INSTITUTIONAL TRADING STRATEGIES AND CONTAGION AROUND CRISIS PERIODS. V. Ravi Anshuman Rajesh Chakrabarti Kiran Kumar

Size: px

Start display at page:

Download "INSTITUTIONAL TRADING STRATEGIES AND CONTAGION AROUND CRISIS PERIODS. V. Ravi Anshuman Rajesh Chakrabarti Kiran Kumar"

Amelia Gibson
5 years ago
Views:

1 INSTITUTIONAL TRADING STRATEGIES AND CONTAGION AROUND CRISIS PERIODS V. Ravi Anshuman Rajesh Chakrabarti Kiran Kumar

2 How do FII Investments affect stock market? April 2, 2012, MINT

3 LITERATURE Brennan and Cao (1997) argue that FIIs use lagged domestic returns as an input in forming expectations about future returns FII flows may be related to lagged domestic returns. Grinblatt, Titman and Wermers (1995), Nofsinger and Sias (1999); Wermers (1999, 2000), Cai, and Zheng (2004); Richards (2005), Bennett et al (2003); Parrino et al (2003)) provide evidence that FIIs resort to positive feedback trading strategies. Jotikasthira, Lundblad, Ramadorai (2012) look at asset fire sales by international fund managers as a channel of contagion Grinblatt and Keloharju (2000) conclude that foreign investors mostly tend to be momentum traders and whereas Finnish domestic investors tend to be contrarian investors. These findings suggest that the FIIs and DIIs use different trading strategies.

4 MOTIVATION Do FIIs and domestic institutions behave the same way in the Indian market? Do institutional trading strategies change during times of crisis? If FIIs are destabilizing, then the crisis period is the best place to examine this hypothesis. Are there any contagion effects? Bringing bad international news to Indian shores? Institutional Trading Strategy?? Crisis Contagion

5 CRISIS PERIOD PRE-CRISIS PERIOD 6 th April, 2007 to December 31 st, The value of the NIFTY rose from to CRISIS PERIOD 1 st January, 2008 to 31 st December, The value of the NIFTY fell from to POST-CRISIS PERIOD 1 st Jan 2009 to 31 st December, The value of the NIFTY rose from to

6 Market and Investor Category Trends across the time periods

7 DATA FIIs and DIIs own a significant proportion of the equity market in India. In the year , an average of 3883 foreign investors were registered with SEBI. Collectively 80% of the assets are under the custodianship of FIIs (25%) and DIIs (55%). In terms of turnover, for the sample period, the FIIs turnover is 28.53% and DII turnover is 11.50%.

8 NET FII (DII) ORDER FLOW

9 Descriptive Stats Panel A : Pre-Crisis Period - 16th April 2007 to 31st Dec 2007 Mean Median Max Min StdDev Skew Kurt Obs FII Net DII Net NIFTY Returns S&P500 Returns Panel B: Crisis Period - 01st Jan 2008 to 31st Dec 2008 FII Net DII Net NIFTY Returns S&P500 Returns Panel C : Post-crisis Period - 01st Jan 2009 to 31st Dec 2009 FII Net DII Net NIFTY Returns S&P500 Returns Panel D: Mean differences between Pre-crisis, Crisis and Post-crisis Mean(pre) = Mean(Cri) Mean (Cri)=Mean(Recov) Mean(Pre)= Mean (Recov) t-stat p-value t-stat p-value t-stat p-value FII Net DII Net Nifty Returns SP500 Returns

10 CORRELATIONS Table 2 : Pair-wise Correlation coefficients between the variables This table reports pair wise correlations between NIFTY daily continuous compounding returns (prior, same and future) and investor category-wise daily net purchases. The correlations are reported separately for full period, Pre-Crisis, crisis and post-crisis period. ***, **, * indicate statistical significance at 1%, 5% and 10% level. PairWise Correlations (1) (2) (3) (4) (5) Panel A : Full Period : 16th April 2007 to 31st Dec 2009 FII Net (1) 1 DII Net (2) -0.54*** 1 NIFTY Returns (t-1) (3) 0.38*** -0.26*** 1 NIFTY Returns (t+1) (4) 0.07* NIFTY Returns (t) (5) 0.49*** -0.16*** Panel B: PreCrisis period : 16th April 2007 to 31st Dec 2007 FII Net 1 DII Net NIFTY Returns (t-1) 0.41*** -0.22*** 1 NIFTY Returns (t+1) NIFTY Returns (t) 0.47*** -0.22*** Panel C: Crisis Period : 01st Jan 2008 to 31st Dec 2008 FII Net 1 DII Net -0.47*** 1 NIFTY Returns (t-1) 0.36*** -0.25*** 1 NIFTY Returns (t+1) NIFTY Returns (t) 0.47*** -0.22*** Panel D: Post-crisis Period : 01st Jan 2009 to 31st Dec 2009 FII Net 1 DII Net -0.43*** 1 NIFTY Returns (t-1) 0.39*** -0.25*** 1 NIFTY Returns (t+1) NIFTY Returns (t) 0.55***

11 Scatter plot of net FII and net DII order flows FII NET DII NET

12 INTENSE TRADING DAYS To identify days of intense trading, we sort the trading days into deciles based on net FII flows. Decile 1 contains days where FII selling was the most intensive, and Decile 10 contains days where FII buying was most intensive. We define intense buying days in two ways: those trading days that lie in the top decile (Decile 10) or those trading days that lie in the top two deciles (Decile 9 and Decile 10). Similarly, we define intense selling days as trading days that lie in the bottom decile (Decile 1) or those trading days that lie in the bottom two deciles (Decile 1 and Decile 2).

13 Market returns around Intense FII trading Panel A : Response of NIFTY to extreme movements of FII_NET Decile k = -5 k = -4 k = -3 k = -2 k = -1 k = 0 k = 1 k = 2 k = 3 k = 4 k= 5 Pre-Crisis (***) (***) (***) 1 & (***) (***) (***) (*) 9& (*) (*) (***) (***) (***) (***) (***) (***) (***) (***) (*) (*) (*) (***) (***) (***) (***) Crisis (*) (***) (***) (***) (***) 1& (***) (***) (***) 9& (***) (***) (***) ((***) (***) Post-crisis (***) (***) 1& (***) (***) 9& (***) (***) (***) (*) (***) (***)

14 Market returns around Intense DII trading Panel B : Response of NIFTY to extreme movements of DII_NET Decile k=-5 k=-4 k=-3 k=-2 k=-1 k=0 k=1 k=2 k=3 k=4 k=5 Pre-Crisis (***) (*) (***) (***) (***) (***) 1 & (***) (***) (***) (***) (***) (***) (***) (***) 9 & (*) (*) (***) (***) (*) Crisis 1 & (***) 9 & (***) (***) (***) (***) (***) (***) Post-crisis (***) (***) (***) (***) (***) (***) (***) (***) 1 & (***) (***) (***) (***) 9 & (***) (*) (***) (***) (***)

15 FII trading behavior around good and bad times of Nifty PreCrisis FII_Net DII Net ret(t-1)< ret(t-1)> Diff (p-value) ret(t+1)< ret(t+1)> Diff (p-value) ret(t)< ret(t)> Diff (p-value) ve feedback Postcrisis FII_Net DII Net ret(t-1)< ret(t-1)> Diff (p-value) ret(t+1)< ret(t+1)> Diff (p-value) ret(t)< ret(t)> Diff (p-value) ve feedback (but reduced) FII_Net DII Net Crisis ret(t-1)< ret(t-1)> Diff (p-value) ret(t+1)< ret(t+1)> Diff (p-value) ret(t)< ret(t)> Diff (p-value)

16 INSTITUTIONAL TRADING AROUND EXTREME MARKET MOVEMENTS What do we know about institutional trading activity around days of extreme market movements. We identify those trading days with extreme market movements. Decile 1 consists of those trading days with the least returns on the NIFTY index and Decile 10 consists of trading days with the highest returns on the NIFTY index. Then we determine the average net FII (DII) flows in the (-5, +5) window around the days of extreme returns on the NIFTY index.

17 Trading patterns of FII around extreme movements of NIFTY Panel A : Trading Behaviour of FII around extreme movements of NIFTY Pre-Crisis Crisis Post-crisis Decile k=-5 k=-4 k=-3 k=-2 k=-1 k=0 k=1 k=2 k=3 k=4 k= (*) (***) (***) (***) (***) 1& (***) (***) (***) (***) (*) (***) 9& (***) (***) (***) (***) (*) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (*) 1& (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) 9& (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (*) 1 & (***) (***) (***) 9& (***) (***) (***) (***) (***)

18 Trading patterns of DII around extreme movements of NIFTY Panel B : Trading Behaviour of DII around extreme movements of NIFTY Pre-Crisis Crisis Post-crisis Decile k=-5 k=-4 k=-3 k=-2 k=-1 k=0 k=1 k=2 k=3 k=4 k= (*) (*) (***) (***) (***) (***) (***) (***) 1 & (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) 9 & (***) (*) (***) (***) (***) (*) (***) (***) (***) (*) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) 1 & (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) 9 & (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) 1 & (***) (***) (***) (***) (***) (***) (***) (***) (***) (***) 9 & (***) (***) (***) (***) (***) (***) (*) (***) (***) (***) (***) (*) (***) (***) (***)

19 Daily VAR estimates R-squared PRE-CRISIS period NIFTY_ RET FII_NET DII_NET *** NIFTY_RET(-1) [ ] [ ] [ ] CRISIS period NIFTY_ RET FII_NET DII_NET *** ** [ ] [ ] [ ] POST-CRISIS period NIFTY_ RET FII_NET DII_NET *** [ ] [ ] [ ] FII_NET(-1) *** *** [ ] [ ] [ ] ** [ ] [ ] [ ] *** *** [ ] [ ] [ ] DII_NET(-1) ** *** [ ] [ ] [ ] *** [ ] [ ] [ ] *** [ ] [ ] [ ] Intercept *** [ ] [ ] [ ] *** *** [ ] [ ] [ ] *** [ ] [ ] [ ]

20 Impulse Response Graphs from VAR analysis - FII Pre Crisis Crisis Post Crisis Response of FII_NET_VALUE to Cholesky One S.D. Innovations Response of FII_NET_VALUE to Cholesky One S.D. Innovations Response of FII_NET_VALUE to Cholesky One S.D. Innovations FII_NET_VALUE DII_NET_VALUE NIFTY_RET (t-1) (t-1) (t-1) FII_NET_VALUE DII_NET_VALUE NIFTY_RET FII_NET_VALUE DII_NET_VALUE NIFTY_RET

21 CONTAGION EXTREME MOVEMENT IN S&P 500 FII (DII) TRADING NIFTY RETURNS

22 Effect of past S&P returns on future Nifty returns Surprise matters more DepVar: Nifty_Fut PreCrisis Crisis Post-crisis Coeff t-stat p-value Coeff t-stat p-value Coeff t-stat p-value Intercept dii_net DUP_DII DD_DII DUP_FII DD_FII fii_net SPRT_PAST AR(1) MA(1) AdjR-sq DW test

23 CONCLUSIONS FIIs seem to be acting as short-term momentum traders who trade in the direction of the market. DIIs act as short-term contrarian traders whose trades dampen the impact of fluctuations in FII trading. DIIs seem to provide a stabilizing influence on the market. At the same time, it is not possible to infer that FIIs are destabilizing the Indian equity markets. No significant evidence to show that FII trades transmit bad news from their home country to the Indian markets. Rather, it is only contrarian (good) news in the home country that finds its way into Indian markets via FII trades. FII trading is associated with contagion only during the crisis period, but in the opposite direction (spreading the antidote rather than the germ).

24 Thank You

FOREIGN FUND FLOWS AND STOCK RETURNS: EVIDENCE FROM INDIA

FOREIGN FUND FLOWS AND STOCK RETURNS: EVIDENCE FROM INDIA Viral V. Acharya (NYU-Stern, CEPR and NBER) V. Ravi Anshuman (IIM Bangalore) K. Kiran Kumar (IIM Indore) 5 th IGC-ISI India Development Policy