联系我们： 手动添加方式: 微信>添加朋友>企业微信联系人>13262280223 或者 QQ: 1483266981

Empirical Finance: Course Work 1
Your team has been recently hired by Imperial Global Asset Management, and your task
is to run a few exercises. Below, you find the guidelines about your assignments, but you
can amend the specifications if you have plausible economic arguments. Finally, write a
brief investment report no longer than about 3,000 words (this is just an indicative) as a
financial economist. Describe what you have done and present the key results. There is no
need to write any formulas. You can also be creative and report whatever may convince
an investor to bet on your strategy. Whether a strategy works or not, try to come up
with an economic story that explains why this is the case. Last but not least, make any
assumptions you need, but please mention them. You do not need to submit codes and/or
excel files.

Data
You have collected the following long-span data from Global Financial Data:
Stock market index for both the US and UK,
Short-term Treasury yield for both the US and UK,
Long-term Treasury yield for both the US and UK,
GBPUSD exchange rate.
Exercise 1 [30%]
Take data for the US (i.e., stock market index, short-term yield, and long-term yield) and
answer the following questions.
Questions
1. Present and comment their summary statistics.
1
2. Present and comment the ACF and PACF.
3. Identify the appropriate time-series specification, and estimate the key parameters.
Present and discuss your results.
Exercise 2 [30%]
Consider the following models:
1. Benchmark model
yt = α + εt
2. Competing model
yt = α + βxt 1 + εt
where
yt
is the monthly stock return between t 1 and t,
xt 1 is the term spread (long-term yield minus short-term yield) observed at
time t 1.
3. Use an expanding window (start with 10 years of data) as well as a 10-year rolling
window to generate out-of-sample (OOS) forecasts for both models using US data
Questions
1. Compute the OOS R-squared or Roos
2
,
2. Test the null hypothesis of equal predictive ability using the Clark and West test,
3. Evaluate the market timing ability as in Henriksson and Merton (1981).
4. Present and discuss your results.
2
Exercise 3 [40%]
Consider the following models:
1. Benchmark model for stock returns
yt = α + εt
2. Competing model for stock returns
yt = α + βxt 1 + εt
3. Benchmark model for exchange rate returns
et = α + εt
4. Competing model for exchange rate returns
et = α + βzt 1 + εt
where
yt
is the monthly stock return between t 1 and t,
et
is the monthly exchange rate return between t 1 and t,
xt 1 is the term spread (long-term yield minus short-term yield) observed at
time t 1.
zt 1 is the interest rate differential (US short-term yield minus UK short-term
yield) observed at time t 1.
5. Use an expanding window (start with 10 years of data) as well as a 10-year rolling
window to generate out-of-sample (OOS) forecasts using all data.
Questions
Consider a portfolio consisting of the US stock market, UK stock market, and the US
short-term bond (a proxy for the riskless rate).
Set σ
= 10% per annum as target volatility, and rebalance your portfolio every
month using your OOS forecasts.
3
Report the portfolio average return and volatility in % per annum,
Report both SR and SO in annualized terms,
Report the performance fee P in basis points per annum,
Plot the cumulative returns, portfolio weights, and the one-year rolling SR.

Good Luck!
4