How can you even have an idea that a trading strategy or system might be profitable if it isn't backtested? Backtesting is not a cure all, but it is a tool to you an idea if you are working with an idea that has a positive expectation. Good trading.
Look at what Mark Mark Carhart at Goldman's Global Alpha does. Carhart, a former assistant professor of finance at the University of Southern California, helps oversee other hedge funds, four mutual funds and scores of separate accounts. In all, he and Iwanowski have $101.5 billion at their command. Carhart and Iwanowski use math-heavy trading tactics that fund consultant Sol Waksman likens to counting cards in a casino. The two lead a corps of computer-loving traders, statisticians and finance and economics Ph.D.s. It works. I use Odds Maker from TI and it gives me that card counting advantage. I think the key thing is that I can test a system versus trying to find a pattern in a stock to trade. I can run a ton of scenarios and tell you what is working right now. I am in like 15 to 20 positions at once all based on real-time odds.
Absolutely, the 'odds' are what it's all about. No system is going to assure that a single trade is going to turn a profit, and too many novice traders think that is the case and get discouraged when it doesn't happen. But through backtesting, one can go a long way towards determining the probability that over a given number of trades, a profit is likely to be realized. That's all one can expect from backtesting. But that in itself is a great advantage.
Now that I trade this way I am convinced that it is the only way to go for the future of the active trading space. All the time things happen and we as traders have to adjust. I think from what I am seeing and talking with others is that scalping is going the way of the Dodo. It is just too random. That is why I switched my style to this odds based position trading. I would never have been in half of the names I am in right now. I feel good that I don't really worry about the futures etc, just follow the system and the rules and it works out. I have down days sometimes several in a row but overall it cooks.
Amazing comments. I gather that you do have some sort of universe that you are rotating through and that you make some money off and on, mostly on. I rarely run into a person who is doing better than their yield on commodities futures as compared to their yield on stock trading. Ordinarily the yield on commoditie runs 10 to 20 times that of stocks per day. I peg stocks at about 2 and 1/2 percent a day on average for position trading (unleveraged) over the trading days of the year.
Jack, I think most of us would wet our pants for a 625% annual return (2.5% x 5 trading days/week x 50 weeks a year with 2 off for good behavior). When I personally sat next to someone who is probably the closest I have ever come to a real trading god, all he was able to clear was 300%. And, by the way, he backtested. Curve fitting, by the way, is not the point of backtesting. I spec plays on macro factors, but if I'm going to trade a specific market, I will backtest it first and see what works. Not that it isn't intuitive anyway, but good to have that confirmation. And once you get your initial run of the backtester, then you have to start thinking about %wins, DD, and all sort of fun stuff. Easier just to punt.
No, it is a pragmatic value for what to expect while monitoring your holdings and the stocks in your batting order. I am a person who is accustomed to dividing my capital into several streams. The basis of the number is mostly related to the hold time of the position. 4 to 8 days is a common hold period so 4 to 8 streams is a good arrangement. I use FA to pick my universe. Each stock in it is ranked. The rank is the % per day that the stock makes in the hold period of its "natural cyle" of price movement. My trading is based upon TA and in particular the P, V relationship. For clearer understanding o the cycle I add the Accumulation/Distribution factor into the P and V mix. I score the stock's cycle using binary values for P, V and A/D in that order of importance because the frequency of change of the values is doubled as you go from Price to Volume to A/D. The trough breakout up is a 7 in decimal base and the peak passes through 4 going to 3. I buy when the volume breakout leads the price breakout going from a score of 0 to 7 in the morning of the breakout. If you read the illustrated transcript of "Putting the Pieces Together" from 22 JUN 07 of the Tucson IBD MeetUp, you can see 31 snagits of the batting order of four streams of capital each of which have three stocks selected for when they are going to breakout during a given week. The 31 snagits show how a day progresses and how a person does the position trading with his capital. The illustration there is one where the trading is being done by a person to use given capital and trade it over a 100 day period to achieve a thing I call OOM. OOM stands for Order of Magnitude. At any given time people who trade PVT style are working on getting their capital to change by an order of magnitude in a period of 100 days. This is called crossover trading and involves optimizing the use of capital. In this reference, you can see how nominal 2 1/2 percent a day is in trading using the PVT method. It is a very low value for a person who is making the money that is available every day. For example the stock at bat in the A stream is sold during the day to have capital for a buy the next day. This stock was finishing its run up and the prior day had made 10% in that day alone. Those two days in the run accounded for 13% of the total run up in a few days. Stocks qualify for the universe by having 5 cycles in 6 months. The major objective of making money in stocks is to have a Universe that is reliable and repeatable. The FA analysis achieves this. An example of repeated trades of a given stock is NTRI in 2006. It was tradable four times, all known to be coming up in advance. Each trade was 3 days duration. The yield per cycle was 30% over three days or about 10% a day. This trading consumed 12 days of 240 trading days in the year. The remaining days (228) were used to trade other stocks in between the frour cycles of NTRI. One of the journals in ET was used by several others to post these concurrent trades that they made at those times. For making money, it is best to set up your activities to know when trades are coming up and then on the day of the breakout to monitor the volume breakout which precedes the prive cycle breakout. In this way the trading time is optimixed. The general expression used to describe this trading method is: Tomorrow's Newspaper Today. If a person were to be assiduous about being wealthy, he could use this concept to estimate where he would be in a while. Sit down and take you present capital and write it on a sheet of paper. List a column of dates that show the passage of 100 days each. There are more than 2 a year. Next to each date put in the future value of your capital by simply writing a number ten times larger than the value above the entry for the prior date. If you have 10,000 dollars now in 100 days you can have 100,000 dollars. Before the end of the year you will move the decimal point again and have 1,000,000 dollars. If you choose not to do this that is fine. On the other hand it is a worthwhile effort. Most people choose to not do such a thing because of the process involved to learn to do it. the fact that you asked a question is the first step to begin learning. Here is a what if for you. there is a person posting in ET who wants to have 10,000,000 to manage. He says he has about 17,500 dollars now. He just finished losing 2,500 dollars last year. A relative of his gave him 200,000 and it makes 10% a year as managed by a broker somewhere. He has an uncle who, his grandmother says can lose a million dollars and shrug it off. His grandmother does not want him managing his own money because of his losses last year. What would it be like for him to learn to trade PVT? He has been in ET and the method is in its third yearly journal. What would his grandmother think if he were to take the 17,500 and make it 175,000 by Summer time this year. then he could take the 210,000 then in his account and combine it with his 175,000 and have 3,850,000 by the end of the year. So how far past 10 million would he be by the end of 2008 when he is approaching 22 years of age? He will probably be focussed on other things by then. If he learns the template for trading stocks, he will quickly find out how to use the same template to trade futures. That jounal just started in January as a spin off of the stock journals. It gets a few hits a day nowadays. The originator of the stock journal now trades ES primarily for the simple reason that it makes more money. About 10 to 20 times more per day as a percentage of capital. What would happen if a person wanted to make 10,000,000 dollars and he already has 17,500 left of his present account? This would be something to see. Actually it is something to see. Others are doing this kind of thing every day of the week. Some people can't do this kind of thing at all. They are choosing to not do it everyday of the rest of their lives. Personally, I like to think that is it a pleasant thing to know is available and waiting for anyone who chooses. It is very pleasant to just sit down a take out a piece of paper and see what is possible and then either decide to go to work or decide to pass and let the opportunity just lay there on the sheet of paper as long as you want before you decide to bear down and get the job done. On the other hand I occassionally sit around with people doing this and we think of what it would be like if Cramer got wind of this or even Stevie Cohen.......LOL.... For 2007, this is going to be a good year. Some more people are going to get to work.