### All Weeks Python and Statistics for Financial Analysis Coursera Quiz Answers

#### Python and Statistics for Financial Analysis Coursera Quiz Answers

#### Week 1: Python and Statistics for Financial Analysis

Q1. Which of the following library has DataFrame object?

**Pandas**- Numpy
- Matplotlib
- Statsmodels

Q2. Which of the following is the correct way to import a library, eg Pandas?

`1.pandas import`

`1.include`

`1.pandas`

**1.import pandas as pd**

Q3. What is the method of DataFrame object to import a csv file?

- import_csv()
**from_csv()**- read_csv()
- csv()

Q4. Which of the following attributes of a DataFrame return a list of column names of this

DataFrame?

**columns**- shape
- dtype
- column

Q5. Which of the following can slice ‘Close’ from ‘2015-01-01’ to ‘2016-12-31’ from data,

which is a DataFrame object?

**1.data.loc[‘2015-01-01’:’2016-12-31’, ‘Close’]**

`1.data.iloc[‘2015-01-01’:’2016-12-31’, ‘Close’]`

Q6. What is the method of DataFrame to plot a line chart?

- scatter()
**plot()**- plot_graph()
- axhline()

Q7. Suppose you have a DataFrame – data, which

contains columns ‘Open’, ‘High’, ‘Low’, ‘Close’, ‘Adj Close’ and ‘Volume’.

What

does data[[‘Open’, ‘Low’]] return?

- All

columns of data except ‘Open’ and ‘High’ - No results are shown
**Columns ‘Open’ and ‘Low’**- The first

row of data which contains only columns ‘Open’ and ‘High’

Q8. Suppose you have a DataFrame ms , which contains the daily data of ‘Open’, ‘High’, ‘Low’, ‘Close’, ‘Adj

Close’ and ‘Volume’ of Microsoft’s stock.

Which of

the following syntax calculates the Price difference, (ie ‘Close’ of tomorrow –

‘Close’ of today)?

**1. ms[‘Close’].shift(1) – ms[‘Close’].shift(1)**

`1.ms[‘Close’].shift(-1) – ms[‘Close’].shift(-1)`

`1.ms[‘Close’].shift(1) – ms[‘Close’]`

`1.ms[‘Close’].shift(-1) – ms[‘Close’]`

Q9. Suppose you have a DataFrame – ms , which contains the daily data of ‘Open’, ‘High’, ‘Low’, ‘Close’, ‘Adj

Close’ and ‘Volumn’ of Microsoft’s stock.

What is

the method of DataFrame to calculate the 60 days moving average?

- rolling().mean(60)
- moving_average(60)
**rolling(60).mean()**- rolling(60).median()

Q10. Which of the following idea(s) is/are correct to the simple trading strategy that we introduced in the lecture video?

**Use longer**

moving average as slow signal and shorter moving average as fast signal- We short

one share of stocks if fast signal is larger than slow signal **If fast**

signal is larger than slow signal, this indicates an upward trend at the

current moment

#### Week 2: Python and Statistics for Financial Analysis

Q1. Roll two dice and X is the sum of faces values. If we roll them 5 times and get 2,3,4,5,6

Which of the following is/are true about X?

- The mean of X is 4.
- X can only take values 2,3,4,5,6
**X is a**

random variable

Q2. Roll two dice and X is the sum of faces values. If we roll them 5 times and get 2,3,4,5,6

What do we know about X?

- The dice is fair.
- Range of X is 6-2=4
- The most likely value of X is 6
**We have 5 observations of X**

Q3. Roll two dice and X is the sum of faces values. If we roll them 5 times and get 2,3,4,5,6

X is a ** __** random variable.

**discrete**- continuous
- None of

the above

Q4. Why do we

use relative frequency instead of frequency?

- Relative
frequency is easier to compute - Frequency

cannot show the number of appearance of outcomes **Relative frequency can be used to compare the**

ratio of values between difference collections with difference number of values- Relative

frequency is easier to compute when the number of observations increases

Q5. What can

we say about relative frequency when we have large number of trials?

**Relative frequency becomes approximately the**

distribution of the corresponding random variable- The

relative frequency of each possible outcome will be the same - The relative

frequency stays constant after a very large number of trials, eg. n=10000 - None of the above

Q6. What is the notion of “95% Value at Risk” ?

- 95% Value

at Risk is 95% quantile - 95% VaR

measures how much you can lose at most - 95% VaR

measures how much you can win at most **95% VaR measures the amount of investment you can**

lose, at the worst 5% scenario

Q7. In the lecture video, we mentioned the calculation of continuous random variable is based on the probability density function.

Given a

probability density function, f(x) = 1/100, what is the probability

P(10<X<20), where X~Uniform[0, 100]?

- f(20) –

f(10) - f(10)
- f(20)
**(20- 10) * 1/100**

Q8. What

methods should we use to get the

cdf and pdf of normal distribution?

**norm.cdf() and norm.pdf() from scipy.stats**- cdf() and

pdf() form numpy - cdf() and

pdf() from pandas - norm.cdf()

and norm.pdf() from statsmodels

Q9. Which additional library should we import when we want to calculate log daily return specifically?

- Pandas
**Numpy**- Statsmodels
- Matplotlib

Q10. What

is the distribution of stock returns suggested by Fama and French in general?

- A perfect normal distribution
**Close to normal distribution but with fat tail**- Arbitrary distribution
- Left-skewed distribution

#### Week 3: Python and Statistics for Financial Analysis

Q1. What is true

about sample and population?

- Population

can always be directly observed - Parameters

from population is always the same as statistics from sample **Sample is a subset of population which is**

randomly draw from population- The size

of population is always finite

Q2. You have a

DataFrame called ‘data’ which has only one column ‘population’.

data = pd.DataFrame()

data[‘population’] = [47, 48, 85, 20, 19, 13, 72, 16, 50, 60]

How to draw sample with sample size =5, from a

‘population’ with replacement?

(Hint: You can modify the code illustrated in the Jupyter Notebook “Population and Sample” after Lecture 3.1)

`1.data[‘population’].sample(5, replace=False)`

`1.data[‘population’].sample(10)`

`1.data[‘population’].sample(5)`

**1.data[‘population’].sample(5, replace=True)**

Q3. Why is the

degrees of freedom n-1 in sample variance?

**The degrees of freedom in sample variance is**

constrained by the sample mean- None of

the above - The extreme value in the sample is removed for fair analysis
- Only n-1

values in the sample is useful

Q4. What does Central Limit Theorem tell you about the distribution of sample mean?

- The
distribution of sample mean follows normal distribution only if the population

distribution is normal - The

distribution of sample mean follows normal distribution with any sample size

only if the population distribution is normal **The distribution of sample mean follows normal**

distribution with very large sample size follows normal distribution regardless

of the population distribution- The

distribution of sample mean with large sample size follows chi-square

distribution regardless of the population distribution

Q5. Suppose we have 3 independent normal random variables X1, X2 and X3:

What is the distribution of X1 + X2 + X3?

- Remains the same even X1, X2 and X3 are added up
**Mean and variance of X1, X2 and X3 are added up**- Mean remains unchanged; variances are added up.
- Mean remains unchanged; variance takes 3 square root.

Q6. Why do we

need to standardize sample mean when making inference?

- Sample

mean becomes normally distributed after standardization - Sample

mean becomes population mean after standardization **The standardized distribution of sample mean follows N(0,1) which is easier to make inference**- None of the above

Q7. What can a 95%

confidence interval of daily return of an investment tell you?

- With 95% chance

your daily return falls into this interval **With 95% chance this interval will cover the mean of**

daily return- With 5% chance your

daily return falls into this interval - None of the above

Q8. Check the

Juypter notebook of 3.3 Sample and Inference. What is the confidence interval of this exercise?

- [0.000015603,

0.001656] **[-0.000015603, 0.001656]**- [-0.0001690,

0.001471] - [0.0001690,

0.001471]

Q9. When do you reject a null hypothesis with alternative hypothesis μ>0 with significance level α?

- p value is larger than α
**p value is smaller than α****z < z_(1-α)**- z > z_(1-α)

Q10. When doing analysis of stock return, you notice that with 95% confidence interval, the upper bound and lower bound are negative.

Base on this data, what can you tell about this stock?

**There is 95% chance of which the mean return of this stock is negative**- We must lose money by investing in this stock
- There is only 5% chance of which the mean return of this stock is negative

#### Week 4: Python and Statistics for Financial Analysis

Q1. Why do you use

coefficient of correlation, instead of covariance, when calculating the association between two random variables?

- None of the above
- Covariance is not

suitable to use when the underlying distribution is not normal - Covariance cannot

address nonlinear relationship but coefficient of correlation can address

nonlinear relationship **Covariance can be affected by the variance of**

individual variables, but coefficient of correlation is rescaled by variance of

both variables

Q2. What is the range

and interpretation of coefficient of correlation?

- From 0 to 1, 0

means perfect negative linear relationship and 1 means perfect positive linear

relationship - From 0 to 100, 0

means perfect positive linear relationship and 100 means perfect negative

linear relationship - From 0 to 100, 0

means perfect negative linear relationship and 100 means perfect positive

linear relationship **From -1 to 1, -1 means perfect negative linear**

relationship and 1 means perfect positive linear relationship

Q3. Refer to the https://www.coursera.org/learn/python-statistics-financial-analysis/notebook/F0Luf/simple-linear-regression-model

Is LSTAT a significant predictor of MEDV at significance level 0.05?

- Yes, because the coefficient b_1 is not zero
- Yes, because the p value of b_1 is larger than 0.05
- No, because the coefficient b_1 is negative
**Yes, because the p value of b_1 is smaller than 0.05**

Q4. To evaluate the performance of linear regression model, we refer to the summary of “model” as seen in https://www.coursera.org/learn/python-statistics-financial-analysis/notebook/F0Luf/simple-linear-regression-model

What is the percentage of variation explained by the model?

- 0.95
**0.54**- 0.46
- 0.829

Q5. How to check if a

linear regression model violates the independence assumption?

- Draw residual

versus predictor plot - Draw scatter plot

of predictor versus target **Durbin Watson test**- QQ plot

Q6. If any of the

assumptions of linear regression model are violated, we cannot use this model

to make prediction.

- True
**False**

Q7. Check the Jupyter Notebook 4.4- Build the trading model by yourself!

We have a variable ‘formula’ which store the names of predictors

and target. How should you modify this ‘formula’ if you want to drop the

predictor ‘daxi’?

`1.formula = ‘spy~aord+cac40+nikkei+dji+daxi’`

`1.formula = ‘spy~aord+cac40+nikkei+dji-daxi’`

**1.formula = ‘spy~aord+cac40+nikkei+dji’**

`1.formula = ‘spy~ -daxi’`

Q8. Check the Jupyter Notebook 4.4- Build the trading model by yourself!

What is the most

significance predictor for ‘SPY’?

- nikkei
- dji
**arod**- cac40

Q9. What does it mean

if you have a strategy with maximum drawdown of 3%?

- During the

trading period, the minimum you lose is 3% - During the

trading period, the maximum gain from the previous peak of your portfolio value

is 3% - During the trading period, the maximum lose of

your portfolio is 3% **During the trading period, the maximum drop from the**

previous peak of your portfolio value is 3%

Q10. How can you check

the consistency of your trading strategy?

- Check if the

return of your strategy is positive using all historical data you have - Define some

metric for evaluating your strategy, eg Sharpe Ratio, maximum drawdown, and

check if your strategy can generate positive return using all historical data

you have **Define some metric for evaluating your strategy, eg**

Sharpe Ratio, maximum drawdown. Then split your data into train set and test

set and check if your strategy can generate positive return using both train

set and test set- There is no way to to check the consistency

#### Get All Course Quiz Answers of **Entrepreneurship Specialization**

Entrepreneurship 1: Developing the Opportunity Quiz Answers

Entrepreneurship 2: Launching your Start-Up Quiz Answers

Entrepreneurship 3: Growth Strategies Coursera Quiz Answers

Entrepreneurship 4: Financing and Profitability Quiz Answers