Markus,

I support the idea of taking pauses between trades (to relieve some of the excitement or frustration - depending if the streak was winning/losing).

However, I'd like to respectfully disagree with the logic you provide here. Elementary statistics' main rule is that "the next toss of a coin is still a 50/50 % chance of heads/tails" - regardless what were the previous 9 tosses. It seems that you are suggesting that if I had 9 tails in a row, my probability of getting a 'head' on the next toss is now more like 90-95% - unfortunately, that is not the case.

Likewise, if the win/loss expectation of the strategy is 60%, it will remain 60% for each new trade, regardless of the previous ones' outcomes. Thus, there is no reason to stop trading given to us strictly by statistics.

However, what happens in reality, since we are all humans and not machines, is that streaks of wins or losses tend to make us emotionally "wound up" (either with positive or negative emotions, but clearly not "emotionless"). I do believe that the chances of winning become skewed towards negative from the theoretical 60 %, precisely because we are not being completely free of emotion when considering the next trade. Thus in my view, the reason to take pauses between trades has little to do with statistics, but it is to allow the mind to calm down and find its emotional balance, so as to best prepare for the next trade.

Sergio

By the way, I was able to catch a nice move in the G6E that day worth $1500 ... so the $1000 was more like $2300 for the week.

Very interesting post and I will take what you said into consideration. I am of the opinion, that when you are HOT you should continue to milk the cow. Very interesting point of view, Markus. I agree with you on some parts, but if you are not overtrading, and are following your trading plan ...

Is this being Greedy?

David AKA Tiger

Is your thesis based on the emotions of trading or statistical facts??? If you're implying that it is statistically based, you either never took a statistics class or were sleeping when they covered the basic parts!

If trading would be pure statistics, then we could simply program computers and let them trade for us. In today's webinar "Conquering Fear and Greed" we talked extensively about the emotional side of trading.

First of all, let me make one thing clear:

Statistically it is correct that the chances of getting tail in a coin toss are still 50%, even if you tossed the coin already 9 times and got 9 times tail. This assumes that the coin doesn't have a "memory". 

The markets are different. Prices are not dictated by "supply" or "demand"; prices are driven by the traders' emotions. An economic report itself does NOT move the markets, it is the reaction of traders to the report that let prices fluctuate. 

Therefore you can not approach the market with pure statistics and mathematics. Many people tried - and failed. Again, if the price action could be determed by mathematical models, there wouldn't be any more traders. We simply would let computer trade against each other, and prices wouldn't move at all since all computer models would eventually use the same information and algorithm.

I believe that trading is more similar to playing blackjack than tossing a coin. You can not change the odds when tossing a coin: It's 50/50. But we all know that it is possible to put the odds in your favor on the black jack table.

Feel free to prove me wrong. But based on my experiences one can improve his trading performance by setting weekly goals and avoid overtrading.

I think the idea is good. One have to stop if the goal is achieved. To reward her/himself, evaluate, and not to get into a trade exited, which is emotion. With emotion comes fear and greed, so he/she most probably will NOT follow the strict rules, so he/she will sole most probably.

The explanation is not so good. If we go further on, we could deduct that after 6 winning trades there are only 4 losing trades left, so let us take the system's trades reverse, and that way we will have 10 winning trades :)

Another analogy: if you play with a dice and number 6 comes out 1:6 chance, then if you want to get a 6 again, skip the next five turn :)

Be carefull however, because if you play russian roulette, and the first 5 shoot was a winner, the sixth will be lethal for sure!

Regards
Endre Zerenyi

Markus,

As a potential customer who is considering signing up for your course I am concerned when I read this post. Either you are misusing the statistial results of your backtesting (or trading results), or you have left some essential information out of your post.

The 6 out of 10 ratio I assume was the result of hundreds (ideally thousands) of trades backtested, or better yet trades executed. Unless you take into account the distribution of those trades and show that the results were non-random within the entire pool, your logic is seriously flawed. ie. How many winners in a row has your system ever produced? How many consecutive losers? What was the win/lose ratio for each block of 10 trades? Why would you assume the gross distribution of winners/losers will translate down to a smaller sample size of 10 trades?

In theory, statistically each trade individually and independent of the trades preceding it should have a 60% chance of success. The market is not like cards (blackjack) where you know what cards are in the deck and which have already been played). So unless you can show non-random trade distributions that prove you will only (and always) get 6 winners & 4 losers each week I don't understand the point of your post. You could just as easily get 8 winners this week and 8 losers next.

So, again unless you can demonstate the results of the gross data pool can be non-randomly applied to smaller subsets with consistent accuracy your basic premise is false.

I am glad to see all these comments. This entry is definitely the most controversial entry in this blog and I feel that I need to clarify some statements.

If you are looking at your backtested results, then you MUST consider the trading system parameters. 

Example: If you backtested your trading strategy over 200 trades under the premise that you only trade the first two hours of the day, then you can NOT extrapolate these results to 24h trading. As you know, the markets have nice moves in the first two hours of trading but remain in a narrow sideways range during the night session. If you would continue trading all day long then you will most probably get different results.

The same applies to trading on different days during the week. If you usually traded your strategy in the first three days of the week, then you can not extrapolate your results to trading all week long.

And if during your backtesting you stopped trading for the week after achieving a weekly goal you might see that the results can be very different for this type of strategy vs. trading the markets all the time.

In addition you need to know that trading is about de-selection. It's about AVOIDING trades that are not 100% according to your rule. It's about avoiding OVERTRADING and emotional trading. And that's why you should use a weekly goal.

You will see that often a trader takes trades that are not according to his rules because he thinks that he MUST trade. "I MUST make $300 per day". But when there's nothing to trade, then you need to sit on your hands. If you abandon your trading plan, you will start losing. 

Back to the presented approach: This weekend I was invited by Intershow and the CME Group to speak at the Forex Trading Expo in Las Vegas. Friday night I was part of a panel of experts who discussed - amongst others - fully automated trading systems. All of the experts agreed that fully automated trading can NOT work. It takes a trader to profitably trade a strategy, not a computer.

Tim Morge, who has been a floor trader on the CME and manages cash forex positions in excess of $2 billion US, summarized it very well: "The markets might not have a memory, but traders do. And the prices move because of the action of the traders." Therefore you can not eliminate the psychological and human aspect of trading.

Though the above mentioned approach might not be based on traditional statistics, it IS the right way to trade the markets. Traders who tried to approach the market with pure logic and mathematical concepts lost their money quickly. Since computers became more powerful in the early 90s, many tried to develop automated trading systems, some even based on Chaos Theory and Neural Networks. Today, 18 years and probably trillions of research dollars later, we still have TRADERS running the big Hedge Funds around the world and not computers.