Summary of the 21-Dec-2015 class:
- Engle
and Granger Cointegration
- Cointegration
of order k,k. CI(k,k).
- Spurious regression.
- Error Correction Model.
- Engle-Granger cointegration statistics
test.
- Normalization (1, β1/β2).
- Lag distribution.
- Johansen and VAR.
- Rank and Eigenvalue.
- Trace Statistics and
Max-Eigen statistic.
-The International
Financial News: (www.cnbc.com)
(1). Bank of Canada chief: Pointless to try
to stop C$ moves.
(2). Dollar falls against yen after Japan stops short of
extra QE.
---------------------------------
Summary of the 07-Dec-2015 class:
- Efficiency: in econometrics, it means the
amount of standard error of statistics estimation. (the larger standard
error may have the White Noise problem which is not the iid and not the BLUE).
- UIP+ Risk premium
- AIC (sample size less than100), SIC &HQC (sample size
more than 100)
- Unit root test for Stationary and non-stationary
-Random walk: Yt=b1Yt-1+
Ut , Ut ~iid
- Matrix
-The necessary and sufficient condition for stability is that all
characteristics roots lie within the unit circle.
-The International
Financial News: (www.cnbc.com)
(1). Currencies of major oil exporting nations such as the
Canadian dollar and the Norwegian crown fell on Tuesday as crude prices hovered
near 7- year lows.
(2). Malaysian ringgit remains the most vulnerable regionally to lower oil prices.
(2). Malaysian ringgit remains the most vulnerable regionally to lower oil prices.
---------------------------------
Summary of the 30-Nov-2015 class:
- DGP (data
generating process), a functional form.
- The general
form: Yt=, when Y0 is known, and t is at any location of time.
- For the AR(1), Yt=a1 Yt-1+ at1Y0
+Ut , Ut ~iid (independent identical distribution).
- The reason for iid is to get BLUE.
- E(Yt) = E(a1Yt-1)+ E(Ut) , E(Ut)=0, the E( ) is an average.
- More samples or longer
term in the observation, more close to the population.
- Stationary definition:
E(Yt)= u for all t, Var(Yt)=σ2, Cov(ut, ut-j)=0 , j is not zero.
- The general form with
intercept: Yt=a0+a1Yt-1+ut
- Mean-reversion.
- AR(2)à Yt=a0+a1Yt-+a2Yt-2+
ut
For long term: u = a0/(1-a1-a2) , but (a1+a2 are not
equal to 1)
- The reason to have lag-periods is
to have iid (by applying white-noise method).
- In time series, it does not mean
it would better when there are more samples.
It needs to approve there is no
structure change over a long period.
- Lag operator
- The general form of ARMA (p,q)
- Trend: Long swing
-Random walk
-The
International Financial News: (www.cnbc.com)
(1). Chinese central bank says welcomes
additional of yuan into IMF currency basket as endorsement of China’s economic
development and reform.
(2). Jan Hatzius, Goldman Sachs's top economist, expects the Federal Reserve to go gradual on rates, and for modest U.S. growth to prevail
(2). Jan Hatzius, Goldman Sachs's top economist, expects the Federal Reserve to go gradual on rates, and for modest U.S. growth to prevail
Summary of the 23-Nov-2015 class:
- Review how to test UIRP: Q-test, Q2-test
- Robust standard error
-Structure change
-Quizlet website
-Stationary and non-stationary time series data
-White noise:
- zero means
-constant variance is constant
- no auto-regression
-AR(1) model (with w.n.)
-The
International Financial News (www.cnbc.com):
(1). The euro hit a seven-month low versus the U.S. currency, on
expectations that the ECB will ramp up its monetary stimulus.
(2). Nikkei
slips as yen strengthens against the dollar.
-About
the final term paper:
-The source from:
-Foreign
exchange rate from the Taiwan Central bank historical data.
(http://www.pxweb.cbc.gov.tw/dialog/statfile9.asp)
-Taiwanese
travelers to Japan data from the statisticsof Tourism Bureau of Taiwan.
(http://admin.taiwan.net.tw/statistics/month.aspx?no=135)
-The
definition and hypothesis:
-Period:
115 monthly data between Apr-2006 and Oct-2015.
-The dependent variable Yt = Taiwanese
traveler numbers to Japan at time T.
-The independent variable Xt = the foreign exchange JPY/NTD
at time T.
-H0: no linear regression
relationship between Yt and
Xt.
Summary of the 16-Nov-2015 class:
- Comments and suggestions to the presenters:
- cointegration must have the EC item (error correction).
- unrelated information to forex can be reduced at the end-term paper.
- the paper that student presented in the class is written by other writer,
so do not say “my paper…” in the presentation.
- find another easier paper because some students may not complete
the end-term paper.
- too many words, re-shooting the video.
-The
International Financial News (www.cnbc.com):
(1). China's
currency gets key backing from IMF.
(2). China, Turkey
renew 12 billion yuan bilateral currency swap deal.
-------------------------------------
Summary of the 02-Nov-2015 class:
- The mid-term presentation(the 1st section):
- Comments and suggestions to the presenters:
- too much words in the
presentation PPT file.
- need to explain the equation
and models.
- Irrelevant data to the topic.
- the paper is unrelated to the course.
- speak too fast.
- don’t understand the content of the presentation.
- wrong translation between English and Chinese.
-The
International Financial News (www.cnbc.com):
1. Japanese
banks count costs of dollars as Fed moves to end easy money.
2. The forex-Dollar falls from multi-month highs in light holiday session.
2. The forex-Dollar falls from multi-month highs in light holiday session.
--------------------------------------
Summary of the 26-Oct-2015 class:
- The difference and explanation between UIP and
CIP.
- The explanation of Naïve expectation:
- Depreciation rate: (Se-S)/S
- If no theorem, then experiment it.
- No Auto regression test: Q test and LM test.
- Homoskedasticity test: White test and Q2
test.
- Normality test: JB test.
-The
International Financial News: (www.cnbc.com)
1. The People's Bank of China (PBOC) announced a quarter-point cut in benchmark
interest rates, also reduction in the amount of deposits banks were
required to hold in reserve by 50 basis points.
2. International Monetary Fund staff are
set to give the all-clear for China's yuan to be included in the lender's
benchmark currency basket.
-------------------------------------
Summary of the 19-Oct-2015 class:
- The difference between stationary an non-stationary
- Cointegration
- Check if the model is health
- OLS: Gauss-Markov theorem
- Indirect
- The term paper can be as simple as OLS.
- Q test.
- UIP: Uncovered Interest rate Parity
- PV: Present Value.
- comparing between (1+r) and
(1/S)(1+r*)Se
- Forecast with Naïve expectation
-The International
Financial News: (www.cnbc.com)
1. Expectations are low
the Fed will provide any new policy insight in the
coming week, and focus should stay on economic reports and a flood of earnings,
from Apple, big oil,
drugmakers and others.
2. The Euro vs. US dollar fell even further in early trading Friday,
trading just above $1.10.
-------------------------------------
Summary of the 12-Oct-2015 class:
- St=
local currency / foreign currency. For
example (bilateral exchange rate) ,
S= NTD/USD, If the S is changed from
33 to 34, then the USD is increased, and NTD
decreased.
decreased.
- The more
practical exchange rate (multilateral rate):
the effective exchange rate or the trade weighted exchange rate.
- Spot rate (within 3 days) vs. Forward exchange rate (more than 3
days).
- Bid vs. Ask rate.
- Bid ask spread. 3 things are related to the Spread: HR costs,
Mobility, Risks.
- Remember the terminology to show the professional.
- Foreign exchange market: OTC,
over-the-counter.
- Speculation is not negative term, but positively for those who predict
the opportunities and make the right investment in advance.
- How to judge the rate is the flowing rate or fixed rate? USA for the flowing, and China for the fixed, …for
example.
- Bretton Woods
international monetary agreement of 1944.
-Reviews of the statistics.
-The International
Financial News: (www.cnbc.com)
(1). The dollar
sank to a seven-week low against a basket of currencies as more signs of
slowing growth in the United States and China raised doubts the U.S.
(2). Brazilian companies stepped up U.S. dollar purchases in
September at the fastest pace in six months, seeking to protect themselves
against a slumping local currency “real” which is down 32 percent this year,
--------------------------------------
Summary of the 05-Oct-2015 class:
-Anchor effect: the 22K case.
-Quantitative analysis v.s. Qualitative analysis
-If the interest rate rise can cause the currency appreciation, then how
much
will the currency be appreciated when the interest is rising by 0.25%?
-The Empirical International Finance: (range)
(1). Theory (Concept).
(2). What are the empirical studies (real world data +
econometrics).
(3). Empirical + International Finance Theory:
--the exchange rate is the core theory of the International Finance
--Cash flow: asset, UIP,
CIP, etc… for those which are effected by interest rate.
--Logistics: LOP, PPP,
etc…for those which are effected by prices.
--Information flow: News
model, Event studies, Structure change…etc.
-
Those which are out-of-scope of International Finance range (common mistakes):
(1).The pure stock market data.
(2). Macroeconomics (without foreign
information)
-Reviews of the Statistics:
--Standard deviation v.s.
Standard error.
--The 6 sigma: Statistical Process Control (99.99966%).
--OLS v.s. Co-integration.
-The International Finance News (www.cnbc.com)::
(1). IMF
sees China slowdown risks, urges Beijing to float
yuan.
(2). Yen gains as
BOJ drops no hints on further easing of policy.
------------------------------------
------------------------------------
Summary of the 21-Sep-2015 class:
- The class introduction:
- class syllabus.
- textbooks
-ilearning information.
- grading standard.
- how to read English paper.
- how to write a paper.
-The
International Financial News (www.cnbc.com):
(1). Euro rises from 2 -1/2-week low against dollar at
$1.11050
(2). Dollar rebounds after Yellen
keeps alive 2015 Fed rate hike view.
-------------
沒有留言:
張貼留言