Hi Floyd,
1) Initially when I was testing the AFL under live market conditions, I was maintaining a trade log. As the AFL was being modified during that testing, The trade log would also be changing and it would have been wrong to use it to evaluate the success rate of the strategy. Altough I do not remember exact proportion of profits I was satisfied with the performance. Later when I started posting the triggers along with daily updates, it became difficult to maintain all those past trades without cluttering the Execl Sheet with that data. So, I stopped updating the trade log. However, now I am planning to start it again with clear demarcation so that it will not create much distraction for the users who are not interested in trade log. But I will start it from May. This is because I will be away during April and therefore there will be discontinuity. After my return in April end I will make a new Excel file to contain the past performance also.
2) In non trending market there will be whipsaws. Any system based on Moving averages suffers from this phenomenon. It is posiible to reduce the whipsaws with longer averaging periods but it will delay the entry and exits resulting in losing many potential profits. Generally, if you find that the system started giving triggers which would reverse too early, it will be a sign of side way price movement. After two or three such failures, keep away till a trend begins. When you have two or three failed trades, you can see on the chart that a channel would have developed within which the price moves. Wait for the price to break out of that channel and then start trading. Although you can use other technical indicators to detect the absence of trend, they are not 100% fool proof. To detect a trend or its absence a few days or weeks data on the chart has to be examined. By the time we establish the presence of a trend or its absence it may be time for the price to break out. So, a visual inspection of chart and identification of the trend would be essential. An experienced trader would detect it by eyeballing the charts better than the indicators.
3) If you see the Stochastics, MACD histogram and RSI you may be able to avoid some false trades. As I mentioned above, it is not 100% fool proof but can be relied upon to some extent. As a thumb rule you can check the following:
- The Moving Average graph would be flat or beginning to flatten.
- The Stochastics, fast and slow, would be moving parallel or criss-crossing in the range between 30 and 70.
- The MACD Histogram would be quite high (above or below zero) or very close to zero line and going frequently on either side of zero line.
- RSI would be neither in oversold nor in overbought region.
Use the following settings for stochastics: 10, 3 and 10, 3, 3
Use RSI(9)
Use default settings for MACD Histogram.
- If fast stochastic is crossing or crossed above slow, in the region below about 20 (can be sometimes up to 30) it is indicating a Buy trigger.
- If MACD histogram has just gone above zero or is just below zero but increasing (going toward zero line) it is also buy signal
- If RSI(9) is above 50 and increasing when the above two conditions are fulfilled it is also a strong indication in support of the first two.
The SL or Sell Trigger takes care of the exit. However if the opposite of above three conditions are observed, it is a strong signal to exit.
As I said earlier it is a thumb rule. Thre will a few failures but we will be protected better.
The market does not read the charts nor follow the indicators. It decides what it wants to do and the charts and indicators simply follow the market.
-Anant
