18
PHI 1 – Unified Algorithmic Trading Sets you free to think markets. Propels you beyond the status quo.

PHI 1 Unified Algorithmic Trading · 2020. 11. 24. · Statistical Arbitrage 1. Cipla and Sun Pharma 2. NestleInd and Dabur (Infy?) 3. M and M Limited and Tata Motors 4. Sbin, Indian

  • Upload
    others

  • View
    2

  • Download
    0

Embed Size (px)

Citation preview

Page 1: PHI 1 Unified Algorithmic Trading · 2020. 11. 24. · Statistical Arbitrage 1. Cipla and Sun Pharma 2. NestleInd and Dabur (Infy?) 3. M and M Limited and Tata Motors 4. Sbin, Indian

PHI 1 – Unified Algorithmic Trading

Sets you free to think markets. Propels you beyond the status quo.

Page 2: PHI 1 Unified Algorithmic Trading · 2020. 11. 24. · Statistical Arbitrage 1. Cipla and Sun Pharma 2. NestleInd and Dabur (Infy?) 3. M and M Limited and Tata Motors 4. Sbin, Indian

Statistical Arbitrage

#Correlation Check

self.corr, self.p_value = pearsonr(self.datas[0].close.array,

self.datas[1].close.array)

Logs('Pearsons correlation coefficient: %.3f' % self.corr)

Page 3: PHI 1 Unified Algorithmic Trading · 2020. 11. 24. · Statistical Arbitrage 1. Cipla and Sun Pharma 2. NestleInd and Dabur (Infy?) 3. M and M Limited and Tata Motors 4. Sbin, Indian

Correlation

Page 4: PHI 1 Unified Algorithmic Trading · 2020. 11. 24. · Statistical Arbitrage 1. Cipla and Sun Pharma 2. NestleInd and Dabur (Infy?) 3. M and M Limited and Tata Motors 4. Sbin, Indian

Correlation

Page 5: PHI 1 Unified Algorithmic Trading · 2020. 11. 24. · Statistical Arbitrage 1. Cipla and Sun Pharma 2. NestleInd and Dabur (Infy?) 3. M and M Limited and Tata Motors 4. Sbin, Indian

Stationarity

Page 6: PHI 1 Unified Algorithmic Trading · 2020. 11. 24. · Statistical Arbitrage 1. Cipla and Sun Pharma 2. NestleInd and Dabur (Infy?) 3. M and M Limited and Tata Motors 4. Sbin, Indian

Stationarity

Page 7: PHI 1 Unified Algorithmic Trading · 2020. 11. 24. · Statistical Arbitrage 1. Cipla and Sun Pharma 2. NestleInd and Dabur (Infy?) 3. M and M Limited and Tata Motors 4. Sbin, Indian

Statistical Arbitrage`

#Stationarity Check

X = sm.add_constant(list(self.datas[1].close.array))

model = sm.OLS(list(self.datas[0].close.array),X, hasconst=True)

self.val = model.fit()

self.adfuller = adfuller(self.val.resid)

Logs('ADF Statistic: %f' % self.adfuller[0])

Logs('p-value: %f' % self.adfuller[1])

Logs('Critical Values:')

for key, value in self.adfuller[4].items():

Logs('\t%s: %.3f' % (key, value))

Page 8: PHI 1 Unified Algorithmic Trading · 2020. 11. 24. · Statistical Arbitrage 1. Cipla and Sun Pharma 2. NestleInd and Dabur (Infy?) 3. M and M Limited and Tata Motors 4. Sbin, Indian

Statistical Arbitrage`

#Trading Logic

self.val1 = self.datas[0].close - (self.datas[1].close * self.val.params[1]) -

self.val.params[0]

self.boll = BollingerBands(self.val1, period=20, devfactor=1.5)

Page 9: PHI 1 Unified Algorithmic Trading · 2020. 11. 24. · Statistical Arbitrage 1. Cipla and Sun Pharma 2. NestleInd and Dabur (Infy?) 3. M and M Limited and Tata Motors 4. Sbin, Indian

Statistical Arbitrage

1. Cipla and Sun Pharma

2. NestleInd and Dabur (Infy?)

3. M and M Limited and Tata Motors

4. Sbin, Indian bank/pnb

5. hdfcbank and axis, icici

More Examples

Page 10: PHI 1 Unified Algorithmic Trading · 2020. 11. 24. · Statistical Arbitrage 1. Cipla and Sun Pharma 2. NestleInd and Dabur (Infy?) 3. M and M Limited and Tata Motors 4. Sbin, Indian

Audience Exercise

mean_p = "Mean :" +str(self.val2)

Logs(mean_p )

std_d = "Standard Deviation :"

+str(self.val3)

Logs(std_d )

def norm(self.val.resid,mean,std):

z_score = [(x - mean)/std for x in data]

return z_score

self.val5 =

norm(self.val.resid,self.val2,self.val3)

normd_s = "Normalized Series :"

+str(self.val5)

Logs(normd_s)

def mean(sample):

return sum(sample) / len(sample)

self.val2 = mean(self.val.resid)

def std_d(data):

mean = sum(data) / len(data)

# Square deviations

deviations = [(x - mean)

** 2 for x in data]

s1=sum(deviations)

s2=s1/len(data)

s3=s2 ** (1/2)

return s3

self.val3 = std_d(self.val.resid)

Bollinger Bands Parameter Tuning

Page 11: PHI 1 Unified Algorithmic Trading · 2020. 11. 24. · Statistical Arbitrage 1. Cipla and Sun Pharma 2. NestleInd and Dabur (Infy?) 3. M and M Limited and Tata Motors 4. Sbin, Indian

Audience Exercise

if (self.val1[0] > self.boll.lines.top[0]) and (self.position.size==0):

buy(data = self.datas[1])

sell(data = self.datas[0])

if (self.val1[0] < self.boll.lines.mid[0]) and (self.position.size!=0):

square_off()

When Residual breaches the Upper Band

Page 12: PHI 1 Unified Algorithmic Trading · 2020. 11. 24. · Statistical Arbitrage 1. Cipla and Sun Pharma 2. NestleInd and Dabur (Infy?) 3. M and M Limited and Tata Motors 4. Sbin, Indian

Audience Exercise

Dollar Neutral

Beta Neutral

Industry/Sector Neutral

Interest Rate Neutral

Oil Price Neutral

………..

Factor Neutrality

Identify all factor risks and neutralize your portfolio (We are

capturing mispricing risk only)

Page 13: PHI 1 Unified Algorithmic Trading · 2020. 11. 24. · Statistical Arbitrage 1. Cipla and Sun Pharma 2. NestleInd and Dabur (Infy?) 3. M and M Limited and Tata Motors 4. Sbin, Indian

Caution

The strategy heavily depends on the mean reversion of prices to their historical or predicted normal. This may not happen in certain cases and the prices can continue to drift away from the historical normal.

Systematically determine when the reversion in spread is unlikely

and exit the trade

Page 14: PHI 1 Unified Algorithmic Trading · 2020. 11. 24. · Statistical Arbitrage 1. Cipla and Sun Pharma 2. NestleInd and Dabur (Infy?) 3. M and M Limited and Tata Motors 4. Sbin, Indian

Further Reading

Practitioners

Page 15: PHI 1 Unified Algorithmic Trading · 2020. 11. 24. · Statistical Arbitrage 1. Cipla and Sun Pharma 2. NestleInd and Dabur (Infy?) 3. M and M Limited and Tata Motors 4. Sbin, Indian

Further Reading

Academics and Educators

Page 16: PHI 1 Unified Algorithmic Trading · 2020. 11. 24. · Statistical Arbitrage 1. Cipla and Sun Pharma 2. NestleInd and Dabur (Infy?) 3. M and M Limited and Tata Motors 4. Sbin, Indian

Further Reading

Skeptics

Page 17: PHI 1 Unified Algorithmic Trading · 2020. 11. 24. · Statistical Arbitrage 1. Cipla and Sun Pharma 2. NestleInd and Dabur (Infy?) 3. M and M Limited and Tata Motors 4. Sbin, Indian

ALG

OR

ITH

MIC

TR

AD

ING

WO

RK

FLO

WCREATION EVALUATION / REFINEMENT DEPLOYMENT

Screening

Basic Screening

Advanced Screening

Strategy Creation

Evaluation

Bulk Backtesting

Scenario Backtesting

Grading on Scenarios

Ideabook Simulation (Paper)

Deployment

Monitoring Scaling

Page 18: PHI 1 Unified Algorithmic Trading · 2020. 11. 24. · Statistical Arbitrage 1. Cipla and Sun Pharma 2. NestleInd and Dabur (Infy?) 3. M and M Limited and Tata Motors 4. Sbin, Indian

AVIRAL [email protected] | +91 95655 13605

GAURAV [email protected] | +91 79000 76658

WEBSITEhttps://www.phi1.io/

OFFICEPowerweave Heuristic Investment Technologies Pvt. Ltd.Plot No. 27, Road No.11, Near Tunga Hotel,Marol Industrial Area, MIDC Andheri (E), Mumbai.

LET’S TALK