To view full questions and answers, please kindly visit our site: http://cfaexampreparation.com/375/top-82-cfa-sample-mock-exams-questions-answers-level-2-ethics/
Concerns 1 period and 1 point in time
Will result in unreliable testing and predictions1. Functional Forms2.
Explanatory Variables are correlated with errors terms in Time Series
models3. Other Time Series that result in non-stationary
From Clients- are OKAY as long as they do not impeded on future
judgement. Must disclose this information.From Brokers- are NOT OKAY.
NEVER accept even if you have permission.
Can only be used within the the time period specified.
1. Standard errors are too small2. Coefficient errors are OKAY3. TYPE 1
ERRORS- standard errors are too small, t statistic is too large and
null will be rejected too often
Measures the variability of actual Y to the estimated Y from the
regression. It is the standard deviation of the error terms in the
regression.
Here, if the client mis-ordered the trade they must incur the cost. It
should NEVER be passed to their clients. They can pay the broker to
incur the cost, but it must come from their own pocket.
No violation when analyst combines public information with items of
NONMATERIAL information (such as seeing two CEOs have lunch).
= Coefficient/standard deviationUsed to test significance of
hypothesized values. H0 = 0 versus Ha =/ 0. If H0 is rejected then we
know the variable is statistically significant.Degrees of Freedom =
n-k-1 (k = number of variables)
Occurs when the variance of the residual is correlated with each
other. Common with time series data.1. Positive- error in previous
period increases prob of positive error in the next2. Negative- error in
previous period increases prob of negative error in next
To test for serial correlation, you must use the T-Test NOT Durbin
Watson.=autocorrelation / standard errorwhere Standard Error = 1/square
root (T)where T = number of observationsand degrees of freedom = T-2
This is where the usage is both for decision making and general
management in the firm (such as bloomberg terminal). Can only allocate
the amount of soft dollar used for INVESTMENT decision making.
1. Lagged dependent is used as independent2. Function of dependent is
used as independent3. Independent variables are measured with errors for
qualitative variables
Total Sum of Squares= total variability Y explained by X. It is equal
to the sum of squared differences between ACTUAL and MEAN of Y.
Owner is defined as applicable fiduciary duty. The duty to a pension fund, the trustees are NOT the actual client.
RSME Compares accuracy of AR models in forecasting out of sample
values. Lower RMSE will have higher predictive power for the future.
Occurs when the variance of the residuals is not constant across
observations.1. Unconditional- variation is NOT related to independent
variable2. Conditional- variation IS related to independent variable and
creates significant problems
This infers that IF client was not notified of a suitable investment.
You must extend the participation to all portfolios meeting the
criteria.
Research includes both Proprietary and Third-Party research and must
directly assist investment manager in decision-making process and NOT in
the general management in the firm (such as preparing investor
statements, etc).
You must promptly disclose any changes that materially affect the
investment process. This does NOT mean you have to discuss prior
investment process, only the CHANGE.1. description of new model2.
limitations of the model
1. Use Hansen method to correct standard errors. If both
Heteroskedasticity and Serial Correlation are present, you can use
Hansen to correct BOTH.2. Use Times Series.
What is the intent of the parties involved?
Explains the variation of Y that is explained by ONE independent variable. =SST-SSE/SST =RRS/SST
Must include written disclosures as to the source and nature of the
performance. Particularly important when included performance of prior
employment, etc.
1. Examine Scatter Plot2. Bruesch-Pagan TestStatistic = n * r^2 where
r^2 = regression of the residual. If statistic is rejected using t test,
then there IS presence of heteroskedasticity.
1. Care is higher2. Skill is higher. Delegation of duties is okay.
Whereas Old rule said NO3. Caution = total return not standalone basis.
Whereas Old Rule said NO growth, preservation only4. Loyalty is
consistent5. Impartiality is consistent
This occurs when you run two time series. If they are NOT covariance
stationary but ARE co-integrated, you CAN USE DATA because they are
economically linked. Look for BOTH variables (x and y) to see if they
have unit root.Example: Beta versus stock return against returns of
market.
Must always be disclosed either on report or website. Recommended to do both.
RRS/K
Explains how well a set of independent variables explains the
variation in the dependent variable. =MSR/MSEThis is a ONE TAILED test
(use t test to test significance)
Occurs when the dependent variable is lagged by values of itself.
Because it is lagged, we MUST TEST for serial correlation.Example: Sales
from prior period will affect next month sales.
= B0/1-b1Note that this is why Random Walk has unit root where b1 = 1. Therefore Random Walk is NON STATIONARY.
Occurs when variance of the residual is dependent from the previous
period. Test by squaring the residual and regressing the data. If the
coefficient of the new regression is statistically different than 0, pr
the p-value states there is significant there is evidence of ARCH(1).
This is because now the residual is going to grow by a constant
amount!Correct it by using Generalized Least Squares.
1. Expected value is constant and finite= mean reverting.2. Variance
is constant and finite.3. Covariance is constant and finite.
1. Conflicting T (statistically significant variables) and F
statistics (or R squared is too high/low). Remember that F statistic
tests OVERALL variables while t tests INDIVIDUAL significance.2. High
correlation among variables MAY indicate a problem.
1. Important variables are omitted2. Variables should be transformed (log vs linear)3. Data is improperly pooled
1. Take the change Ln(sales) = Ln(x) - Ln(y)2. Plug the change into
the equation = b0 + b1Change Ln(sales)3. Since the equation is equal to
change in Ln sales, add previous Ln4. Now take the natural log e^ to
remove logAfternoon 3 #98
Additional compensation must have WRITTEN CONSENT from all parties
involved. Only necessary if the position is in direct competition of
your employer. Working at the bar (limited hours so that you do not
deprive your time) is OKAY.
Arrangement under which client tells the manger to execute trades under its account with a specific broker.
1. First subtract x(t-1) from both sides of AR(1)2. Test if (b-1) is
different from 0 = H03. If it is not significant from 0 then there is
UNIT ROOT because this means b=1!
Sum of Error Squares= variability of Y UNEXPLAINED by X. Difference between actual and predicted Y.Degrees of freedom= n-2
No comments:
Post a Comment