Banks as Liquidity Provider of Second to Last Resort Til Schuermann* Federal Reserve Bank of New York Q-Group, October 2008 * Any views expressed represent those of the author only and not necessarily those of the Federal Reserve Bank of New York or the Federal Reserve System. Filename
Talk based on joint work with Evan Gatev and Phil Strahan (Boston College, Finance) First: how it s supposed to work Then: how it seems not to be working now. Filename 1
3M TED Spread Filename 2
2 1.8 1.6 LIBOR - OIS Spread 2 Jan 2002-25 Sep 2008 LIBOR-OIS Spread (bp) 1-month 3-month avg (thru July 2007) 8.7 10.9 std dev (thru July 2007) 3.1 3.5 Max (9/25/08): 1.860%, 1.966% 1.4 1.2 1 0.8 0.6 0.4 0.2 0 Jan-02 May-02 Sep-02 Jan-03 May-03 Sep-03 Jan-04 May-04 Sep-04 Jan-05 May-05 Sep-05 Jan-06 May-06 Sep-06 Jan-07 May-07 Sep-07 Jan-08 May-08 Sep-08 % avg (since Aug 2007) 50.7 70.0 std dev (thru July 2007) 25.0 20.6 (Jan 2, 2002 - Sept. 25, 2008) 1-month 3-month Filename 3
Some data What s a few trillion between friends.. Early 2007: ABCP + SIV + ARS + TOB + VRDN $2.2 trn O/N tri-party repo: $2.5 trn Hedge funds AUM: $1.8 trn Assets of 5 i-banks: $4 trn Assets of 5 U.S. BHCs: $6 trn Assets of all U.S. banks: $10 trn Meanwhile, sum of write-offs to date (> $500bn) exceeds cost of S&L crisis (~ $250bn in current $) Filename 4
Bank liquidity management A bank offers two short-term liquidity contracts Loan commitments A L Transaction deposits E Seems very unstable What if demand spikes for both at the same time? And what if that happens systematically (affecting all banks) Worry about bank runs Filename 5
Bank liquidity management A bank offers two short-term liquidity contracts Loan commitments A L Transaction deposits E Other sources of bank liquidity Hold cash and liquid assets Access to the inter-bank market Borrow from the central bank Filename 6
But maybe combining the 2 contracts reduces risk... Diversification synergy Combining transactions deposits and loan commitments reduces idiosyncratic risk (Kashyap, Rajan & Stein, JF 2002) Transaction deposits hedge the systematic liquidity risk exposure of loan commitments Flight to quality Banks can bear systematic shocks to liquidity demand due to funding inflows (Gatev and Strahan, JF 2006) Deposit-lending synergy is stronger in a liquidity crisis (e.g. Fall 1998) Seems related to government safety net Funding flows not related to bank solvency or size Effects absent prior to FDIC (Pennacchi JME 2006) Filename 7
Research questions How does bank risk (stock volatility) vary with liquidity exposure? Exposure from deposits Exposure from loan commitments Is there evidence of a natural hedge to mitigate liquidity risk? Does the hedge become more evident when liquidity becomes scarce? Case study: Fall 1998 (Gatev, Schuermann & Strahan, NBER 2005) Filename 8
Sample: Time-Series / Cross-Section Data Largest (based on market cap) 100 US banks each year, 1990-2002 Drop bank-years when M&As occur In 1990 leaves 85 banks Number of banks ranges between 98 (2002) and 68 (1996) Market data (weekly stock returns) and call report data Almost 50,000 bank-week observations Cluster data (errors) by bank to avoid assuming independence over time for each bank: 170 unique banks Filename 9
Loan-Deposit synergy: early evidence Unused Commitments / (Commitments + Loans) (LC) Transactions Deposits / Total Deposits (TD): bottom third bottom third middle third top third Stock-return Volatility 28% 29% 32% Assets (Billions of $s) 10.58 10.20 7.14 Equity / Assets 8% 8% 11% middle third Stock-return Volatility 29% 29% 30% Assets (Billions of $s) 17.85 21.64 16.53 Equity / Assets 8% 8% 7% top third Stock-return Volatility 36% 32% 31% Assets (Billions of $s) 34.63 89.10 83.36 Equity / Assets 8% 8% 8% Mean Commitments Ratio 0.30 0.31 0.37 Filename 10
Research design to address questions Dependent variable = stock-return volatility (weekly) Conditional return volatility: GARCH(1,1) Realized volatility (total or residual) Modeling Bank Risk Volatility = α + β 1 LoanCommitments t-1,i + β 2 DepositBase t-1,i +β 3 (LoanCommitments t-1,i *DepositBase t-1,i ) + OtherControls + ε i,t β 1, β 2 > 0 (Exposure); β 3 < 0 Filename 11
Control variables Market conditions Volatility of S&P500 Paper-bill spread (3M non-financial) Yield on 3M T-bill Bank characteristics Size: Log of assets Capital ratio: Capital/assets Inter-bank access: Fed funds purchased/assets Liquid assets: (cash + securities)/assets Other risks (market, credit risk) Trading assets/assets C&I loans/assets CRE loans/assets NPL/assets Loan-loss provision/assets Net charge-offs/assets Credit rating Filename 12
0.8 Low TD banks: risk as LC Figure 1a: Stock Return Volatility for Low Transactions Deposit Banks 1990-2002 0.7 0.6 0.5 0.4 0.3 slope = 0.28 (5.55) 0.2 0.1 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Unused commitments / (commitments + loans) Time average of annualized bank stock return volatility and commitment ratio for bank with below-median levels of transaction deposits for 170 largest U.S. banks (plot is for 85 banks). Source: Volatility based on authors' calculations using data from CRSP. Commitment ratio is from Call Reports. Filename 13
0.8 High TD banks: risk unchanged as LC Figure 1b: Stock Return Volatility for High Transactions Deposit Banks 1990-2002 0.7 0.6 0.5 0.4 0.3 0.2 slope = -0.10 (-1.22) 0.1 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Unused commitments / (commitments + loans) Time average of annualized bank stock return volatility and commitment ratio for bank with above-median levels of transaction deposits for 170 largest U.S. Filenamebanks (plot is for 85 banks). Source: Volatility based on authors' calculations using data from CRSP. Commitment ratio is from Call Reports. 14
Results of multivariate regressions Direct exposure coefficients (β 1 & β 2 ) positive By themselves, more exposure to LC & TD increases risk Hedge coefficient (β 3 ) negative Results insensitive to volatility measure: GARCH or realized (total or residual) Results robust when controlling for market and credit risk Filename 15
Reverse causality? Why are there some banks on the off-diagonal? E.g. Low LC exposure but high TD (upper right corner) Smallest banks, bank-dependent clientele but little liquidity insurance provided Still, reverse causality is possible Risk mgmt motive drives bank choice of TD and LC rather than other way around Maybe (otherwise) safe banks choose to expose themselves to greater liquidity risk (high LC, high TD) Filename 16
Idiosyncratic vs. systematic liquidity demands During normal times, diversification synergy comes from reducing effect of idiosyncratic liquidity demands What if there is a systematic shock to liquidity? All borrowers show up demanding liquidity But: supply of TD increases too Hedging effect should be even stronger... And it is! Look at the times of low liquidity (top 5% of paper-bill spread distribution: >75bp; avg. = 40bp) Hedging term (β 3 ) nearly triples in size Also consider Fall 1998 liquidity (flight to quality) crisis Filename 17
3M non-fin CP spread (basis points) weekly, Jan - Dec 1998 140 120 Oct. 16, 1998: 125bp 100 80 60 40 20 0 J F M A M J J A S O N D Filename 18
Filename 19
Filename 20
0.8 Low TD banks: risk 2x faster as LC Figure 3a: Stock-Return Volatility for Low-Transactions Deposit Banks Fall 1998 0.7 0.6 slope = 0.46 (2.69) 0.5 0.4 Old slope: 0.28 0.3 0.2 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Unused Commitments / (Commitments + Loans) Time average of annualized bank stock return volatility and commitment ratio for bank with below-median levels of transaction deposits for 64 largest U.S. Filenamebanks (plot is for 32 banks). Source: Volatility based on authors' calculations using data from CRSP. Commitment ratio is from Call Reports. 21
High TD banks: risk unchanged as LC 0.8 Figure 3b: Stock-Return Volatility for High-Transactions Deposit Banks Fall 1998 0.7 0.6 0.5 slope = -0.16 (-0.66) 0.4 0.3 0.2 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Unused Commitments / (Commitments + Loans) Time average of annualized bank stock return volatility and commitment ratio for bank with below-median levels of transaction deposits for 64 largest U.S. Filename banks (plot is for 32 banks). Source: Volatility based on authors' calculations using data from CRSP. Commitment ratio is from Call Reports. 22
Conclusions (so far) Deposit-loan combination reduces bank risk Idiosyncratic liquidity demands Risk-reducing synergy more powerful when paper-bill spreads are wide Systematic liquidity demands Helps with causality Results not due to other risks (market, credit) Filename 23
180 160 3M non-fin CP spread (basis points) weekly, Jan 1997 - Sept 2008 avg (thru July 2007) 32.0 std dev (thru July 2007 22.7 162 140 120 avg (since Aug 2007) 75.1 std dev (thru July 2007 37.8 (Jan 2, 1997 - Sept. 19, 2008) 100 80 95th %-ile (91bp) 60 40 20 0 Jan-97 Jul-97 Jan-98 Jul-98 Jan-99 Jul-99 Jan-00 Jul-00 Jan-01 Jul-01 Jan-02 Jul-02 Jan-03 Jul-03 Jan-04 Jul-04 Jan-05 Jul-05 Jan-06 Jul-06 Jan-07 Jul-07 Jan-08 Jul-08 Filename 24
It s good to be a (commercial) bank When short term funding, e.g. CP, in the capital markets dries up, go to your bank If you no longer wish to place your short term funds in ABCP, go to your bank How long can this go on? Until balance sheet can grow no more Where does this leave investment banks? Filename 25
What s going on now? Banks have been hoarding liquidity Especially European banks Deposit flows Foreign/domestic.. New Fed facilities And credit spreads? Filename 26
So what s happening to bank deposits? Deposit Growth Rates Quarter-over-quarter, all commercial banks Domestic Foreign Total 2001-2005 1.96 1.32 1.86 2006q1-2007q2 1.16 6.30 2.05 2007q3 0.82 6.97 2.07 2007q4 4.20 4.23 4.21 2008q1 2.23-0.38 1.70 2008q2-1.01 3.08-0.16 Entire period 1.75 2.60 1.91 Filename 27
High overnight agency and MBS spreads to Treasury 2.5 2.0 March 18: Agency: 1.65% MBS: 2.1% (!) Spread (bp) Agency MBS avg (thru July 2007) 1.8 5.2 std dev (thru July 2007) 11.8 11.4 avg (since Aug 2007) 26.6 31.6 std dev (thru July 2007) 44.7 46.5 Spread % 1.5 1.0 (May 21, 1991 - Sept. 23, 2008) Sept. 17: Agency: 1.75% MBS: 2.0% 0.5 0.0 2/1/08 3/1/08 4/1/08 5/1/08 6/1/08 7/1/08 8/1/08 9/1/08 Source: Bloomberg Agency Spread MBS Spread Filename 28
Abnormally low overnight Treasury repo rates 4.0 3.5 3.0 2.5 Rate % 2.0 1.5 1.0 0.5 0.0 2/1/08 3/1/08 4/1/08 5/1/08 6/1/08 7/1/08 8/1/08 9/1/08 Overnight Treasury Repo Fed Funds Target Source: Bloomberg, FRBNY Filename 29
Abnormally low overnight Treasury repo rates 4.0 3.5 3.0 2.5 Rate % 2.0 1.5 1.0 0.5 0.0 2/1/08 3/1/08 4/1/08 5/1/08 6/1/08 7/1/08 8/1/08 9/1/08 TSLF dates Overnight Treasury Repo Fed Funds Target Filename Source: Bloomberg, FRBNY 30
Problems addressed by new lending facilities Backstop Standing Facilities Depository Institutions Discount Window Primary Dealers Primary Dealer Credit Facility (PDCF) Auction Facilities Term Auction Facility (TAF) Term Securities Lending Facility (TSLF) TAF: illiquid term markets and the stigma that accompanies discount window borrowing. TSLF: illiquid functioning in repo funding markets illustrated by abnormal rates and high haircuts. PDCF: the lack of market-based back-stop credit in repo markets. Filename 31
What can you pledge at the TSLF & PDCF? TSLF: OMO collateral plus investment grade securities: private label RMBS, CMBS, Agency CMOs, ABS such as CDOs, CLOs,corporates, munis, MBS (R and C), ABS So long as it can be priced by the clearing banks PDCF: above plus sub-investment grade securities plus equities Importantly, previously repo-able securitized instruments are no longer stuck on firms balance sheets Facilities designed as liquidity vehicles Filename 32
Baa and Aaa Spread (to Treasury) weekly, Jan. 2, 1998 - Sept 19, 2008 450 400 Baa spread 11 Oct, 2002 Aaa spread 350 300 250 95th %-ile 200 150 100 21 Sept. 2001 50 0 Jan-98 Jul-98 Jan-99 Jul-99 Jan-00 Jul-00 Jan-01 Jul-01 Jan-02 Jul-02 Jan-03 Jul-03 Jan-04 Jul-04 Jan-05 Jul-05 Jan-06 Jul-06 Jan-07 Jul-07 Jan-08 Jul-08 Filename 33
Thank You! http://nyfedeconomists.org/schuermann/ Filename 34