easy money management advice?

Wow, this last one was like a very long paper for me, so thank you. I will have to digest it a little bit at a time. Also thanks for the books in the earlier post. That will take even longer to process.

Knowing myself I doubt that I will read them though - I wanted a quick and simple answer, otherwise I would have aced the formulas already. That's right I don't feel I have that much energy, or call me lazy. Which brings me to a question to the author of the paper-post. Why do these people he talks about keep failing? I can tell you that if they are like me, we might fail because we think that after all we should have a 50% chance of winning without having any knowledge of anything, just randomly trading, so how is it possible that we have to put all these efforts into money and risk management on top of all the efforts to observe the markets and having built, tested and coded several trading systems? I mean - when is it going to be enough work before I can make some money, considering I have a 50% chance of winning to start with?

Thanks for all the advice, though - it helps me reason with basic premises, and I hope it also helps everyone else the fact that I am asking everyone to keep it simple. Sometimes people mention things that not everyone knows about, and the thread doesn't go anywhere. Maybe asking you for the concepts of money management rather than the famous formulas (kelly, and so on), will make the thread more useful for everyone.

Sometimes it's even easier for people to say "use this formula", than having to explain what the basic principle of that formula is, or their opinion on the principles of money management.

I figure it's better not to use any formulas at all, rather than using them without being able to control what you are doing completely. Unless I fully understand them first, I won't use them. So far all I can say I am positive about is that 1 + 1 = 2 and similar math. Even the RSI formula is too complicated for me so I never used it.

Oh, by the way, I am still reading your long post as I write. Now I came across the "fixed fractional" sentence. My idea (excuse me for daring to express opinions on this with all my ignorance) is I will most definitely use all money all the time. I don't agree with any theories or formulas that are telling me to invest only a part of my account, because first of all they will slow me down like crazy, and I want to become rich quickly, but because most of all, if I diversify enough (and that is my objective), I will be able to re-invest all my capital all the time (including gains) without risking financial ruin. That is why I built 20 systems. Of course I would have survived so far if I hadn't re-invested everything, but also I wouldn't have been able to get to 25k from 4k twice. My point is that I want to learn a way to money management that will allow me to reinvest my all my capital all the time. That is why I am struggling, because all the formulas I came across disagree with this idea, which I still think is right. Seriously, if you diversify enough, there is no reason why you shouldn't invest all your capital.

Now, in the last part of your long post, you seem to agree with me. Screw the books, the theories and all the technicalities. Don't trust all these formulas and find your own way, with live trading in mind. "Energizer bunny" sounds like what I want - something that will trade all the time all the capital.

Also one important part that sounds familiar to me is this:

- Highest % wins mean nothing in live trading.
- Highest return system means nothing in live trading.
- Max drawdown is something most automated traders screw up on.

I don't want to use all these data from tradestation unless I can ace them. What if they are wrong? I am gonna fail just based on some tradestation report. I wanna use more my common sense and what i see in real trading than formulas that process data from tradestation reports.

 

I have traded with Tradestation these 12 years since before 2000I and I can tell you - you will never ace the market (I don’t mean that personally) and there are no quick simple answers – just hard work running a trading business. It's just a fact we must accept or we will probably never be able to trade automated systems. A traders job is to take what the market gives us and use our wits to retain as much as reasonably can be accepted.

An automated System must contend with markets that have characteristics of:
- Direction of price – Up, down or sideways.
- Volatility of price – High, medium or low.

The automated system you build only gets a piece of one or may be two combinations of these directional and volatility types. Then it changes to new direction and new volatility. When this change happens to you, you either shut your system down or you take the big drawdown until the market comes back. Guess what? Some time the market doesn’t come back to what it was any time soon.

No one has ever designed the magic system or the “energizer bunny” system that just keeps going and going and going. To trade with automation you have to be prepared to shift to handle all these changes in market conditions. So traders have to put damage control in their systems and business procedures in their plans to recognize when their systems don’t work. This includes manual procedures to stop the automation when you recognize the markets have changed but your software doesn’t.

Take a look at Tradestation data (that is stats) when you run a test. They are going to be wrong in live trading even if you forward test. How are you going to protect your trading business when the same stats don’t come up in live trading? You won’t find that in a book. You build it from hard bloody trading of this automation under different market conditions until you understand your systems. And there is a price to pay for learning these facts. Of the 800 or so systems I built I have 9 I now trade. I know in depth when and where they trade and under what market conditions. That’s what an automated trader must learn.

You can trade all of your money… if you can sleep at night. When your $100,000 account using max margin is trading full on and you now have a $40,000 drawdown…wheee. Then it gets tougher maintain your composure and you may drop the idea of trading all of your account next time. That’s part of learning the systems do’s and don’ts. But you don’t have to pay in blood unless you want to. That’s your choice.

To get to your own money management rules I often work in reverse. For example I set a max drawdown of 6% for a month. Then I look at the swing trading system it trades 4 trades a month. That means the most I can take on a single trades position size per trade of 1.5%. Then I run that through testing. I find average loses are 1.8% because volatility is higher. So the position size is kicked lower to 1.18% to compensate for the volatility. This is the kind tweaking process you build to get the system in line before live trading. If it does not match at the first performance review it gets tweaked again until you get it right. It is lots of work. There is no “magic” bullet from a book.

 

I understand and thanks for the advice.

I have been going through the same things lately - getting to know my systems. Until now, it was like magic (a magic that I had created), and I didn't even realize which one worked best.

Now, after a year of automated trading (with some discretionary here and there, due to an addiction I am trying to get rid of), and two financial ruins (in a relative sense, since I was starting from just 4 thousands dollars), I can grasp more or less which ones have a lower drawdown, and also, very important, which ones are correlated among one another.

The first reason of my two rapid financial ruins is that in both cases I was trading very correlated systems. They brought me to zero in just a few weeks, when it had taken me months to get where I was. Also, partly it was caused by my discretionary trading, but that's another problem (not entirely, because I may do discretionary trading because I don't feel I can totally trust my automated systems).

 

My excel file breathes very heavy by the way, having all these macros (28 systems running on one file). Does anyone know a way to make it run more lightly, other than programming it in a different language?

Wow, I've got to tell you - I just found this great link:
http://www.ozgrid.com/VBA/SpeedingUpVBACode.htm

 

Not that anyone might care anymore about this, since I admitted how bad I am. But just for future reference I want to state as follows.

I made a bit of money, and went from 4000 to 5000 so I am quite in a good mood. Also, I feel that the money I made is somehow a consequence of all this debate and thinking on money management, because it forced me to focus and find out which were the systems that performed best, then I used them, and I made money immediately.

I kept on working on my very basic rough draft of a money management excel worksheet. I am attaching how far I got.

There are no money management formulas because as I said if I don't fully understand them I do not trust using them.

What I did is setting up, for each of the 28 systems, a given (capital) value necessary to add an additional contract. This value obviously corresponds to at least the margin required by the broker.

For the trading system I am the most confident about (overnight ZN), that capital value from which I derive each additional contract is very close to the margin necessary, since I don't believe I will ever go broke using my whole capital on it. There's a remote chance I may be wrong, but I feel it's worth taking the risk. There's also a remote chance I will die before getting rich, so I'd rather take that risk to get rich a bit quicker. Of course if I lived forever I would take my time.

For the CL trading system, instead, which is based on the OIL future, and which will bring me down 2000 dollars each time it incurs a loss, that value is fixed at more than twice the margin required by the broker. First of all because of the huge losses, but also because I feel that system doesn't win as often.

Anyway, I know I may sound like a dreamer, but just as I admitted that I doubled the capital a few times and then I lost everything on two occasions in 2008, I want to predict that this time I will triple my capital every month for a few months, before going broke... (just kidding for the last part).

In the file I am going to attach, I ordered the 28 trading systems from 1 to 28 in order of my preference for them, which is roughly also their estimated quality and profitability (no formulas, or ratios here - just my estimates, which partly are based on formulas here and there).

Also, I divided the 28 trading systems in five classes with different colors: red for the most trusted and invested in, orange a little less, yellow even less... and so on (up to no color for those I almost don't want to invest in, yet).

The only value one has to change is the capital available, to the right, and that in turns tells you (and will tell me, because I inserted it in my own system already) how many contracts your excel file will trade for every different trading system (there are 28 of them, based on a total of 9 different securities).

 

Travis,

You have a good energy despite what you stated about yourself. I would like to point out the following:

1. One does NOT have a 50% chance to win. This is one of the biggest mistakes people make. The chance to win is LESS than 50%.

2. Point 1. and the fact that you have a higher win ratio (assuming that your reward is higher than your risk) suggests to me that your tests may have been made WITHOUT a stop loss.

Did you implement a stop loss in your tests?

3. Money management is really simple. Assume you are accepting to lose 20% of your capital. Then divide the losses such that on any given trade you will lose 2%, if the trade is stopped.

After the 20th trade, evaluate your capital, and repeat step 3.

4. The reason why you choose 20 trades, is that 20 is a number that is practically enough to filter out the noise.
It is like selecting 20 guys randomly, and measuring their height. The average should roughly be the average height of humans. If you select a higher number of guys, you will get a better estimate, but it will be roughly close to the average human height.

So setting a loss the same for each trade will allow you to know the average gain of a trade. If the average is negative then you have a losing method. If it is positive, you have a winning system.

5. I hope it helps.

I am writing a series of lessons on how to set the whole trading business. In it, I prove why the win/loss ratio is less than 1., meaning one loses more oftern than one wins if the trade are selected randomly, and there is a stop loss.

 

How did you guess? "Higher win ratio" you say that I said. I am not sure I was clear. All I meant to say is that my winners yield on average more than my losers (in absolute values obviously, because of course the winners win money and the losers lose money). I don't know how you gathered from this that I do not use stoplosses, but it is true - I do not use any stoploss (only on 9 of the 28 systems, the 9 I didn't even test on tradestation because they rely on the book, but those I am not trading with money yet). Anyway, I enter a trade at a given time and exit at another time, regardless of gain or loss - which means no stoploss and no takeprofit. I had to program the whole thing in vba on excel and it was hard enough to code what I coded for me. Let alone that I don't want to add too many features to my systems to avoid "curve fitting" and "over optimization". Also my easylanguage code for backtesting the same systems on tradestation is kept simple by avoiding stoploss and takeprofit.

---

Aside from this everything else you said makes sense, as far as I could understand it (not completely).

On the point regarding the 50% chance of winning, I agree with you, but I also have to explain that for practical purpose it sounds reasonable to me to still say 50% considering my broker is IB and I trade heavily traded futures with 1-tick spread usually, so I only spend about 5 dollars round trip in commissions and 1-tick in spread. Now, since my trades (when they are profitable) yield on average 300 dollars (per contract), and the costs of each trade are about 15 dollars... you know? It's like what... when you win your gain is 300-15=285 and when you lose your loss is 300+15=315... so ok you don't win 50% of the time, but randomly trading you can win 285/600 = 47.5% of the time. (Excuse me if the formula is wrong, but as I said I am not that good with formulas). So, anyway, that's what I meant when I said 50%, I really meant "slightly less than 50%".

------

Regarding the lessons you are writing... if it is free, can I have a link to your web site or similar? You know, for us lazy people I would advise not just you but everyone to not write but record lessons on youtube, with charts, tables and such... here's an example:
http://www.youtube.com/watch?v=RlRE3qtvd0Q

It's easier for the teacher and for the student that way.

 

I mean this is totally sick man...Sorry to say this, I don't mean to offend you...In another post you said you doubled your money to 25k...This means you started with something like 12K. If you have a $2,000 stop which is about 17% of your capital to start with, do you expect to make any money like that?

Just take a coin out of your pocket and flip it. Sooner or later you will see 5 heads or tails in a row. You are done...I mean done...

 

Well, as I said I am not very knowledgeable on this subject. I am aware that I don't sound like a good system trader, and least of all like a good programmer. Most of you here are rocket scientists to me.

Having said this, it is not all entirely as reckless as it may seem to you.

The CL system, which is the riskiest one, will be traded simultaneously with all the others. Also, I should mention that I will start trading it once I get back up to 20,000.

As you correctly say, if it loses five times in a row and each time it loses 2000, and it might happen (improbable though), I will go back down to 10,000, losing 50% of my capital, but it is unlikely that it will have time to lose 5 times in a row (it makes one trade every 2 days) and that the other systems won't make any money in the meanwhile (since they are using the same capital at different times, and will make over 50 trades in the meanwhile).

Then of course if you push your hypothesis as far as saying that all 20+ systems will do nothing but lose for two straight weeks, then you are right - I will lose everything in just two weeks.

But then again it's more likely that I will be struck by lightning. Or rather, in other words, I am willing to take the risk of losing once again all my capital (for the third time).

For the record, this is what I went through in my first year of automated trading - from march 2008 to march 2009.

I brought 4,500 to 23,000 in four months, then brought it down to zero (1000, not enough to trade anymore) in two months (with a combination of system and discretionary trading).

Then - with a loan - I brought 8,000 to 26,000 in one month, paid the loan back, and then brought 18,000 down to zero again a little bit later (once again with the same combination of losses from system and discretionary). On both occasions - I guess I could say - I lost everything faster than I had made it.

Now I am starting for the third time, after going in the red in my other bank account, and wiring my broker with a starting capital of 4,000, which has gone to 5,000 yesterday. Maybe I am dreaming, but I hope to at least double my capital every month and reach a million within a few months. You see, because I am always starting with just a few thousands I hope to stay reckless (so to speak) all the way to 1 million (if I ever get to it). Death is approaching and there is no time to waste.

 

You are a very courageous guy. There good things in what you are doing, but there are also some very subtle mistakes you are making. It is hard to explain in writing.

Your drawdown should never be more than 20%, because loss and gain are not symmetrical in % terms. This is pronounced if you have more than a 20% loss.

Other remarks:

1. I wrote about the 50% win probability because you wrote in one of your earlier posts that 50% is what one would expect. This is true only if you do not have a stop loss.

Test your systems with a stop loss, and you would notice that the win rate would drop very fast.

2. You need a take-off monney mangement strategy, and a cruising altitude money making strategy.

3. Could you say what is your holding time when you lose, and your holding time when you win? I am guessing that your holding period when you win is larger than your holding period when you lose, for a system that is winning.