International finance (3 credits)

Professor Michel Robe

Group Assignment #2
General Instructions.
Please form groups of three to five people in order to solve either of the two minicases below. You have a month to complete the work: the due date for this group assignment
is the last day of the final exam period at 5PM.
You have two choices:
(i)

Case Choice #1: real exchange rates and competitive positions of US, Australian and
Brazilian companies.

(ii)

Case Choice #2: drivers of nominal vs. exchange rate changes in Turkey.

Completing the second choice requires carrying out regression analyses. That case is thus more
demanding quantitatively, so that successful completion yields a 10% grade bonus (i.e., your
raw grade is multiplied by 1.1). There is a similar 10% bonus if you solve part (2) of Choice #1.
In order to handle this second assignment, please remember that no collaboration is allowed
between groups (of course, maximum collaboration is expected within a given group). If you
have any doubts about the honor code that governs the completion of this assignment, please
consult the course syllabus or talk to me!
Data sources:
Data on the relevant interest rates, CPI and FX rates are provided on the courses website.
Additional series you may wish to use is likely available as part of the IMFs International
Financial Statistics (IFS). IFS series are available through the AU Library website.
Alternatively, IR and FX rate figures can be obtained from Bloomberg.

Case Choice #1 USD, Commodity Currencies and Purchasing Power Parity.

Mr. Henry Guessright (a principal of Veritas Emerging Market Fund, LLC) is interested in
commodity currencies. He foresaw, back in 2002, that a commodity super-cycle would soon
materialize amid strong growth in emerging economies (especially China) and output constraints
in commodity-producing countries. He correctly guessed that the resulting commodity price
IM-1

boom would drive up local currencies values in commodity-producing countries in particular

Australia (coal, iron), Brazil (iron, soybeans, sugar), Russia (oil) and South Africa (coal, gold).
Back in 2007, when commodity prices reached unheard-of heights, and in 2010, after they
rebounded once the Great Recession had ended, Mr. Guessright became concerned with changes
in the competitive standing of non-commodity firms in commodity-producing countries (i.e.,
companies that produce goods or services outside of the primary sector of those economies).
We are now at the end of 2014, and this time Mr. Guesswright thinks that the opposite
situation is taking place amid a sharp drop in commodity prices (almost 50% for oil and iron
ore). As members of his team, he has tasked you with preparing a report on real exchange rate
changes in Australia and Brazil (the two countries where hes thinking of making a move there is
a substantial non-commodity manufacturing sector).
(Q1) Mr. Guessright feels that the major worry for non-commodity companies in
commodity-exporting countries is competitiveness with China. He believes that China has
managed its currency with respect to the U.S. dollar in the past and will, in the future, continue to
do so. Hence, he would like you to do the following:
a. Using monthly data (the highest frequency at which inflation information is published),
create plots of the past (nominal) exchange rate levels and percentage changes against the
monthly inflation differential between the USA and, respectively, Australia or Brazil.
Hint 1: These plots should provide a hint as to whether there are one or more structural breaks in
the relevant FX and inflation time series (i.e., whether there are time periods before and after
which you see clearly different levels of inflation or FX rate patterns).
Hint 2: In addition to possible structural breaks, comment on outliers in the monthly (or quarterly)
observations and any possible structural breaks in the observed trends.

b. Using monthly CPI and FX data (provided on the course website), plot the time series of the
real exchange rate in the past 24 years (Brazilian Real vs. USD, and Turkish Lira vs. USD).
Hint 1: for each country where you found a structural break in the FX and/or inflation time series in
part a, please consider two alternative reference months to compute the real exchange rate Jan.
1991 and whatever month marks the start of the structural break.
Hint 2: for accuracy, compare your effective real exchange rates to the indices that the Bank for
International Settlements (BIS) provides for each country discuss any major differences.

c. Using the above graphs, comment on possible changes in the competitive positions of,
respectively, Australian and Brazilian firms compared to U.S. firms.
Hint: Do you see a change in either RER in 2003-2007? After Lehman Brothers demise? Following
the economic recovery in 2010? After the big commodity price drop that started in June 2014?).

(Q2 5% grade bonus) Mr. Guessright is considering investing in a Brazilian company that
exports 50% of its output to the USA and 50% to Australia. He wonders how FX rate changes
will affect the companys position against US and Australian competitors in 2015. Please help!
Hint: Compute the trade-weighted (i.e., weighted-average) real exchange rate for this company from 2003
to present. Where is the RER as of end-2014? What is your forecast for the RER in 2015? On that
basis, would you expect the Brazilian companys performance to improve or worsen in 2015?
IM-2

Case Choice #2 USD, Turkish Lira and Purchasing Power Parity.

Veritas Emerging Market Fund, LLC specializes in investing in the worlds emerging stock
markets. Mr. Henry Guessright (the same person as in Choice # 1) is an experienced hand at the
firm; he is also your boss. He is currently interested in the Turkish stock market. He correctly
anticipated, back in 2004-2005, that Turkey would be invited to negotiate its membership in the
European Union and that this development would boost the stock market in Turkey.
Ten years later, as of late Fall 2014, he is quite concerned with the potential for exchange
rate volatility in Turkey and would like to understand what has historically driven Turkish
exchange rates. Since the inflation rate has often, in the past, been higher (or much higher) in
Turkey than in the USA, he is guessing that purchasing power parity has prevailed at least to
some extent. As members of his team, he has tasked your group with checking this out.
In other words, you have to prepare a report on the following question: Does purchasing
power parity typically hold for the Turkish lira-U.S. dollar exchange rate?
As a first pass, Mr. Guessright would like you to do the following:
a. Using monthly data (the highest frequency at which inflation information is available from
public sources), please plot the Turkish lira-U.S. dollar exchange rate percent changes
against the inflation-rate differential between Turkey and the U.S. (Jan. 1991 Nov. 2014).
Hint 1: This plot should provide a hint as to whether there are one or more structural breaks in the
FX and inflation time series (i.e., times before and after which you see clearly different levels
of inflation or patterns of FX rate changes). Such a structural break may help provide a clue
as to whether PPP holds during the entire time period, or only during some time sub-period(s)
but not during some other time sub-period(s).
Hint 2: Comment on outliers in the monthly (or quarterly) observations and any possible structural
breaks in the observed trends.

b. Regress percentage exchange rate changes on the inflation differential and provide estimates
of the intercept and of the slope coefficient. Interpret the regression results.
Hint 1: before regressing, run a simple correlation analysis is the correlation high? Is it positive?
Hint 2: if the graph in part a. suggests structural breaks, does it make sense to run the regressions for
the entire sample or would it make more sense instead to run regressions for sub-periods?

c. Even if PPP did not hold perfectly every month since 1991, it is still possible that inflation
differentials nevertheless drove the bulk of nominal exchange rate changes over time. The
way to check is to see if the real exchange rate (RER) went up or down substantially for long
periods of time. Using Jan. 1991 and Jan. 2002 as alternative reference months, plot the RER
(Turkish Lira vs. USD) (i) since 1991, (ii) in 1991-2001 and (iii) in 2002-2014. Discuss!
d. Compare the respective explanatory powers of PPP vs. uncovered IRP in the 1991-2001 and
2002-2014 sub-periods. Have PPP and/or UIRP held at different horizons (say, 1 month, 3
months, 6 months, 1 year, or even 2-3 years)?
Hint: Several interest-rate series are provided which would be most relevant to UIRP computations?
IM-3