Discussion in 'Trading' started by shanoballs, Aug 20, 2006.

  1. <div style="text-align: justify;"><span style=";font-family:trebuchet ms;font-size:100%;" >Randomness:</span>

    <span style=";font-family:trebuchet ms;font-size:100%;" >How do I know that my trading performance is not just due to luck? If the last 50 trades I took had made money for me consistently, then should I worry about any random component that might be contributing to my performance? For example consider the following equity curve that was generated by taking random trades with fixed holding period of 5 days.</span>

    <span style=";font-family:trebuchet ms;font-size:100%;" >I’m making the computer pick a number between 0 and 10 and if the number is higher than 5, I go long, if the number is lower than 5 then I go short. Here is how the equity curve looks on a single run of this strategy with a fixed holding period of 5 days:</span>
    <center><img src="http://img226.imageshack.us/img226/7411/random1wq8.gif" /></center>
    <div style="text-align: justify;"><span style=";font-family:trebuchet ms;font-size:100%;" >Now, take a look the first square. For roughly 30 or so trades the strategy appears to be making consistent money. The curve is rising on a nice upward slope. Also notice the second square where the strategy makes money consistently for roughly 50 or so trades.

    What if I were to SELL the signals of this strategy to someone (without disclosing the actual logic), preferably a retail trader, right around the time when it took the 100th trade? The person would have taken the next couple of trades and thought about what she is going to do with her first million in profits. Most likely she would have called me and thanked me and even advised her trading buddies to seek my services.

    What if a new trader started trading and was met with a nice winning streak as shown in the first or second square simply due to luck? We know that it is possible for this to occur, and what would the new trader attribute her success to, her strategy?

    Yet there is no edge here, absolutely nothing backing these trades aside from randomness. Other than speaking for the importance of the number of trades taken (sample size), I am not certain on how to interpret these results given that I only have a mild understanding of statistics. However, if I had to guess,<span style="font-weight: bold;font-size:100%;" > </span><span style="font-weight: bold;font-size:100%;" >I would say that there is at least a slight</span><span style="font-weight: bold;font-size:100%;" > </span><span style="font-weight: bold;font-size:100%;" >random component in the financial markets</span><span style="font-weight: bold;font-size:100%;" > </span><span style="font-weight: bold;font-size:100%;" >and anyone can easily be fooled by it</span><span style="font-weight: bold;font-size:100%;" >.</span> I think if we accept that as a fact, then we would have to be very cautious when we are looking at performance reports. This means that we cannot come to any conclusions about our methodologies based on how well they perform, particularly if the strategy only has a couple of trades under its belt.

    I think at this point it is very clear why it would be impractical for a new trader to use a discretionary approach to analyze their trading strategy. For how many trades can a trader apply a strategy by hand when it is losing money trade after trade? How many trades would it take for someone to lose interest? Yet, without doing so, until they have a large enough sample size, they would not be able to tell if their performance is due to complete randomness or not. Any conclusions that they come to, based on the performance of their method, would be completely meaningless. I’m just guessing here, but this might be the reason why there are so many stories of traders who make money for a good period of time and then one day blow up and lose everything. Perhaps they were getting lucky the entire time and one day their luck just ran out. So I guess this would mean that my very successful trader friend could be a complete moron that just happened to have gotten lucky. But I probably shouldn’t try to tell him that.
    </span><div style="text-align: right; font-style: italic; font-weight: bold;font-family:trebuchet ms;"><span style="font-size:100%;">