Analysis of Financial Data
1
Course materials
• Textbook
1. Koop, Gary (2006), Analysis of Financial Data, John Wiley & Sons
• Additional references:
1. Chris Brooks (2019), Introductory Econometrics for Finance,
Cambridge University Press.
2. Tsay, Ruey S (2013), An Introduction to Analysis of Financial Data with
R, John Wiley & Sons
• Slides, problem sets and datasets: uploaded weekly through LMS
2
Assessment Requirements
Criteria Requirement Weight Frequency
• Criteria
• Requirement
Attendance 10% Weekly
• Weight
Empirical project 40% Week 13-14
Mid-term Students are required to apply the techniques
learned in the course with a financial data set.
Details of the project will be discussed in the
lectures.
Final exam Examination with different types of questions 50% According to university’s
schedule
3
Topics
• Chapter 1: Basic data handling
• Chapter 2: Exploratory Financial Data Analysis
• Chapter 3: A Review of Regression for Financial Data
• Chapter 4: Regression with dummy variables
• Chapter 5: Regression with lagged explanatory variables
• Chapter 6: Univariate time series analysis
• Chapter 7: Forecasting
4
Chapter 1
Basic data handling
5
1.1 Types of data
6
1.1.1 Time series, cross-sectional and panel data
• Time series data
- A time series data set consists of observations of one or more variables
over time.
- Time series data are typically presented in chronological order.
- Can be observed at many frequencies: annual, quarterly, monthly,
weekly, daily and hourly.
Example:..………………
- Notation:
+) Yt for t =1,…,T
+) Yt-s: variable is lagged s periods
+) Yt+k : variable is leading k periods 7
1.1.1 Time series, cross-sectional and panel data
- Key features of time series data
➢Observations are commonly dependent across time: most
economic time series are closely related to their recent histories.
➢Exhibit strong seasonal pattern.
- Problems that could be tackled using time-series data
➢How the value of a country’s stock index has varied with that
country’s macroeconomic fundamentals.
➢How the value of a company’s stock price has varied when it
announced the value of its dividend payment.
➢The effect on a country’s exchange rate of an increase in its trade
deficit.
8
1.1.1 Time series, cross-sectional and panel data
• Cross-sectional data
- A cross-sectional data set consists of a sample of individuals,
households, firms, cities, countries, regions or any other type of unit at
a specific point in time.
- The ordering of the data typically does not matter (unlike time series
data).
- Example:……………………………………..
- Notation: Yi for i=1,…,N
9
1.1.1 Time series, cross-sectional and panel data
- Problems that could be tackled using cross-sectional data
➢The relationship between company size and the return to investing
in its shares.
➢The relationship between a country’s GDP level and the probability
that the government will default on its sovereign debt.
10
1.1.1 Time series, cross-sectional and panel data
• Panel data
- Data sets have both a time series and a cross-sectional component.
- Example: ……………………………
- Notation: Yit for i=1,..,N and t=1,…,T
11
1.1.2 Quantitative and qualitative data
● Quantitative data
- is numerical
- Example: share price is $25
● Qualitative data
- data does not come in the form of a single number
- Example: in a survey of companies ask if investment financed through
debt (as opposed to equity or retained earnings). Answer is Yes/No.
● Binary variables
- variables can take on only the values 0 or 1.
- used for turning qualitative data into quantitative
- Example: Yes =1, No = 0
12
1.1.3 Continuous and discrete data
• Continuous data
- Can take on any value and are not confined to take specific numbers.
- Their values are limited only by precision.
- Example: the rental yield on a property could be 6.2%, 6.24%, or
6.238%.
• Discrete data
- Can only take on certain values, which are usually integers and are
often defined to be count numbers.
- Example: the number of people in a particular underground carriage or
the number of shares traded during a day.
13
1.1.4 Cardinal, ordinal and nominal numbers
- Cardinal numbers are those where the actual numerical values that a
particular variable takes have meaning, and where there is an equal
distance between the numerical values.
- Example: price of a share
- Ordinal numbers can only be interpreted as providing a position or an
ordering.
- Example: position of a runner in a race
- Nominal (categorical) variables: arise when numerical values are
arbitrarily assigned, or when codings are assigned to qualitative data.
- Example: when describing the exchange that a US stock is traded on,
‘1’ might be used to denote the NYSE, ‘2’ to denote the NASDAQ
and ‘3’ to denote the AMEX). 14
1.1.5 Financial versus economic data
• Financial data often differ from economic data in terms of:
• Frequency and quantity of data
- Observed at much higher frequencies (daily to minute-by-minute)
- Very large number of observations available for analysis
• Quality
- Prices reflect actual trades or quoted prices → measurement error
and revisions are limited
- Financial data are “noisy”: it is more difficult to separate
underlying trends or patterns from random and uninteresting
features.
- Almost always not normally distributed
15
1.2 Basic data handling
16
1.2.1 Data transformations
• Changing the frequency of time series data: reduce or increase the
frequency
- Example: change the frequency from monthly to quarterly figures
• Nominal versus real data: convert nominal series to real terms using
an appropriate price deflator (CPI, PPI, GDP deflator…)
17
1.2.1 Data transformations
• Logs:
- Effectively linearize the exponential trend of the series.
- Example: (natural) log of GDP will appear on a graph as a roughly
straight line, rather than the exponential curve exhibited by the raw
GDP series.
- Be used to linearize a model that is non-linear in its
parameters.
- Example: Cobb–Douglas production function
- Allow the regression coefficients to be interpreted as elasticities
- Example: In the log-linear production function
18
1.2.1 Data transformations
• Differencing: to remove the trend component from a (time) series
entirely
- First-order differencing: ∆𝑌𝑡 = 𝑌𝑡 - 𝑌𝑡−1
- Second-order differencing: ∆2 𝑌𝑡 = ∆(𝑌𝑡 - 𝑌𝑡−1 )
= ∆𝑌𝑡 - ∆𝑌𝑡−1
= (𝑌𝑡 - 𝑌𝑡−1 ) − (𝑌𝑡−1 − 𝑌𝑡−2 )
= 𝑌𝑡 − 2𝑌𝑡−1 + 𝑌𝑡−2
19
1.2.1 Data transformations
• Growth rate:
- In many cases, it is economically more meaningful to analyse data in
growth rates rather than levels
- Focus on changes over time, allowing analysis of change–change
relationships
- Two types of growth rate:
𝑌𝑡 − 𝑌𝑡−1
𝐺𝑟𝑜𝑤𝑡ℎ 𝑟𝑎𝑡𝑒 𝑜𝑓 𝑌𝑡 = x100%
𝑌𝑡−1
or
𝑌𝑡
𝐺𝑟𝑜𝑤𝑡ℎ 𝑟𝑎𝑡𝑒 𝑜𝑓 𝑌𝑡 = 𝑙𝑛 𝑥100%
𝑌𝑡−1
= 𝑙𝑛𝑌𝑡 − 𝑙𝑛𝑌𝑡−1 x100%
20
1.2.2 Returns in financial modelling
• Asset returns: It is preferable not to work directly with asset prices, so we
usually convert the raw prices into a series of returns.
Simple returns or Continuously compounded returns
pt − pt −1 pt
Rt = 100% Rt = ln 100%
pt −1
pt −1
where, Rt denotes the return at time t
pt denotes the asset price at time t
ln denotes the natural logarithm
• The formula for a return has to be modified to include all income the
investor would receive from holding the asset.
21
Closing price of S&P500
Daily returns of S&P500
6,000 12
5,000 8
4,000 4
3,000 0
2,000
-4
1,000
-8
0
-12
88 90 92 94 96 98 00 02 04 06 08 10 12 14 16
88 90 92 94 96 98 00 02 04 06 08 10 12 14 16
22
1.2.2 Returns in financial modelling
• Portfolio return
- The simple return on a portfolio of assets is a weighted average of the
simple returns on the individual assets
N
R pt = wip Rit
i =1
- But this does not work for the continuously compounded returns.
23
1.2.2 Returns in financial modelling
- Excess return: the difference between an asset’s return and that of some
benchmark asset which is usually defined as a safe, low-risk asset (e.g. the
return on a government bond)
𝐸𝑅𝑡 = 𝑅𝑡 − 𝑅0𝑡
where R0t is the return on the safe asset.
24
1.2.3 Indices and base dates
• Index numbers:
- An index is a number that expresses the relative change in value (for
example, price or quantity) from one period to another.
- Example: The Dow Jones Industrial Average (DJIA); Standard and
Poor’s composite share index (S&P500) are stock price indices;
consumer price index (CPI).
- The change is measured relative to the value at a base date.
- It should not be interpreted as an indicator of whether value is “high”
or “low”.
25
1.2.3 Indices and base dates
Price Price
index index Standardized price
• Change the base date of an index: two Year (2005 (2010 index (2010 base
base base year)
indices may be compared directly only year) year)
if they have the same base date, which
may lead to the need to change the base 2005 100 45.9
date of an index. 2006 132 60.6
• E.g. Suppose we have the following 2007 196 89.9
data
2008 213 97.7
- How to change the base year to 2010?
2009 258 118.3
- How obtain a single series in 2005
prices? 2010 218 100 100
2011 85
2012 62 26
1.3. Obtaining data
27
1.3. Obtaining data
• Some free sources
- [Link] [Link] [Link]
[Link] [Link] [Link]
- Federal Reserve Bank of St Louis at
- [Link]
- [Link]
- US Treasury Department at [Link]
- The World Bank at [Link]
- The IMF at [Link]
28