Comments on The Fd Federal lr Reserve s Primary Dealer Credit Facility Tobias Adrian and James McAndrews

Transcription

1 Comments on The Fd Federal lr Reserve s Primary Dealer Credit Facility Tobias Adrian and James McAndrews SGE Session on The Fed s New Lending Facilities ASSA Meetings San Francisco John B. Taylor Stanford University January 4, 29

2 Summary Excellent paper on economic rationale for PDCF Analogy: Broker Dealers (PDCF), Depository Institutions (PCF) Focus on role of the Repo market Haircuts, margins, leverage, downward spirals (33:1 to 2:1) Leverage first rose in recent crisis (SIVs, large rate cuts) Structural changes over time: Growth of broker dealers and hedge funds relative to depository institutions Trend toward tradable assets and liabilities Repo runs or margin spirals rather thanclassic bank runs Repo runs or margin spirals rather than classic bank runs Protections: Penalty rates, regulation, moral hazard Main Comments: Empirical Impact Slippery slope to more facilities

3 Empirical Findings credit default swap prices of primary dealers fell after the introduction of the PDCF To date, the PDCF seems to be having measurable effects on the repo and other funding markets, as many primary dealers found that their perceived risk of default, as measured by credit default swap prices, declined after the facility s introduction. Yet no evidence presentedin the paper, so let s take a look at CDS rates for Merrill Lynch and Goldman Sachs (for example)

4 millions of dollars basis points , 125, 1, 75, 5, 25, PDCF CDS Merrill Lynch CDS Goldman Sachs Feb 26 Mar 23 Apr 21 May 18 Jun 16 Jul 13 Aug 1 Sep 8 Oct 5 Nov 3 Dec 31 Dec Weekly averages Weekly of daily average CDS data CDS and and weekly PDCF average data PDCF loans

5 millions of dollars basis points 6 15, 125, 1, 75, 5, 25, 3 Mar 24 Mar PDCF CDS Merrill Lynch CDS Goldman Sachs 14 Apr 5 May 26 May 16 Jun 7 Jul 28 Jul 18 Aug 8 Sep 29 Sep 2 Oct 1 Nov 1 Dec Daily CDS data with daily PDCF date interpolated from weekly average data 2 1

6 Examine Dynamic Cross Correlations Graphical evidence is difficult to interpret CDS rates appear to increase when PDCF loans rise! But there are cause and effect issues to sort out We can apply time series methods to sort out the timing differences Estimate bi variate vector auto regressions (VARs) between CDS rates (on Merrill Lynch and Goldman Sachs) and PDCF loans Compute impulse response functions from the VARs Look for significant effects Here are the results

7 PDCF and Goldman Sachs CDS Impulse Responses Based on Weekly Data Response to Cholesky One S.D. Innovations ± 2 S.E. Response of CMER1U5 to CMER1U5 Response of CMER1U5 to PDCF Response of PDCF to CMER1U5 Response of PDCF to PDCF 3, 3, 2, 2, 1, 1, -1, -1, -2, ,

8 PDCF and Merrill Lynch CDS Impulse Responses Based on Weekly Data Response to Cholesky One S.D. Innovations ± 2 S.E. 8 Response of CGS1U5 to CGS1U5 8 Response of CGS1U5 to PDCF , Response of PDCF to CGS1U5 3, Response of PDCF to PDCF 2, 2, 1, 1, -1, -1, -2, ,

9 Evidence of Effectiveness of PDCF PDCF loans Granger cause CDS rates negatively Inthissense sense there isevidence ofsignificanteffectiveness of the PDCF CDS rates Granger cause PDCF loans positively Likely reflecting the increased demand for PDCF Likely reflecting the increased demand for PDCF loans at the time of market stress

10 A Slippery Slope to More Facilities? Securities (Treasury and Agency) held outright Repos Loans from the TAF Other Loans Primary Credit Facility (discount window) Primary Dealer Credit Facility Asset Back Commercial Paper Money Market Mutual Fund Liquidity Facility Loans to AIG Term Asset Backed Securities Loan Facility (credit card, student, auto)* Private Portfolio holdings Commercial Paper Funding Facility Maiden Lane I (Bear Stearns) Maiden Lane II (AIG) Maiden Lane III (AIG) Money Market Investor Funding Facility* Mortgage Backed Securities Purchase Program* Other Federal Reserve Assets (very large loans to foreign central banks) Note: Minimal change items are gold stock, SDRs, Treasury currency outstanding, seasonal and secondary discounts, float

11 billions of of dollars Reserve Balances of of Depository p y Institutionsat at Federal Reserve Banks December 31, 28 $848 Billion September 1, 28 $8 $8 Billion

12 billions of dollars Reserve Balances Weekly average Wednesday level c c c 3 Jul 6 Aug 13 Aug 2 Aug 27 Aug 3 Sep 1 Sep 17 Sep 24 Sep 1 Oct 8 Oct 15 Oct 22 Oct 29 Oct 5 No 12 No v v 19 No 26 No 3 De 1 De v v c c 17 De 24 De 31 De

13 Impact of PCDF on Monetary Policy? billions of dollars 2 16 Wednesday Levels PDCF Reserve Balances 12 8 Change from Sept 1 to Sept 17: PDCF +$59.78B Reserve Balances +$57.728B 4 3 Jul 6 Aug 13 Aug 2 Aug 27 Aug 3 Sep 1 Sep 17 Sep 24 Sep 1 Oct