| A Quant Strategy to trade single stocks |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
A colleague proposed an interesting strategy:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
In this example I used a fixed basket of 15 stocks for my universe to
choose from.
|
|
|
|
|
|
|
|
|
|
|
|
|
| I
tried to pick a basket that I felt I could stick with over a long period
without having to refresh the universe. |
|
|
|
|
|
|
|
|
|
|
| I
tried to pick all the giant leaders of tech, plus where I see growth in the
technology space (3D printing and cloud computing in my opinion) |
|
|
|
|
|
|
|
|
I
didn’t go as deep as chip manufactures/part suppliers etc.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 7
are tech giants |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
-Apple |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
-Google |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
-Microsoft |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
-Rim |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
-Yahoo |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
-Amazon |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
-Internation Tech ETF (to give exposure to Samsung, approx 20% weight in this
ETF) |
|
|
|
|
|
|
|
|
|
|
|
| 2
are 3D Printing leaders |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
-Stratasys |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
-3D Systems |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| The
Remaining 6 are Cloud computing leaders (some of the Tech giants also fall
into this category, google, Amazon) |
|
|
|
|
|
|
|
|
|
|
-SalesForce.com |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
-VMWare |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
-Verizon |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
-Citrix |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
-Red Hat |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
-Rackspace |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| The
portfolio holds no core positions, everything is based off a trend capture. |
|
|
|
|
|
|
|
|
|
|
|
|
| YTD
performance on this basket with the highly data mined constraints is 43% to
the end of October, with a max drawdown of -8%. |
|
|
|
|
|
|
|
|
| Charted
against Managed Volatility it does appear to add some uncorrelated returns. |
|
|
|
|
|
|
|
|
|
|
|
| There
are no fundamental P/E ratios used so some of these companies trade on some
pretty lofty earning multiples, but at least there is an exit plan should
momentum disappear. |
|
|
|
|
|
| Earnings
releases can be a bit risky though as there is potential to gap through a
stop overnight. |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
The results are pretty impressive. However with single stocks the
dilemma in a backtest is would you have ever picked these same stocks a year
ago, so results must be taken with a grain of salt.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Data mining is very difficult to avoid in any backtest like this
especially when many of these stocks have had a rip roaring year (excluding
RIM or course).
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
It still needs some tweaking around the edges to reduce the trading
activity, however I found that by taking profits at a fast rate during the
trends I can often lock in decent profits without giving too much back as the
momentum fades off.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| Backtested performance as of October 2012: |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
-
I will provide an update soon.
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|