Short Term Trading Analysis. V. Daragan

Stock prices go up and down. It would seem that it is simple to make a fast profit from day trading. You need to buy stocks after a price drop and sell them after some time with a profit. Unfortunately, this is a minus sum game. Bad stock selection can easily kill you. The faster you trade, the faster your losses become large. Every transaction has a cost: bid-ask spread and brokerage commissions. Very tiny profits will disappear in a long run when you take into account these transaction costs. Nevertheless, day traders still exist and many of them are very rich people. How did they do that? If you are a professional and trading on the floor of the exchange you can look for micro trends and catch the leaving train right on time. Big capital and zero commissions allow you to make micro profits constantly.

These notes are for nonprofessionals. Can you make profits by short term trading? Yes, you can. You should be very selective when buying stocks for trading and try to anticipate the stock price moves before professionals. Is this possible? The answer is yes. Let us show how to do that and what is the risk of short term trading. We will consider trading stocks. We suppose that you are using on-line discount brokers and you are paying small commissions (10 dollars or less). We also suppose that you are able to watch stock quotes a few times a day and you are able to buy or sell stocks during the trading day. These notes are not about day trading. Our time frame is 2 - 5 days. Computer analysis and our own experience has shown that this period of stock holding is long enough to make profits with reasonable risk level.

Consider the situation: an amateur is trading stocks. His choices are terrible and his style of trading is not getting better. What can we do for this guy? How can we improve his return? Suppose the average growth probability of the stocks selected by this amateur is 40%. He cannot believe that his choice is always bad and he just wants to have some trading strategy to improve his return. He likes the stop-limit strategy described in the previous section and he wants to know what stops and limits are best for him. The equations for calculating the optimal stop and limit levels are rather complicated, and here we will just give you the answer: L/S ratio should be more than 10 in this case. So, the stop order should be very close to the stock price at the moment of purchase.

This is the only strategy that can help this guy to survive any longer in the market.
The same idea can be applied to a bear market. If you buy stocks when market is declining your stop orders should be as close to the purchase price as possible. For experienced professional traders, when stock picks are excellent and the growth probability of these stocks is high (for example, 60%) the stop order level should be far away from the purchase price. The same rule can be applied to a bull market. Do not place stop orders close to the purchase price if you are sure about your stocks or when the market is rising. For example, if you expect a profit about 30% your stop orders should be about 20% less than the purchase price. We will consider this problem in detail later, when some specific trading strategies will be analyzed.

It is so easy to lie using statistical analysis. Probabilities, distributions, standard
deviations, ... - all this stuff is made specially to leave the truth in the dust. Give us any chart and we will develop an excellent trading system with 90% growth probability. The analysis will show a 20% return on each trade, and we can write a nice book about a new revolutionary trading system. When can one believe in the results of computer statistical analysis? Consider an example. You can start your own analysis. Take the chart of some stock and find the patterns when the stock is ready to rise or to fall. Write down these patterns. Take another chart. Check all previously selected patterns and see what happened. Take another chart... Take another chart... Do it for thousands of stocks. Check your patterns for bull and bear markets. Look for what is working and what is not working.

And give us the percentage of events when your patterns are working. If this number is more than 60%, we will use your patterns for trading. It can be profitable. So, you have to analyze a large number of stocks over a long period of time to draw a conclusion. However, even after such analysis you can not give 100% guarantee of future results. Your statistical analysis will give you the standard deviation of your expected returns, and you should use these numbers to estimate possible troubles. What patterns do we believe in? We like trading ranges if we can call them as patterns. There is a real crowd psychology behind these patterns. All traders know about trading ranges, and this fact makes trading ranges rather reliable. Our computer analysis has shown: yes, they work.

How do trading ranges work? What average profit can you obtain using trading ranges as a basic of your trading strategy? What time scale should you consider to obtain maximal profit? How about the trends? These are the questions we are going to answer. We performed a computer analysis for
thousands of US stocks for a period of almost 5 years. For some stocks, an 11 year trading history has been considered. Now, let us say a few words about general problems of computer analysis. Every day you can load down the files from all the US stock exchanges. You can see in these files the opening price, maximum and minimum prices during the trading day, closing price and trading volume for thousands of stocks.

Прикрепленный файлРазмер
Short_Term_Trading_Analysis.zip369.7 кб

Основы