Slow Moving Capital Mark Mitchell CNH Partners Lasse Heje Pedersen NYU, NBER, and CEPR Todd Pulvino CNH Partners
Motivation Arbitrageurs normally provide liquidity by buying low and selling high This requires capital What happens if arbitrageurs loose capital? Frictionless economy: –New capital arrives instantly, e.g. Lucas (1978) Frictions matter: –Arbitrageurs depend on investors, Shleifer and Vishny (1997) –Margins may increase, Brunnermeier and Pedersen (2006) –Other traders lack infrastructure and information, Merton (1987)
This Paper: Empirically Instigate the Effect of Capital Outflow from Arbitrageurs Strategy: Identify markets in which we can estimate “Fundamental value” and market price Capital flows to and from natural liquidity providers Actions and realized returns of liquidity providers We Analyze Three Cases 1.Capital outflow due to redemptions: –Convertible bonds, Capital outflow due to exit of large trader caused by losses in other markets –Convertible bonds, Capital outflow due to losses –Merger arbitrage, 1987
Findings Large capital shocks can lead to: Natural liquidity providers become liquidity demanders New capital arrives only after months Prices drop -- and rebound slowly Realized returns initially negative, then turn positive
Convertible Bonds Convertible bond: –Corporate bond + call option (+ more) Theoretical value can be inferred from –Issuer stock price –Stock price volatility –Option-implied volatility –Risk-free interest rates –Credit spreads
Convertible Bond Arbitrage Buy convertible bond if it trades at a discount Short the issuers stock Potentially: –Short risk-free bonds –Short non-convertible bonds (or buy CDS) –Short stock options
Analysis 1: Convertible Bond Arbitrage Capital Outflows in 2005 Natural liquidity providers: Convertible Bond Arbitrage Hedge Funds (HFs) Capital outflows in 2005: –2005Q1: 20% capital redeemed –2005Q1 – 2006Q1: assets fell by half Convert Arb HFs sold convertible bonds
Redemptions in 2005
Adjusted Holdings of Convertible Bonds
Qtr Convert Arb HFsMulti-strategy HFs Convert Mutual Funds
Convertible Bond Arbitrage Returns and Market Price / Theoretical Value
Interpretation Prices drop and rebound Price-to-fundamentals lowest around redemption notices (45 days before end of June and end of December) Returns negative, then positive Response by other traders: –Multi-strategy hedge funds –Mutual funds
The Case of Amaranth In 2005, Amaranth had –Losses in convertible bonds –Profits in energy trading –Overall profit and no capital problems Decided to liquidate convertible bonds at time of significant cheapness Collapsed in 2006 due to losses in energy
Analysis 2: LTCM Blowup in Implications for Convertible Bonds Large hedge fund LTCM had losses due to Russian default, option positions, etc. Had to liquidate large position in convertible bonds
Convertible Bond Arbitrage Returns and Market Price / Theoretical Value
Analysis 3: Merger Arbitrage and the 1987 Crash In a merger, “target” is bought at a premium. At announcement, target increases in value, typ % But, there remains a “deal spread,” typically around 3% Due to –Risk of deal failure –Selling pressure: Mutual funds sell after announcement Merger arbitrageurs buy target –Stock deal: hedge by shorting acquirer –Cash deal: no hedge
Merger Arbitrage and the 1987 Crash
Interpretation Oct : U.S. House Ways and Means Committee proposed legislation Oct. 19 (Black Monday) and 20: crash Oct: 21-31: –Stock market rebounds –Congress backs off proposed legislation –But, prop traders kept selling Berkshire Hathaway Annual Report (Warren Buffett): “During 1988 we made unusually large profits from [risk] arbitrage … the trick, a la Peter Sellers in the movie, has simply been ‘Being There.’ ”
Conclusion: The Speed of Arbitrage Findings –Liquidity providers can become demanders –New capital arrives slowly (new and old arbitrageurs) –Prices drop and rebound –Realized returns are initially negative and later turn positive and large Conclusions: –Frictions matter –The process of arbitrage is far from instantaneous –Effect could be due (in part) to reduced market liquidity