You are on page 1of 4

Assignment #3 for Week 7 - FA17- 9572

Fall 2017

TEAM 8
Mike Buckner
Yan Oo
Gabrielle Rubenstein
Alex Senatore
George Batty
1. First, calculate the mean annual return and its standard deviation for every column.

stoktr smalls corptr ltgovttr ltgovtin ltgovtca itgovtr itgovin itgovca tbills infl inflbond
Mean 12.16% 17.41% 5.73% 5.16% 4.94% -0.26% 5.24% 4.58% 0.29% 3.74% 3.22% 5.22%
Std 20.20% 34.31% 8.45% 8.68% 2.92% 7.31% 5.63% 3.16% 4.05% 3.27% 4.57% 4.57%
Dev

2. Next, calculate the correlation coefficient of returns for all pairs of assets (you can
skip columns F, G, I, J, and L for this particular task).

stoktr smalls corptr ltgovttr itgovtr tbills inflbond


stoktr 1 0.814293 0.224227 0.147038 0.072347 -0.05148 -0.0177
smalls 0.814293 1 0.10094 0.015218 -0.04238 -0.09982 0.045232
corptr 0.224227 0.10094 1 0.935069 0.897015 0.211766 -0.1495
ltgovttr 0.147038 0.015218 0.935069 1 0.902118 0.236173 -0.14845
itgovtr 0.072347 -0.042382 0.897015 0.902118 1 0.502539 0.020849
tbills -0.05148 -0.099816 0.211766 0.236173 0.502539 1 0.417271
inflbond -0.017698 0.045232 -0.149502 -0.148446 0.020849 0.417271 1

3. Do the mean returns line up the way that you expect that they would? If not, would
you propose any subjective adjustments before using these estimates in a portfolio
optimization program? To put this another way, you are going to use the historical
means, standard deviations, and correlations to mathematically optimize your
portfolio. Do you think you should make adjustments to any of these inputs? If so,
why?
The optimal thing when reviewing the results of the spreadsheet:
Asset 2 has the highest average return but also the highest sigma. Asset 1 has the
second highest average return but also the second highest sigma. Asset 5 has
slightly average return than asset 4 but asset 4 has higher sigma than asset 5. When
examining the correlation asset 5 then 4 provide more diversification. Considering
the above factors, these are driving the weights to invest more in asset 5 and 4 than
other assets. Therefore, the adjustments to be made to the inputs are the following
in order to invest more in assets 5 and 4.
Input
Asset Portfolio
1 10.00%
2 5.00%
3 5.00%
4 30.00%
5 60.00%
4. What lessons do you learn from looking at the mean and standard deviations for
columns F, G, I, and J?
ltgovtin ltgovtca itgovin itgovca
Mean 4.94% -0.26% 4.58% 0.29%
Std
Dev 2.92% 7.31% 3.16% 4.05%

The Interest components have comparable average return with comparable sigma.
The interest components have higher average return and lower sigma compared to
capital gain components.

5. What is the historical risk premium between large stocks and Tbills?
Historical Risk Premium 8.43%

6. In column M we have a simple estimate of what the return on a TIPS might have
looked like had TIPS existed since 1926. By construction, the assumption in column
M is that TIPS pay a 2% real return. What is the correlation coefficient between
the hypothetical TIPS and Tbills?
TIPS - Tbills 0.417271

7. Now open the portfolio optimization spreadsheet that I posted. Assume that there
is a risk-free asset that offers a nominal return of 3% per year. Next, use the data
you calculated above to find the optimal portfolio containing large stocks, small
stocks, LT government bonds, corporate bonds, and TIPS. Keep in mind that in the
optimization spreadsheet, the graph is scaled to accept monthly data, but the
optimal portfolio weights in column J are still fine. (If you want to see how the
graph looks, divide the mean returns by 12 and the standard deviations by the
square root of 12, leaving the correlations unchanged, and the scale will be about
right). What are the optimal portfolio weights? Can you provide an intuitive
explanation of why the optimal weights turned out the way that they did.
Opt. Port.
Weights
Large Stock 8.06%
Small Stock 2.06%
LT Govt Bond -11.33%
Corp Bond 33.40%
TIPS 67.81%

The optimal portfolio weights swayed heavily towards TIPS and Corporate Bonds,
which is logical given that these two securities have the smallest variation (measured
as standard deviation), making them the least risky asset in the portfolio. In
contrast, Small Stock and Large Stock securities have the largest variation, making
them the riskiest assets, and then smallest weight. LT Government Bonds has a
negative weight because it has the smallest return and is an opportunity to short the
position to free up capital for the rest of the portfolio.

8. Using the data in the optimization spreadsheet, calculate the Sharpe ratio of the
optimal portfolio, each of the 5 risky asset classes, and an equally-weighted
portfolio.
Optimal Portfolio
0.12738809

5 Risk Asset Classes


1 0.4535632
2 0.4200983
3 0.2487469
4 0.32273415
5 0.48651553

Equally Weighted
0.116231433

9. To create an historical TIPS series, we added 2% to the historical inflation series.


Of course, one might argue that the actual real return on TIPS might be higher or
lower than the 2% that we assumed. Other than that issue, is there something about
the way that we constructed the TIPS series that makes them look more or less
attractive than we really expect them to be in practice? Notice what I am NOT
asking here. It is possible to make errors in our input assumptions regarding asset
classes other than TIPS that might indirectly cause an over or under weighting in
TIPS in the optimal portfolio. I dont want to focus on that. I want you to think
specifically about the way that our TIPS historical series is constructed and
determine whether or not that makes TIPS looks better or worse than they really
are.

You might also like