my journal 2

This is a discussion on my journal 2 within the Trading Journals forums, part of the Reception category; Just thinking out loud and summarizing some things I learned. There are two methods for trading that I can think ...

Closed Thread
LinkBack Thread Tools Search this Thread
Old Feb 28, 2010, 12:19pm   #400
travis's Avatar
Joined Mar 2003
principles that lead to profitability

travis started this thread Just thinking out loud and summarizing some things I learned.

There are two methods for trading that I can think of:

1) Automated trading. If your systems are back-tested, automated and forward-tested and they're profitable, then there's no discussion needed here, as they most likely will work in the future. If they don't, you fix them until they do.

2) Discretionary trading. Here things are much more complicated because you don't know what you will do exactly and therefore you don't know if your method (since you don't have a univocal method) has worked in the past. You can only have guiding principles, and they are as follows:

1. longer timeframe and larger profit targets: commissions and spread costs may cause you unprofitability: increase your timeframe and the length of your trades, until those costs are irrelevant. Then you will have a 50% chance of being right on every trade.

2. stoploss (small): if you're always betting your whole capital on the next trade, this will first of all affect your judgment, but most of all, you sooner or later will incur into an accident that will wipe out your account, no matter how good you are. Since leverage makes things more complicated (your margin may be a fraction of your capital, but you may be still risking your whole capital), what should be said is that you shouldn't risk more than a small fraction of your capital on each trade. In other words, you should use a stoploss, a small one. Meaning that your stoploss is not when your account is blown out.

3. fewer trades. The more you trade the more irrational you get (not for everyone, but for me). When you win, you have a tendency to be more carefree on your next trade, and more likely to screw up. When you lose, your tendency is to be blinded by vengefulness. These tendencies can be kept under control more easily if you haven't traded a lot, because trading wears you out.

4. be more selective. Trading less does not mean you shouldn't monitor the markets. The opportunities are many, but if you don't monitor the markets and make a random trade each time you peek at it, you'll still be unprofitable even if you just trade once a day. On the other hand, if you make one trade out of every ten opportunities you see, selecting the best one, then you can make several trades per day. But that's going to be tiring, so you're likely to only be able to make two or three good trades per day. On the chart game this means, twenty good trades per session, after which you should stop because you're worn out.

If you follow the above principles, your edge should work out and make you profitable. If you're not profitable despite following the above principles, then it means you don't have an edge. But since it's easy to have an edge, you're probably not following one of those principles, which can be summarized again as follows:

1) elimination of fixed costs by increasing timeframe and profit target
2) limited loss per trade (stoploss which is made possible by a limited leverage)
3) lower frequency of trading to avoid irrational decisions
4) longer monitoring of markets to pick only the best opportunities

To summarize it all in one sentence: longer timeframe, longer monitoring, fewer trades, with limited losses, should make you profitable.
travis is offline  
Old Feb 28, 2010, 12:22pm   #401
travis's Avatar
Joined Mar 2003
re: my journal 2

travis started this thread
Originally Posted by FXTrend240 View Post
lol i'll have to wait until i get a dma brokerin this case
Well, I advise you to simply open an account with Interactive Brokers and focus on futures. On the other hand, this did not make me profitable, whereas you're profitable without it. I would say that I have a better weapon but you shoot better, so you hit more targets. But IB is definitely the best weapon there is as far as I know.
travis is offline  
Old Feb 28, 2010, 3:31pm   #402
Adamus's Avatar
Joined Mar 2008
mechanical system

Originally Posted by travis View Post
Sounds like a great offer, except that we'd have to figure out a way so that I don't have to hand over to you all my systems just like that, because I'd like to keep them for myself after all the work I've done on them. Let's discuss about it further, even here on the forum.
It's obvious you don't want to hand over your systems and to be brutally honest I don't actually want to wade through a heap of excel code for a system that a user on an internet forum claims 'works'. Not that you haven't built up a bit more credibility than that with your journal but I just wanted to illustrate the point.

What I would have to do to satisfy myself that something is worth trading is:

(a) comprehend the actions of the system and why its signals should have an edge
(b) code up the system in TradeStation Easy Language to look at the signals and the equity curve in different markets and market conditions
(c) run the system against all the data I've got using IB commissions, 1 tick for slippage and a tick for the spread for every trade
(d) look at what markets it works on and what markets it doesn't work on and may be optimise any parameters for different markets
(e) examine the stats - avg win, avg loss, % winning trades, drawdowns, biggest losses, geometric average trade, etc
(e) do some walk-forwards
(f) do some simulated trading

So it's probably easiest if I just ask you loads of questions.

Yet, another thing to consider is that (with a complex set of functions) I'd be trading more the systems with a low drawdown and a high return than the others, so I might end up not trading about 10 of the 40 systems, and trading 2 contracts of the best 10 systems. I am sure you know what I mean. I've got a whole sheet on my excel workbook dedicated to these money management and risk management formulas.
I really don't know what to think about that kind of approach because it strikes me as inherently dangerous when you chop and change the markets you trade like that.

I have a basic fear that if you have a system that works on one market but not on another market, and you start trading the market that works, aren't you effectively running the risk that the market you trade might start behaving like the market that doesn't work? Perhaps a stupid thing to believe, a bit irrational maybe but I never found anything to allay that fear.

For example, one column of formulas makes sure that all systems that are trading have not lost over 5% of present capital in their worst trade. Say I got an awesome CL (oil) system which has a worst loss of 2000 dollars (which could be the case). Even if it had the greatest profit (Return On Account), I will not allow that system to trade until I'll have 40k (the worst loss of 2k has to be only 5% of the account). Then I have other systems with a biggest loss of 200 dollars, and those will be able to trade up to 10 contracts with the same margin. But then I've got another set of formulas that keep the maximum capital per system to a maximum of 33% or so. Then there's another set of formulas that decide the amount invested based on ROA (return on account, which is the profit in forward testing, divided by the biggest drawdown in back-testing, or forward testing if higher). The objective would be to use all capital at all times, but respecting these rules at the same time.
That sounds pretty much standard practice from my POV as well. However I don't trust the biggest loss to be simply the biggest loss that occurred in back-testing. I have a look over the whole history of the future (is that an oxymoron?) for the biggest gap and generally take that as my largest daily loss, and then multiply it by the average number of days for losing trades. Kind of random but I like it.

What the systems trade is these 9 futures (one contract per system, as you guessed):
I monitor 25 markets or so:

Currencies: Aussie $, Can$, Euro/$, BP/$, Yen/$, Euro/Yen, BP/Euro
Indices: SNP, FTSE, DAX, Nikkei, Hang Seng
Bonds: US TBonds, Gilts, Bunds
STIRs: Eurodollar, Short £, Euribor
Ags: Cotton, Wheat
Softs: Cocoa, Coffee, Sugar
Metals: Gold, Copper
Energy: Crude, NG

Your advice and ideas are welcome. I've heard similar offers from a couple of other people, but until now I haven't found any way to make this work. Years ago on elitetrader another user asked me for my systems and told me he'd share the profits with me. I was in a similar mental and financial state as now (desperate) and I stupidly sent them to him (LOL). He disappeared and I never heard from him ever since. Not that he disappeared. He said they weren't good enough, and I haven't heard from him since. I don't know if he's using them right now or not, but luckily over a year ago the systems weren't as good nor as many as now.
I guess that person probably just couldn't do much with your system. Although he might have stolen half of your ideas. It sucks that you didn't get continuous dialog to satisfy yourself about the trader's actual situation re your system.

So basically what I'd like to know is more about your system's back-testing results. Or in fact forward-test results. Have you got any forward test results? Or simulations?

And btw, what time frame does the system work on?
What matters most is how well you walk through the fire.
Adamus is offline  
Old Feb 28, 2010, 4:18pm   #403
travis's Avatar
Joined Mar 2003
re: my journal 2

travis started this thread I'll reply as I read, in order not to forget anything.

By what I am reading you're not asking for my code, but you are nonetheless asking to know exactly how my systems work (so you can back-test them), and to me that represents exactly the same problem. I don't mean to be rude or offend you. I am giving you the same answer I gave to other people before.

You're right about that idea that I should not stop trading the systems that don't work or increase the capital allocated to those that work best, if I did that with a short forward-testing period. But it makes sense to do so if they don't work for a period of 7 months of forward testing (dozens of trades per system). After all, that's the principle by which we discard systems that don't work in back-testing. We don't keep them and say "they didn't work but they might start working", just as we don't discard those that worked expecting them to stop working. We assume that what worked will continue to work and viceversa (given a reasonable test period).

Here's the equity line of all my systems put together, with 1 contract traded per system, regardless of how good that system is. You'd need about 20k to trade all systems together, but then again you could incur its worst drawdown (cfr. fall from 31k to 17k) and lose all 20k immediately, so it's best to start with 40k.


To trade the systems with 1 contract per system, you'd need the capital needed to trade them all (20k) plus the capital you can lose at their worst drawdown (14k), so you'd need about 40k. Unless of course you're counting on getting lucky and not incur the worst drawdown immediately.

A more likely picture is weighing the contracts according to each systems' maximum loss. I used this formula to adapt my equity line to a number of contracts allocated according to the each system's maximum loss:
where the K column is the column of the maximum losses for each of the 40 systems, and an arbitrary 5 is the highest number of contracts that a system can trade (to avoid problems due to lack of forward testing data).


The systems this way have a highest drawdown of 15k, which is close to the other way, but which happens when they're at 46k already. This 15k drawdown means 15% if you start with 100k, and 25% if you start with 60k (which is roughly what you'd need to trade all the contracts needed). That's still almost 100% per month, if you can stomach a potential minus 25% in your capital. If you can't, then we need 100k, and the system will be much less performing.

I don't remember what an oxymoron is, but I remember a movie that talked about it (last part of this video):

Interesting: the markets you monitor practically include all the markets I monitor and that my automated systems trade.

As far as the forward tests results (unlike the systems), I can send them to you, because they're not telling you exactly what my systems do. Please send me your email address via a private message in order to send you that file (also, I'd like to continue this discussion via email). I'd have a problem with sending you the back-testing results, because then we'd be only a small step away from giving you the formulas. Besides, in most of my tradestation reports (exported in excel format), I included my code, so I'd have to open dozens of such reports and get rid of the easylanguage code.

In fact, I realize you can't do much without the formulas, and yet you'll agree that you'd get the same answer, were you to ask anyone else for their systems. So I guess this explains why this thing cannot work, as I said from the start. I said I was skeptical, because I haven't found a way to make it work before, except by getting ripped off, giving my systems and getting nothing in return.

Sure, if anyone offered to send me money, I'd accept it, but it wouldn't be convenient for them - so I wouldn't accept it, and I actually refused it several times before. I'd accept it if it were a (non-refundable) gift, most likely coming from my father. Sure, if I offered to send my systems, anyone would accept them, but it wouldn't be convenient for me. The signals would be something in between, but it is not practical, because I don't have the software infrastructure for sending them, and I don't want to have to build the reputation nor go through the paper-work needed to sell signals. So, nothing seems to work. I could even take a loan and do this myself, but I don't want to take that risk, and as a consequence I couldn't even accept capital from anyone else (since I don't even want to risk the money from a loan). I guess I'll have to keep doing this all by myself, while at the same time answering questions from users like you, who help me simply by asking me questions, because they make me think about things. For example, many posts ago, Weighbridge helped me solve a problem in calculating the maximum drawdown via an excel formula, as opposed to computing it manually each weekend. It was a great improvement, which wouldn't have happened without this forum. So, thanks to all of you for writing on my journal.

Your last question: the systems trade for several hours: from a minimum of 3 hours to a maximum of 3 days.

Last edited by travis; Feb 28, 2010 at 8:34pm.
travis is offline  
Old Feb 28, 2010, 8:03pm   #404
Adamus's Avatar
Joined Mar 2008
re: my journal 2

I know where you're coming from but I guess I expressed myself badly. I don't actually want your raw backtesting results or your forward test results, especially not in a format I'm unfamiliar with. I was just asking what you had.

I just want to find out what sort of results you've got. The interpretation of backtesting stats is totally affected by your/my perception of risk. You might have results that convince me to get together with you to trade, but I think you're more likely to have results that I won't like and it'd be interesting to find out why you like them when I don't.

How someone else interprets a set of trade result statistics is interesting. I often suspect I'm falling into the trap of searching for the Holy Grail and rejecting trading systems I should be taking risks with.

I also didn't realise you had TradeStation and IB TWS running! And you're still using that chart game. You could be doing simulated trading on TWS! Man oh man.
What matters most is how well you walk through the fire.
Adamus is offline  
Old Feb 28, 2010, 8:41pm   #405
travis's Avatar
Joined Mar 2003
chart game vs. IB paper trading

travis started this thread Well, as I wrote before, TWS paper trading might be better than the chart game, in that it's much closer to my real trading, but it's so boring that I won't practice with it. So, since I want to practice but find TWS paper trading too boring, I do it on the chart game, where speeding up time makes it acceptable for me (but i still have a hard time learning to use the stoploss and patiently waiting for the right opportunity, even when "patiently waiting" only means I have to click a few weeks ahead). It's like for weight-lifting. If you won't go to the gym because it's too boring, you might as well look for another activity which is not as good but which you will do for sure. As we say in Italy, "il meglio Ť nemico del bene", "doing things better is the enemy of doing things well", which means that you shouldn't be such a perfectionist as to not do something at all. Each time I set out to practice on IB TWS paper trading account, I quit almost immediately because of how boring it is. Whereas the chart game (despite having similar problems as I do in real trading: impatience and not cutting losses), I can play for hours per week. So, it's better than nothing.

I guess what makes it impossible for me to practice on IB paper trading is that each time I really think a trade is good (with fake money but prices in real time), I feel like doing it with real money. So on that paper trading platform, I just screw around (and lose), or if I see good trades, I immediately move to the real platform and place them with real money. So I basically can't practice. I guess this is why Splitlink said here that paper trading doesn't work for him (instead it works for me, but only on the chart game, because it's not in real time, and also because you can speed it up).

Anyway, it's better to move ahead by one small step (chart game) but nonetheless move forward, than to stall because you want to make a huge step (TWS paper trading) but can't. I got into trading only because I want to make money, and not because I have to practice some virtues and do thing perfectly. And, accordingly, I am trying to get that money in the fastest possible way.

I applied this same principle in my trading systems. If I wanted to do them perfectly, I never would have started, but I've always followed the principle of taking one step at a time, one little improvement, as long as I kept going, never undertaking efforts too big to finish, because then I might have quit.

Of course, if I was in a good trading school (I don't know if there are any), and if I trusted the teachers, I'd follow everything they told me (even things that don't make sense to me), and then I might do things your way. If you told me there's a guarantee of success by doing IB's paper trading... and followed my progress step by step until I reach success, I'd do what you're telling me and use IB paper trading. But everyone on forums tells you "do this" and "do that", and then they disappear. So I might as well do things my way, as long as I'm seeing even the smallest improvements, which I am. By teaching myself things, I am like an ignorant professor teaching a student, but we've got continuity: the professor and the student will always be there. I don't trust trading schools, at least not those available online. And mentors, I have never met any. I might be teaching myself nonsense, but I'll be here tomorrow. These people on forums telling you "you should do it this way" will not be here tomorrow to tell you what the next step is.

Last edited by travis; Mar 1, 2010 at 8:55am.
travis is offline  
Old Mar 1, 2010, 12:06pm   #406
Adamus's Avatar
Joined Mar 2008
Re: chart game vs. IB paper trading

I am slowly beginning to see the method in your madness.

Your lack of self-control when you make a good trade on the TWS simulated trading platform is quite surprising really. You must be the polar opposite character type from myself. I'm going back to read your previous post because I see you put some graphs in there since I replied to it!
What matters most is how well you walk through the fire.
Adamus is offline  
Closed Thread

Thread Tools Search this Thread
Search this Thread:

Advanced Search

Similar Threads
Thread Thread Starter Forum Replies Last Post
say hello to my little journal gozinsky Trading Journals 13 Sep 27, 2011 9:37pm
my journal travis Trading Journals 2334 Jan 18, 2010 11:57am
The EA Journal! Victor90 Trading Journals 1 Oct 28, 2008 12:47am

LinkBacks (?)
LinkBack to this Thread:
Posted By For Type Date
Science Forum - why Standard deviation is better than Avarage Deviation This thread Refback May 22, 2010 2:34am

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)