Archive for April, 2010

Trading System BackTest Rules and Guides

Monday, April 12th, 2010

One of the most important aspects of any trading system whether used for stocks or futures trading is a “thoroughly” tested trading system. This process is done in paper and no real money involved. The first stage of the testing process is called “backtesting”.

Backtesting and Paper trading will put the formulated trading system to test under the assumption of:

Randomness of Historical Stocks or Futures Data = Randomness of Future stocks or futures data

It means that the trading system will be executed on past data and assess results. Backtesting can be done with stock trading software like E-signal or done in an Excel spreadsheet. Of course, doing it in Excel software is somewhat a bit complicated and manual.

So what are rules of backtesting? The rules should conform to the standard and acceptable rules in statistical sampling.

1.) The sample size of back test is important. It is recommended that if you have full access to past trading data, a minimum of 5 years time span is recommended. So if you are in year 2010, you should gather past trading data to be used as back testing sample from 2004 to year 2009.

2.) Even though you have a past five years of historical data, you should make sure that in those 5 years, you have at least 300 trades. So it means that in a month, you should execute at least 5 trades. If the studied trading system does not able to produce 300 trades in 5 years. Then you might consider using a longer time span say 10 years. (more…)

Calculate Stock Volatility: Avoid stocks with highest volatility

Saturday, April 3rd, 2010

Volatility as applied to stocks trading is a measure of risk of a certain stocks. Technically, volatility measures the dispersion or variation of the stocks that corresponds to the variation of the expected return.

In this short guide I will illustrate how to compute stocks volatility so that you might avoid trading stocks with the highest volatility. Take note that the higher the volatility, the higher the risks associated with the stocks.

In statistics, there is a technical measurement called %CV or coefficient of variation. You can apply this same concept in measuring the volatility of stocks. The formula for %CV is as follows:

%CV = (Standard deviation / Mean) x 100% (more…)