All my thoughts that I bring here are my own. I donít say they are original, all Iím saying that I didnít read or heard them. In some cases it was the opposite. Some one gave a statement and after I thought about it I decided itís wrong.
I encourage you to think about it and make your one opinion.
I put some questions in this essay and they are great questions. By great I mean I got the answers, but try to solve them to your self.
1. What is trading?
a. Itís a huge casino. You place your bets when you look at some indicators or price action. You win or lose and move to the next trade. Now you place your bet on some different criteria. In the long run you know who wins in a casino.
b. Itís a huge poker table. You are more serious, you calculate some probabilities, you make research, you try some setups in sim before you commit in real money. Itís not for me Iím a lousy poker player, but there are professional poker players who make money. As a trader you can discus your strategies with pals, but when you are in a position you shouldnít discus it with any one, other wise you will credit yourself for the wins and blame your pal for the lose.
c. Itís a card counting game, like Black Jack. You know that you can be profitable in Black Jack, by counting cards. The problem is the casino will throw you out. There is a difference to trading. First no one gives you the rules when to bet, and not like in Black Jack with 6 packs of cards that your betting probability is correct, in trading the probability is correct for past period and unknown for the future. But any one who relies on technical analysis believes that what happened in the past will happen in the future. This type of trading is ďSystem tradingĒ. There are lots of arguments why it is more profitable. No emotions, back testing confirms your setups,Ö Its better to have several pals developing strategies together. Each person has his strengths and weaknesses. Each suggestion can be tested and it can only improve the setup. There are traders that mutilate this method by changing parameters of their setups, run beck tests without counting for commissions and slippage and more. But this kind of trading is for me and other points that I discus here come from it.
2. What is a setup?
A setup is an objective condition for entry, a stop loss and a win exit. If you change even one of this points (entry, SL, Target) you have a new setup. By new I mean you have different back test results. My strategies are different from each other mostly on entry conditions. All my strategies have at least 5 exit conditions built in and are shared by all strategies. They are stop loss, break even, trailing stop, profit target and end of day exit. They are types, and each type has several entities and not all must be in play in specific setup. For example SL may be: 1. ticks. 2. previous bar. 3. current bar. 4. above/below MA. And more. Stop loss has two more parameters: 1. Ticks/Max Loss/% of daily range. This means that if a SL type is ďticksĒ then here I enter the number of ticks. If itís some thing else then this number of ticks determines if I take this trade. So if my stop type is previous bar of bigger interval and this parameter is 30, that means that for long trade I want to put my stop below the bar but if it will be more than 30 ticks donít enter. The same with other exits. I may have two or more profit targets in a setup, but their meaning is: exit when target1 or target2 occur.
I donít like stops and targets which are ďticksĒ. Not that I donít use them, but I donít like them. Why do you think that you have to enter a trade on some price pattern or action and to exit a trade on some number? Better place for stop and target is a price action but itís more difficult.
3. Trade statistics.
You should know that a trade setup has two ratios.
a. Wining percentage.
b. Win/Reward ratio.
So a question goes like this:
How can you compare two setups? Which is better?
Setup 1: wining percentage of 68% and 1:1 win/reward.
Setup 2: wining percentage of 48% and 2:1 w/r.
Please try to come to an elegant answer.
4. To scalp or not to scalp?
I do not scalp. Iím not saying that people can not make money this way, but I believe itís much more difficult. There is one guy with a trading room on the internet and he says: ďWhat is more difficult to take 1 point from the market or to take 10 points?Ē His answer is that it is easier to take 1 point, but after thinking on this question I realized that 10 points are easier. Here is why.
a. It is a wrong question, the right question should be: What is more difficult to win 100 trades of 1 point net, or to win 10 trades of 10 points each?
b. Lets play a little game. We will enter a market by flipping a coin. Tales will be Long and heads -> Short. We will make our stop equal to the target. So 1 point stop and 1 point target against 10 point stop and 10 point target. Those both trades have same loss/reward ratio right? It is 1:1.
Let examine the scalp trade on the ES. Lets say the price now is 900 and we want to go long. We hit market order and in 99.9% we will be filled 900.25. We place our stop 1 point behind at 899.25. For as itís a 1 point stop, but from the market perspective itís a 0.75 points stop. Now to take 1 point profit the market has to go to 901.5 and if we hit market sell we will be filled with 901.25 which is our goal of 1 point profit. (You can think of placing limit orders it will be the same). So from market point of view our trade was 1:2 win/reward and not 1:1!!!
5. To scale or not to scale?
I will cover here one type of scaling, but they all are same to me and all are explained the same. So letís talk about scaling out. I think that psychologically it is O.K. to do it, but logically itís wrong! I want to explain it with a little story.
There was a great professional trader and he was a nice guy. He traded in a trade room with some trainees. One day two of them, Tom and Jerry, come to him and ask if they can show them how he trades. He had a lot of money so he traded 2 contracts each time. Tom and Jerry had money to trade only one contract. They thought how to follow him if they can not trade exactly as he. After a wile that had an idea. Jerry who was not so brave decided that he will follow the pro but exit on the first target and Tom who was brave decided that he will place same trade with same stop loss but exit on the second target of the pro. So thatís how they traded for a period of time. The pro announced when to enter a trade, where to put a stop and what are his targets. Jerry took the small target and Tom the second. After a period of time the Pro ended winning 10K.
Question: 1. How much Tom and Jerry won together?
2. What is a probability of each winning the same amount?
Without knowing this setup I bet that it is more likely that Jerry end up with 1K or even with -2K.
The point of this exercise is to show that a trade with a scaling can be separated in two trades and each can be back tested on its own. And let the best trade win.
6. To Optimize or not to Optimize?
I think that optimization on its own is wrong! The optimizer run maybe millions of iterations and finds the best parameter values for a given data. The proper use should be to see how whose values affect the future. It is called Walk Forward! Without it its useless. Donít believe me try the SampleMACrossOver in NT. You will get astonishing results. I run for half a year and got Net Profit of 11K+. So maybe itís a holy grail?
7. Optimize on..
Because the usage of an optimizer is not to get the best result on a given data set, but to predict the best result in the future, I donít know what the best parameter for the optimization is. On the other hand if you have a talk with God and he told you that what have been will be duplicated in the future, then I think the best parameter for optimization is an answer to the following question.
You have two setups. Setup 1 gives you 40K net profit with 6K drawdown.
Setup 2 gives you 10K profit with 1K drawdown.
Which is better? Why?
Thatís all for now. I encourage you to contradict me I will appreciate it. I prefer to eat my hat and not lose money because of wrong conceptions.
The following 7 users say Thank You to baruchs for this post:
I won't contradict you, but I will play along for the sake of a discussion since you spent a lot of time with a well thought out post that I enjoyed.
1. This is true to a degree but there is a big difference. Casino is fixed odds(sportsbetting aside for the moment)
For example in poker we have 2 spades in our hand..
2 spades hit the flop.. we know we have roughly a 36% chance to make the flush..
Thus we need to see 3.5 or 4 to 1 to make the call to make it worth our wild.
This is a fact, this is static, forever until the destruction of earth this is true. So with that we can figure out our EV and never have to worry about this again.
Trading you know nothing.
You can uptest, forward test, down test, or back test it will never be forever true that you will hit X% of your trades. It gives you an idea but the reality is you never know. It's dynamic thus making it different from casino games and closer to sportsbetting.
In sportsbetting we have odds that are not universal. We can come up with formulas or whatever but the fact is you have to beat the oddsmakers and the oddsmakers make adjustments, thus so do you. True you get the posted odds, but as far as coming up with a universal method of calculating the "correct universal odds" there is none. Thus it's dynamic bookmakers and bettors alike make adjustments to get the best "number" they call it bets that have "value".
What are you doing when you trade?
You take trades because you think it has value in that area.
Hmm value in that area... where have I head that before
3. I understand where you are going
but as stated above your 48% is not absolute. Next year your 48% could be 68% and vice versa, you don't know that it won't and you can't know that it won't.
4. If I told you I make 3 trades a year and my average hold time is 7 months, then that would make everything probably look more like a scalp. It's irrelevant to put a label on it.
Again we can talk R:R all you want, fact of the matter order flow change I'm getting fvck out .. I don't give a damn what happen in the past on that setup, or on that R:R ratio. Guy in front of me is slowing down , I can say well it shows 75% of the time it I don't have to slow down and I'll save gas by not doing so. Yeah ok your world.
5. " I will cover here one type of scaling, but they all are same to me and all are explained the same. So let’s talk about scaling out. I think that psychologically it is O.K. to do it, but logically it’s wrong!"
You say you think that psychologically it is O.K, well that's it, end of conversation. Personal preference and sleeping better at night is normally better than math in this game. Sorry, and I'm a math guy.
I will give you 2 million or you can have 10% shot at 500 million. Mathmatically what is better? Yeah ok, you know the answer but fact of the matter is 2 million ain't nothing to sneeze at! Fvck math, I'm taking the 2 million and sleeping good.
6. Optimize? Optimize what? You have no clue what tomorrow will bring in this game so you are optimizing on past results that may or may not be true next year or for that fact tomorrow. Take a look at the indicator game, tweaking, changing, setting, etc...... you'll kill yourself going crazy with that. Auctions are dynamic why try to apply static things? If they were that static, it wouldn't be any trading happening.
Anyway, thought you had a good post and was bored so thought I would make a little conversation with you. Seems like you are well on you're way to being successful in whatever you choose to do in life.
I liked your thinking and I understand that your arguments were just for the sake of a discussion so here are my answers.
1. I said already that I don't play poker, but I like to watch it on TV. As I understand the player doesn't play his hand, he plays his opponent’s hand. You can have pocket Aces and each hand you will play them differently. Same in discretionally trading, you are influenced by what you did before. Did you have several wins or loses prior to the current bet. Did you have an argument with your wife etc.
It's not exactly true that in trading you know nothing, in regards of a discretionally or system trader. They believe in technical analysis. Technical analysis says only one thing: There is a positive probability that a price pattern or an indicator behavior that happened in the past will happen in the future. If you don't believe in it then you think that technical analysis is crap.
3. I didn't intend to say that past statistic is a guarantee for the future. Still if I tell you that I have a great strategy and I want you to trade it. You ask me what the becktest results are, and replay that in past 6 month you would have lost 20K, but I strongly believe that it's a best strategy in the world. Would you trade with real money? The point is that although we don't know what the future will bring, we trade on results of beck testing. Still you didn't solve the question.
4. By scalping I mean want to take up to two points in a trade as a setup.
5. I would take 500 million. If I lose than next day I go for 500 million again and again. Now tell me in 100 days who will have more money? I hope you didn't mean to say that in trading you make one bet and retire to Bahamas.
6. My point was that by optimization alone you achieve nothing, but by Walk forward you know that in the past your optimization worked, and you believe it will work in the future.
The following 2 users say Thank You to baruchs for this post:
1. Yes no limit poker is slightly more comparable to trading.... Limit poker less comparable...
Your analogies were great, I just wanted to give an example that was more closely related to supply\demand, which sports betting.. book makers move the lines based on the order flow from bettors....they don't want to be to heavy on TEAM A .so they adjust the spread..... just like the locals don't want to be too "long" or too "short" so the adjust
Correct it's not that you know nothing.. it's that you know nothing about your expected results compared to the casino games.. you can't figure out what your EV is... you can't because your results are not static..their estimates..
3. "The point is that although we don't know what the future will bring, we trade on results of beck testing."
No not everyone mate... not everyone
Correct, I didn't answer your question, it's on T for me to knock out the park--
... no need to answer it.
Actually you left out the biggest bit of information for me to make a decision.
How many reps does the 48% system get in? How many reps does the 60% system
If the 60% system gets 1500 trades per year, but the 48% system gets 300 a year, well then things change now don't they?
5. Yes you will have more money.. no doubt, if you get the other 9 chances.
But who says you get the other 9 chances to hit for the money?
You just assumed that you will.
Why because we are talking about %'s and theoreticals.. ...
And that's the overall point I was trying to make.
Which I'm sure I didn't because I'm not a very good orator.
Everything you are saying is correct, but don't get too "boxed in" and trapped in your thinking. The market, life, everything is not a test lab.
Don't fear uncertainty ...embrace..enjoy it, use it to your advantage to extract money from the weak.
Sometimes I can't answer why I did that and went against the "odds"..
For the persentage question if you want to compare the results as they were and not make a prediction in to the future, it doesn't matter how many trades each setup makes. For the argument sake they bouth have 434 trades.
On the matter of 500M, you gave me the statistics and again if you don't believe in statistics, you don't believe in technical analysis. If you gave this example as a one time bet, then it has no relevance to trading. In trading you will win and you will lose. The target is to win more than you lose.
The following user says Thank You to baruchs for this post:
I have a feeling you need to optimise againts recent days, as market has its own mood, and keeps that mood for certain period, if optimisation is capable of catching it - you have winning strategy. Of cource it could be a transition period, and then you fail.