. Select the stocks you intend to short and the stocks you intend to buy.

Report the
tickers, your investment in each stock, and the values of the characteristics that led
you to believe that these stocks should be bought/sold. Make sure that at least two
characteristics, beyond the one that classies rms as winners/losers, are the ones
we discussed in class. Describe the reasoning why these characteristics make your
rms a good subsample for exploiting the momentum effect
ect. (40 points)

Long Position Stocks:
ETE
TRGP
TARO
AMP
CP
I think so,jaren will be able to tell us when he puts in the orders:

Short Position Stocks:
UNXL
UPI
ICLD
BAGR
TEVE

Reasons:
(small cap firms have empirically demonstrated a stronger momentum effect)
(large beta)
(a loss of around 30 to 50 % in a 52 week period (past 52 weeks)


When deciding which stocks to short, we specifically focused on small cap firms, firms with a
high negative beta, and firms which had a loss of around 30-50% for the past 52 weeks.

The reasoning behind picking small cap stocks to short is due to the fact that they have
empirically demonstrated a stronger momentum effect. This is because they are neglected and
hard to trade which means the mispricing should be stronger.

We picked a large beta (negative for shorts) because the higher the beta, the more volatile each
individual stock is. We are going ahead and implying that both winners and losers are extremely
volatile, and we are looking to short losers and buy winners. The reasoning behind shorting
losers with high betas is because the beta for losers is higher for distressed firms. This therefore
making the higher negative beta a great characteristic for picking our shorting targets. On the
flip side, however, we chose a high positive beta for our longs to choose winners, but not
distressed winners.

Lastly we chose to short losers and buy winners. The characteristic to help us distinguish
winners and losers would be a loss (for shorts) or gain (for longs) of around 30-50% in a 52
week period (past 52 weeks).

ii. Explain how you use what you learned about the short-term reversal to make your
strategy more profitable. (10 points)

The short term reversal effect suggests that recent winners (top 10% based on last
months return) exhibit a 1% loss in the following month, while recent losers (bottom
10% based on last months return) exhibit a 1% gain in the following month. The short
term reversal effect only lasts for one month and creates a window of initial losses for
the momentum strategy. In order to make the momentum strategy more profitable and
minimize this early downside potential of the short term reversal effect, an investor can
generate a list of recent winners to take a long position in and recent losers to take a
short position in and wait one month, so that the short term reversal effect will subside.
After the short term reversal effect subsides, the long and short positions can be taken
and the momentum strategy will not be subject to the downside caused by short term
reversal. (Decrease in value of asset underlying the long position/ Increase in value of
the asset underlying the short position). An investor can also absorb the negative effect
of short term reversal by sorting on last years returns instead of last months returns
because the bounce back effect will be noticeably smaller: from 0.5% to 1% to less than
.25% based on the short term reversal effect.

Short- term reversal makes our strategy more profitable because it is the fundamentals of the
momentum strategy. When stocks go up people will buy but at one point the price will go back
down. If you sort on last month when the price bounces back the return will be large. It makes
our strategy profitable because we are investing in small firms, and they tend to be extreme
winners or extreme losers and extreme winners and extreme losers is what makes up the
momentum strategy.


iii. Compute the return to S&P 500 in the past 36 months (Hint: you need one very
simple formula). Looking at the return to S&P 500, do you predict your momentum
strategy to be profitable? (10 points)
The return for S&P 500 was 15.59% based on the return formula. Return = ((buy price -
sell price) / buy price) * 100. Base on this we predict that our momentum strategy is
profitable. This is because the earnings momentum is strong and according to Chan,
Jegadesh, and Lakonishok, we should keep these stocks for a few quarters before
selling them.
Start date: Sep-2011
Start price $1,321.12
End date: Sep-2014
End price $1,817.04
Number monthly prices 36
Minimum price occurred on Sep-2011
Minimum price $1,173.88
Maximum price occurred on Jan-2014
Maximum price $1,822.36
Total gain or loss (Start date to End date) 37.54%
Annualized gain or loss (Start date to End date) 11.21%
Total dollar value on End date of a $10,000 purchase made on Start
date
$13,754
Average percent return for all possible buy and sell combinations 15.59%
Number returns 666
Number profitable returns (Winners) 602
Percent profitable returns 90.39%
Number unprofitable returns (Losers) 64
Percent unprofitable returns 9.61%
Winners to losers ratio 9.41


iv. Use the regression from Chordia and Shivakumar (JF 2002) to predict the return to
the momentum strategy in the next month (you do not need to run a regression, you
can borrow the coeffcients from the slides). You may use the August 2014 values
of the variables you need. Assume that the dividend yield is 2% and obtain the
value of all other variables from FRED database at the St. Louis Fed branch. If
you are doing the calculations right, in November 2011, DEF=1.27%, TB=0.01%,
TERM=1.9%. You should use the returns as percentages, not as decimals (i.e., 2%
is 2, not 0.02). (15 points)

Data from FRED Database:
Regression Terms Values
Term Premium (10 year treasury constant maturity rate August 2014)-(1
year treasury constant maturity rate August 2014)= 2.4%-
0.1%=2.3%

Default Premium Baa-Aaa = 4.7% (August 2014 value) 4.1% (August 2014
value) = 0.6%
1 month TB rate 4 week treasury bill for August 2014 = 0.03%
Dividend Yield 2%


Substituting the values of the business cycle variables: Term Premium, Default
Premium, 1 month TB Rate and Dividend Yield found on the FRED Database into the
regression equation:
MOM
t
= -5.60 6.35[DEF
t-1
]

0.42[DIV
t-1
]

+ 2.10[TB
t-1
]

+ 2.94[TERM
t-1
]
Yields an expected return of:
MOM
t
= -5.60 6.35(0.6%)

0.42(2.0%)

+ 2.10(0.03%)

+ 2.94(2.3%)
= -3.425%
Therefore, we expect to lose -3.425% on the momentum strategy in the following
month. However, since August 2014 data was used, this regression predicts the
expected return to the momentum strategy in September.





v. Take the smallest winner stock in your portfolio and the smallest loser stock for your
portfolio (smallest means the lowest market cap). Calculate their trading costs as
the bid-ask spread plus the price impact. Assume a hedge fund intends to trade $2
million in each of the two stocks and faces the same price impact as anyone. The
fund expect to complete the opening/closing of the positions in one day and intends
to hold the positions for 6 months. How large the return to the strategy involving
only those two stocks has to be before the hedge fund starts making prot? (25
points)

v.
Short Position: UPI | Market Capitalization: 55.03M
Roll Measure = 2 * sqrt( Cov(R
t
, R-
t-1
) )
= 2 * sqrt(.991105)= 1.9910%
Amihud Measure =
30.72
Trading Cost =
2.00 million x (.01991/2) + 2.00 million *(.3072 * 2 / 2)
= .01991 + .6144 = .6341
= .6341 Million
One Way Trading Cost (Percentage):
= .6341 Million / 2.00 Million
= 31.75%

Long Position: TRGP Market Capitalization: 5.78B
Roll measure = 2*sqrt(.00000482444)= .43929%
Amihud Measure =
.000384296

Trading Cost =
2.00 million x (.0043929/2) + 2.00 million *(.000384296 * 2 / 2)
=.0043929+.0007686 = .0051615 Million
One Way Trading Cost (Percentage):
= .0051615 Million/ 2.00 Million
= .258%

Total Trading Cost (One Way) =
.6341 Million + .0051615 Million = .6393 Million

Total Trading Cost (Two Way) =
= .6393 million * 2 = 1.278 Million
Total Trading Cost (Two Way) Percentage =
1.278 Million / 4.00 Million = 31.97%
The return on the momentum strategy over the 6 month investment horizon would have
to exceed 31.97% in order for the strategy to be viable.