Some years ago I had bought
this options trading system where I had to manually calculate the signal for
the next day. After a while it became tedious and decided to put the formula on
Excel. I had so much fun doing this that I got out my back copies of the magazine
"Technical Analysis of Stocks and Commodities” and looked for indicators
with the Excel formulas and started working it on Excel. Indicators such as
Bandpass indicator, Polarized Fractal Efficiency, there was an article on how
to create sophisticated indicators using ideas from Physics. I was very excited
about this and started working on an “ultimate indicators”, combining the
various indicators and using conditional formatting on Excel to flash buy/sell
signals.
I have also bought various
investment books and working papers over the years. These were mostly on
technical topics, like Chaos Theory, Neural Networks, Game Theory, GARCH, etc.
Most of the time I didn't really had the technical know-how but was interested
and managed to get some ideas. The most useful book for me was "The
New Technical Trader" by Tushar Chande and Stanley Kroll which have lots
of Excel formulas.
In the end, I decided that I
wanted to create my own stuff and started dropping all the other indicators. My
background came in handy. I had studied Electrical and Electronic Engineering
before I dropped out after a year to study Economics. In my Economics course I
had gone for electives such as Econometrics and Mathematical Model. It was
time to put it to use. Combining theory with real world applications I set upon
developing my own trading model, using mostly my ideas. I called it Logical
Reasoning, which is basically using the logical functions in Excel and to model
the market.
I used to create some
sophisticated worksheets back then and remembered use to run out of
columns when Excel stop at IV. Sometimes I didn't had the computer power to
continue with a project. I must have developed over 100 models over the years.
One of my favourite one was where I had to define
a quantitative pattern and then search historical data for the given
pattern. It returns the percentage of the match and indicates what the market
did in the next 1-3 days for each of the day in the history. For example I can
sort and filter for 90% and above and if majority shows an up day tomorrow, I
can then look at the current indicators to see if it supports the result. There
were some interesting result such as seeing a particular day of the month in
history matching the search day and month. Still work in progress as it is not
very reliable.
Getting data into excel is
another challenge. I was analysing Forex data but noticed chunks of data
were missing in the historical data. For the type of quantitative model I was
developing data integrity is paramount. I decided to switch to
Indices and Stocks with data from Yahoo.
I get to see many data
vendors advertising streaming quotes in Excel but I don't know how they expect
you to use it. I am assuming most who use Excel for trading have some kind of
trading indicator, where you require historical prices and today's prices to be
able to see your indicator live. When you are trading there is no time for
cutting and pasting data. You want something fast and something you can use
with ease.
I found eSignal OnDemand
Qlink to be the best for this. The quotes are on fixed cells (rolling basis),
automatically merging the historical data with today's prices. All you have to
do is change the ticker and up comes the quotes. I was on the delayed data plan
which wasn't very convenient as I can't afford the real time data plan for
now. With the real time data plan I should be able to spot the
opportunities quickly. I am sure it will be good for trading weekly options on
stocks.
The next step will then be to
automatically route the trade to the broker. Mexcel Trader does this for Forex.
It will be nice to see something similar and affordable for stocks
and options. I will have to look into Interactive Brokers TWS and see how that
can work with my model.
I enjoy using Excel for my
trading as it gives me the freedom to be innovative. Over the years I have
developed lots of know-how and I am quite comfortable using it. I
have plans to add a real time data plan in the near future which will
bring the model to live. When the model is live I should be able to spot more
opportunities and act on them quickly.
No comments:
Post a Comment