apr 6 1 May 6 1 Jun 6 1 Jul 6 1 Sep 6 1 Oct 6 1 Nov 6 1 Dec 6 Mean Range

10 Is TBTF Getting Worse? Greater importance of finance to the economy Credibility problem has its roots in concern about spillovers Greater concentration in the financial and banking system Morgan+Bear+WaMu Bank of America+Countrywide+Merrill Lynch Wells Farg+Wachovia

11 Finance value added increasing: Finance / GDP (%)

12 Concentration increasing: Top 1 Bank Share

13 But post crisis adjustments have made system more robust More Capital More Liquidity More Stable Funding Less Off Balance Sheet Leverage

14 Capital ratios have increased at large banks 9.5% 9.% 8.5% 8.% 7.5% 7.% 6.5%

15 Large banks hold more cash.%.% 4.% 6.% 8.% 1.% 1.% 14.%

16 More stable funding: Deposit/assets at large banks 75.% 73.% 71.% 69.% 67.% 65.% 63.% 61.% 59.% 57.% 55.%

17 Have expectations of bailouts have declined? CDS Markets now price large bank risk Natural experiment: Removal of TAG at the end of 1 Money flowed into small banks (not large one)

18 CDS Spread pre v. post crisis, Top 1 banks 5.% 4.5% 4.% 3.5% 3.%.5%.% 1.5% 1.%.5%.% Median Interquartile Range

19 Changes in Transactions Deposits with the Expiration of TAG (Kroszner, 13) 7% 6% 5% 4% 3% % 1% % -1% 6.1%.3% -.4% GSIB CCAR (excluding GSIB) Other banks

20 Conclusion TBTF rooted in hard to solve credibility problem Markets rationally expect large, interconnected banks to be bailed out Expectations of bailouts are not static High in the 198s, lower in the 199s, higher in s Post crisis Financial system is stronger And expectations of bailouts seem diminished New tools to deal with distress Political pressure not to bailout (Tea Party) Fiscal pressures (too big to save?)

21 THANK YOU!