Group member:

- Nguyễn Ngọc Thanh Trúc_MAMAIU21113

- Vũ Thị Mai Phương_MAMAIU21006
- Phan Thị Khánh Nguyên_MAMAIU21046

QUIZ WEEK 2
1. c
2. b
3. a
4. b
5. d
6. c
7. c
8. d
9. c
10. c
11. c
12. d
13. d
 HOMEWORK CHAPTER 1
7. Explain the differences between the following pairs of terms

a) Continuous and Discrete data

- Continuous data: take any value in a range.

For example, in an interval [1,9], we can take values 2.45, or 7.92; weight of a child.

- Discrete data: only take on certain specific values (integer values).

For example, in an interval [1,9], we can only take values 1, 2, 3, ..., 9; number of
students in a class.

b) Ordinal and Nominal data

- Ordinal data: a variable is limited to defining only a position or ordering.

Example: Graduation ranking of a department in International University.

- Nominal data: unordered values.

Example: ‘gender’ of male or female.

c) Time-series and panel data

- Time-series data: data collected over a period of time on one or more variables.

Example: A series of real estate in HCMC prices collected weekly over 5 years.
- Panel data: contains observations about different cross sections over time.

Example: the daily prices of blue-chip stocks over 10 years.

d) Noisy and clean data

Noisy data:

- Noisy data contains random errors, outliers, or inconsistencies.

- Noisy data is data with a large amount of additional meaningless information in it

called noised. This includes data corruption.

- Noisy data makes it difficult separating underlying trends or patterns from random
and uninteresting features. User system cannot comprehend and interpret correctly.

- Noisy data can adversely affect the results of any data analysis and skew
conclusions if not handled properly.

- Noisy data unnecessarily increases the amount of storage space required.

- Noisy data can be caused by hardware failures, programming errors, faulty data
collection instruments, data entry problems, data transmission problems, technology
limitation, inconsistency in naming convention.

- Dealing with noisy data often involves preprocessing steps such as data cleaning,
outlier detection, error correction to improve the quality of the dataset before
analysis.

For example: Financial data

Clean data:

- Clean data refers to data where the amount of noise is at a minimal level and the
data are at least free of errors, outliers, and inconsistencies.

- Clean data is accurate and consistent. It is considered high quality and accurately
represents the underlying trends or patterns being studied.

- Clean data is crucial for building reliable and accurate models as it helps in
identifying meaningful patterns and making accurate predictions.

- Clean data is desirable since to achieve it may require significant effort in data
collection, data preprocessing, and quality control measures.

e) Simple and continuously compounded returns

The key distinction between simple return and log return lies in their mathematical
properties and interpretations.

Simple returns:

- Simple return directly measures the absolute change in value.

- Simple return, also known as arithmetic return or absolute return, is a
straightforward method to measure the gain or loss of an investment relative
to the initial investment.
- It is expressed as a percentage and represents the difference between the final
value (or terminal value) and the initial value of an asset, divided by the initial
value.
- The formula for calculating simple return is as follows:
 𝑃𝑡 − 𝑃𝑡−1
𝑅𝑡 = 𝑥 100%
𝑃𝑡−1

- Simple return is intuitive and widely used in financial analysis due to its
simplicity.
- Simple returns are additive across assets, thus we have to careful when build
a portfolio using simple returns.
- It is suitable for shorter periods and does not account for the compounding
effect.

Continuously compounded return:

- Log return measures the relative change on a logarithmic scale.

- Continuously compounded return, also known as log return or natural logarithmic

return, addresses the limitation of simple return by considering the compounding
effect.

- It measures the relative change in the logarithm of the investment’s value over a
given period.

- The formula for calculating continuously compounded return is as follows:

𝑃𝑡
𝑟𝑡 = ln ( ) 𝑥 100%
𝑃𝑡−1

- Log return is expressed as a percentage, just like simple return.

- Log return is commonly used in academic research, portfolio management, and risk
modeling.
- It accounts for the compounding effect and enables accurate aggregation of returns
over multiple periods.

- Continuously compounded returns are additive across time

In the limit, as the frequency of the sampling of the data is increased so that they are
measured over a smaller and smaller time interval, the simple and continuously
compounded returns will be identical.

f) Nominal and real series

Nominal series:

- Nominal series, also known as nominal data, refers to data that has not been
adjusted for inflation or other factors.

- It represents values at current prices or nominal values.

- Nominal series reflect the actual prices or values of goods, services, incomes, or
other economic variables without considering changes in purchasing power over
time.

- For a series of nominal values in successive years, different values could be because
of differences in the price level. But nominal values do not specify how much of the
difference is from changes in the price level.

- Nominal series are affected by changes in prices due to inflation, deflation, or other
factors, making it difficult to compare values over different time periods without
adjusting for these changes.
Real series:

- Real series, also known as real data or real values, are adjusted for inflation or other
factors to account for changes in purchasing power over time.

- Real series represent values in terms of constant prices or real values, which remove
the effects of inflation or other price fluctuations.

- By adjusting for changes in prices, real series provide a more accurate

representation of the underlying economic trends or changes in the quantity of goods
and services produced or consumed.

- The real return on an asset is equal to the nominal return on the asset minus the
inflation rate.

8) Present and explain a problem that can be approached using time-series

regression, another one using cross-sectional regression, and another using
panel data.

a) Time-Series Regression Problem:

Time Series Analysis is a statistical technique widely used in diverse fields such as
finance, economics, engineering, and social sciences to analyze and forecast data
over time. For instance, time series data include daily stock prices, hourly weather
data, or monthly sales data.

Problem: Predicting Monthly Sales for an E-Commerce Website

Approach: Time-series regression can be applied to analyze historical sales data to

predict the monthly sales for the upcoming year. Moreover, by applying time series
analysis, we can identify seasonal patterns, such as increased sales during holiday
seasons or dips in sales during certain months. This information can help the
business optimize inventory management and plan marketing strategies accordingly.
Time is regarded as the independent variable, whereas we can select such dependent
variable – sales and predictor variables (explanatory variables) that could possibly
affect sales. This may include variables such as marketing expenses, website traffic,
promotions, or any other features that seem to influence sales. We can build a
regression model that captures trend, seasonality, and other time-dependent patterns
in the data. Once we obtain a model, use it to make predictions for the upcoming
year. We select the relevant predictor variables for the future periods and obtain the
forecasted sales values.

b) Cross-Sectional Regression Problem:

Cross-sectional regression is a statistical technique used to analyze the relationship

between variables across multiple observations at a specific point in time, rather than
over a period. In the context of event studied, cross-sectional regression can be
employed to examine the relationship between abnormal returns and other factors,
such as firm characteristics or event-related variables, for a sample of firms
experiencing the same event.

Problem: Exploring Determinants of Stock Prices

Approach: We can apply cross-sectional regression on stock data and analyze the
factors that influence stock prices (a basket of stocks such as VN30 stocks). The
dependent variables would be the stock prices, and the independent variables could
include variables such as company financial measures (earnings, revenue), industry
performance, market indicators (interest rates, inflation), and some certain company-
specific features. By conducting a cross-sectional regression analysis, we can
identify the determinants that significantly affect stock prices in the given sample.

c) Panel Data Analysis Problem:

Panel data analysis, also known as cross-sectional time series analysis, is a statistical
method used to analyze data collected on the same entities (such as individuals,
households, firms, or countries) over multiple time periods. This approach allows
researchers to study the differences between individual subjects and the changes
within the same subjects over time.

Problem: Examining the determinants of stock returns or bond yields across

different companies or financial assets over time.

Approach: Gather a panel dataset that includes information on stock returns or bond
yields for different companies or financial assets over multiple time periods.
Additionally, collect relevant variables that may affect stock returns or bond yields,
such as financial ratios, macroeconomic indicators, industry-specific factors, and
market conditions. We can identify the features that result in variation in asset returns
over time, while accounting for certain firm characteristics and market-wide trends.

10)

Submit in attached excel file “Homework Chapter 1_Q10”