Implied Funding Liquidity

Size: px

Start display at page:

Download "Implied Funding Liquidity"

Eleanore Stewart
5 years ago
Views:

1 Implied Funding Liquidity Minh Nguyen Yuanyu Yang Newcastle University Business School 3 April / 17

2 Outline 1 Background 2 Summary 3 Implied Funding Liquidity Measure 4 Data 5 Empirical Results 2 / 17

3 Background I Arbitrage is a trading strategy (Sharpe et al, 1998) The put-call parity is a no-arbitrage relationship When put-call parity holds, implied volatilities should be the same When there is a violation of put-all parity: caused by some specific market features, such as short-sales restrictions and non-synchronicity, which means that the put-call parity deviations do not represent profitable arbitrage opportunities (Easley et al, 1998; Cremers and Weinbaum, 2010) The implied volatility spreads contain critical information about the future change for the price of the underlying stock which has not yet been reflected in stock price quotes (Cremers and Weinbaum, 2010) 3 / 17

4 Background II The put-call implied volatility spreads are positively relative to future returns of the underlying stock (Bali and Hovakimian (2009), Cremers and Weinbaum (2010), Doran et al (2013) and Muravyev et al (2013)) Bali and Hovakimian (2009): imbalance in put-call volatilities can have predictive power in explaining stock returns Option prices might also rely on supply and demand of the securities in financial markets Investors buy put to protect against downside risk, which will increase put option demand, and influence their prices and put-call parity The option demand move option prices evidence: Bollen and Whaley (2004) and Garleanu, Pedersen, and Poteshman (2009) In normal market conditions, financial intermediaries can engage in arbitrage trades to remove anomalies 4 / 17

5 Background III In reality, these institutions face significant costs and funding constraints preventing arbitrage trades to happen and therefore leading to deviations from the fundamental arbitrage-free relationships Brunnermeier and Pedersen (2009) provide a model that links an asset s market liquidity and funding liquidity Traders provide market liquidity, and their ability to do so depends on their availability of funding Conversely, traders funding, ie, their capital and margin requirements, depends on the assets market liquidity Moreover, under certain conditions, margins are destabilizing and market liquidity and funding liquidity are mutually reinforcing, leading to liquidity spirals Despite significant theoretical and empirical works, the literature remains mute whether option prices carry any information that explains the variations of the average excess returns 5 / 17

6 Summary This study proposes a market-wide liquidity measure by linking systematic deviations from Put-Call parity in the equity option markets to funding constraints We show that this implied funding liquidity measure carry additional information about stock returns, which are not captured by conventional liquidity proxies or due to established factors in the US equity markets We show that investing in the stocks with the largest exposure to the innovations in implied funding liquidity and shorting the stocks with the smallest generate significant returns even controlling for transaction costs 6 / 17

7 Measure Construction The implied funding liquidity, IFL t, is measured as follows: IFL t = VW i,t IV C i,t IV P i,t N i=1 VW i,t Where IVi,t C and IV P i,t are the implied volatilities obtained from the call and put options of the same strike price, VW j,t is the market capitalization of stock i at time t, N is the number of stock at that time The IFL innovation is the negative log change of the IFL: IFL t = (ln(ifl t ) ln(ifl t 1 )) 100 where IFL t denotes the innovation of the value weighted absolute implied volatility spreads IFL t 7 / 17

8 Data In this study, we use the historical price and implied volatility data for the US equity options markets obtained from OptionMetrics This dataset includes call and put option mid prices, their strike prices, open interest, volume, remaining time to expiration of the option (expressed in years) We consider a sample ranging from 04 January 1996 to 31 August / 17

9 Implied Funding Liquidity (a) Monthly Level (b) Monthly Innovations This figure presents the time series of the monthly level and innovations of the standardised implied funding liquidity measure over the sample from 04 January to 31 August / 17

10 Market Return Predictability Panle A: CRSP Value Weighted Returns Horizon (in months) IFL t 1 MKTEX t 1 Adj R-Sq *** (t-stat) (-268) (132) ** *** (t-stat) (-223) (1250) *** (t-stat) (-075) (751) Panle B: S&P 500 Returns Horizon (in months) IFL t 1 MKTEX t 1 Adj R-Sq ** (t-stat) (-245) (100) * 11631*** (t-stat) (-196) (1191) *** (t-stat) (-053) (742) This table shows the IFL predictability with horizons of one month, three months, and six months The full sample period spans from January 1996 to August 2015the significance level is labelled in / / for 10%, 5% and 1% level, respectively 10 / 17

11 IFL Innovation Betas Panel A: 25 Portfolios Formed on Size and Book-to-Market Betas Low B/M High B/M Small 029*** 022*** 018*** 017*** 020*** (469) (418) (430) (433) (458) 2 023*** 019*** 017*** 017*** 019*** (448) (453) (434) (435) (435) 3 024*** 018*** 017*** 018*** 019*** (499) (469) (466) (489) (503) 4 021*** 020*** 020*** 019*** 017*** (488) (575) (539) (535) (430) Big 015*** 014*** 016*** 012*** 017*** (467) (478) (492) (377) (434) Panel B: 30 Industry Portfolios Industry Betas Industry Betas Industry Betas Autos 021*** Clths 016*** Food 011*** (364) (337) (386) Beer 010*** Cnstr 023*** Games 025*** (281) (500) (497) Books 022*** Coal 027*** Hlth 013*** (524) (288) (443) BusEq 022*** ElcEq 021*** Hshld 008** (373) (437) (251) Carry 018*** FabPr 023*** Meals 014*** (413) (459) (427) Chems 016*** Fin 016*** Mines 023*** (376) (376) (389) Oil 016*** Steel 029*** Servs 020*** (397) (466) (414) Other 015*** Telcm 015*** Util 011*** (353) (391) (373) Paper 012*** Trans 011*** Smoke 008 (325) (288) (155) Rtail 015*** Txtls 017*** Whlsl 014*** (413) (272) (410) 11 / 17

12 Fama-MacBeth Cross-Sectional Regression Results Panel A: 25 Portfolios Model λ 0 λ IFL λ MKT λ SMB λ HML λ Mom R 2 Univariate 10189*** 01876*** 857% (2587) (2960) CAPM 03976*** 00194*** 10287*** 6939% (1046) (583) (3458) FF 02299*** 00111** 09965*** 05188*** 02734*** 9034% (635) (257) (7711) (389) (354) FF, Mom 02469*** 00113** 09868*** 05220*** 02645*** *** 9048% (760) (259) (8456) (392) (346) (-313) Panel B: 30 Industry Portfolios Model λ 0 λ IFL λ MKT λ SMB λ HML λ Mom R 2 Univariate 09197*** 01708*** 6% (2155) (1832) CAPM 03420*** 00145** 09564*** 5047% (892) (276) (1717) FF *** 00140** 10040*** 00740** 03474*** 5810% (533) (262) (1897) (211) (560) FF, Mom 02731*** 00145** 09804*** 00817** 03258*** *** 5901% (697) (276) (1886) (232) (515) (-278) 12 / 17

13 Realised vs Predicted Average Returns (c) 25 Portfolios (d) 30 Industry Portfolios This figure presents the actual versus predicted average returns from the one factor model with implied funding liquidity for the 25 Fama and French s portfolios and 30 industry portfolio, respectively The sample spans from 04 January 1996 to 31 August / 17

14 Comparison with other Liquidity Measures Pastor and Stambaugh (2003) measure market liquidity of an individual stock based on the impacts of orders on stock returns and show that the innovations in the liquidity average across all of the stocks can explain the variations in stock returns Brunnermeier et al (2008) use the Treasury-LIBOR interest rate spread (TED spread) to capture funding constraints in financial markets and show that the Ted Spread can affect currency returns In the Treasury markets, Fontaine and Garcia (2012) develop a liquidity variable obtained from the new and old bond spread while Hu et al (2013) obtain the noise measure from the deviations between the observed bond yields and the predicted yields calibrated from Nelson and Siegel (1987) model 14 / 17

15 Comparison with other Liquidity Measures Panel A: 25 Portfolios Model λ 0 λ IFL λ MKT λ PS λ TED λ FG λ Noise R 2 +PS 04153*** 00240*** 10273*** % (1087) (659) (3182) (053) +TED 06529*** 00227*** 10219*** *** 6993% (769) (651) (3267) (-319) +FG 04037*** 00238*** 10304*** *** 6977% (1083) (672) (3430) (-505) +NOISE 05015*** 00253*** 10255*** % (471) (732) (3062) (-081) +ALL 05370*** 00212*** 10267*** *** *** % (502) (785) (2901) (015) (-341) (-294) (091) Panel B: 30 Industry Portfolios Model λ 0 λ IFL λ MKT λ PS λ TED λ FG λ Noise R 2 +PS 04132*** 00150** 09512*** 22133* 5089% (964) (260) (1739) (182) +TED 06883*** 00138** 09462*** *** 5088% (814) (236) (1684) (-412) +FG 03589*** 00155** 09571*** *** 5073% (872) (264) (1723) (-483) +NOISE 05674*** 00181*** 09478*** ** 5083% (695) (311) (1669) (-268) +ALL 06054*** 00120** 09440*** *** ** % (630) (223) (1711) (135) (-292) (-267) (027) The basic model is return = IFL + MKT This table reports s R-squares which is the cross-sectional R-squares, and the Newey West t-stat is reported in the bracket and the significance level is labelled in / / for 10%, 5% and 1% level, respectively 15 / 17

16 Economic Value We examine a portfolio strategy by investing in the stocks with the largest exposures to implied funding liquidity innovations and shorting the stocks with the lowest exposures to these liquidity innovations We sort stocks into ten portfolios based on these exposures obtained by regressing the daily stock returns to the daily liquidity innovations over the past two years We keep the composition of the portfolios constant over the horizon of one year and rebalance the portfolios once a year 16 / 17

17 Economic Value Panel A: Portfolio Sorted Returns Portfolio Average Return CAPM Alpha FF3 Alpha Carhart4 Alpha MR p-value < 0001 (-17936) (-68892) (-67489) (-67942) < 0001 (-14023) (-68440) (-67031) (-67482) < 0001 (-10626) (-68291) (-66879) (-67329) < 0001 (-71428) (-68200) (-66786) (-67237) < 0001 (-17108) (-68115) (-66698) (-67150) < 0001 (41460) (-68040) (-66622) (-67074) < 0001 (102758) (-67960) (-66542) (-66995) < 0001 (146820) (-67866) (-66447) (-66902) < 0001 (167680) (-67695) (-66277) (-66735) < 0001 (138144) (-67011) (-65588) (-66058) Panel B: Portfolio Sorted Returns with Transaction Cost Portfolio Average Return CAPM Alpha FF3 Alpha Carhart4 Alpha Sharpe Ratio H-L 05031*** 05029*** 05022*** 05002*** (197003) (193753) (195009) (192194) H-L (Adjusted) 04671*** 04667*** 04665*** 04637*** (181681) (178821) (181850) (177255) 17 / 17

Online Appendix to. The Structure of Information Release and the Factor Structure of Returns

Online Appendix to. The Structure of Information Release and the Factor Structure of Returns Online Appendix to The Structure of Information Release and the Factor Structure of Returns Thomas Gilbert, Christopher Hrdlicka, Avraham Kamara 1 February 2017 In this online appendix, we present supplementary