Abstract
Although the direct effect of lender-of-last-resort (LOLR) facilities is to forestall the default of financial firms that lose funding liquidity, an indirect effect is to allow these firms to minimize deleveraging sales of illiquid assets. This unintended consequence of LOLR facilities manifests itself as excess illiquid leverage in the financial sector, can make future liquidity shortfalls more likely, and can lead to an increase in default risks. Furthermore, this increase in default risk can occur despite the fact that the combination of LOLR facilities and reduced asset sales raises the prices of illiquid assets. The behavior of U.S. broker-dealers during the crisis of 2007–09 is consistent with this unintended consequence. In particular, given the Federal Reserve’s LOLR facilities, broker-dealers could afford to try to wait out the crisis. Although they did reduce traditional measures of leverage to varying degrees, they failed to reduce sufficiently their illiquid leverage, which contributed to their failures or near failures. Several mechanisms to address this unintended consequence of LOLR facilities are proposed.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
Merrill Lynch, Q2 Earnings Call, July 17, 2008
Broker-dealers normally rely on private markets to finance their positions, but these markets became severely impaired in early 2008. Lenders of funds through repurchase agreements became increasingly cautious, worrying both about the liquidation value of collateral and about the credit risk of counterparties. These lenders reacted by increasing haircuts—reducing the cash they were willing to lend against a given amount of collateral—and by refusing to lend at all against certain types of collateral. See, for example, Copeland, Martin, and Walker (2011), Gorton and Metrick (2012), and Krishnamurthy, Nagel, and Orlov (2014).
Although the initial terms of the PDCF limited eligibility to securities “for which a price is available” (that is, quoted in the tri-party repo system), the eligible investment-grade corporates, municipals, and private MBS/ABS were certainly not all liquid in the sense that broker-dealers could move large positions at quoted prices during and after the fall of Bear. In fact, anecdotal evidence indicates that, toward the end, Bear Stearns had trouble funding even its agency securities, which were eligible for the more restrictive initial TSLF and which constituted the most liquid tranche of the PDCF.
See Fischer (1999) for an excellent survey of LOLR policy prescriptions and the literature that evolved from these prescriptions. Although Fischer's focus is on the role that can be played by the International Monetary Fund (IMF) as the international LOLR when sovereign or banking crises need to be contained from spreading across borders, he also succinctly presents the argument underlying the moral hazard induced by LOLR, including domestic LOLR. He recognizes that, while moral hazard needs to be contained, it is unlikely to be eliminated entirely through the design of LOLR facilities.
In a particularly startling example, Acharya and Steffen (2014) show that the Bank of Cyprus, using ECB financing, appears to have quadrupled its holdings of Greek debt 2010-11.
Caballero, Hoshi, and Kashyap (2008), in the context of the Japanese banking crisis of the 1990s, attribute the phenomenon of “zombie banks” lending to “zombie firms,” along with the resulting credit crunch, to the excessive forbearance of the Bank of Japan. Diamond and Rajan (2011) argue that delaying fire sales in expectation of central bank or government support can increase the returns to liquidity (that is, to the capacity for acquiring assets that are eventually sold in fire sales) and lead to an ex-ante freeze in credit markets.
In the following numerical examples,
, so the relevant solvency condition is 
.In certain parameterizations, for example, in which the bank is very highly levered, a U-shaped bank supply curve and a downward-sloping buyer demand curve give rise to multiple equilibria.
The collateralized agreements are allocated to the various activities as follows. Shorts of $171.111 billion require that amount of collateralized agreement assets, leaving the total collateralized agreements, $386.792 billion, minus $171.111 billion, or $215,681 billion, as matched-book assets. By definition, matched-book liabilities equal matched-book assets, so subtracting $215,681 of collateralized agreement liabilities from total collateralized agreement liabilities of $288,135 billion, leaves $72,454 for funding.
Level 1 assets are almost always more liquid than Level 2 and Level 3 assets, but grouping Level 2 and Level 3 assets together makes sense because assets migrate more fluidly between these categories than between Level 1 and Level 2. During the crisis, sales of particular assets in a particular quarter provided pricing benchmarks for other assets, which could then move from Level 3 to Level 2. Similarly, a dearth of sales and, therefore, benchmark prices in a particular quarter, would push various Level 2 assets to Level 3.
The analysis of this section attributes all changes in assets to purchases or sales while, in fact, some of the changes may very well be due to changes in the market prices of existing assets. Furthermore and unfortunately, data are not available to distinguish purchases and sales from valuation changes. But, to the extent that observed deleveraging was due to generally declining prices, broker-dealers sold assets even less aggressively than claimed here, which actually strengthens the broad conclusions of the paper.
There are some operational hurdles in providing LOLR loans to institutions that do not usually interact with the Federal Reserve. But these can be overcome, as they were with the CPFF, by having banks and investment banks act as agents for these loans. See, for example, Adrian, Kimbrough, and Marchioni (2011). Note that this agency model is very different from having banks and investment banks use their own balance sheets to provide indirect LOLR to their customers, which transmission mechanism is not, as discussed earlier, very reliable in a crisis.
For an account of the history of TARP, see, for example, Paulson (2010).
, so the relevant solvency condition is 
.References
Acharya, V.V. and D. Backus, 2009, “Private Lessons for Public Banking: The Case for Conditionality in LOLR Facilities,” in Restoring Financial Stability: How to Repair a Failed System, ed. by V.V. Acharya, and M. Richardson (New York: John Wiley & Sons), Chapter 14, March.
Acharya, Viral V., Irvind Gujral, Nirupama Kulkarni, and Hyun Song Shin, 2011, “Dividends and Bank Capital in the Financial Crisis of 2007–2009,” Working Paper No. 16896 (National Bureau of Economic Research).
Acharya, Viral V. and Sascha Steffen, 2014, “The “Greatest” Carry Trade Ever? Understanding Eurozone Bank Risks,” forthcoming Journal of Financial Economics.
Acharya, Viral V., Michael J. Fleming, Warren B. Hrung, and Asani Sarkar, 2014a, “Dealer Financial Conditions and Lender-of-Last Resort Facilities,” Working Paper (NYU Stern School of Business).
Acharya, Viral V., Hanh Le, and Hyun Song Shin, 2014b, “Bank Capital and Dividend Externalities,” Working Paper No. 19707 (National Bureau of Economic Research).
Adrian, Tobias, Karin Kimbrough, and Dina Marchioni, 2011, “The Federal Reserve’s Commercial Paper Funding Facility,” FRBNY Economic Policy Review, May.
Bagehot, W., 1873, Lombard Street, A Description of the Money Market (Homewood, IL: Richard D. Irwin Inc).
Brunnermeier, Markus and Lasse Pederson, 2009, “Market Liquidity and Funding Liquidity,” Review of Financial Studies, Vol. 22, No. 6, pp. 2201–38.
Caballero, R.J., T. Hoshi, and A.K. Kashyap, 2008, “Zombie Lending and Depressed Restructuring in Japan,” American Economic Review, Vol. 98, No. 5, pp. 1943–77.
Copeland, Adam, Antoine Martin, and Michael Walker, 2011, Repo Runs: Evidence from the Tri-Party Repo Market, Staff Reports 506 (Federal Reserve Bank of New York).
Diamond, D. and R. G. Rajan, 2005, “Liquidity Shortages and Banking Crises,” Journal of Finance, Vol. 60, No. 2, pp. 615–47.
Diamond, D. and R. G. Rajan, 2011, “Fear of Fire Sales, Illiquidity Seeking, and Credit Freezes,” Quarterly Journal of Economics, Vol. 126, No. 2, pp. 557–91.
Drechsler, I., T. Drechsel, D Marques-Ibanez, and P. Schnabl, 2013, “Who Borrows From the Lender of Last Resort?” Working Paper (NYU Stern School of Business).
Federal Reserve Bank of New York. 2008, Press Release, March 16.
Fischer, Stanley, 1999, “On the Need for an International Lender of Last Resort,” Speech delivered at the joint luncheon of the American Economic Association and the American Finance Association, New York, January 3, 1999, www.imf.org/external/np/speeches/1999/010399.HTM.
Fleming, M.J., 2012, “Federal Reserve Liquidity Provision during the Financial Crisis of 2007–2009,” Annual Review of Financial Economics, Vol. 4, pp. 161–77.
Gorton, G. and A. Metrick, 2012, “Securitized Banking and the Run on Repo,” Journal of Financial Economics, Vol. 104, No. 3, pp. 425–51.
He, Zhiguo, In Gu Khang, and Arvind Krishnamurthy, 2010, “Balance Sheet Adjustments in the 2008 Crisis,” IMF Economic Review, Vol. 58, No. 1, pp. 118–56.
Krishnamurthy, A., S. Nagel, and D. Orlov, 2014, “Sizing up Repo,” Journal of Finance, forthcoming.
Paulson, Henry, 2010, On the Brink (New York: Hachette Book Group).
Rochet, Jean-Charles and X. Vives, 2004, “Coordination Failures and the Lender of Last Resort: Was Bagehot Right After All?,” Journal of the European Economic Association, Vol. 2, No. 6, pp. 1116–47.
Additional information
Supplementary information accompanies this article on the IMF Economic Review website (www.palgrave-journals.com/imfer)
This paper was prepared for the IMF Economic Review Conference in Honor of Stanley Fischer.
*Viral V. Acharya is the C. V. Starr Professor of Economics in the Department of Finance at the NYU Stern School of Business, a Research Affiliate of CEPR, and a Research Associate of the NBER. Bruce Tuckman is a Clinical Professor of Finance at the NYU Stern School of Business and a Senior Fellow at the Center for Financial Stability. The authors are grateful to Pierre-Olivier Gourinchas and Ayhan Kose (editors), Ricardo Caballero (discussant), Morgan Ricks (discussant), three anonymous referees, participants at the International Monetary Fund’s 14th Jacques Polak Conference in honor of Stanley Fischer on November 7–8, 2013 and at the 2014 NYU Law School Conference on Banking Reform, and Sveriges Riksbank for helpful comments. The authors also thank Katherine Waldock for outstanding research assistance. All errors are of the authors’.
Electronic supplementary material
Appendices
Appendix I
Appendix I Proofs
Proof of Proposition 1:
From Equation (6) in Section “The Bank’s Optimal Deleveraging Policy,” we have that:
In order to proceed, we must first establish the SOC for a local maximum:
Differentiating (∂E)/(∂α) wrt α, we obtain:
Owing to the solvency condition, only a negative shock can render the bank insolvent at date 1. Because u is mean-0, g′(u B ) > 0, and so both (1−l)x2 and (g(u B )/g′(u B )) have positive signs. It is necessary, therefore, to verify the SOC on a case-by-case basis.
Lemma 1:
-
Unless α*=0, α* is interior and the SOC is satisfied. See Online Appendix A for proof.
We would then like to obtain an expression for (du B /dl). We know from Equation (7) in Section “The Effect of LOLR on Deleveraging and Default Probabilities” that:
We obtained that (∂u B /∂l) < 0 and that (∂u B /∂α) < 0 by the solvency condition, but we do not know the sign of (dα*/dl). We do know by the FOC that at the optimal level of α*, if we change l, (∂E/∂α)=0 must still hold. That is,
Assuming the SOC holds, (∂2E/∂α2) < 0. To obtain the sign of (dα*/dl), therefore, we must sign the expression (∂2E/∂α∂l).
Solving for the partial derivatives within this expression:
Substituting in signs yields our result:
Proof of Proposition 2: So far, we have:
From the proof of Proposition 1, we can substitute expressions for (∂2E/∂α2) and (∂2E/∂α∂l) to solve for (dα*/dl) then fill in the remaining terms:
Note that (du B /dl)) > 0 if the term inside the brackets of (du B /dl) is greater than 1. This will be true if
This holds by the solvency condition and the fact that (eI−α)(1−l)x2 > 0. Thus, unless we are in a corner region in which α*=0 and ((dα*)/(dl))=0, we have our result.
Appendix II
Additional Quotations
Quotations Illustrating the Focus on the Quality of Assets on the Balance Sheet
During the quarter, we sold a variety of assets, not just the most liquid. We sold $4.2 billion of loans, of which 45 percent were mezzanine loans and 55 percent were senior loans … [W]e sold approximately $3.5 billion of level 3 assets and also had additional writedowns of approximately $2 billion … however, this reduction will be offset by net transfers in and other activity of approximately $3.5 billion.
—Ian Lowitt, CFO, Lehman Brothers, Q2 2008 Earnings Call
Prashant Bhatia (Citigroup): The gross long CDO exposure was down about $6 billion. Can you breakout what drove the decline and the same on the short side that was down by about $4 billion?
John Thain (Chairman and CEO, Merrill Lynch): There are sales, but it is mostly markdowns.
—Merrill Lynch, 2008:Q2 Earnings Call
We continued to reduce concentrated risk positions including leverage in Real Estate related loans. These asset classes represented 57 percent of tangible common equity at year-end down from 85 percent in the third quarter and 224 percent at year-end 2007. Legacy leverage loan exposure now stands at $7 billion, down from $52 billion at its peak last year. Our commercial Real Estate portfolio declined by approximately 25 percent to $10.9 billion in the fourth quarter alone.
—David Viniar, CFO, Goldman Sachs, 2008:Q4 Earnings Call
On the Importance of the Existence of LOLR Facilities for Private Funding Markets
In addition to [our] conservative risk framework, the Federal Reserve’s announcement on Sunday has introduced policies that go even further in mitigating our liquidity risk. The Fed has agreed to accept in the PDCF a broad range of collateral … So while our access to funding continues to be quite robust, the Fed’s actions greatly diminished the liquidity risk associated with our secured funding book …
[I]t is not like we have assets sitting there that we need to take to the Fed. But I think the Fed being there is really a big statement about liquidity risk. I’ve seen reports. I’ve seen articles about tri-party repo is too risky; it is going away. I have never given any of them any weight. But with the Fed doing that it kind of takes away most of that risk—if you thought it was there.
—David Viniar, CFO, Goldman Sachs, Q3 Earnings Call, September 16, 2008
The most difficult to fund non-central bank eligible assets which includes capital commitments and other receivables represents roughly 10 percent of our funding needs and had a weighted average maturity of greater than 90 days.
—Colm Kelleher, CFO, Morgan Stanley, Q3 Earnings Call, September 16, 2008
Investment Banks Felt Free to Optimize Risk and Return Without Incorporating the Goals of and the Risks to the Liquidity Facilities
[The] Federal Reserve’s decision to create a lending facility for primary dealers and permit a broad range of investment-grade securities to serve as collateral improves the liquidity picture, and, from my perspective, takes the liquidity issue for the entire industry off the table.
—Dick Fuld, “Wall Street Watches Lehman Walk on Thin Ice,” MarketWatch, The Wall Street Journal, March 17, 2008
Meredith Whitney (Oppenheimer): Could you set a market by hitting whatever cash bid there is out there and just get it over with?
John Thain (Chairman and CEO, Merrill Lynch): No, I do not think we want to do dumb things and so we have been balanced in terms of what we sold and at what prices we sold them. We have not simply liquidated stuff at any price we could get. At some point some of the return profiles that people want … you would not want us to sell the assets. We will continue to sell assets but in a way that makes sense from generating returns to our shareholders.
—Merrill Lynch, 2008:Q2 Earnings Call, July 17, 2008
Michael Hecht (Bank of America): [S]hould we expect leverage to continue to fall? What … [are] the implications for the type of ROE you guys can earn through this cycle? Are you seeing any pressure here from regulators, rating agencies, investors, to bring leverage down further?
Colm Kelleher (CFO, Morgan Stanley): Well we are obviously in constant touch with a broad array of regulators … what we’ve been doing is taking down the balance sheet because on a risk-adjusted basis, that’s what we want to do … So the answer is we’re in constant dialogue but we’re not getting any pressure in terms of the cycle ROE.
—Morgan Stanley, 2008:Q2 Earning Call, June 18, 2008
We are still placing a significant emphasis on our capital and liquidity to ensure that we have enough dry powder to continue investing in our businesses and to take advantage of attractive risk-adjusted opportunities …
People cannot gloss over lightly the event that took place in March and the effect that had on the market … So that is what made us pull down the sails, sail close to shore, preserve our ammunition. We do have excess capital, we do have excess liquidity, we do have leverage if we want to where we can take risk …
So I’m not saying we’re in risk reduction model; we clearly have reduced the risk, we’ve reduced the balance sheet, we’re liquid, we’ve got capital, we clearly feel we can make money through bear cycles and bull cycles and we’re just waiting for the right risk-adjusted opportunity to come along …
I kind of think we’re in the right sort of spot at the moment in terms of giving us the optionality we need to be opportunistic and to be defensive. Obviously we’d like a more normalized market to get rid of some of this legacy portfolio so we can optimize return to the balance sheet. But I think we feel comfortable that we’ve got optionality sitting where we are at the moment given all the uncertainty around it …
Some of these [market] prices, frankly, have got to silly and irrational levels so we’re going to have the ability to take advantage of that …
—Colm Kelleher, CFO, Morgan Stanley, 2008:Q2 Earnings Call, June 18, 2008
… what we will do with our exposures and our risk will really depend on the opportunities we see in the market … if we see opportunities we will take advantage of it and increase risk and if we think opportunities are not as good, we’ll decrease risk.
—David Viniar, CFO, Goldman Sachs, Q2 Earnings Call, June 17, 2008
Given our significant reduction of legacy assets and our lack of direct consumer exposure, we believe that our balance sheet is strong. Importantly, we have significant capital to take advantage of market opportunities as they arrive in 2009 … As a result of the broader market dislocation, the competitive landscape has changed. Across many of our businesses, trading margins are robust and the premium on risk capital is higher than we’ve seen in years. In this type of environment return on assets is improving.
—David Viniar, CFO, Goldman Sachs, Q4 Earnings Call, December 16, 2008
Global Markets Financing and Services revenues increased to a record level, up approximately 25 percent from the prior-year period, as the firm took advantage of opportunities to both add clients and increase average balances.
—Merrill Lynch, 2008:Q2 Press Release, July 17, 2008
Prashant Bhatia (Citigroup): On the prime brokerage side, it looks like … record revenues even coming off of seasonally strong last quarter, how much of that is share gain versus pricing?
Colm Kelleher (CFO, Morgan Stanley): Pretty much pricing to be honest; not share gain.
—Morgan Stanley, 2008:Q3 Earnings Call, September, 16, 2008
[Morgan Stanley was] targeting … capital to businesses where [it] has leading positions and where the Firm believes it will have better risk-adjusted returns … engaging in a deliberate and focused reduction of balance sheet-intensive businesses including a resizing of Prime Brokerage, the exit of select Proprietary Trading strategies, the reduction of Principal Investments and the closure of Residential Mortgage Origination.
—Q4 Earnings Press Release, Morgan Stanley, December 17, 2008
Roger Freeman (Barclays Capital): With respect to the balance sheet decline this quarter can you help us think about some of the buckets there?
Colm Kelleher (CFO, Morgan Stanley): Our adjusted leverage has come down as well. So it was broad. It is easier to reduce a matched book which we did but if you think about it in terms of where we are it was pretty broad based. About half the reduction came from a reduction in prime brokerage balances and the rest is pretty much spread out.
—CFO, Morgan Stanley, 2008:Q4 Earnings Call, December, 18, 2008
