Kiwi, I have been a trader for 4 years (not very long in my view). How long have you been trading online (please don't include buying and holding shares through paper methods)? What did you learn from Monte Carlo analysis and why did you abandon it? It might be helpful to other taders to explain why it provides no benefit. You mention money management - how do you do this, or do you do it innately through experience?
Trading 10 yrs+ Primary Income 5 yrs+ Futures (US, Currencies, European, Asian). Where did I say I abandoned Monte Carlo analysis. Why do you think it has no benefit? Is trading your primary income? Feel free to prove my hypothetical courageous traders above would fail to outperform the whimps (but don't use any assumptions that I'll come back and refute tomorrow morning after I've had a few more beers ) Whats important: - survival through different markets and market conditions. - courage because mediocre consistency will keep you poor. - knowledge to gather the tools you need to be effective. - understanding to learn and adapt and discipline to deliver the promise. And achieving your goals and being happy as you do
As someone who has been thru the mechanical system mill from inside out, I'll me chime in a bit on <b>kiwi's</b> points. Monte Carlo risk of ruin, percent correct, profit factors are all critical in building a productive system. That is true because systems are highly inefficient compared to <i>emotionless</i> discretionary trading. Any & all mechanical systems are curve-fitted to some degree. It is that fitness itself which creates profit during certain market behaviors and drawdowns during others. Entry point strategy may always remain robust... it is usually the management - exit rules of a system that create profit or drawdowns. * I myself trade using chart indicators and patterns. I only trade during peak productive hours for my preferred market (eminis). What I do would be considered discretionary, in what I label as "methodical". There is no way in the world to back-test or crunch data on my performance or results, because what I do cannot be written as a system. Only after x-years performance could I look backward and analyze actual results. Aspiring traders who believe they can write up code for any approach and quant it from there have not yet learned all aspects to this profession. Likewise, I've written numerous simple and complex mechanical systems that passed all the tests with flying colors, including monte carlo from all angles. None of them, not one system has walked forward in real money as highly profitable since. Some of them would be profitable over the course of time, but a good discretionary trader can outperform those with ease. ** Aspiring traders can fixate too much on the probabilities of success and miss real-money rewards in the process. I know of two very public system writing "pros" who do not trade a single dime of real money. They market their wares with high price tags, the systems pass every mathematical test for success yet the originators refuse to trade any of them. Why do you think that is? Mechanical systems are highly inefficient due to lack of intelligence = ability to adjust in changing conditions. Discretionary methods <i>can be</i> managed inefficiently due solely to lack of emotional management. Successful trading is a lot more than pure math & stats... those can often be an illusion until experience lends clear understanding to it all. As for me, pattern = indicator trading is my only personal choice. Get good at that, and there is no need for monte carlo testing. The only chance for ruin comes from within oneself... via emotional management.
Backtesting , Monto carlo simulation, data mining is all useful concept. They are framework to build your systemic approach but they are not gospel truths. I had a long conversation with someone who has worked closely and observed a market wizards trading. Now this particular market wizard in his public writings has talked and written about a lot on risk management, risk of ruin and so on. But many of his system traded were based on rules which were different from what was publicly said if you strictly interpreted the writings. The explanation was it is more convenient in real trading and second the tools tell me what is worst possible scenario but I have higher risk tolerance.
System versus discretionary trading is really a red herring. The real topic is risk versus reward. In kiwi's two examples, we do not know the risk for each system. He has not given the maximum peak-to-peak drawdown of either system. Thus, determining the risk:reward ratio is not possible. I prefer to start with a safe, consistent system and ratchet up the margin to a level that allows maximum profitability while still being beneath that which would have a high probability of a blowout. -Raystonn
At least you realised that those numbers were not as weak in the way Richbynature thought they were but they are only numbers. The point behind them is that the statement "Consistency is the measure of success" is somewhere between wrong and simply inadequate. Lots of people can do the numbers (hell I've got 3 degrees, trust me, so can I) but delivering more than low performing consistency is required to really succeed. There are a lot of people who can be consistent but few who can really deliver ... like lescor (way to go on last month!) For a real trader (rather than someone trying to analyse historical system numbers or historical fund performance figures) the measures of success are more likely to be found in: Ask anyone here who hires traders.
Back-testing data results is only possible when we have at least one constant to build a basis from. A mechanical system is assumed to be constant unless someone changes all of the rules. A system will execute entries, management and exits the same, every time without fail until otherwise intervened. In this case, the only variable is future price action in its chosen market of operation. Discretionary or methodical trading is exactly as described. There is no fixed constant... only degrees of variables. The market itself remains a variable, but so does any individual trader. Monte Carlo stuff is only useful for measuring a squre peg system in the round hole marketplace to see how its "fitted" for probable ruin. It assumes random market action will be interpreted by the blind system exactly the same, all of the time. However, it must also be assumed the individual trader will use some degree of discretion over time. People change. They learn new patterns, tips and tricks. They visually interpret subtle changes in the market and adjust accordingly. People are flexible... able to bend with various market conditions. Systems are rigid, they break instead of bending. Therefore, it can be assumed that an individual trader following any methodical approach to trading can be more productive in sustained fashion AND operate at a much higher degree of efficiency than mechanical systems. If said individual trader manages emotions correctly and adjusts to fluctuating market conditions, that person can experience sharper swings with less risk of ruin than a system. The individual efforts are all variables to some degree... there are no constants in this equation. ** Back-testing a discretionary method is still important, but much different than the process for quantifying a system. Measuring individual performance based on risk of operation is likewise different as well. A skilled trader can operate on a finer edge over time than any system can, directly due to the human flexibility factor.
Kiwi, With reference to the 2 approaches you have given as examples my only point with regard to 'consistency' is the probability of account extinction. If your account is only £10K then in period 1 there is a strong chance of the account going broke. If your account is £10m then I would consider both approaches to be conservative and lead to consistency (in terms of an equity curve's growth). Would you be able to give us a feel for the account size on which you based these figures? I think it would be helpful to specify as it gives a feel for the risk tolerance per trade / period. (PS what were your 3 degrees in, I have never met anyone with that many!?)
Discussing stuff in any forum is difficult. When I said your postings were way off the mark I was not differentiating between opinions and facts. You have expressed both opinions and facts. And, in either case I'm sure you could give references as a consequence of your finding what you state through following the path you have followed to get to the point you are at. Your thesis: consistency: the measure of success and how you deal with its foundation and development is what I was referring to. You are way off the mark with regard to each of these three things. When I express a view, I have reasons for doing so. Consistency as you describe it and measure it is not a key ingredient to determining success or the degree of anyone's success. Consistency would not qualify as a determinent of success no matter who devised the way to measure consistency. Any trader who is concerned with understanding his degree of success would go elsewhere to make that determination. So that is probably OT in this thread. With regard to consistency of trading success, it is not a constant but varies ever upward as time passes simply because of the broad and pervasive factors guiding it. They fall into two categories: market perfomance and trader performance. You left out market performance. Markets yield capital to those who trade. The yield varies over time and the variation is greater than what is spoken in ET as the best a trader can expect at an expert level. Let A be the ET maximum performance on record level that is accepted as real by persons like you or posters who have 7000 or so posts. Let B be the variation in yield of the markets. I am saying that B is greater than A. What you think you are measuring as the best possible is less than the variation in what the market offers over time. I conclude that what you deem to be success and the details of measuring it , is not in the ball park of the potential a person has for true success. Few persons in ET deal with the actual real levels of success that are possible and are attained by some traders. I hold the highest possible standard as the standard of success. I feel this standard is well defined, is measurable and has nothing to do with the measures you have posted. You should use yours, but they are not my cup of tea simply because I am interested in making money and everything subtends this. I examine what is possible in the context of what the market is offering to traders to take out of the market at any given time. How does a person determine his consistency? And what is the framework for looking at the array of possible ways to improve consistency? I feel there are 8 categoies and within each category the range of improvement is a doubling of performance. You measure a trader curve of profits and ignore the variation in the market's yield performance. I measure the trader profits and how he is utilizing 8 factors to close the gap between the current market yield performance (a major variable). The trader for stocks sees the current market yield performance through two reduction filters: 1. stocks that qualify to be traded by quality, repeatability and reliability ; and 2. by the rank of the stocks (the unit of rank is: percent gain/day). The standard, then is a curve drawn as a compounding of 240 days in the year (the exponent used) using a daily profit of the average gain of the better stocks whose number is equal to the streams of investment the trader is running. A typical % would be between 4 and 7 %/day. In any interval on the curve (say it is done weekly), there is an impact caused by the market current yield. When long and short trading is considered, the balance of profits per day is affected. Greater %'s make up the short portion of the contribution. The trader of indexes, intraday, sees the cuurent market yield performance in terms of the daily range and the portions of the day that are determined by the differing potentials as the days goes along. I use four levels of differing potential within the day. The curve of the potential through the year is determined by compounded segments. As a nominal valuation of the curve, I use periodic dumping of profits (weekly) while running 10 to 25 contracts (this depends upon the number of weeks the trader has been running the method) and the base daily profit is 3 times the H-L. To get to the value of 3 times the H-L, it is necessary to be in the market during all times that the market is offering profits (this is when price is changing). The basis for determining a market orientation is to continually evaluate what the right side of the market is and, further to be in a trade on that side of the market. This is not a well documented orientation in terms of the historical record of people doing this. It probably is foreign to those who use conventional wisdom as you do. Conventional Wisdom is not going to change soon. Nevertheless, as you say, dealing with the facts of the matter is important and it does require dealing with the yield performance of the markets as the absolute reference for any measures of trader performance. The conventional wisdom standard for traders is often determined by a relative non functional standard. Your work is an example of this. You are not able to work with any values that deal with things that are going on day after day in markets. Were I you I would not be concerned by my comments. It is way past the time when you were able to consider how things work in markets. You are tied to what you call facts (not opinions) and those facts are a set of facts that relate to a convention that is broadly and widely used. At 73, I have learned that making money is most closely related to the money that is available to be made at any given time. What is available is the standard which I require myself to use to measure consistency and success. It looks like you use track records for sometning to draw conclusions about other things. You are not looking at anything related to either consistency nor success.
How do you explain Jim Simons' Renaissance? Aren't they the very successful example of pure math and quant?