Asymmetric Information and Inventory Concerns in Over-the-Counter Markets

Transcription

1 Asymmetric nformation and nventory Concerns in Over-the-Counter Markets Julien Cujean U of Maryland (Smith) jcujean@rhsmith.umd.edu Rémy Praz Copenhagen Business School rpr.fi@cbs.dk Thematic Semester on Commodity Derivatives Markets HP March 26, 2015

2 Context Over-the-counter (OTC) markets Decentralized trading Trade details negotiated in bilateral meetings Trading risks: timing, quantity, price 1/35

3 Research Question How does information about the trading needs of your counterparties affect an OTC market? Costs of trading Market participation Allocative efficiency and welfare 2/35

4 Relevance Regulators introduce post-trade transparency TRACE, Dodd-Frank Act, MiFD Benefits: Better valuation of asset BESSEMBNDER, MAXWELL, AND VENKATARAMAN (2006) GOLDSTEN, HOTCHKSS, AND SRR (2007) EDWARDS, HARRS, AND PWOWAR (2007) Costs: Reduced liquidity provision Lobbying material by SFMA, surveys DUFFE (2012) ASQUTH, COVERT, AND PATHAK (2013) 3/35

5 Relevance Transparency and commodity derivatives Risk magazine (January 16, 2015): liquidity at the back-end of the futures curve has dried up Southwest Airlines: Spread on long-term jet fuel swaps up by 35bps. Timothy Massad, CFTC chairman (February ): n an illiquid market, real-time reporting can hurt that ability [the ability of commercial end-users to hedge] 4/35

6 Reduced Liquidity Provision? a a r huge buy by a dealer quote? P = 95 b V = 100 quote? P = 90 b V = 100 5/35

7 Main Findings Transparency affects allocative efficiency (%) inventory costs (%) dispersion of transaction prices (%) market participation (ambiguous and fragile) welfare (ambiguous and fragile) 6/35

8 Literature Review OTC markets DUFFE, GÂRLEANU, AND PEDERSEN (2005, 2007) LAGOS AND ROCHETEAU (2007, 2009) OTC markets and asymmetric information BLOUN AND SERRANO (2001) DUFFE AND MANSO (2007), DUFFE, MALAMUD, AND MANSO (2009, 2010, 2014) nventories HO ANDSTOLL (1980, 1981) GROSSMAN AND MLLER (1988) Formalism DAMOND (1982) HUANG, MALHAMÉ, AND CANES (2006), LASRY AND LONS (2007) 7/35

9 Outline Model Market equilibrium Market participation 7/35

10 Assets 1. Risk-free rate r > 0 2. Risky asset paying dividends at the rate dd d = m d dt + d db t Model 8/35

11 nvestors Continuum of agents Endowment at the rate Time-varying exposures d a t = Z a t dd t dz a t = a db a t see LO, MAMAYSKY, AND WANG (2004) Model 9/35

12 Trading / Risky asset traded on an illiquid over-the-counter (OTC) market Entry costs apple 1 Expected search time is (# market participants) Bargaining over the transaction details see DUFFE, GÂRLEANU, AND PEDERSEN (2005, 2007) Model 10/35

13 Trading /: Bargaining (i) a asks b for a quote. (ii) f b finds it optimal b receives a signal with X B(1, ), µ b quotes a price p a chooses a quantity q s a = Xz a +(1 X), is the transparency of the market Model 11/35

14 Trading /: Bargaining (i) a asks b for a quote. (ii) f b finds it optimal b receives a signal with X B(1, ), µ b quotes a price p a chooses a quantity q s a = Xz a +(1 X), is the transparency of the market Model 11/35

15 Trading /: Bargaining (i) quoted price P(s) Signal s about a a b Transparency is P[signal is correct] (ii) traded quantity Q(P) Model 12/35

16 Preferences nvestors maximize expected CARA utility from consumption V t = sup (c s) s t E t apple Z 1 t e (s t) e c s ds Model 12/35

17 Timeline Entry Decision initial exposures entry costs apple Trading Endowment shocks 0 t 1 Model 13/35

18 Outline Model Market equilibrium Market participation Market equilibrium 13/35

19 Market Equilibrium Take as given the investors in the market trading by others = flow equation my trading distribution of exposures HJB Market equilibrium 14/35

20 HJB Equation V (w, z) =sup {U ( c s) V w (w, z) c} c + V w (w, z)(rw + zm d ) + 1 V ww (w, z)z V zz(w, z) z 2 2 "!# sup V (w qp (z q, s z), z + q) + E L(zq,sz ) 1 {zq2a} q V (w, z) 2 " EL(za,sa) sup E L(za,sa) [ V (w + Q (z a, p) p, z Q (z a, p)) s a] p V (w, z) # Market equilibrium 15/35

21 Type Dynamics dz t = z db t + 0 X r,t Q(z t, P(z q, z t )) +(1 X r,t) Q(z t, P(z q, )) X q,t Q(z t, P(z t, z r )) +(1 X q,t) Q(z t, P(z t, )) z t dn r t 1 A dn q t, Market equilibrium 16/35

22 Market Equilibrium Proposition There exists an equilibrium for which the value functions have the form V (t, w, z) = exp r v 0 (t)+v 1 z + v 2 z 2. Exponential convergence at rate 4 9 (1 + 2 ) Steady-state variance of exposures Var [ z 1]= 2 a 4 9 (1 + 2 ) Market equilibrium 16/35

23 Transparency Makes nventories Costly loghvalue functionl = vhzl t = 0.87 t = 1 -shz 0 L shz 0 L exposure z stationary density of exposure 2 1 t = 0.87 t = 1 sht=1l exposure z Market equilibrium 17/35

24 Transparency Makes nventories Costly reservation spread -shz 0 L exposure z t = 0.87 t = Corollary When the transparency increases, Trades become smaller Cross-sectional dispersion of prices increases Consistent with Southwest Airlines complaints. Market equilibrium 18/35

25 Outline Model Market equilibrium Market participation Market participation 18/35

26 Endogenous Market Participation Net benefits to joining the OTC market (z) = a (z E [ z 0 ]) 2 + b Var [ z 0 ] + c 2 a apple gross benefits won t join! entry costs 0 E[z] time 0 exposure Liquidation a is % in & in ntermediation b is (& 0) in % in entrants are substitutes Anticipated risk-sharing c is % in % in entrants are complements Market participation 19/35

27 Rational Market Participation E = {investors who enter the OTC market} E expected = E realized (, (z)) Solution Methods 1. Homogeneous initial exposure 2. Cases for which most investors enter the market 3. Numerics Market participation 20/35

28 Rational Market Participation E = {investors who enter the OTC market} E expected = E realized (, (z)) Solution Methods 1. Homogeneous initial exposure 2. Cases for which most investors enter the market 3. Numerics Market participation 20/35

29 Market Participation: Method 1 Homogeneous initial exposure Proposition f Var[ z 0 ]=0 No participation is an equilibrium Full participation is an equilibrium if (1) apple Partial participation is an equilibrium if (p) =apple, p 2 (0, 1) benefits costs t=0.87 t=1 0 1 proba of entry Market participation Participation and Welfare weakly decreasing in transparency. 21/35

30 Market Participation: Method 2 Solution available when there is full participation entry benefits bhzl entry costs k Most investors enter the market exposure z entry benefits bhzl entry costs k δ exposure z Market participation 22/35

31 Market Participation: Method 2 Proposition Around the full participation case, when a is large enough, two equilibrium paths for > full market participation and welfare can & in discontinuous participation drop when < full Ambiguous effect of transparency on market participation, trading delays, welfare Economy is fragile in the transparency Market participation 23/35

32 Market Participation: Method 3 Numerical solution: Assume strong enough complementarity Market Participation Realized Rational Expectations Beliefs Market participation 24/35

33 Transparency can Decrease liquidity Market Participation Trading Volume Welfare Transparency Market participation 25/35

34 Transparency can Decrease Liquidity increased transparency either increased trading costs or reduced liquidity reduced market participation increased market participation increased intermediation revenues improved liquidity Market participation 26/35

35 Transparency can Decrease Liquidity Consistent with Southwest Airlines claims. Consistent with CFTC s view that trade dissemination can hurt hedgers. Not consistent with BESSEMBNDER, MAXWELL, AND VENKATARAMAN (2006) GOLDSTEN, HOTCHKSS, AND SRR (2007) EDWARDS, HARRS, AND PWOWAR (2007) but common value effects drive these studies. Market participation 27/35

36 Transparency can ncrease Market Participation Market Participation Trading Volume Welfare Transparency Market participation 28/35

37 Transparency can ncrease Market Participation increased transparency either increased trading costs or reduced liquidity reduced market participation increased market participation increased intermediation revenues improved liquidity Market participation 29/35

38 Transparency can ncrease Market Participation Some swap execution facilities (SEF) have a name give-up on their order-book, and some do not. Dealers favor SEF with name give-up. Risk Magazine (February 11, 2015): Name give-up provides a vehicle for credit allocation Market participation 30/35

39 Equilibrium is Fragile Market Participation Trading Volume Welfare Transparency Market participation 31/35

40 Equilibrium is Fragile increased transparency either increased trading costs or reduced liquidity reduced market participation increased market participation increased intermediation revenues improved liquidity Market participation 32/35

41 Equilibrium is Fragile Empirical Evidence ASQUTH, COVERT, AND PATHAK (2013) document bond trading volumes being reduced by up to 40% after post-trade transparency was introduced. According to our model, this drop in trading volume was accompanied by a drop in welfare. Market participation 33/35

42 Equilibrium Selection Policy Recommendation Subsidizing liquidity provision eliminates the low participation equilibria. Norwegian central bank ensures a liquid sovereign bond market with attractive interest rates for dealers. Market participation 34/35

43 Wrapping up Trading is more costly in a transparent market but less trading in equilibrium Market participation ambiguous and fragile in transparency Welfare ambiguous and fragile in transparency Market participation 35/35

44 Thank you! 35/35

45 35/35

46 Parameter Values Notation Parameter Value r interest rate 0.01 volatility of dividends 1 a volatility of exposure 0.52 scaling of matching function 1 risk-aversion 1 35/35