Here is the most current update on the VL5xZ5 strategy:
I've been playing with a mechanical shorting
strategy, with the following parameters:
Rather than using relative strength to sub-select
from the VL stocks with timeliness rank 5, I sub-
select using the Zacks rankings (that is, I choose
the stocks which are #5 in both VL and Zacks).
Then, since I am interested in using this strategy as
a hedge or market-neutral strategy, I combine the
short sales with the purchase of an equivalent amount
of SPYders.
I reported on this strategy previously on AOL MF
boards and in the shorting threads on Silicon
Investor. This is an update at month 4 (as of closing
prices 8/26).
In my first test (beginning 4/27, not with real
money), I sold short 18 stocks which were ranked #5
in both Zacks and VL, and bought the equivalent
amount of SPY. It is now four months since I began
this test. For the period since 4/27, SPY is down
0.7% and the shorts are down 27.8% - so the total
return from a completely hedged portfolio for the
last 4 months would be 27.1% (these are period
returns over four months). All positions were
purchased/sold short at the worst price of the day,
no commissions were included, and dividends were
ignored. From 4/27 to the present, the Russell 2000
lost 20.8%, so a portfolio constructed with Russell
2000 on the long side, and the VL 5 shorts on the
short side would be up 7% over the past 4 months.
Here are the returns by month (cumulative):
2 mos 3 mos 4 mos
Shorts +12.4% +19.7% +27.8%
SPY + 4.0% + 4.0% - 0.7%
Rus2000 - 6.3% -10.4% -20.8%
In a second test (beginning 6/10, using real money),
I sold short 10 stocks which were ranked #5 in both
Zacks and VL and had dividend yields below 3%. I left
out 2 more for which I could not borrow shares that
day. I have since added to this portfolio, shorting 5
more stocks on 6/29, and 5 more on 7/13. On 7/13, I
included stocks with higher dividend yields (to 4%)
to my immense regret, since one such stock (Betz-
Dearborn) was taken over at a large premium. I have
also covered 2 of the original (6/10) shorts at small
losses based on positive news reports, one of which
(Boeing) subsequently went down sharply :-). The
short sales from 6/10 have made 20.9% (including the
losses from the covered positions), those from 6/29
have made 19.4% (including one stock covered at a 9%
loss due to a takeover), and the ones from 7/13 have
lost 6.1% (including the stock on which I lost 75%
due to a takeover). The total market neutral portfolio with SPY
bought in equivalent amounts each time I shorted
stocks, would be up 10.2% (the shorts have gained
14.4%, the SPY has lost 4.1%). If instead the Russell
2000 were bought in equal amounts, the market-neutral
portfolio would have lost 1.2% (in other words, my
real-money shorts didn't drop quite as much as the
Russell 2000 over the same periods).
Don Katz
|
