cointigration

Question 1

What is a hedge ratio ?

Answer 1

The independent variable Uses X times by the dependent variable

Question 2

Which library has hedge ratio function in?

Answer 2

From statsmodels.api import OLS
Parameters independent dependent
output modelled.params for the hedge ratio

Question 3

OLS stands

Answer 3

Ordinary least squares regression

Question 4

How to set model

Answer 4

model=sm.OLS(y.iloc[:90],x.iloc[:90])

Question 5

how to find the hedge fund ratio

Answer 5

model = model.fit()

print(model.params[0])

Question 6

ADF

Answer 6

augmented Dickey–Fuller test (ADF) tests the null hypothesis that a stationarity unit root is present in a time series sample.

Question 7

If ADF is lower that crital leve

Answer 7

We can reject the null hypothesis that the test data is not stationary, and say it has stationarity

Question 8

CADF test results are lower than the 99% SDT statistically.

Answer 8

We reject the null hypothesis,

Therefore the data has stationarity, cointegrated augmented Dickey Fuller,

Question 9

buy hedge ratio and then work out CADF

Answer 9

portfolio = y - model.params[0] * x
cadf = ts.adfuller (portfolio.Close)

Question 10

What are the shortcomings of the ADF test, how are they solved by using the Johansen test?

Answer 10

The Johansen test is not order related when considering cointegration, it is able to take more time series.

Question 11

What indicates stationarity using the Johansen Test

Answer 11

If some eigenvalues are negative, then it is possible to create a linear stationarity

Question 12

Makeup df with columns x y

Answer 12

df = pd.DataFrame( {‘x’: df_x[“Adj Close”] , ‘y’: df_y[‘Adj CLose’] } )

Question 13

Johanson test and null hypothesis

Answer 13

Reject the null hypothesis if the probability value is less than or equal to 0.05. thus stationarity is found

Question 14

test statistic = -1.480

critical value = {‘1%’: -3.434, ‘5%’ : -2.86, ‘10%’ : -2.56}

Answer 14

The test statistics is greater than critical values at 99%, 95%, and 90% confidence levels. So, we cannot reject the null hypothesis that the CL and BZ are not cointegrated from Oct 10, 2011 to Set 06, 2017.

Question 15

Why is it an issue to have a negative regression coefficient for a stock, while building a custom basket for an Index Arbitrage Strategy?

Answer 15

We will be long a stock in both the index and the basket.

Question 16

Covariance for one stocking: itself ?

Answer 16

Is the variance of that stock

Question 17

What is Systematic risk

Answer 17

Systematic risk is Market risk doesn’t make any difference to diversification.

Question 18

Can OTM options be exercised?

Answer 18

Yes, OTM options expire without worth.

Question 19

Can the option pricing model be a helpful guide during expiry trading?

Answer 19

Yes, the option pricing model is used to assess the option fair value.

Question 20

Sharpe ratio is:

Answer 20

The Sharpe ratio is equal to the strategy returns minus the risk-free ratio divided by the standard deviation. The Sharpe ratio allow us to assess the excess of returns compared with its risk.

Question 21

Sortino ratio is

Answer 21

The Sortino ratio is in essence the same as the Sharpe ratio although the standard deviation is computed just only with the downside trades.

Question 22

Calmar ratio is:

Answer 22

The Calmar ratio measures the risk-adjusted returns, divide by the maximum drawdown.

Question 23

K-S test is:

Answer 23

The Kolmogorov-Smirnov test or K-S test measures the distance between the cumulative distribution of the strategy returns and the reference distribution.

Question 24

What you need to know about a strategy

Answer 24

i) If you haven’t tested an idea, you are just guessing.
ii) Academic research paper might provide you with a trading idea.
iii) If you don’t know what to expect out of a trade, you can’t know if a trade is working.
iv) Back-testing is an integral part of strategy building

Question 25

What is true regarding strategy ideation and evaluation?

Answer 25

Knowing numerical results won’t really lead to obvious Improvements and Knowing non-quantitative results can help. E.g. trading on Fed days.

Question 26

Important aspects of trading

Answer 26

i) Pre-trade planning and expectations
ii) Finding a trade with an edge
iii) Execution and hedging
iv) Evaluation

Question 27

Does bagging help fix the overfitting problem in decision trees?

Answer 27

Yes, but random forest is better

Question 28

what is type I error ?

Answer 28

If we reject the null hypothesis when it is true, then we made a type I error.

Question 29

Type II error.

Answer 29

If the null hypothesis is false and we failed to reject it, a Type II error.