LIQUIDITY PROVISION DURING THE CRISIS OF 1914: PRIVATE AND PUBLIC SOURCES BY MARGARET M. JACOBSON* ELLIS W. TALLMAN** For presentation at: the Workshop on Monetary and Financial History held at the Federal Reserve Bank of Atlanta, June 24-27, 2013 *Federal Reserve Bank of Cleveland ** Oberlin College and Federal Reserve Bank of Cleveland Disclaimer: These are the views of the authors and not those of the Federal Reserve Bank of Cleveland or the Federal Reserve System.
INTRODUCTION Compares 1914 to the outcomes of the National Banking Era panics of 1873, 1893, and 1907 Explores why crisis prevention mechanisms were better in 1914 than in the other crises 1914 supported stable deposit levels to promote the expansion of the aggregate money supply Investigates the previously overlooked palliative role of clearing house loan certificates in the New York City
OUTLINE Background Data Empirical Methods and Results Discussion/Extensions Conclusions
BACKGROUND Crisis of 1914 in the United States Not a major event But could have been Before the Federal Reserve System was operational Fed operational in mid-november 1914 Liquidity crisis but no banking panic Intervention by policymakers and market participants Policymaker intervention (public liquidity provision) Issuance of Aldrich-Vreeland emergency currency Market participants (private liquidity provision) Clearing house loan certificates (CHLCs)
CONDITIONS IN JULY 1914 Trepidations in European financial markets Political situation impending war July 27-July 30 European stock markets, bourses close Prevent stock market dislocations and gold outflows July 31 New York Exchange closed by Secretary of the Treasury William McAdoo After consulting with and with support from the New York Stock Exchange
NEW YORK STOCK EXCHANGE Financial conditions in New York were benign Compared to Europe Stock market closure prevented: Liquidation of the US gold supply Throughout July, foreign investors were liquidating US stocks into gold and shipping the gold to London Gold standard implied that gold outflows could induce a contraction the base money supply Stock liquidations Policymakers feared fire sales of assets Silber (2007) suggests that fears were unwarranted
CRISIS CONDITIONS Structure of the National Banking System No explicit lender of last resort No mechanism to adjust base money supply rapidly Financial system prone to runs by banks, depositors New York national banks - reserve depositories Pyramid structure of bank reserves New York City national banks held their excess reserves in the call loan market at the New York Stock Exchange Demand loans as liquid Closure of the New York Stock Exchange Made call loans illiquid clear risk of cash scramble
RESERVES IN THE NATIONAL BANKING ERA Country Banks (6% in vault 9% as deposit in approved reserve city bank) Reserve City Banks (12.5% in vault 12.5% in central reserve city bank) New York City central reserve city banks (25% in vault) Stock market Reserves Call loans Reserves of interior banks held as deposits at NYC national banks NYC banks fund call loans with these deposits Deposit behavior during National Banking Era crises Removal of deposits by depositors and by interior banks led to issues of CHLCs and during major crises, a suspension of cash payments Money supply contracted as credit contracted
RESERVES IN 1914 Reserves Country Banks Reserve City Banks New York City banks Stock market Stock market closure in 1914 prevented New York City banks from adjusting short term balances using call loans. No federal funds market to adjust reserves Commercial paper market was still small Call loans But there was no banking panic and money supply grew
REAL NET FLOWS TO THE INTERIOR BANKS FROM NEW YORK CITY BANKS 800 600 400 200 0-200 Millions of Dollars -400-600 1893 1907-800 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Note: series are deflated using nominal GDP for each respective year Source: Commercial and Financial Chronicle, various volumes Johnston and Williamson, MeasuringWorth, 2013
REAL NET FLOWS TO THE INTERIOR BANKS FROM NEW YORK CITY BANKS 800 Millions of Dollars 600 400 1914 200 0-200 -400-600 1893 1907-800 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Note: series are deflated using nominal GDP for each respective year Source: Commercial and Financial Chronicle, various volumes Johnston and Williamson, MeasuringWorth, 2013
AGGREGATE US MONEY SUPPLY FOLLOWING CRISIS OR DISTRESS Percent change in money supply at initial crisis month 1.15 1.10 1.05 From August 1914 1.00 0.95 From October 1907 0.90 0.85 1 2 3 4 5 6 7 8 9 10 11 12 13 14 14 Months following the start of crisis
QUESTIONS TO BE ADDRESSED New York City banks were shut off from their main source of financial liquidity in 1914 Why was there no banking panic in 1914? Why was the outcome of 1914 so different from the previous panics of the National Banking Era?
EXISTING LITERATURE The success in 1914 associated with: Actions of policymakers (Silber 2007) The invention of emergency currency (Friedman and Schwartz 1963, Wicker 2005, Silber 2007) Clearing house loan certificates seen as a relics But contemporary observers of the 1914 distress attribute success to both forms of liquidity Noyes (1916) Commercial and Financial Chronicle (1915)
CURRENCY HELD BY THE PUBLIC 2300 2200 2100 2000 1900 1800 1700 1600 Millions of Dollars 1500 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 Note: Shaded areas indicate crisis periods Currency held by the public is defined as currency in circulation outside of the Treasury and Federal Reserve banks minus vault cash of all banks Source: Friedman and Schwartz (1970)
CURRENCY HELD BY THE PUBLIC Millions of Dollars 2300 2200 500 400 2100 2000 Emergency currency (right axis) 300 200 1900 1800 1700 1600 Currency 100 0-100 -200 1500 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916-300 Note: Shaded areas indicate crisis periods Currency held by the public is defined as currency in circulation outside of the Treasury and Federal Reserve banks minus vault cash of all banks Source: Friedman and Schwartz (1970)
DATA New York City clearing house member banks New York Clearing House banks represented: Nearly half of U.S. banking assets in 1914 Most acutely affected by stock market closure What we have: Daily data on emergency currency and clearing house loan certificates requested and received by bank Weekly balance sheet items (a subset) for each bank leading up to and following the onset of the distress Aggregate banking data weekly for clearing house member banks during the distress
COMPARING CLEARING HOUSE LOAN CERTIFICATES AND EMERGENCY CURRENCY Clearing house loan certificates 6% interest to holder Guaranteed by entire clearing house Could not circulate to public as currency State banks, trusts, and national banks Issued throughout national banking era Aldrich Vreeland Emergency Currency 3% interest for 3 months, 0.5% after Issued by the U.S. Treasury Could be paid out to depositors Only national banks Only issued in 1914 Borrow up to 125% of capital and surplus
TOTAL BORROWING NEW YORK CITY: BY LIQUIDITY PROVISION 46% 54% $269,670,960 Total Emergency currency Clearing house loan certificates
TOTAL BORROWING NEW YORK CITY: BY INTERMEDIARY TYPE 10% 12% $269,670,960 Total 78% National banks Trust companies State banks
TOTAL BANKING ASSETS NEW YORK CITY: BY BANK TYPE 12% 38% 50% $3,885,130,000 Total National banks Trust Companies State banks
COMPARISON OF LIQUIDITY PROVISIONS, NEW YORK CITY 160 140 120 Millions of dollars 100 80 Emergency currency 60 40 20 Clearing house loan certificates 0 Aug-1914 Sep-1914 Oct-1914 Nov-1914 Dec-1914 Jan-1915 Feb-1915 Note: emergency currency cancellations are linearly interpolated from aggregate values
CLEARING HOUSE LOAN CERTIFICATES, NEW YORK CITY 80 70 Millions of dollars 60 50 National Banks 40 30 20 10 State banks and trust companies 0 Aug-1914 Sep-1914 Oct-1914 Nov-1914 Dec-1914 Jan-1915 Note: Cancellations from November 20-28 are estimated to match the Clearing House Loan Committee Report of November 30, 1914.
HYPOTHESIS National Banks borrowed for different reasons: Aldrich-Vreeland emergency currency to satisfy cash withdrawal demands from depositors or interior banks Clearing house loan certificates to offset adverse balances at the clearing house (local bank drains) State bank and trust companies Had no alternative liquidity mechanism at hand Used clearing house loan certificates as a means to accommodate withdrawal demands from any source Liquidity enhanced with clearing house loan certificates Institutions that borrowed CHLCs were in better position to increase deposits and loans than otherwise
Base Money Supply Constraint Cash for depositors 1. No borrowing 2. Borrowing only clearing house loan certificates 3. Borrowing both clearing house loan certificates and emergency currency 1. 2. 3. Clearing balances at the clearing house
EMPIRICAL METHODS Simple comparison of grouped deposit level Deposit levels before distress and after it Four groups; banks that borrow: NEITHER form of temporary liquidity ONLY CHLCs (state and trusts) ONLY ECs emergency currency (national banks) BOTH emergency currency and clearing house loan certificates (again, national banks)
EMPIRICAL METHODS Hypothesis: 1: Borrowers of temporary liquidity in either form should maintain deposit levels in midst of distress. Examine data for 1914 to assess evidence Further: Compare with earlier NBE panics Compare pre- and post-distress deposit levels to analogous measures from previous NBE crises
NET DEPOSITS FOR NATIONAL BANKS 1700 Millions of Dollars 1600 1500 Net Deposits 1400 1300 1200 1100 Loans 1000 900 800 Jul-1914 Oct-1914 Jan-1915 Apr-1915 Jul-1915 Note: Shaded area indicates duration of crisis
CLEARING HOUSE TRUSTS AND BANKS IN NEW YORK 2.9 Millions of Dollars 2.8 Loans 2.7 2.6 2.5 Net Deposits 2.4 Jun-1914 Jul-1914 Aug-1914 Sep-1914 Oct-1914 Nov-1914 Dec-1914 Jan-1915 Note: Shaded area indicates duration of crisis
NET DEPOSITS FOR NATIONAL BANKS 1700 1600 Millions of Dollars 500 400 1500 1400 1300 1200 1100 1000 900 Net Deposits Loans Total borrowing (right axis) 300 200 100 0-100 -200-300 800 Jul-1914 Oct-1914 Jan-1915 Apr-1915 Jul-1915-400 Note: Shaded area indicates duration of crisis
CHANGE IN NET DEPOSITS CLEARING HOUSE BANKS 40 20 0-20 -40 Millions of dollars* -60-80 -100-120 Change from start of crisis to minimum value Change from the start and end of crisis 1873 1884 1890 1893 1907 1914 *Scaled by nominal GDP for each respective year
CHANGE IN LOANS AMONG CLEARING HOUSE BANKS Millions of dollars* 120 100 80 60 40 Change from start of crisis to minimum value Change from the start and end of crisis 20 0-20 -40-60 1873 1884 1890 1893 1907 1914 *Scaled by nominal GDP for each respective year
DATA SUMMARY Banking activity and intermediation Contracted during the three major NBE panics Remained flat in 1914. Is borrowing status associated with deposit level changes? Examine net deposit levels from the August and December weekly reports in the Commercial and Financial Chronicle Group data by borrowing status four groups
TOTAL NET DEPOSITS BY BORROWING STATUS 20 15 December 1914 August 1914 percent change 15.28 % 10 9.29 % 5 2.09 % 0-5 Emergency currency only Clearing house loan certificates only -3.06 Both % No borrowing
DEPOSIT CHANGE BY BANK SIZE 1000 900 800 700 600 500 400 300 200 100 0 December 1914 total net deposits, millions of dollars No borrowing, $13.01 mil. average assets AVEC only, $60.36 mil. average assets CHLC only, $40.70 mil. average assets Both, $58.39 mil. average assets Deposit growth Deposit contraction 0 100 200 300 400 500 600 700 800 900 1000 August 1914 Net deposits, millions of dollars Note: the area of circle represents the average bank assets in that respective borrowing category
DISCUSSION During National Banking Era panics Clearing house loan certificates were only choice Unable to generate an increase in the money supply In 1914, a different association CHLCs taken out by many intermediaries Even those able to borrow emergency currency CHLCs were associated with deposit growth Trust and state banks (no alternative) National banks - a secondary source of liquidity with less stringent collateral requirements
EXTENSIONS Compare the roles of core and peripheral liquidity creation to the 2008 crisis Comparison of 1893, 1907 and 1914 Isolate where the breakdowns begin And how 1914 avoided the breakdown Other ideas Formalize CHLCs and emergency currency as a composite good -- a closer substitute for legal tender Further empirical work (more technical) Tobit regression using balance sheet measures to predict those banks that borrow clearing house loan certificates, emergency currency or both
CONCLUSION Financial intermediaries in NYC maintain deposit levels, despite the closure of the NYSE in 1914 The Distress of 1914 stands out among National Banking Era panics because deposits, loans, and the aggregate money supply all increased Temporary liquidity was issued to a wide range of financial institutions Emergency currency and CHLCs Borrowers maintained and increased deposit levels to a greater extent than non-borrowers Counter factual thought experiment would deposits have grown at non-borrowing banks if there was no borrowing by any banks?