2023 CFA L2 Book 1 Quants Eco Multiple

FinTree
JuiceNotes 2023
Quantitative Methods |Economics
Chartered Financial Analyst - Level II

© 2023 FinTree Education Pvt. Ltd.
INDEX
Quantitative Methods
1 Basics of Multiple Regression & Underlying Assupmtions 05
2 Evaluating Regression Model Fit and Interpreting Model Results 06
3 Model Misspecification 09
4 Extensions of Multiple Regrission 13
5 Time-series Analysis 15
6 Machine Learning 22
7 Big Data Projects 31
Economics
1 Currency Exchange Rates: Understanding Equilibrium Value 39
2 Economic Growth and The Investment Decision 46
3 Economics Of Regulation 51
Quantitative Methods
Notice : The recipient of this publication is strictly prohibited by law to circulate. We have inserted a
concealed code in the document, which will lead to identification of the user to whom this document was
issued. If this documents is found to be circulated on internet, social media sites and other mode thereto,
the user identification will be reported to CFA Institute and strict legal action will be initiated.
Unless otherwise stated, copyright and all intellectual property rights in all the course material(s)
provided, is the property of FinTree Education Private Limited. Any copying, duplication of the course
material either directly and/or indirectly for use other than for the purpose provided shall tantamount to
infringement and shall strongly defended and pursued, to the fullest extent permitted by law.
The unauthorized duplication of these notes is a violation of global copyright laws and the CFA Institute
code of Ethics. Your
3 assistance in pursuing potential violators of this law is greatly appreciated. If any
violation comes to your notice, get in touch with us at admin@fintreeindia.com
R
Financial Risk Manager
R
ª The Financial Risk Manager (FRM ) designation is offered by the Global Association of Risk Professionals
(GARP) and is the globally recognized standard for those who manage risk. Certified FRMs are part of an
elite, global network, and are valued by top employers across diverse businesses.
ª The global FRM community is growing dramatically, with Certified FRMs represented at nearly every
major banking institution, government regulator, consulting firm and financial services institution around
the world.
R R
FRM Part I FRM Part II
ª Foundations of risk management concepts ª Market risk measurement and management
ª Quantitative analysis ª Credit risk measurement and management
ª Financial markets and products valuation and ª Credit risk measurement and management
risk models
ª Risk management and investment
management
ª Current issues in financial markets
R R
Footer : Global Association of Risk Professionals, FRM , and Financial Risk Manager are trademarks owned by
Global Association of Risk Professionals.
To Know more, visit www.fintreeindia.com
FinTree
Basics of Multiple Regression &

Underlying Assumptions
Describe the types of investment problems addressed by multiple linear
regression and the regression process
1. Identify relationships between variables
2. Forecast variables
3. Test existing theories
A time series is covariance stationary if it satisfies the following three conditions:
Formulate a multiple linear regression model, describe the relation between

the dependent variable and several independent variables, and interpret
estimated regression coefficients
Refer to the Spreadsheet
Explain the assumptions underlying a multiple linear regression model

and interpret residual plots indicating potential violations of these
assumptions
Assumptions of a multiple regression model
Œ Relationship between dependent and independent variable is linear

Independent variables are uncorrelated with the error term and there is no exact
linear relation between two or more independent variables
Ž Expected value of the error term is zero
Variance of the error term is constant (NOT ZERO) for all observations. The
economic relationship b/w variables is intact for the entire time period (eg.
change in political regime)
Error term is uncorrelated with other observations (eg. seasonality)
‘ Error term is normally distributed
Residual plots allow analysts to get a preliminary indication of violation of regression

assumptions
Evaluate Regression Model Fit

& Interpreting Model Results
ANOVA table
Source of variation DoF Sum of squares Mean sum of squares
Regression
k RSS MSR = RSS/k
(explained)
Error
n−k−1 SSE MSE = SSE/n − k − 1
(unexplained)
Total n−1 SST
F-statistic = MSR/MSE with ‘k’ and ‘n − k − 1' DoF
R² and adjusted R²
R2: % variation of dependent variable explained by % variation of all the independent variables
R2 = RSS/SST
R2 = Explained variation/Total variation
Adjusted R2 = 1− ]) n−1
n−k−1 ) ]
× (1 − R2)
Adjusted R2 < R2 in multiple regression
Eg. k=6 n = 30 R2 = 73%
k=8 n = 30 R2 = 75%
Adjusted R21 = 1− ]) 30 − 1
30 − 6 − 1 )
× (1 − 0.732) ] 41.1%
Adjusted R22 = 1− ]) 30 − 1
30 − 8 − 1 )
× (1 − 0.752) ] 39.58%
Adding two more variables is not justified because adjusted R22 < adjusted R21
For evaluating a regression model, regression output may include the Akaike's
information criterion (AIC) and the Schwarz's Bayesian information criteria (BIC).
Both AIC and BIC evaluate the quality of model fit among competing models for the
same dependent variable
For evaluating a regression model, regression output may include the Akaike's
information criterion (AIC) and the Schwarz's Bayesian information criteria (BIC).
Both AIC and BIC evaluate the quality of model fit among competing models for the
same dependent variable
Lower values indicate a better model under either criterion.
AIC is used if the goal is to have a better forecast, while BIC is used if the goal is a
better goodness of fit.
These metrics can be calculated as follows:
AIC = n
x
(
1 n SSE)2 ( K + 1 ) n=30
n +2
x
BIC = n
x
(
1n SSE
n )
+ 1 n (n) ( K + 1) LN ( 30 ) = 3.40
3.40
Where:
K = number of independant variables
K is a penalty parameter in both criteria: higher values of k result in higher

values of the criteria.
Because ln(n) is greater than 2 for even small sample sizes, the BIC metric
imposes a higher penalty for overfitting
adding more Ks
Formulate hypotheses on the significance of two or more coefficients in a multiple

regression model and interpretthe results of the joint hypothesis tests
In addition to AIC and BIC, we can use a formal F-test to evaluate nested models.
Nested models are models such that one model, called the full model or unrestricted
model, has a higher number of independent variables while another model
called the restricted model, has only a subset of the independent variables.
Consider a full model with three independent variables that is evaluated relative to
a more parsimonious restricted model, which includes only the first variable as the
independent variable.
unrestricted model: Yi = b0 + b1 X1 + b2 X2 + b3 X3 +Σ i
restricted model: Yi = b0 + b1 X1 + Σ i
we want to test the following hypothesis:

H0: b2 = b3 = 0. vs. Ha: b2 or b3 = 0.
we calculate the F-statistics to test this hypothesis as:
( SSER – SSEU ) / q
F= with q and (n-k-l) degree of freedom
( SSEU ) / ( n-k-1)
where:
R and U represent the restricted and unrestricted models, respectively

Q = number of excluded variables in the restricted model
K = independent variables in the full model
Decision rule: reject H0 if F (test-statistic) > Fc ( critical value )

The F-test evaluates whether the relative decrease in SSE due to the inclusion
of q additional variables is statistically justified.
Model Misspecification
Calculate and interpret a predicted value for the dependent variable, given the estimated
regression model and assumed values for the independent variables.
Describe how model misspecification affects the results of a regression

analysis and how to avoid common forms of misspecification
Misspecification Description Effect

Omission of Based on economy theory, Biased and incoinsistent
important one or more variables that regression parameters
independent should have been included are May lead to serial correlation
variables (s) omitted. or heteroscedasticity in the
residual
Inappropriate The relationship between the May lead to heteroscedasticity
variable form dependent and independent in the residual
variables may be non-linear.
Inappropriate Variables may need to be May lead to heteroscedasticity

variable scaling transformed before estimating in the residual or
the regression. multicollinearity
Data improperly Samples has periods of May lead to heteroscedasticity

pooled dissimilar economic or serial correlation in the
environments ( that should not residual
be pooled )
Explain the types of heteroskedasticity and how it affects statistical inference
BP chi-square test statistic = n X R2resid with k degrees of freedom
Where:
n = the number of observation
2
R resid = R2 from a second regression ( of the squared residuals from the
The first regression ) on the independent variables
K = the number of independent variables
This is a one-tailed test, because heteroskedasticity is only a problem if the R2 and the BP
Test statistic are too large.
Detecting Serial Correlation
Residual serial correlation at single lag can be detected using the Durbin-Watson (DW)
Statistic. A more general test (which can accommodate serial correlation at multiple lags)
Is the Breusch-Godfrey (BG) test. The BG test regresses the regression residuals against
The original set of independent variables, plus one or more additional variables
Representing lagged residual(s):
Σt = b0 + a1 X1t + a2 X2t + … + ak Xkt + P1 Σt-1 + P2 Σt - 2
We then test whether each of the slope coefficients of the lagged residuals is statistically
Significantly different from 0.
H0: p1 = 0 vs . H2: p1 = 0
The BG test statistic has an F-distribution with p and n – p – k – 1 degrees of freedom,

Where p = the number of lags tested. Most software packages provide the BG test
Statistic.
Correction of serial correlation:
Robust standard errors

Newey west corrected standard errors or heteroskedasticity consistent
standard errors
Explain multicollinearity and how it affects regression analysis

2
Detection: High R , significant F and insignificant t results
Variance Inflation factor
More formally, we can quantify multicollinearity using the variance inflation factor (VIF)
For each of the independent variables. We start by regression one of the independent
2
Variable “J” against the remaining independent variables. The Rj from that equation is
Then used to calculate the VIF of that variable.
VIFj = 1 / (1 - Rj2 )
High values of Rj2 signal that the variable is well explained by other variables, and
Indicates that the variables will have a high VIF. A VIFj value of 1 (i.e., R2 = 0) indicates
that the variable j is not highly correlated with other independent variables. VIF values
2
Greater than 5 (i.e., R > 80%) warrant further investigation, while values above 10
2
(i.e.,R > 90%) indicates severe multicollinearity.
Describe influence analysis and methods of detecting influential data points
High-leverage points are the extreme observations of the independent (or 'X') variables.
High Leverage points - would be identified using a measure called leverage (Lij) which
can be provided by statistical packages.
Leverage: Distance between variable and sample mean, scaled to be between 0 and 1
The higher the value of leverage, the greater the distance—and hence the higher the
potential influence of the observation—on the estimated regression parameters.
The sum of the individual leverages for all observations is k + 1.
If a variable's leverage is higher than three times the average, [3(k + 1) / n], it is
considered potentially influential.
Outliers are extreme observations of the dependent (or 'Y') variable
We can identify outliers using the studentized residuals. The steps below outline the
procedure:
Estimate the regression model using the original sample of size n.
Delete one observation and re-estimate the regression using (n – 1) observations.
Perform this sequentially, for all observations, deleting one observation at a time.
Compare the actual Y value of the deleted observation i to the predicted Y-values using
the model parameters estimated with that observation deleted.
* *
Ei = Yi – Yi
The studentized residual is the residual in step 2 divided by its standard deviation.
*
ei*
t = i
si*
We can then compare this studentized residual to critical values from a t-distribution
with n – k – 2 degrees of freedom, to determine if the observation is influential.
Detecting Influential Data Points
Cook's distance (Di) is a composite metric (i.e., it takes into account both the leverage
and outliers) for evaluating if a specific observation is influential.
[ ]
2
ei hii
Di = 2
(1 – hii )
k x MSE
Where:
ei = residual for ith observation
k = number of independent variables
MSE = mean square error of the regression model
hii = leverage value for ith observation
Di values greater than √k/n indicate that the ith observation highly likely to be an
Influential data point. Generally, a value greater than 1 indicates high likelihood of an
Influential observation, while values above 0.5 merit further investigation.
Once influential observation are identified, we need to determine whether this was the
Result of an input error (in which case, the error should be rectified or the observation
deleted), or if the observation is valid but the model is incomplete (i.e., important
independent variables are omitted).
Extensions of Multiple Regression

Formulate and interpret a multiple regression model that includes qualitative independent
variables.
A dummy variable can be an intercept dummy, a slope dummy, or a combination of the two
INTERCEPT DUMMY SLOPE DUMMY
Y = b0 + d0 D + b1X + Σ
Y = b 0 + b1 X + d1 ( D x X ) + Σ
Y = b0 + d0 D + b1X + Σ
This regression becomes:
And the regression then becomes:
Y = b 0 + b1 X + Σ (if D=o)
Y = (b0 + d0) + b1 X + Σ (ifD=1) Y = b0 + b1X + Σ (ifD=0)
In other words, the intercept of Y = b0 + (b1 + d1)X + Σ (ifD=1)

the line shifts from b0 to (b0 + d0) if D=1.
The value of This shift (d0) can be positive A slope dummy is the interaction term;
or negative. it captures the interaction between
Variable and the continuous variables.
Formulate and interpret a logistic regression model.
Financial analysis often calls for the use of a model that has a qualitative dependent
variable—a binary or dummy variable, which takes on a value of either zero or one.
An example of an application requiring the use of a qualitative dependent variable is a

model that attempts to estimate the probability of default for a bond issuer.
In this case, the dependent variable may take on a value of one in the event of default
and zero in the event of no default.
An ordinary regression model is not appropriate for situations that require a qualitative dependent
variable, because the forecasted values of y using the model can be less than 0 or greater than 1,
which are illogical values for probability.
Returns on *
Nifty = 5% + 1.2 GDP growth + Σ
Prob.of Bankruptcy
Instead, we transform the probability values of the dependent variable into odds: p / (1 – p).
For example, if probability = 0.80, then odds = 0.80 / 0.20 or 4 to 1.
A logistic transformation involves taking the natural log of the odds: ln [p / (1 – p)].
Logistic regression (logit) models use log odds as the dependent variable.
In ( 1-p
P
) = b0 + b1X1 + b2X2 + … + Σ
The coefficients of the logit model are estimated using the maximum likelihood estimation methodology.
The slope coefficients in a logit model are interpreted as the change in the “log odds” of the
event occurring per one unit change in the independent variable, holding all other independent
variables constant.
Once the coefficients are estimated, using the regression equation, the predicted value
Λ
Of y (i.e., y) can be calculated from the values of the X variables. The odds are then
Calculated as:
Λ
y
odds = e
And the probability of the event is calculated as:

Λ
P = odds / (1 + odds) = 1/(1 + e -y)
Similar to the joint F-test to evalute nested models, a likelihood ratio (LR) test is used
For logistic regression.
LR = -2 (log likelihood restricted model – log likelihood unrestricted model)
Where the restricted model has “q” fewer independent variables.
LR always has a negative value. Higher values (closer to 0) indicate a better-fitting

Restricted model.
The test statistic LR has a Chi-square distribution with q degrees of freedom.
While traditional R2 is not available for logit models, software packages often report
Pseudo-R2 values. These Pseudo-R2 values should only be used to compare competing
Models using the same dependent variable.
Time-series Analysis
LOS a Predicted trend value for a time series
Time series: Set of observations on a variable’s outcomes in different time periods

Used to explain the past and make predictions about the future
Linear trend Log-linear

models trend models
Log-linear trend is a trend in

Linear trend is a trend in which the dependent variable
which the dependent variable changes at an exponential
changes at a constant rate rate with time
with time
Used for financial time series
Has a straight line
Has a curve
Upward-sloping line:
+ve trend Convex curve:
+ve trend
Downward-sloping line:
−ve trend Concave curve:
−ve trend
Equation:
yt = b0 + b1t + εt Equation:
ln yt = b0 + b1t + εt
LOS b How to determine which model to use
Plot the data
y y
x x
Linear trend Log-linear

model trend model
Limitation of trend models is that they are not useful if the error terms are serially correlated.
Can be tested with DW test
Requirement for a time series to be covariance stationary
A time series is covariance stationary if it satisfies the following three conditions:
Constant and finite mean
Constant and finite variance (same as homoskedasticity)
Constant and finite covariance of time series with itself
Eg. Xt = b0 + b1 Xt−1
Xt = 5 + 0.5 Xt−1
Xt − 1 = 6 Xt = 8 Xt − 1 = 20 Xt = 15
Xt − 1 = 8 Xt = 9 Xt − 1 = 15 Xt = 12.5
Xt − 1 = 9 Xt = 9.5 Xt − 1 = 12.5 Xt = 11.25
Xt − 1 = 10 Xt = 10
If Xt − 1 = 10, then Xt = 10, Xt + 1 = 10, Xt + 2 = 10 and so on

This is called constant and finite mean
b0 5
Mean of the time series = = = 10
1 − b1 1 − 0.5
For a model to be valid, time series must be covariance stationary
Most economic and financial time series relationships are not stationary
The model can be used if the degree of nonstationarity is not significant
Autoregressive (AR) model
AR model: A time series regressed on its own past values
Equation AR(1): Xt = b0 + b1Xt − 1 + εt
Equation AR(2): Xt = b0 + b1Xt − 1 + b2Xt − 2 + εt
Chain rule of forecasting: Calculating successive forecasts

LOS e Autocorrelations of the error terms
If the error terms have significant serial correlation (autocorrelation), the AR model
used is not the best model to analyze the time series
Procedure to test if the AR model is correct:
Step 1: Calculate the intercept and slope using linear regression

Step 2: Calculate the predicted values
Step 3: Calculate the error terms
Step 4: Calculate the autocorrelations of the error terms (correlation with lag terms)
Step 5: Test whether the autocorrelations are significantly different from zero
If the autocrrelations are not statistically If the autocrrelations are statistically

significantly different from zero (if the significantly different from zero (if the decision
decision is FTR): Model fits the time series is reject): Model does not fit the time series
Test used to know if the autocorrelations are significantly different from zero: t-test
Autocorrelation
t statistic (DoF: n-2) =
Standard error
Standard error = 1/ √ n
n = Number of observations
LOS f Mean reversion

It means tendency of time series to move toward its mean
b0
Mean reverting level =
1 − b1
If Xt > mean reverting level, then Xt+1 will be lower than Xt
If Xt > mean reverting level, then Xt+1 will be higher than Xt
LOS g In-sample and out-of-sample forecasts and RMSE criterion
Sample Predicted Squared

Eg. Xt − 1 Error
value (Xt) value errors
200 - - - -
220 200 216.5 3.5 12.25
215 220 227.8 −12.8 163.84
205 215 225 −20 400
235 205 219.4 15.6 243.36
250 235 236.4 13.6 184.96
1004.41
In-sample root mean SSE 1004.41

squared error (RMSE) √ n √ 5
= 14.17
Eg. Actual Predicted Squared

Error
value value errors
215 - - -
235 225 10 100
220 236.4 −16.4 268.96
240 227.9 12.1 146.41
250 239.2 10.8 116.64
632
Out-of-sample root mean SSE 632

squared error (RMSE) √ n √ 4
= 12.57
Select the time series with lowest out-of-sample RMSE
LOS h Instability of coefficients of time-series models

One of the important issues in time series is the sample period to use
Shorter sample period → More stability but less statistical reliability
Longer sample period → Less stability but more statistical reliability
Data must also be covariance stationary for model to be valid
LOS i Random walk

Random walk
with a drift
A time series in which

A time series in which predicted value of a dependent
predicted value of a dependent variable in one period is equal
variable in one period is equal to the value of dependent
to the value of dependent variable in previous period
variable in previous period plus or minus a constant
plus an error term amount and an error term
Equation: Equation:
Xt = Xt − 1 + εt Xt = b0 + Xt − 1 + εt
ª Both of the above equations have a slope (b1) of 1
ª Such time series are said to have ‘unit root’
ª They are not covariance stationary because they do not have a finite mean
ª To use standard regression analysis, we must convert this data to covariance stationary.
This conversion is called ‘first differencing’
LOS j & k Unit root test of nonstationarity
Autocorrelation Dickey-Fuller
approach test
If autocorrelations do not exhibit More definitive than

these characteristics, it is said to be a autocorrelation approach
nonstationary time series:
Xt − Xt − 1 = b0 + b1Xt − 1 − Xt − 1 + εt
Autocorrelations at all lags are
g
statistically not different from zero
Xt − Xt − 1 = b0 + (b1 − 1)Xt − 1 + εt
or
If null (g) = 0 can not be rejected,
As the no. of lags increases, the the time series has a unit root
autocorrelations drops down to zero
First differencing
Eg. Sales Lag 1 First difference
∆ sales ∆ sales
- -
(current year) (previous year)
230 - - -
270 230 40 -
290 270 20 40
310 290 20 20
340 310 30 20
^ ^ ^
Equation: y = 30 − 0.25x Equation: y = 30 − 0.25(340) y = (55)
Forecasted sales: 340 − 55 = 285
If time series is a random walk then we must convert this data to covariance stationary.
This conversion is called first differencing
LOS l How to test and correct for seasonality
Seasonality can be detected by plotting the values on a graph or calculating autocorrelations
Seasonality is present if the autocorrelation of error term is significantly different from zero
Correction: Adding a lag of dependent variable (corresponding to the same period in previous year)
to the model as another independent variable
LOS m Autoregressive conditional heteroskedasticity (ARCH)
ARCH exists if the variance of error terms in one period is
dependent on the variance of error terms in previous period
Testing: Squared errors from the model are regressed on the first
lag of the squared residuals
^2 ^2
Equation: εt = a0 + a1 εt − 1 + μt
Intercept Predicted
error term of
last period
Predicted Slope Error term of

error term of errors
current period
LOS n How time-series variables should be analyzed for nonstationarity

and/or cointegration
To test whether the two time series have unit roots, a Dickey-Fuller test is used
Possible scenarios:
Œ Both time series are covariance stationary (linear regression can be used)
Only the dependent variable time series is covariance stationary (linear regression
should not be used)
Ž Only the independent variable time series is covariance stationary (linear regression
should not be used)
Neither time series is covariance stationary and the two series are not cointegrated
(linear regression should not be used)
Neither time series is covariance stationary and the two series are cointegrated (linear
regression can be used)
Cointegration: Long term economic or financial relationship between two time series
LOS o Appropriate time-series model to analyze a given investment problem

ª Understand the investment problem you have and make a choice of model
ª If you have decided to use a time-series model plot the values to see whether the time series looks
covariance stationary
ª Use a trend model, if there is no seasonality or structural shift
ª If you find significant serial correlation in the error terms, use a complex model such as AR model
ª If the data has serial correlation, reexamine the data for stationarity before running an AR model
ª If you find significant serial correlation in the residuals, use an AR(2) model
ª Check for seasonality
ª Test whether error terms have ARCH
ª Perform tests of model's out-of-sample forecasting performance (RMSE)

All queries/doubts about this reading can be posted on Watch video with important
FinTree Forum for the reading testable concepts here
Machine Learning
LOS a Machine Learning makes no assumptions about distribution of underlying data
Target Variable - This is the dependent variable (i.e, the y variable)
Features - Independent variables (i.e., the x variables)
Training data set - This is the sample used to fit the model
Hyperparameter - This is a model input specified by the researcher
Difference between Supervised & Unsupervised Learning & Deep Learning
Machine learning : Machines display intelligent decision making ability through activities such
as sensing, reasoning, and understanding
Surpervised Unsupervised Deep
learning learning learning
ª Deep learning algorithms are

ª Uses labeled training data ª Doesn’t use labeled data used for complex tasks such
as image recognition, natural
ª It is the process of training ª In this we have inputs X language processing and so on
on algorithm to take a set of that are used for analysis
inputs X and find out a without any targets Y being ª Programs that learn from their
model that best relate them supplied own prediction errors are
to the output Y called reinforced learning
algorithms
ML Algorithm Type
Supervised Unsupervised
Variables
(Target Variable) (No Target Variable)
Continuous Regression Dimensionality Reduction

● Linear; Penalized Regression/LASSO ● Principal Components Analysis
● Logistic (PCA)
● Classification and Regression Tree Clustering
(CART) ● K-Means
● Random Forest ● Hierarchical
Categorical Classification Dimensionality Reduction
● Logit ● Principal Components
● Support Vector Machine (SVM) Analysis (PCA)
● K-Nearest Neighbor (KNN) Clustering
● Classification and Regression Tree ● K-Means
(CART) ● Hierarchical
Continuous or Neural Networks Neural Networks
Categorical Deep Learning Deep Learning
Reinforcement Learning Reinforcement Learning
Dim
STEP 1 en
Complex Database sio
nR
ed
uct
ion
Simple Database
STEP 2
Classification Numerical
STEP 3 Problem Prediction Problem
Data is Non-linear /
Linear Complex Data
Penalized Regression à CART,

à Random Forests
à Neural Network
Supervised Unsupervised
Classification Classification No of Categories
Known
Linear K-Means
STEP 4
No. of Categories
Complex unknown
Linear Non-Linear Non-Linear Hierarchal Clustering
Data Data
KNN à CART
Neural
SUM à Random Forests
Network
à Neural Network
LOS b Describe Overfitting and identify methods of addressing it
Over Fitting
ª Randomness is misperceived to be a pattern
ª Large number of features (i.e., independent variables) are included in the data sample
ª Overfit models do not generalize well to new data (i.e. out-of-sample R-squared will be low)
ª Decreases the accuracy of model forecasts on other (out-of-sample) data

Training datasets
Machine learning
Input given Output given
Algorithm learns
the relationship Validation datasets
Algorithm use is validated
Input Output
Produces
in Sample Test Sample
Errors
Applied
On
Input Output
In Sample errors
Out of
Sample Errors
Bias Error à Model with poor fit
Variance à Out of Sample errors
Error à Due to over fitted
models
Base Error à Residual errors due to
random noise
A learning curve plots the accuracy rate (i.e. - error rate) in the validation or test
sample versus the size of the training sample
Accuracy
Rate
Size of
test Sample
High Bias Error High Variance Error Robust Model

Desired
accuracy
rate
Accuracy Rate
Sample Size
Out of sample
In sample
ª Variance error increases with model complexity, while bias error decreases with complexity
ª Data scientists often express this as a tradeoff between cost and complexity.
ª An optimal level of complexity minimizes the total error and is a key part of successful model
generalization.
Data Scientists use the following methods to reduce the problem of overfitting
Complexity Cross
Reduction Validation
ÜIn complexity reduction, a penalty ÜFor a model to learn sufficiently,

is imposed to exclude features that researchers must ensure that the
are not meaningfully contributing to training data set is both large and
out-of-sample prediction accuracy. representative of the population.
ÜThis penalty value increases with ÜThe validation sample, similarly

the number of independent should be large and representative to
variables (features) used by the property test the model
model
ÜA sampling technique known as cross
validation estimates out-of-sample
error rates directly from the
validation sample.
K Fold Cross Validation
ª In a k-fold cross validation the sample is randomly divided equally into k part.
ª The training sample comprise (k-1) parts.
ª With one part left for validation
ª Error is then measured for the model in each of the parts.
ª This process is repeated k times, and the average in-sample and out-of-sample error
rates are compiled
LOS c Supervised machine learning algorithms - including penalized regression,
support vector machine, k-nearest neighbour, classification and regression
tree, ensemble learning and random forest-and determine the problems for
which they are best suited.
Supervised Machine Learning Algorithms
Supervised ML are trained using labeled data and can be

divided into two groups.
èRegression for a continuous target variable, which includes
ª Penalized regression
ª LASSO
èClassification for a categorical or ordinal target variable, which

includes
ª Support vector machine (SVM)
ª K-Nearest neighbour (KNN)
ª Classification and regression tree (CART) algorithms
Supervised Machine Learning Algorithms
Penalized regression Support Vector Machine (SVM) K-Nearest Neighbor
Is useful for reducing a large

number of features into a SVM is a supervised algorithm
manageable set can help avoid the used for classification, regression
overfitting problem. and outlier detection.
Least Absolute Shrinkage and SVM is a linear classifier which

Selection Operator (LASSO) determines the hyperplane that
optimally separates the
è A Type of penalized regression
observations into two sets of
è Penalty term has the following data points.
from with >0:
K
Uses of SVM Is used to classify a new
ª Penalty term = l å k =1 b k observation by finding
ª Suitable for small-to medium
similarities between the new
ª Is only added during model size but highly complex high observation and existing data.
building process dimensional data sets.
è LASSO eliminates less important ª Is used to predict company

features failures
è Ensures that the Variable is only ª Can be used to classify text

included if sum of squared from documents into useful
residuals declines by more than categories
the increase in penalty term
n K
å ( Y i - Y1 ) 2 + l å b$ k
i=1 k =1
Classification and Regression Tree Ensemble Learning and Random Forest
CART can be applied to: The practice of combining many predictions

from many models and averaging the result to
ª Predict a categorical Target variable to reduce noise.
produce a classification tree.
Ensemble learning is divided into the following
ª Predict a continuous target variable to categories:
produce a regression tree, or
ª Category 1 : An aggregation of
ª To binary classification and regression heterogenous learners
ª Category 2: An aggregation of
homogeneous learners
Bootstrap Aggregating
Voting Classifiers Random Forest
(Bagging)
A collection of a large number of decision

trees trained using the bagging method.
Advantages:
ª Protects against overfitting on the

Ÿ Majority voting classifiers trapping data
will assign the predicted ªThe original training data ª Reduces the ratio of noise to signal
label with the greatest set is used to generate n
number of votes to a new new training data sets or Drawback:
data point. bags of data.
ª Individual trees cannot be interpreted
Ÿ The greater the individual ªAdvantages of bagging with relatives ease
models which are trained, include model stability
the higher the accuracy of and protection against Uses:
aggregated prediction up to overfitting the model.
a certain point. ª In factor-based investment strategies
for asset allocation and investment
selection
ª In predicting whether an IPO will be

successful given attributes
LOS d Unsupervised machine learning algorithms including principal components

analysis, k-means clustering, and hierarchical clustering and determine the
problems for which they are best suited.
Unsupervised Machine Learning Algorithms
Unsupervised ML
è Does not use labelled data
è Algorithms include:
ª Dimension reduced based on principal components analysis
ª Hierarchical clustering
Principal Component
Clustering
Analysis (PCA)
PCA is a statistical method for reducing highly Cluster Contains a subsed of observations from
correlated features into a few main, uncorrelated the data set which are similar
composite variables.
PCA involves two key concepts : Investment uses of clustering :
ª Eigenvectors : New, mutually uncorrelated
ª For grouping companies based on financial
composite variables that are linear
statement items or financial ratios
combinations of the original features
ª Eigenvalue : Proportion of total variance in the ª Improving portfolio diversification
initial data that is explained by each
Popular clustering approaches include:
eigenvector
Drawback of PCA: Principal components cannot ª K-means clustering
be easily labelled or directly interpreted by the
analyst ª Hierachical clustering
Hierarchical Clustering:
K-Means Clustering Dendrograms
Agglomerative and Divisive
Involves partitioning the Hierarchical clustering results in the

data into a fixed number, k, creation of intermediate rounds of
of non- overlapping clusters, clusters of increasing in
k, & the number of clusters, agglomerative) or decreasing
represents a model (conglomerative) size until a final
parameter clustering is reached.
Advantages: Differences between agglomerative Dendrograms highlight the
and divisive hierarchical clustering hierarchical relationships
K-means algorithm is fast Agglomerative Clustering: among the clusters.
and works well with
ª Bottom-up approach
hundreds of millions of
ª Well suited for identifying small
observations
clusters:
Limitation:
Hyperparameter, k, & must ª Top-down approach
be decided before k-means is ª Well suited for identifying large
run. clusters
LOS e Neural networks deep learning nets and reinforcement learning
Neural Networks, Deep Learning Nets, & Reinforcement Learning
Neural Networks Deep Learning Nets Reinforcement Learning
DLNs comprise of a minimum of 3

hidden layers but often more than 20
hidden layers.
Neural networks (also
Uses
known as artificial neural
networks or ANNS) are ª Pattern recognition problems RL involves an agent that
highly flexible and are ª Credit card fraud detection should perform actions with
commonly used for: the objective of maximizing
ª Vision and control problems in
ª Classification learning its rewards over time taking
autonomous cars environmental constraints
ª Regression supervised ª Natural language processing into consideration.
learning
ª Pricing options
ª Reinforcement learning
ª Predicting corporate fundamental
Ÿ reinforcement learning factors and price-related technical
factors
Watch video with important

All queries/doubts about this reading can be posted on
testable concepts here
FinTree Forum for the reading
Big Data Projects

LOS a State and explain steps in a data analysis project
Big data or alternative data encompasses data generated by financial markets, businesses
and many other sources.
Big Data in Investment Management
Difference between big data and traditional data sources rests on 3 V’s
ŒVolume (or quantity of data)

Variety - Array of available data source
ŽVelocity -The speed at which data are created
Steps in Executing a Data Analysis Project : Financial Forecasting with Big Data
Traditional ML Model building steps using

structured data includes:
ŒConceptualization of the modelling task Text ML Model Building steps include four
involves determining what the output of the steps:
model should be, how the model will be used
and by whom, have it will be embedded in Œ Text problem formulation
existing or new business processes
Data collection Data used for financial Data curation
forecasting tasks are mostly numeric data
derived from internal and external sources. Ž Text preparation and wangling
ŽData preparation and wrangling, involves Text exploration
cleansing and processing of raw data
Data exploration includes exploratory data
analysis, feature selection and feature
engineering
‘Model training: Involves selecting the
appropriate ML method, evaluating
performance of trained models, and model
tuning.
LOS b Objectives and steps of preparing and wrangling data

Data Preparation & Wrangling
External data vs Internal data

ª Saves time & resources ª Time - consuming and resource intensive
ª Useful when a project requires generic data ª Useful when a project requires internal data
ª Information edge (or alpha) may be lost
ª During the cleansing process
Structured Data Unstructured Data

Structured Data & Unstructured Data
Structured data are organized in a systematic format

Possible errors in a raw dataset include:
Œ Incompleteness error : Missing or non-present data.
Invalidity error : Where the data are outside of a meaningful range, resulting in invalid data.
Ž Inaccuracy error : Is where data are not a measure of true value
Inconsistency error : Is where data conflicts with corresponding data points or reality
Non-uniformity error : Is where data is not present in an identical format
‘ Duplication error : Is where duplicate observations are present
Data Wrangling (Preprocessing) : Transformation processes for structured data include
ª Extraction
ª Aggregation
ª Filtration
ª Selection
ª Conversion
Outliers can be detected using: Standard deviation: A data value outside 3 standard deviation from
the mean may be an outlier.
Interquartile range (IQR): Difference between the 75th and 25th

percentile data values
Outliers can be handled using: Trimming (extreme values are removed)
Winsorization (extreme values are replaced)
Scaling: The process of adjusting the range of a feature by shifting and adjusting the scale of data.
Two common ways of scaling include:
èNormalization : The process of rescaling numerical variables in the range of [0,1]
X i - X min
X i ( normalized ) =
X max - X min
èStandardization: The process of centering and scaling the variables.
LOS e Preparing, wrangling and exploring text-based data for financial forecasting
Text processing : Transforming unstructured data into structured data and includes two task :
cleansing and preprocessing.
Text Preparation (Cleansing) : Involves cleaning text to remove non-useful elements from raw data.
Text cleansing process includes the following basic operations:
Œ Remove html tags

Remove punctuations
Ž Remove numbers
Remove white spaces
Normalization process in text processing includes the following steps:
u Lowercasing the alphabet

v Stop words such as ‘the’, ‘is’ and ‘a’
w Stemming : A rules-based approach for converting inflected forms of a word into its base
word.
x Lemmatization : Process of converting inflected forms of a word
ç Bag-of-words (BOW), a procedure for analyzing text using a collection of a distinct set of tokens
from all the texts in a simple dataset
. BOW is memory efficient and easy to handle for text analyses but does not capture the position or
ç
sequence of words present in the text.
Document term matrix (DTM) :
è Structure of a DTM:
ª Rows =# off documents in sample dataset.
ª Columns = # of tokens from the BOW that is built using all the documents in a
sample dataset.
è Drawback of BOW : Does not represent word sequences or positions limiting its use for
advanced ML training applications.
è How to overcome drawback of BOW : Using n-grams, to build a BOW

LOS c & f Objectives, methods, and examples of data exploration.

Methods for extracting selecting and engineering features from textual data.
Data Exploration Objectives & Methods
Data exploration involves three tasks:

ª Exploratory data analysis (EDA) : Summarizing data
ª Features election: Selecting relevant features for ML model training
ª Feature engineering : Process of creating new features
Unstructured Data :
Structured Data
Text Exploration
EDA : It is useful to perform EDA of text data by

computing term frequency on the tokens.
Terms frequency=
No. of times a given token occurs in all text in dataset

Total no. of tokens in dataset
Feature Selection : Involves selecting a subset of terms

EDA can be performed on either: or tokens in the dataset which serve as features for the
ML model training.
Ÿ One dimension with summary statistics
(mean, median, quartiles, ranges, Features selection methods in text data include:
stan d ard d eviatio n , skewn ess an d
kurtosis). Ÿ Frequency measures can remove noisy features by
filtering tokens with high and low TF values across all
Ÿ Two dimensions with a summary statistic texts.
of relationships such as correlation
matrix. Ÿ Chi-square test : Tests the independence of two
events.
Feature selection : Data columns in table or
matrix represent features of structured ª Occurrence of token
data.
ª Occurrence of class
Feature Engineering : Helps to optimize and
further improve the features such that they Ÿ Mutual information (MI) measures how much
can describe the structures inherent in the information is contributed by a token to a class of
dataset. texts.
Feature Engineering Techniques include:
Numbers
‚ N-grams:
ƒ Name entity recognition (NER)
„ Part of speech (POS)
LOS d Model Training
ML model training includes the three tasks which may be repeated several
times until desired ML model performance is attained:
ª Method selection
ª Performance evaluation
ª Tuning
A good model fit performs well and can be validated using out-of-sample data.
Types of model fitting:
ª Overfit model may generate no errors with respect to the training data and has best
accuracy, it fits training data too well and is unlikely to perform on future test cases
ª Underfit model does not fit the training data well and it produces misclassification errors.
ª A good fit model may fit the training data well but may not generalize well to out-of-sample
data.
Model fitting errors can be caused by:
ª Dataset size – small datasets may lead to underfitting as small datasets are not sufficient
to expose patterns in the data
ª Number of features: Smaller number of features can lead to underfitting
Method Selection
Supervised or unsupervised learning
ª Supervised models bring a structure that may or may not be supported by data.
ª Unsupervised ML modelling is challenging because of the absence of ground truth (i.e., no

target variable).
Type of data: For
ª Numerical data: CART methods may be suitable.
ª Text data: GLMs and SVMs are commonly used.
ª Image data: Neural networks and deep learning methods are better
ª Speech data: Deep learning methods can offer promising results.
Size of data. A dataset has two basic characteristics:
ªNumber of instances
ªNumber of features
LOS g Calculate the fit of a machine learning algorithm
Performance Evaluation: Techniques for measuring model performance include:
Error analysis : For classification problems, error analysis involves computing four basic
evaluation metrics
ª True positive (TP)
ª False positive (FP)
ª True negative (TN)
ª False negative (FN)
‚ Receiver Operating Characteristic (ROC) : Uses a plotted curve to show trade-off between the
false positive rate (x-axis) and true positive rate (y-axis) for various cutoff points.
ƒRoot Mean Squared Error (RMSE) is :

ª Appropriate for continuous data prediction
ª Mostly used for regression methods
ª Captures all the prediction errors in the data
ª Smaller RMSE indicates better model performance
ª Formula n ( Perdicted i - Actuali ) 2

å i =1 n
Financial Forecasting Project: Classifying & Predicting Sentiment for Stocks
Robo-readers:
ª Are being used to examine how views expressed in text relate to future
company performance.
ª Analyze sentiment polarity
Economics
Notice : The recipient of this publication is strictly prohibited by law to circulate. We have inserted a
concealed code in the document, which will lead to identification of the user to whom this document was
issued. If this documents is found to be circulated on internet, social media sites and other mode thereto,
the user identification will be reported to CFA Institute and strict legal action will be initiated.
Unless otherwise stated, copyright and all intellectual property rights in all the course material(s)
provided, is the property of FinTree Education Private Limited. Any copying, duplication of the course
material either directly and/or indirectly for use other than for the purpose provided shall tantamount to
infringement and shall strongly defended and pursued, to the fullest extent permitted by law.
The unauthorized duplication of these notes is a violation of global copyright laws and the CFA Institute
code of Ethics. Your assistance in pursuing potential violators of this law is greatly appreciated. If any
violation comes to your notice, get in touch with us at admin@fintreeindia.com
FinTree
TM
CommuterNotes
A normal person spends about two-three hours travelling
every day – going to work, walking down the street to buy stuff
or just going for a walk. Wondering how to utilize this travel
time? The geeks at FinTree have the perfect solution for you!
FinTree's Commuter Notes! Helping you study when you can't!

Commuter Notes are short audio clips that can be
downloaded on any smart phone. These audios are an
interaction between the faculty and 2-3 candidates discussing
a topic and will help you learn subconsciously!
Currency Exchange Rates:

Understanding Equilibrium Value
Introduction
Exchange rate Price of one unit of currency in terms of another
Spot exchange rate Exchange rate for immediate delivery
Forward exchange rate Exchange rate for a transaction to be done in future
$3 Price currency
€ Base currency
€ - Depreciated $2 $3 $4 € - Appreciated
$ - Appreciated € € € $ - Depreciated
Eg. ZAR 52 ZAR 57

$ $
Closing value Opening value
% Appreciation: − 1 % Depreciation: − 1
Opening value Closing value
57
$ - Appreciated: −1 = 9.62%
52
52
ZAR - Depreciated: −1 = 8.77%
57
LOS a Bid–ask spread on spot or forward foreign currency quotation
Bid Ask
Bank will buy Bank will sell
ª Ask > Bid
ª If base currency is bought, party is said to have paid the offer

ª If base currency is sold, party is said to have hit the bid
Interbank market: Where dealers buy and sell foreign exchange among themselves
Bid-ask spread provided by a dealer to clients is wider than bid-ask spread used in
the interbank market
Spreads are stated as ‘PIPs’. 1 PIP = 1/10,000

Factors that affect bid-ask spreads
Bid-ask spread Bid-ask spread in

quoted to clients interbank market
Spread in the interbank market for two

currencies involved Currency pair involved
Size of the transaction Time of the day

Higher size ~ Larger Spread
Market volatility
Relationship between dealer and client
LOS b Triangular arbitrage opportunity

Two arbitrage constraints
Bid shown by a dealer in Cross-rate bids must be

the interbank market can lower than the implied
not be higher than ask cross-rate asks
Eg.
Bid/Ask:
₹ = 65.1020/65.2030
$
= 1.2125/1.3135
$ €
Calculate market-implied bid-ask quote on €/ ₹
1 Bid × Bid = Bid Ask × Ask = Ask
2 Inverted bid = New ask Inverted ask = New bid
New bid New ask
65.2030 × 1.3135 = 85.6441 65.1020 × 1.2125 = 78.9361
1 1
= 0.0116 = 0.0126
85.6441 78.9361
Triangular arbitrage
£ £ $
Eg. Bid/Ask: = 1.1189/1.1213 = 0.7526/0.7545 = 1.1820/1.1824
€ $ €
€ 891,822 $ 1,325,381
Price currency
Base currency
Multiply with Divide by

bid ask
(Up the Quote) (Down the

Quote)
£ 1,000,000 $ 1,054,133 £ 1,000,000 € 1,120,924
£ 793,340 £ 1,254,202
✗ ✔
LOS c Forward premium/discount for a given currency
Eg. #1 Bid/Ask: Spot $/€ = 1.1820/1.1824 Forward points (3 months) = −15.2/−14.6
15.2
3 month forward bid rate = 1.1820 − = 1.1804
10,000
14.6
3 month forward ask rate = 1.1824 − = 1.1809
10,000
Eg. #2 Spot rate MXN/USD = 19.26 Forward rate MXN/USD = 18.35

Calculate forward premium/discount
0.91
USD discount: = -4.72%
18.35
0.91
MXN premium: = 4.95%
19.26
Forward contract: Any exchange rate transaction that has a settlement date longer than T + 2
Forward premium/discount = Forward rate – Spot rate
LOS d Mark-to-market value of a forward contract

Mark-to-market: Profit/loss that is realizable from closing out a position
Eg. Forward contract: 1 mln GBP Rate: 1.3000 USD/GBP Term: 6 months
Spot rate after 90 days: 1.3100/1.3105 90-day forward points : +120/+125 90-day LIBOR: 4%
All-in three-month forward rate: 1.3100 + (120/10,000) = 1.3220
CF at the settlement date: (1.3220 − 1.300) × 1,000,000 = USD 22,000
USD 220,000
Mark-to-market value: = USD 217,821
1 + 0.04 × (90/360)
Factors that affect the bid-ask spread:
Spread in the interbank market for two currencies involved

Size of the transaction
Relationship between dealer and client
Term of the forward contract (longer the term, wider the spread)
LOS e & f International parity conditions
International Fischer relationship (precise)
1 + Nominal interest rate = (1 + Real interest rate) × (1 + Expected inflation)
Determining forward rate

USA India
2% ₹50 10%
$
$1mln ₹ 50mln
$1mln 10%
+ 2% int.
$1.02mln
₹55mln 53.92
55
₹55mln
53.92 1.02
(1 + Int. rate)n
Forward rate = S ×
(1 + Int. rate)n
(1 + 10%)1
Forward rate = 50 ×
(1 + 2%)1
= 53.92
Interest rate parity
Int. rate (India) = 20%

Int. rate (USA) = 10%
(1 + Int. rate)n
₹50
F = S ×
(1 + Int. rate)n = ₹54.54
$
Expected
(1.1538)
spot rate = 50 ×
(1.0576) = ₹54.54
Real int.
Inflation rate
rate = 4%
(1 + 20%)
India = = 15.38%
(1 + 4%)
(1 + 10%)
USA = = 5.76%
(1 + 4%)
Covered interest
Forced by arbitrage. It is always true
rate parity
Uncovered interest
Not forced by arbitrage. It may not be true
rate parity
Eg. Spot rate: $1.2/€ 1-year forward rate: $1.3/€
USD interest rate: 9% Euro interest rate: 7% Determine if an arbitrage opportunity exists
Forward rate 1 + USD int. rate

=
Spot rate 1 + Euro int. rate
Forward rate
× 1 + Euro int. rate Vs 1 + USD int. rate
Spot rate
1.3
× 1 + 7% Vs 1 + 9%
1.2
1.1591 Vs 1.09
Invest Borrow
Covered interest rate parity:

Nominal interest rateA − Nominal interest rateB = % forward premium/discount
Uncovered interest rate parity:

Nominal interest rateA − Nominal interest rateB = Expected % ∆ spot rate
Both covered and uncovered interest rate parity:

Forward exchange rate will be an unbiased predictor of the future spot rate
Ex ante purchasing power parity:

% ∆ spot rate = Expected inflationA – Expected inflationB
Fisher effect:
Nominal interest rateA − Nominal interest rateB = Expected inflationA – Expected inflationB
Both ex ante PPP and Fisher effect:

(Nominal interest rateA − Nominal interest rateB) + % ∆ spot rate = Expected inflationA – Expected InflationB
LOS g & h Use of international parity conditions
Future spot rates Forward exchange rates
Estimated using purchasing power

parity/uncovered interest rate parity Estimated using covered interest parity
Not forced by arbitrage and does not Forced by arbitrage and is always true
work in the short term
If uncovered interest rate parity holds, then covered interest rate parity holds
(forward exchange rate is an unbiased predictor of the future spot rate)
LOS i Carry trade
Carry trade: Investor borrows in lower yielding currency (funding currency) and invests in
higher yielding currency and hope that high yielding currency will depreciate less than
interest rate differential
It assumes that uncovered interest rate parity does not work in the short term
Carry trade generates positive returns during the periods of low volatility
Crash risk: Probability of substantial losses due to high volatility and/or perceived risk in
financial markets
Distribution: Non-normal, more peaked, fatter tails and −ve skewness
LOS j Impact of balance of payment flows on currency exchange rates
Flow supply/demand Portfolio balance Debt sustainability

channel channel channel
Countries with current

Current account surplus: account deficits usually have Capital account surplus means
Appreciation of currency capital account surpluses borrowing > lending
Current account deficit: Investor countries’ portfolio For deficit countries, rising
Depreciation of currency composition is dominated by Debt/GDP ratio will lead to
few investee countries depreciation of currency
Appreciation/depreciation of
currency would help eliminate If investor countries decide to For surplus countries, rising
the initial imbalance in the reduce the holdings, it can Assets/GDP ratio will lead to
long run lead to depreciation of appreciation of currency
investee countries’ currencies
Excessive emerging market capital inflows create problems such as:
ª Unwarranted appreciation of the EM currency

ª Increases in external debt
ª An asset bubble
ª Excessive consumption that contributes to huge growth in
domestic credit and/or the current account deficit
ª Overinvestment in risky projects
LOS k Potential effects of monetary and fiscal policy

Mundell-Fleming model
Capital mobility
Monetary policy Fiscal policy
High Low
Expansionary Expansionary Indeterminate Depreciation
Expansionary Contractionary Depreciation Indeterminate
Contractionary Expansionary Appreciation Indeterminate
Contractionary Contractionary Indeterminate Appreciation
Pure monetary model Dornbusch’s modified monetary model
X% increase in money supply leads to an Short run:

x% increase in price level and then an x% Prices have limited flexibility and,
depreciation of domestic currency Domestic currency depreciates. PPP does not
hold
Shortcoming:
Assumption that PPP holds both in short Long run:
term and long term Prices will increase and,
Domestic currency will appreciate and move
toward values predicted by conventional
monetary approach
LOS l Objectives of central bank or government

intervention and capital controls
Capital controls and central bank intervention aim to reduce excessive

capital inflows which could lead to speculative bubbles
Developed countries Emerging market countries
Foreign exchange reserves Foreign exchange reserves

Currency trading volume Currency trading volume
Low Relatively high
Less ability to manage Some ability to manage

exchange rate exchange rate
LOS m Warning signs of a currency crisis

Deterioration in terms of trade (ratio of exports to imports)
Dramatic decline in foreign exchange reserves
Real exchange rate substantially higher than its mean-reverting level
Increases in inflation rate
Boom-bust cycle in equity markets

Economic Growth And The

Investment Decision
LOS a Factors favoring and limiting economic growth
in developed and developing economies
Key requirements for growth
Savings and investment: High savings High levels of investment High GDP growth
Financial markets and Ÿ Financial sector channels savings to projects that offer highest
intermediaries: risk -adjusted returns
Ÿ Financial sector encourage savings by creating financial
instruments that facilitate risk transfer and enhance liquidity
Ÿ Aggregating small amounts of savings into a larger pool
enables intermediaries to finance larger projects
Political stability, rule of Ÿ Countries that have stable and effective government, and well
law, and property rights: developed legal, regulatory and property rights system have
higher economic growth
Ÿ Factors such as wars, corruption and political instability raise
investment risk and weaken economic growth
Education and Ÿ Basic education raises the skill level of the workforce which
healthcare systems: contributes to potential economic growth
Ÿ Education can also raise growth by increasing the productivity
of existing physical capital
Ÿ Empirical studies show that poor health has resulted in slowing
down of economic growth
Tax and regulatory
systems: Limited regulations More growth and productivity
Free trade and Ÿ Foreign investments break out the cycle of low income, low
unrestricted capital flows: savings, and low investment
Ÿ It can be direct (FDI) or indirect (buying equity/debt issued by
domestic companies)
Ÿ Free trade benefits a country’s economy by providing more
goods at lower costs
LOS b Relation b/w long-run rate of stock market appreciation

and sustainable growth rate of the economy
∆P = ∆GDP + ∆(E/GDP) + ∆(P/E)
Aggregate price Aggregate

of equities earnings
Over the short and medium term all three factors contribute to increase/decrease
in stock market, but in the long-term growth rate of GDP dominates
(E/GDP) and (P/E) can’t rise forever or can’t decline forever
Over the long-term, (E/GDP) and (P/E) will be zero. Therefore:
∆P = ∆GDP
LOS c Why potential GDP and its growth rate matter for equity and fixed
income investors
Higher potential GDP
Consumers expect income

to rise
They increase current

consumption and save less
Higher real interest rates

to encourage savings
ª Actual GDP > Potential GDP = Inflationary gap. This results in higher nominal interest rates
(restrictive monetary policy) and fiscal surplus
ª Higher potential GDP growth rate improves credit quality of fixed income securities
ª Higher potential GDP growth rate reduces expected credit risk
LOS d & e Cobb-Douglas production function
Absolute Relative
Y = T Kα L(1 − α)
Known as growth accounting equation
Y = GDP, T = Total factor productivity,
K = Capital, L = Labor, α = Share of capital %∆Y = %∆TFP + α × %∆K + (1 − α) × %∆L
in GDP
Cobb-Douglas production function exhibits constant returns to scale

Dividing both sides of the equation by L, we can obtain output per worker. Y/L = T × (K/L)α
Increasing all inputs by a fixed % leads to same % change in the output
α = r × (K/Y)
Economies will increase as long as MPK > r
Higher α suggests higher importance for capital for developing
Capital deepening
It is an increase in capital-to-labor ratio
It is the movement along the productivity curve
Once capital-to-labor ratio becomes very high, further additions to capital

have relatively less impact on GDP (diminishing marginal productivity)
Technological progress enhances the productivity of both capital and labor
Long-term growth rate can be increased by technological progress
Technological progress will cause shift in productivity curve
Labor productivity accounting growth equation:

Growth rate in potential GDP = long-term growth rate of labor force
+ long-term growth rate in labor productivity
Advantage: No need to estimate capital input and compute TFP
Disadvantage: Capital deepening and TFP progress can be difficult to

analyze and predict over long-term
LOS f Effect of natural resources on economic growth

Ownership of natural resources is not necessary for growth
Reasons for slow economic growth in countries with abundant natural resources:
Œ Natural resources may fail to develop economic institutions necessary for growth
Countries may suffer Dutch disease (high demand for natural resources results in appreciation
of currency, which makes other domestic industries uncompetitive in global markets)
LOS g Effect of following on the rate and sustainability of economic growth
Labor force Average hours

Demographics Immigration
participation worked
Highly sensitive to
Population growth is business cycle
determined by fertility Labor force
rates and mortality participation rate: Long-term trend in
It is a possible solution
rates average hours worked
Labor force to declining labor force
is downwards
Working age population growth which is
Population growth may
experienced by
increase growth rate of Causes of this
Increase in this rate developed countries
the overall economy development:
may increase per capita with low birthrates
but it has no impact on Legislation, growth of
the rate of increase in GDP part-time and
per capita GDP temporary work,
wealth effect etc.
LOS h How investment in physical capital, human capital, and

technological development affects economic growth
ª Human capital: Knowledge and skills that workers possess
ª Physical capital: ICT (infrastructure, computers, and telecommunications equipment) and non-ICT
ª There is high correlation b/w investment in physical and human capital, and economic growth
ª Introduction of technology results in an upward shift in the production function
ª Improvements in infrastructure generally boosts the productivity of private investments

LOS i
Classical Neoclassical Endogenous
growth theory growth theory growth theory
Key assumption: Assumption: Constant

Population growth Both capital and labor returns to capital
accelerates when the are variable inputs and
subject to diminishing Technological progress is
level of per capita
marginal productivity treated endogenous
income rises above the
(coming from within)
subsistence level
Growth rate of output Economy does not
Technological progress per capita: necessarily converge to a
and land expansion ð steady state growth rate
Increase in labor Φ
Increase in savings
productivity ð Higher 1−α
permanently increases
population growth
growth rate
Growth rate of output:
In the long run, adoption Human capital and R&D
of new technology Φ spending are factors of
+ ∆L
results in larger but not 1−α production just like
richer population capital and labor
Capital deepening has no
Spending on human
Growth is temporary effect on growth rate or
capital and R&D
MPK once the steady
generates benefits to the
It was labeled ‘Dismal state is reached
economy as a whole
science’
LOS j Convergence hypotheses
Absolute convergence Conditional convergence Club convergence
Only rich and middle-income

countries that are members of
Convergence is conditional on the club (having similar
Developing countries will
the countries having same institutional structure) are
match per capita output of
saving rate, population converging to the richest
developed countries
growth rate, and production countries in the world
function
It implies convergence of per Countries with the lowest per
capita growth rates among all capita income in the club
It implies convergence of per
countries but does not imply grow at the fastest rate
capita output as well as
convergence of level of per
convergence of steady state
capita income Poor countries can join the
growth rate
club if they make appropriate
institutional changes
LOS k Economic rationale for governments to provide incentives to

private investment in technology and knowledge
Under endogenous growth theory, private sector investments in
R&D and human capital benefits the society overall
Government incentives can increase private sector investments
in R&D, which can lead to overall growth in the economy
LOS l Expected impact of removing trade barriers
As per neoclassical growth theory:

Convergence will be more quick if economies are open and there is free trade
Developing countries can grow more rapidly if there is free trade and unrestricted capital flows
As per endogenous growth theory:

More open trade will permanently increase growth rate
International trade increases global output through selection effect, scale effect, and
backwardness effect
Economics Of Regulation
LOS a Economic rationale for regulation
Pareto Optimal : Regulation is often required when markets cannot provide efficient solution
Which means that one cannot make any participants better off without making some other
participants worse off
Regulations are needed in the presence of:
Informational frictions : Regulation are put in place in an attempt to ensure that no

participant is treated unfairly or is at a Disadvantage
Externalities : Cost or benefit that affects the party that did not choose to incur that cost or
benefit. For example a polluter may not bear the full cost of their action
Weak Competition : Fewer choices, higher prices and lack of innovation
Social Objectives : Public good is a resource that can be enjoyed by a person without making
it unavailable to others.
Since people share in consumption, regulations are necessary to ensure an optimal level of
production of such public goods
LOS b Explain the purposes of regulating commerce and financial markets
Regulating commerce Regulating Financial Markets
ª Company Laws, tax laws, contract laws, ª To prevent failures of the financial system
competition laws, labour laws
ª Maintain the integrity of markets
ª Regulation may facilitate or hinder commerce.
For example, protections of intellectual property ª Objective : 3 Interelated Goals
facilitate long-term investments in research ŒProtect investors
ŽCreate confidence in the markets
ŽEnhance capital formation
Regulation of Security markets

Several observations can be made about securities markets regulation:
Disclosure : Disclosure provide transparency (i.e, reduce information asymmetry) in financial markets
and hence promote investor confidence
Agency problems : Regulation imposing fiduciary duties seek to mitigate such agency problem
Focused on protecting small investors : Hence the relatively lax regulatory coverage of hedge funds
and private equity funds that are marketed only to qualified investors
Regulation of Financial Institution
Prudential Supervision :
Ÿ Prudential supervision refers to the monitoring and regulation of financial
institutions to reduce system-wide risk and to protect investors
Ÿ Prudential supervision focuses on diversification of assets, an adequate

capital base, and risk management activities of financial institutions
Ÿ The cost benefit analysis of financial market regulation should also include
hidden costs. For example, FDIC insurance for banks may incentivize them
with excessive risk-taking (a moral hazard problem)
LOS c Describe anticompetitive behaviours targeted by antitrust laws globally and

evaluate the antitrust risk with a given business strategy
Anititrust Regulation
ª Antitrust laws work to promote domestic competition by monitoring and restricting activities
that reduce or distort competition
ª Regulators often block a merger that leads to an excessive concentration of market share
ª Anticompetitive behaviour such as price collusion, discriminatory pricing, bundling, and

exclusive dealing is often also prohibited
LOS d Classifications of regulations and regulators
Regulations
Administrative
Statutes Judicial law
regulations
Rules issued by
Laws made by government Interpretations of
legislative bodies agencies or other courts
regulators
Regulators
Government Independent
agencies regulators
Independent regulators are given recognition by

government agencies and have power to make
rules and enforce them
However, independent regulators are usually not

funded by the government and hence are
politically independent
SRB’s
Industry self-regulatory bodies(SRBs) are private

organization that represent as well as regulate
their members
SRBs may have inherent conflicts of interest
SRBs nonetheless are attractive in that they

increase the overall level of regulatory resources,
utilize the industry professionals with the
requisite expertise, and allow regulators to
devote resources to other priorities
SRBs + Govt Recognition = SRO
Enforcement Powers
LOS e Describe uses of self-regulation in financial markets
ª FINRA is an SRO recognized by the SEC in the United States. FINRA’s primary objective is to
protect investors by maintaining the fairness of the U.S. capital markets. FINRA has the
authority to enforce security laws and regulations
ª However, the use of SROs in civil-law countries is not common; in such countries, formal
government agencies fulfill the role of SROs
ª In common-law countries such as the United Kingdom and the United States, SROs have
historically enjoyed recognition
LOS f Describe regulatory interdependencies and their effects
Regulator capture : The Regulatory capture is based upon the assumption that,
regardless of the original purpose behind its establishment,
a regulatory body will, at some point in time, be influenced
or even possibly controlled by the industry that is being
regulated
Regulatory capture is more likely to be concern with SROs

than with government agencies
Regulatory competition & Arbitrage : Regulatory differences between jurisdiction can lead to
regulatory competition, in which regulators compete to
provide the most business-friendly regulatory environment
Regulatory arbitrage occurs when business shop for a

country that allows a specific behaviour rather than
changing the behaviour
Regulatory arbitrage also exploiting the difference between

the economic substance and interpretation of a regulation
LOS g Describe tools of regulatory intervention in markets
Tools of Regulatory intervention
Price Mechanisms Restricting or requiring Provision of public goods

certain activities or financing of private
projects
Regulatory may ban
certain activities (eg. : Use Regulatory may provide
of Specific chemicals) or public good (e.g. National
For Example : Sin Taxes require that certain Defence) or Fund private
activities be performed projects (e.g. Small-
SROs and outside bodies business loans, student
are least likely to use Filling of 10-K reports by loans) depending on their
price mechanisms publicly listed companies political priorities and
objectives
SROs and outside bodies
are least likely to use price SROs and outside bodies
mechanisms are least likely to use price
mechanisms
Regulatory tools developed in response to past events may not necessarily

work well under a different set of circumstances in the future; the jury is still
out on the bail-in process
LOS h
Describe benefits and costs of regulation
Regulatory Burden : Regulatory burden (also known as government burden)

refers to the cost of compliance for the regulated entity.
Regulatory burden minus the private benefits of the
regulation is known as the net regulatory burden
Regulators should be aware of unintended consequences

of regulations
Sunset Clause : Regulatory costs are difficult to assess before a regulation

is out in place. For this reason, many regulatory provisions
include a sunset clause that requires regulators to revisit
the cost benefit analysis based on actual outcomes before
renewing the regulation
LOS i
Describe the considerations when evaluating the effects of regulation on
an industry
ªRegulation can help or hinder a company or industry. Regulation may shrink the size of
one industry( e.g. if it is heavily taxed) while increasing the size of another (e.g. an
industry receiving subsidies).
ªRegulation are not always costly for those that end up being regulated. If the regulator is
captive, regulation may end up benefiting the regulated entities
ªRegulation may introduce inefficiencies in the market. For example, past government
bailouts of financial institutions have conveyed a message of future implicit guarantees.
For this reason, the credit spreads on bonds issued by the financial sector may not fully
reflect their risk

Z- TABLE (COMPLEMENTARY CUMULATIVE)
StandardNormal Distribution
P (Z ::;; z) = N(z) for z ::;; 0
z
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09

0 0.5000 0.4960 0.4920 0.4880 0.4840 0.4801 0.4761 0.4721 0.4681 0.4641
-0.1 0.4602 0.4562 0.4522 0.4483 0.4443 0.4404 0.4364 0.4325 0.4286 0.4247
-0.2 0.4207 0.4168 0.4129 0.4090 0.4052 0.4013 0.3974 0.3936 0.3897 0.3859
-0.3 0.3821 0.3783 0.3745 0.3707 0.3669 0.3632 0.3594 0.3557 0.3520 0.3483
-0.4 0.3446 0.3409 0.3372 0.3336 0.3300 0.3264 0.3228 0.3192 0.3156 0.3121
-0.5 0.3085 0.3050 0.3015 0.2981 0.2946 0.2912 0.2877 0.2843 0.2810 0.2776
-0.6 0.2743 0.2709 0.2676 0.2643 0.2611 0.2578 0.2546 0.2514 0.2483 0.2451
-0.7 0.2420 0.2389 0.2358 0.2327 0.2296 0.2266 0.2236 0.2206 0.2177 0.2148
-0.8 0.2119 0.2090 0.2061 0.2033 0.2005 0.1977 0.1949 0.1922 0.1894 0.1867
-0.9 0.1841 0.1814 0.1788 0.1762 0.1736 0.1711 0.1685 0.1660 0.1635 0.1611
-1 0.1587 0.1562 0.1539 0.1515 0.1492 0.1469 0.1446 0.1423 0.1401 0.1379
-1.1 0.1357 0.1335 0.1314 0.1292 0.1271 0.1251 0.1230 0.1210 0.1190 0.1170
-1.2 0.1151 0.1131 0.1112 0.1093 0.1075 0.1056 0.1038 0.1020 0.1003 0.0985
-1.3 0.0968 0.095'1 0.0934 0.0918 0.0901 0.0885 0.0869 0.0853 0.0838 0.0823
-1.4 0.0808 0.0793 0.0778 0.0764 0.0749 0.0735 0.0721 0.0708 0.0694 0.0681
-1.5 0.0668 0.0655 0.0643 0.0630 0.0618 0.0606 0.0594 0.0582 0.0571 0.0559
-1.6 0.0548 0.0537 0.0526 0.0516 0.0505 0.0495 0.0485 0.0475 0.0465 0.0455
-1.7 0.0446 0.0436 0.0427 0.0418 0.0409 0.0401 0.0392 0.0384 0.0375 0.0367
-1.8 0.0359 0.0351 0.0344 0.0336 0.0329 0.0322 0.0314 0.0307 0.0301 0.0294
-1.9 0.0287 0.0281 0.0274 0.0268 0.0262 0.0256 0.0250 0.0244 0.0239 0.0233
-2 0.0228 0.0222 0.0217 0.0212 0.0207 0.0202 0.0197 0.0192 0.0188 0.0183
-2.1 0.0179 0.0174 0.0170 0.0166 0.0162 0.0158 0.0154 0.0150 0.0146 0.0143
-2.2 0.0139 0.0136 0.0132 0.0129 0.0125 0.0122 0.0119 0.0116 0.0113 0.0110
-2.3 0.0107 0.0104 0.0102 0.0099 0.0096 0.0094 0.0091 0.0089 0.0087 0.0084
-2.4 0.0082 0.0080 0.0078 0.0075 0.0073 0.0071 0.0069 0.0068 0.0066 0.0064
-2.5 0.0062 0.0060 0.0059 0.0057 0.0055 0.0054 0.0052 0.0051 0.0049 0.0048
-2.6 0.0047 0.0045 0.0044 0.0043 0.0041 0.0040 0.0039 0.0038 0.0037 0.0036
-2.7 0.0035 0.0034 0.0033 0.0032 0.0031 0.0030 0.0029 0.0028 0.0027 0.0026
-2.8 0.0026 0.0025 0.0024 0.0023 0.0023 0.0022 0.0021 0.0021 0.0020 0.0019
-2.9 0.0019 0.0018 0.0018 0.0017 0.0016 0.0016 0.0015 0.0015 0.0014 0.0014
-3.0 0.0013 0.0013 0.0013 0.0012 0.0012 0.0011 0.0011 0.0011 0.0010 0.0010
56
Z-TABLE (CUMULA TIVE)
StandardNormal Distribution
P (Z :-s; z) = N(z) for z 2 0
z
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09

0 0.5000 0.5040 0.5080 0.5120 0.5160 0.5199 0.5239 0.5279 0.5319 0.5359
0.1 0.5398 0.5438 0.5478 0.5517 0.5557 0.5596 0.5636 0.5675 0.5714 0.5753
0.2 0.5793 0.5832 0.5871 0.5910 0.5948 0.5987 0.6026 0.6064 0.6103 0.6141
0.3 0.6179 0.6217 0.6255 0.6293 0.6331 0.6368 0.6406 0.6443 0.6480 0.6517
0.4 0.6554 0.6591 0.6628 0.6664 0.6700 0.6736 0.6772 0.6808 0.6844 0.6879
0.5 0.6915 0.6950 0.6985 0.7019 0.7054 0.7088 0.7123 0.7157 0.7190 0.7224
0.6 0.7257 0.7291 0.7324 0.7357 0.7389 0.7422 0.7454 0.7486 0.7517 0.7549
0.7 0.7580 0.7611 0.7642 0.7673 0.7704 0.7734 0.7764 0.7794 0.7823 0.7852
0.8 0.7881 0.7910 0.7939 0.7967 0.7995 0.8023 0.8051 0.8078 0.8106 0.8133
0.9 0.8159 0.8186 0.8212 0.8238 0.8264 0.8289 0.8315 0.8340 0.8365 0.8389
1 0.8413 0.8438 0.8461 0.8485 0.8508 0.8531 0.8554 0.8577 0.8599 0.8621
1.1 0.8643 0.8665 0.8686 0.8708 0.8729 0.8749 0.8770 0.8790 0.8810 0.8830
1.2 0.8849 0.8869 0.8888 0.8907 0.8925 0.8944 0.8962 0.8980 0.8997 0.9015
1.3 0.9032 0.9049 0.9066 0.9082 0.9099 0.9115 0.9131 0.9147 0.9162 0.9177
1. 4 0.9192 0.9207 0.9222 0.9236 0.9251 0.9265 0.9279 0.9292 0.9306 0.9319
1.5 0.9332 0.9345 0.9357 0.9370 0.9382 0.9394 0.9406 0.9418 0.9429 0.9441
1.6 0.9452 0.9463 0.9474 0.9484 0.9495 0.9505 0.9515 0.9525 0.9535 0.9545
1.7 0.9554 0.9564 0.9573 0.9582 0.9591 0.9599 0.9608 0.9616 0.9625 0.9633
1. 8 0.9641 0.9649 0.9656 0.9664 0.9671 0.9678 0.9686 0.9693 0.9699 0.9706
1.9 0.9713 0.9719 0.9726 0.9732 0.9738 0.9744 0.9750 0.9756 0.9761 0.9767
2 0.9772 0.9778 0.9783 0.9788 0.9793 0.9798 0.9803 0.9808 0.9812 0.9817
2. 1 0.9821 0.9826 0.9830 0.9834 0.9838 0.9842 0.9846 0.9850 0.9854 0.9857
2.2 0.9861 0.9864 0.9868 0.9871 0.9875 0.9878 0.9881 0.9884 0.9887 0.9890
2.3 0.9893 0.9896 0.9898 0.9901 0.9904 0.9906 0.9909 0.9911 0.9913 0.9916
2.4 0.9918 0.9920 0.9922 0.9925 0.9927 0.9929 0.9931 0.9932 0.9934 0.9936
2.5 0.9938 0.9940 0.9941 0.9943 0.9945 0.9946 0.9948 0.9949 0.9951 0.9952
2.6 0.9953 0.9955 0.9956 0.9957 0.9959 0.9960 0.9961 0.9962 0.9963 0.9964
2.7 0.9965 0.9966 0.9967 0.9968 0.9969 0.9970 0.9971 0.9972 0.9973 0.9974
2.8 0.9974 0.9975 0.9976 0.9977 0.9977 0.9978 0.9979 0.9979 0.9980 0.9981
2.9 0.9981 0.9982 0.9982 0.9983 0.9984 0.9984 0.9985 0.9985 0.9986 0.9986
3 0.9987 0.9987 0.9987 0.9988 0.9988 0.9989 0.9989 0.9989 0.9990 0.9990
57
STUDENTS T-DISTRIBUTION
Level of significance for One-Tailed Test
df l 0.1 1 o.o5 1 o.o25 1 o.o1 1 o.oo5 1 o.ooo5
Level of significance for Two-Tailed Test

df 0.2 0.1 0.05 0.02 0.01 0.001
1 3.0777 6.3138 12.7062 31.8205 63.6567 636.6192
2 1.8856 2.9200 4.3027 6.9646 9.9248 31.5991
3 1.6377 2.3534 3.1824 4.5407 5.8409 12.9240
4 1.5332 2.1318 2.7764 3.7469 4.6041 8.6103
5 1.4759 2.0150 2.5706 3.3649 4.0321 6.8688
6 1.4398 1.9432 2.4469 3.1427 3.7074 5.9588

7 1.4149 1.8946 2.3646 2.9980 3.4995 5.4079
8 1.3968 1.8595 2.3060 2.8965 3.3554 5.0413
9 1.3830 1.8331 2.2622 2.8214 3.2498 4.7809
10 1.3722 1.8125 2.2281 2.7638 3.1693 4.5869
11 1.3634 1.7959 2.2010 2.7181 3.1058 4.4370

12 1.3562 1.7823 2.1788 2.6810 3.0545 4.3178
13 1.3502 1.7709 2.1604 2.6503 3.0123 4.2208
14 1.3450 1.7613 2.1448 2.6245 2.9768 4.1405
15 1.3406 1.7531 2.1314 2.6025 2.9467 4.0728
16 1.3368 1.7459 2.1199 2.5835 2.9208 4.0150

17 1.3334 1.7396 2.1098 2.5669 2.8982 3.9651
18 1.3304 1.734 1 2.1009 2.5524 2.8784 3.9216
19 1.3277 1.7291 2.0930 2.5395 2.8609 3.8834
20 1.3253 1.7247 2.0860 2.5280 2.8453 3.8495
21 1.3232 1.7207 2.0796 2.5 176 2.8314 3.8193

22 1.3212 1.7171 2.0739 2.5083 2.8 188 3.7921
23 1.3195 1.7139 2.0687 2.4999 2.8073 3.7676
24 1.3178 1.7109 2.0639 2.4922 2.7969 3.7454
25 1.3163 1.7081 2.0595 2.4851 2.7874 3.7251
26 1.3150 1.7056 2.0555 2.4786 2.7787 3.7066

27 1.3137 1.7033 2.0518 2.4727 2.7707 3.6896
28 1.3125 1.7011 2.0484 2.4671 2.7633 3.6739
29 1.3114 1.6991 2.0452 2.4620 2.7564 3.6594
30 1.3104 1.6973 2.0423 2.4573 2.7500 3.6460
40 1.3031 1.6839 2.0211 2.4233 2.7045 3.5510

60 1.2958 1.6706 2.0003 2.3901 2.6603 3.4602
120 1.2886 1.6577 1.9799 2.3578 2.6174 3.3735
200 1.2858 1.6525 1.9719 2.345 1 2.6006 3.3398
00 1.2816 1.6449 1.9600 2.3264 2.5759 3.2906
58
F-TABLE AT 5 PERCENT (UPPER TAIL)
Degrees of freedom of numerator along the top most row

Degrees of freedom of denominator along the left most column
df 1 2 3 4 5 6 7 8 9 10 12 15 20 24 30 40
1 161 199 216 225 230 234 237 239 241 242 244 246 248 249 250 251
2 18.5 19.0 19.2 19.2 19.3 19.3 19.4 19.4 19.4 19.4 19.4 19.4 19.4 19.5 19.5 19.5
3 10.1 9.55 9.28 9.12 9.01 8.94 8.89 8.85 8.81 8.79 8.74 8.70 8.66 8.64 8.62 8.59
4 7.71 6.94 6.59 6.39 6.26 6.16 6.09 6.04 6.00 5.96 5.91 5.86 5.80 5.77 5.75 5.72
5 6.61 5.79 5.41 5.19 5.05 4.95 4.88 4.82 4.77 4.74 4.68 4.62 4.56 4.53 4.50 4.46
6 5.99 5.14 4.76 4.53 4.39 4.28 4.21 4.15 4.10 4.06 4.00 3.94 3.87 3.84 3.81 3.77
7 5.59 4.74 4.35 4.12 3.97 3.87 3.79 3.73 3.68 3.64 3.57 3.51 3.44 3.41 3.38 3.34
8 5.32 4.46 4.07 3.84 3.69 3.58 3.50 3.44 3.39 3.35 3.28 3.22 3.15 3.12 3.08 3.04
9 5.12 4.26 3.86 3.63 3.48 3.37 3.29 3.23 3.18 3.14 3.07 3.01 2.94 2.90 2.86 2.83
10 4.96 4.10 3.71 3.48 3.33 3.22 3.14 3.07 3.02 2.98 2.91 2.85 2.77 2.74 2.70 2.66
11 4.84 3.98 3.59 3.36 3.20 3.09 3.01 2.95 2.90 2.85 2.79 2.72 2.65 2.61 2.57 2.53
12 4.75 3.89 3.49 3.26 3.11 3.00 2.91 2.85 2.80 2.75 2.69 2.62 2.54 2.51 2.47 2.43
13 4.67 3.81 3.41 3.18 3.03 2.92 2.83 2.77 2.71 2.67 2.60 2.53 2.46 2.42 2.38 2.34
14 4.60 3.74 3.34 3.11 2.96 2.85 2.76 2.70 2.65 2.60 2.53 2.46 2.39 2.35 2.31 2.27
15 4.54 3.68 3.29 3.06 2.90 2.79 2.71 2.64 2.59 2.54 2.48 2.40 2.33 2.29 2.25 2.20
16 4.49 3.63 3.24 3.01 2.85 2.74 2.66 2.59 2.54 2.49 2.42 2.35 2.28 2.24 2.19 2.15
17 4.45 3.59 3.20 2.96 2.81 2.70 2.61 2.55 2.49 2.45 2.38 2.31 2.23 2.19 2.15 2.10
18 4.41 3.55 3.16 2.93 2.77 2.66 2.58 2.51 2.46 2.41 2.34 2.27 2.19 2.15 2.11 2.06
19 4.38 3.52 3.13 2.90 2.74 2.63 2.54 2.48 2.42 2.38 2.31 2.23 2.16 2.11 2.07 2.03
20 4.35 3.49 3.10 2.87 2.71 2.60 2.51 2.45 2.39 2.35 2.28 2.20 2.12 2.08 2.04 1.99
21 4.32 3.47 3.07 2.84 2.68 2.57 2.49 2.42 2.37 2.32 2.25 2.18 2.10 2.05 2.01 1.96
22 4.30 3.44 3.05 2.82 2.66 2.55 2.46 2.40 2.34 2.30 2.23 2.15 2.07 2.03 1.98 1.94
23 4.28 3.42 3.03 2.80 2.64 2.53 2.44 2.37 2.32 2.27 2.20 2.13 2.05 2.01 1.96 1.91
24 4.26 3.40 3.01 2.78 2.62 2.51 2.42 2.36 2.30 2.25 2.18 2.11 2.03 1.98 1.94 1.89
25 4.24 3.39 2.99 2.76 2.60 2.49 2.40 2.34 2.28 2.24 2.16 2.09 2.01 1.96 1.92 1.87
26 4.23 3.37 2.98 2.74 2.59 2.47 2.39 2.32 2.27 2.22 2.15 2.07 1.99 1.95 1.90 1.85
27 4.21 3.35 2.96 2.73 2.57 2.46 2.37 2.31 2.25 2.20 2.13 2.06 1.97 1.93 1.88 1.84
28 4.20 3.34 2.95 2.71 2.56 2.45 2.36 2.29 2.24 2.19 2.12 2.04 1.96 1.91 1.87 1.82
29 4.18 3.33 2.93 2.70 2.55 2.43 2.35 2.28 2.22 2.18 2.10 2.03 1.94 1.90 1.85 1.81
30 4.17 3.32 2.92 2.69 2.53 2.42 2.33 2.27 2.21 2.16 2.09 2.01 1.93 1.89 1.84 1.79
40 4.08 3.23 2.84 2.61 2.45 2.34 2.25 2. 18 2.12 2.08 2.00 1.92 1.84 1.79 1.74 1.69
60 4.00 3.15 2.76 2.53 2.37 2.25 2.17 2.10 2.04 1.99 1.92 1.84 1.75 1.70 . 1.65 1.59
120 3.92 3.07 2.68 2.45 2.29 2.18 2.09 2.02 1.96 1.91 1.83 1.75 1.66 1.61 1.55 1.50
00 3.84 3.00 2.60 2.37 2.2 1 2.10 2.01 1.94 1.88 1.83 1.75 1.67 1.57 1.52 1.46 1.39
F-TABLE AT 2.5 PERCENT (UPPER TAIL)
Degrees of freedom of numerator along the top most row

Degrees of freedom of denominator along the left most
column
df 1 2 3 4 5 6 7 8 9 10 12 15 20 24 30 40
1 648 799 864 900 922 937 948 957 963 969 977 985 993 997 1001 1006
2 38.51 39.00 39.17 39.25 39.30 39.33 39.36 39.37 39.39 39.40 39.41 39.43 39.45 39.46 39.46 39.47
3 17.44 16.04 15.44 15.10 14.88 14.73 14.62 14.54 14.47 14.42 14.34 14.25 14.17 14.12 14.08 14.04
4 12.22 10.65 9.98 9.60 9.36 9.20 9.07 8.98 8.90 8.84 8.75 8.66 8.56 8.51 8.46 8.41
5 10.01 8.43 7.76 7.39 7.15 6.98 6.85 6.76 6.68 6.62 6.52 6.43 6.33 6.28 6.23 6.18
6 8.81 7.26 6.60 6.23 5.99 5.82 5.70 5.60 5.52 5.46 5.37 5.27 5.17 5.12 5.07 5.01
7 8.07 6.54 5.89 5.52 5.29 5. 12 4.99 4.90 4.82 4.76 4.67 4.57 4.47 4.41 4.36 4.31
8 7.57 6.06 5.42 5.05 4.82 4.65 4.53 4.43 4.36 4.30 4.20 4.10 4.00 3.95 3.89 3.84
9 7.21 5.71 5.08 4.72 4.48 4.32 4.20 4. 10 4.03 3.96 3.87 3.77 3.67 3.61 3.56 3.51
10 6.94 5.46 4.83 4.47 4.24 4.07 3.95 3.85 3.78 3.72 3.62 3.52 3.42 3.37 3.31 3.26
11 6.72 5.26 4.63 4.28 4.04 3.88 3.76 3.66 3.59 3.53 3.43 3.33 3.23 3.17 3.12 3.06
12 6.55 5.10 4.47 4.12 3.89 3.73 3.61 3.51 3.44 3.37 3.28 3.18 3.07 3.02 2.96 2.91
13 6.41 4.97 4.35 4.00 3.77 3.60 3.48 3.39 3.31 3.25 3.15 3.05 2.95 2.89 2.84 2.78
14 6.30 4.86 4.24 3.89 3.66 3.50 3.38 3.29 3.21 3.15 3.05 2.95 2.84 2.79 2.73 2.67
15 6.20 4.77 4.15 3.80 3.58 3.41 3.29 3.20 3.12 3.06 2.96 2.86 2.76 2.70 2.64 2.59
16 6.12 4.69 4.08 3.73 3.50 3.34 3.22 3.12 3.05 2.99 2.89 2.79 2.68 2.63 2.57 2.51
17 6.04 4.62 4.01 3.66 3.44 3.28 3.16 3.06 2.98 2.92 2.82 2.72 2.62 2.56 2.50 2.44
18 5.98 4.56 3.95 3.6 1 3.38 3.22 3.10 3.01 2.93 2.87 2.77 2.67 2.56 2.50 2.44 2.38
19 5.92 4.51 3.90 3.56 3.33 3.17 3.05 2.96 2.88 2.82 2:72 2.62 2.51 2.45 2.39 2.33
20 5.87 4.46 3.86 3.51 3.29 3.13 3.01 2.91 2.84 2.77 2.68 2.57 2.46 2.41 2.35 2.29
21 5.83 4.42 3.82 3.48 3.25 3.09 2.97 2.87 2.80 2.73 2.64 2.53 2.42 2.37 2.31 2.2?
22 . 5.79 4.38 3.78 3.44 3.22 3.05 2.93 2.84 2.76 2.70 2.60 2.50 2.39 2.33 2.27 2.21
23 5.75 4.35 3.75 3.41 3.18 3.02 2.90 2.81 2.73 2.67 2.57 2.47 2.36 2.30 2.24 2.18
24 5.72 4.32 3.72 3.38 3.15 2.99 2.87 2.78 2.70 2.64 2.54 2.44 2.33 2.27 2.21 2.15
25 5.69 4.29 3.69 3.35 3.13 2.97 2.85 2.75 2.68 2.61 2.51 2.41 2.30 2.24 2.18 2.12
26 5.66 4.27 3.67 3.33 3.10 2.94 2.82 2.73 2.65 2.59 2.49 2.39 2.28 2.22 2.16 2.09
27 5.63 4.24 3.65 3.31 3.08 2.92 2.80 2.71 2.63 2.57 2.47 2.36 2.25 2.19 2.13 2.07
28 5.61 4.22 3.63 3.29 3.06 2.90 2.78 2.69 2.61 2.55 2.45 2.34 2.23 2.17 2. 11 2.05
29 5.59 4.20 3.6 1 ·3.27 3.04 2.88 2.76 2.67 2.59 2.53 2.43 2.32 2.21 2.15 2.09 2.03
30 5.57 4.18 3.59 3.25 3.03 2.87 2.75 2.65 2.57 2.51 2.41 2.31 2.20 2.14 2.07 2.01
40 5.42 4.05 3.46 3.13 2.90 2.74 2.62 2.53 2.45 2.39 2.29 2.18 2.07 2.01 1.94 1.88
60 5.29 3.93 3.34 3.01 2.79 2.63 2.51 2.41 2.33 2.27 2.17 2.06 1.94 1.88 1.82 1.74
120 5.15 3.80 3.23 2.89 2.67 2.52 2.39 2.30 2.22 2.16 2.05 1.94 1.82 1.76 1.69 1.61
00 5.02 3.69 3.12 2.79 2.57 2.41 2.29 2. 19 2.11 2.05 1.94 1.83 1.71 1.64 1.57 1.48
60
CHI -SQUARED TABLE
Values of X 2 (degrees of freedom , level of significance) probability in right tail.
df 0.99 0.975 0.95 0.9 0.1 0.05 0.025 0.01 0.005

1 0.000157 0.000982 0.003932 0.015791 2.705544 3.841459 5.023886 6.634897 7.879439
2 0.020101 0.050636 0.102587 0.210721 4.60517 5.991465 7.377759 9.21034 10.59663
3 0.114832 0.215795 0.351846 0.584374 6.251388 7.814728 9.348404 11.34487 12.83816
4 0.297109 0.484419 0.710723 1.063623 7.77944 9.487729 11.14329 13.2767 14.86026
5 0.554298 0.831212 1.145476 1.610308 9.236357 11.0705 12.8325 15.08627 16.7496
6 0.87209 1.237344 1.635383 2.204131 10.64464 12.59159 14.44938 16.81189 18.54758

7 1.239042 1.689869 2.16735 2.833107 12.01704 14.06714 16.01276 18.47531 20.27774
8 1.646497 2.179731 2.732637 3.489539 13.36157 15.50731 17.53455 20.09024 21.95495
9 2.087901 2.70039 3.325113 4.168159 14.68366 16.91898 19.02277 21.66599 23.58935
10 2.558212 3.246973 3.940299 4.865182 15.98718 18.30704 20.48318 23.20925 25.18818
11 3.053484 3.815748 4.574813 5.577785 17.27501 19.67514 21.92005 24.72497 26.75685

12 3.570569 4.403789 5.226029 6.303796 18.54935 21.02607 23.33666 26.21697 28.29952
13 4.106915 5.008751 5.891864 7.041505 19.81193 22.36203 24.7356 27.68825 29.81947
14 4.660425 5.628726 6.570631 7.789534 21.06414 23.68479 26.11895 29.14124 31.31935
15 5.229349 6.262138 7.260944 8.546756 22.30713 24.99579 27.48839 30.57791 32.80132
16 5.812213 6.907664 7.961646 9.312236 23.54183 26.29623 28 .84535 31.99993 34.26719

17 6.40776 7.564186 8.67176 10.08519 24.76904 27.58711 30.19101 33.40866 35.71847
18 7.014911 8.230746 9.390455 10.86494 25.98942 28.8693 31.52638 34.80531 37.15645
19 7.63273 8.906517 10. 11701 11.65091 27.20357 30.14353 32.85233 36.19087 38.58226
20 8.260398 9.590778 10.85081 12.44261 28.41198 31.41043 34.16961 37.56623 39.99685
21 8.897198 10.2829 11.59131 13.2396 29.61509 32.67057 35.47888 38.93217 41.40106

22 9.542492 10.98232 12.33801 14.04149 30.81328 33.92444 36.78071 40.28936 42.79565
23 10.19572 11.68855 13.09051 14.84796 32.0069 35.17246 38.07563 41.6384 44.18128
24 10.85636 12.40115 13.84843 15.65868 33.19624 36.41503 39.36408 42.97982 45.55851
25 11.52398 13.11972 14.61141 16.47341 34.38159 37.65248 40.64647 44.3141 46.92789
26 12.19815 13.84391 15.37916 17.29189 35.56317 38.88514 41.92317 45.64168 48.28988

27 12.8785 14.57338 16.1514 18.1139 36.74122 40 .11327 43.19451 46.96294 49 .64492
28 13.56471 15.30786 16.92788 18.93924 37.91592 41.33714 44.46079 48.27824 50.99338
29 14.25645 16.04707 17.70837 19.76774 39.08747 42 .55697 45.72229 49.58788 52.33562
30 14.95346 16.79077 18.49266 20.59923 40 .25602 43.77297 46.97924 50.89218 53.67196
50 29.70668 32.35736 34.76425 37.68865 63.16712 67.50481 71.4202 76.15389 79.48998

60 37.48485 40.48175 43 .18796 46.45889 74.39701 79.08194 83.29768 88.37942 91.9517
80 53.54008 57.15317 60.39148 64.27785 96.5782 101.8795 106.6286 112.3288 116.3211
100 70.0649 74.22193 77.92947 82.35814 118.498 124.3421 129.5612 135.8067 140.1695
61
Financial Modelling
What is Financial Modelling?

ª Financial Modelling involves modelling Financial Data for Decision Making
ª Financial Modelling Skills are applied to variety of scenarios like Equity Research, Mergers and
Acquisition, Project Finance etc.
ª Financial Modelling Certification at FinTree equips candidates to develop a model from scratch without
using ready-made templates.
What is the Course Content?
We have divided Financial Modelling

Course into Four Parts:
R
Part I: Advance Excel Training
Part II: Building Financial Model

Infrastructure
Part III: Forecasting
Part IV: Valuation
What is the duration of the Course?

ª The duration of one batch is roughly three months. The Certification is provided by FinTree after the
completion of the batch.
ª For classroom, we operate on a club Membership model, wherein, in the same fees, candidates are
allowed to (and encouraged to) attend three more (1+3) subsequent batches. Every batch we pick up
models from different sectors and that provides deeper understanding to the participants.
ª Online course validity : 1 year
To Know more, visit www.fintreeindia.com
FinTree
FinTree
CFA® Level II JuiceNotes 2022
© 2022 FinTree Education Pvt. Ltd., All rights reserved.
FinTree Education Pvt. Ltd. FinTree Education Pvt. Ltd.

1 Muktali Building, First floor, Lane 148, 3rd Floor, 60 Feet Rd, KHB
16, Bhandarkar Rd, near by TVS Colony, 5th Block,
Showroom, Pune, Koramangala, Bengaluru,
Maharashtra 411004 Karnataka 560034
Disclaimer: CFA Institute does not endorse, promote, review, or warrant the accuracy or quality of the
products or services offered by FinTree Education Pvt. Ltd. CFA Institute, CFA®, and Chartered Financial
Analyst® are trademarks owned by CFA Institute.
63

2023 CFA L2 Book 1 Quants Eco Multiple

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

2023 CFA L2 Book 1 Quants Eco Multiple

Uploaded by

Copyright:

Available Formats

FinTree

Quantitative Methods |Economics

Chartered Financial Analyst - Level II

ª Quantitative analysis ª Credit risk measurement and management

ª Current issues in ﬁnancial markets

Basics of Multiple Regression &

Formulate a multiple linear regression model, describe the relation between

Refer to the Spreadsheet

Explain the assumptions underlying a multiple linear regression model

Assumptions of a multiple regression model

Œ Relationship between dependent and independent variable is linear

Residual plots allow analysts to get a preliminary indication of violation of regression

Evaluate Regression Model Fit

Total n−1 SST

F-statistic = MSR/MSE with ‘k’ and ‘n − k − 1' DoF

R2 = Explained variation/Total variation

Adjusted R2 < R2 in multiple regression

Eg. k=6 n = 30 R2 = 73%

Lower values indicate a better model under either criterion.

These metrics can be calculated as follows:

K is a penalty parameter in both criteria: higher values of k result in higher

Formulate hypotheses on the signiﬁcance of two or more coeﬃcients in a multiple

we want to test the following hypothesis:

R and U represent the restricted and unrestricted models, respectively

K = independent variables in the full model

Decision rule: reject H0 if F (test-statistic) > Fc ( critical value )

Describe how model misspeciﬁcation aﬀects the results of a regression

Misspeciﬁcation Description Eﬀect

Inappropriate Variables may need to be May lead to heteroscedasticity

Data improperly Samples has periods of May lead to heteroscedasticity

Explain the types of heteroskedasticity and how it aﬀects statistical inference

BP chi-square test statistic = n X R2resid with k degrees of freedom

Detecting Serial Correlation

Σt = b0 + a1 X1t + a2 X2t + … + ak Xkt + P1 Σt-1 + P2 Σt - 2

The BG test statistic has an F-distribution with p and n – p – k – 1 degrees of freedom,

Correction of serial correlation:

Robust standard errors

Explain multicollinearity and how it aﬀects regression analysis

Describe inﬂuence analysis and methods of detecting inﬂuential data points

The sum of the individual leverages for all observations is k + 1.

Outliers are extreme observations of the dependent (or 'Y') variable

Estimate the regression model using the original sample of size n.

Delete one observation and re-estimate the regression using (n – 1) observations.

Detecting Inﬂuential Data Points

Extensions of Multiple Regression

INTERCEPT DUMMY SLOPE DUMMY

Y = (b0 + d0) + b1 X + Σ (ifD=1) Y = b0 + b1X + Σ (ifD=0)

In other words, the intercept of Y = b0 + (b1 + d1)X + Σ (ifD=1)

Formulate and interpret a logistic regression model.

An example of an application requiring the use of a qualitative dependent variable is a

And the probability of the event is calculated as:

P = odds / (1 + odds) = 1/(1 + e -y)

LR = -2 (log likelihood restricted model – log likelihood unrestricted model)

Where the restricted model has “q” fewer independent variables.

LR always has a negative value. Higher values (closer to 0) indicate a better-ﬁtting

The test statistic LR has a Chi-square distribution with q degrees of freedom.

Time series: Set of observations on a variable’s outcomes in diﬀerent time periods

Linear trend Log-linear

Log-linear trend is a trend in

LOS b How to determine which model to use

Plot the data

Linear trend Log-linear

Constant and ﬁnite mean

Constant and ﬁnite variance (same as homoskedasticity)