Triple moving average...

Discussions about the testing and simulation of mechanical trading systems using historical data and other methods. Trading Blox Customers should post Trading Blox specific questions in the Customer Support forum.
illuminati
Full Member
Full Member
Posts: 24
Joined: Sun Apr 15, 2012 12:52 am

Post by illuminati » Wed Apr 18, 2012 1:46 pm

I think another drawback with having a risk limiter is that it runs counter to the idea of maximizing returns. You never know when the next ten bagger will come along and if you miss it because you have too much risk, then thats just unfortunate.

Again, its all about trade offs in systematic trading. We are all trying to optimize and find an area, set of entry exit mm rules, that we are comfortable with, but in the end of the day, we are all to certain extent curve fitting...

Mike Cautillo
Senior Member
Senior Member
Posts: 27
Joined: Thu Mar 15, 2012 2:59 pm

Post by Mike Cautillo » Wed Apr 18, 2012 2:35 pm

Ill...I just quickly ran the risk manager and plugged a 30% maximum risk threshold and in comparison to no risk manager the results came out much better....it raised my CAGR% and reduced my draw down....is this normal.

illuminati
Full Member
Full Member
Posts: 24
Joined: Sun Apr 15, 2012 12:52 am

Post by illuminati » Wed Apr 18, 2012 2:49 pm

I don't know the nature of the system but I would not arrive at the conclusion that the risk limiter is desirable because it increase MAR. You should understand how the risk limiter works with the system and if it makes sense to have it.

I spent some time last year doing this, by simply adding on risk control. On the surface, it increased profitability, but under the hood, some aspect of the system changed (trade entered exited, winning months, etc). I forgot where I read this, but I think Bob Spear somewhere demonstrated that curve fitting also applies to MM rules, not just entry exit. It is that change that you should look at and question whether its a good tradeoff. (im not saying there has to be a change as some good systems just comes from you mashing ideas together)

What I am saying is that you should know "why" it improved.

hope this helps,
il

Mike Cautillo
Senior Member
Senior Member
Posts: 27
Joined: Thu Mar 15, 2012 2:59 pm

Post by Mike Cautillo » Wed Apr 18, 2012 3:20 pm

True that...I definitely will evaluate why and how it will improve my system...always do.

Not sure is you saw one of my last post...how are you evaluating how much capital is required to start trading your system.

Thanks.

Toosday
Roundtable Fellow
Roundtable Fellow
Posts: 70
Joined: Fri Jun 03, 2011 11:21 am
Location: Austin, TX

Post by Toosday » Wed Apr 18, 2012 4:00 pm

This seems like a good discussion. I have to admit, I am a bit confused about your question with regards to starting capital. In my mind, that is really a loaded question.

Is trading your only source of income?
Do you trade outside money?
What is your slippage assumption?
How much $ do you want to make?
What margin to equity does your system trade at?
What is your budgeted MDD?
Etc. etc.

I find this to be one of the more intimate questions to answer as it really depends on your situation and risk tolerance. I may be misreading your question and please tell me if I am.

illuminati
Full Member
Full Member
Posts: 24
Joined: Sun Apr 15, 2012 12:52 am

Post by illuminati » Wed Apr 18, 2012 4:18 pm

Toosday wrote: In my mind, that is really a loaded question.
I must say i agree, its highly subjective and different for everyone.

May I add that individual traders face huge problems with implementing MM algos because of their limited trading capital. Cant really get fancy when starting capital is less than 1 million. (in my opinion) and not to mention trading multiple systems. This business is not welcoming at all. All in all, I am assuming we are only talking about futures trading, you can start out with less for stocks and forex.

AFJ Garner
Roundtable Knight
Roundtable Knight
Posts: 2040
Joined: Fri Apr 25, 2003 3:33 pm
Location: London
Contact:

Post by AFJ Garner » Wed Apr 18, 2012 4:57 pm

Mike Cautillo wrote:AFJ...are you saying that you will need to establish how and what markets are traded when the the total equity risk has been hit....if so would it not be on a first come first serve basis....random??
You have 30, or 60 or 150 markets in your portfolio. Depending on your system, you may be able to take every trade and still not hit sector or overall risk limits if you have a very large account size indeed and keep the bet size on initiation very low. Most mere mortals are likely to have to put in place risk limits which will be hit day to day so that trades will get refused because you have enough risk in a particular sector or overall.

One of the standard risk limiters in TB allows you to take every trade and then sells off partial positions when the risk limits are breached. This is expensive in terms of slippage and effort and sheer number of trades and may be difficult from a "granularity" point of view for a small account. The more useful (in my mind) of the TB risk limiting Blox is that produced by Sluggo. It refuses ab initio to take trades which would take you over the limit.

But potential trades have to be presented to the risk management Blox in some sort of order. You might choose to present trades to the risk manager in alphabetical order based on symbol; you may wish to present trades based on sectors and then alphabetically within sectors. You may wish to present the best or worst performing instruments first. In reality you can't present trades to TB on a random basis if you use TB for order generation but you certainly can do so in back testing for fun and interest.


There are so many choices to be made once you get into system design it becomes mind boggling. The longer you mess around the more choices present themselves.

Mike Cautillo
Senior Member
Senior Member
Posts: 27
Joined: Thu Mar 15, 2012 2:59 pm

Post by Mike Cautillo » Wed Apr 18, 2012 5:12 pm

Definitely a loaded question Toosday....I apologize for this but let me clarify where i was going with this....after you have performed your back test and let us assume you have plugged in all the appropriate MM rules....would the starting equity you started with be the appropriate number that would be sufficient.

I assume if the system tested all the entire length of the trial ...it would not run out of money...again assuming all the MM rules are set or are their other things you need to look at.

Ill...I agree that TF for futures is very challenging with limited resources but having said that...I feel you will be able to trade with 500K easily....you do need to be quite diligent on the MM side of things but then again shouldn't you always.

Mike Cautillo
Senior Member
Senior Member
Posts: 27
Joined: Thu Mar 15, 2012 2:59 pm

Post by Mike Cautillo » Wed Apr 18, 2012 5:27 pm

AFJ...I am testing 50 markets...but you confused me, if it takes all the trades initially and the risk limit hasn't been hit...then isn't it doing what its supposed to do to....the fact is it should not not take any additional positions if the total risk is hit...how does the standard setting in the risk manager perform the test.

I know that positions are reduced once the total risk is hit...I guess this is if the losses in the open positions are exceeding the limit.....the hardest thing thus far for me with the testing is you make a lot of assumptions mentally and tend to forget the system works on specific instructions so bridging the disconnect is crucial.

I need to learn- not to assume the system is doing anything but instead needing an exact command or looking at the command it function/rule on.

Make any sense.

AFJ Garner
Roundtable Knight
Roundtable Knight
Posts: 2040
Joined: Fri Apr 25, 2003 3:33 pm
Location: London
Contact:

Post by AFJ Garner » Thu Apr 19, 2012 3:45 am

Perhaps we are wandering off track. Yes, sure, if all trades are taken and the risk limit is not hit then all is fine. And indeed if you keep your bet size and or your portfolio size at a certain level you can so organise things that your chosen risk limits will never, or rarely be hit. A risk limiter Blox can work in a number of ways. The Blox I was referring to does not exit positions as risk grows: it merely stops you taking on new positions if the existing risk is too high. This can work well with profit taking exits which reduce risk on profitable positions.

The point I was trying to make is that almost every little alteration you make in the design of a system has far reaching consequences. Which may or may not lead to curve fitting. The Blox I was referring to will take all orders on a given day until your risk thresholds are hit and then it will reject any further orders for that day. Imagine you have 6 potential new positions: long gold, short corn, long the 3m ED, short wheat, long the Vix, short Crude. You may have room for only two new trades. Which are you going to take? Which are you going to reject? TB will take trades in the order in which they are presented as it scrolls through your portfolio. Which trades are taken and which are rejected will have consequences. By altering the order in which potential new trades are presented to the risk manager you can influence which trades are taken and which are rejected.

By definition, TB has to process instruments in a portfolio in some order. I assume you have simply left the default procedure in place. You can find details of the default in the manual.

The risk manager Blox I am talking about can be found here:
viewtopic.php?t=4389&highlight=thermal+bbbo

You can change the order in which instruments in the portfolio are presented to the risk manager by the use of the following TB functions in the following scripts:

Code: Select all


Rank Instruments Script
instrument.SetLongRankingValue( myRankingSystem )

Before Trading Day Script
System.SortInstrumentListByLongRank


Mike Cautillo
Senior Member
Senior Member
Posts: 27
Joined: Thu Mar 15, 2012 2:59 pm

Post by Mike Cautillo » Thu Apr 19, 2012 7:53 am

AFJ.... interesting piece .....thanks so much. This is definitely becoming more of the case...needing to looking at the specifics of the rule and its consequences... by the sound of it you been through this once or twice.

As I said I need to train my brain-no assumptions.

I appreciate you taking the time to help me out.

Good trading!!

Toosday
Roundtable Fellow
Roundtable Fellow
Posts: 70
Joined: Fri Jun 03, 2011 11:21 am
Location: Austin, TX

Post by Toosday » Thu Apr 19, 2012 8:39 am

Mike,

If I can offer some advice, please focus on some of the softer issues as well. You really need to spell out your trading goals. Your assumed risk and the money you would like to make from trading really need to overlap. For me personally I have been making trading decisions for years and have developed a knack for knowing what I am comfortable with. If you know yourself well enough then your decisions are easier. If you are not accustomed to risk based decisions then you may want to be more conservative until you become proficient in analyzing risk return trade-offs.

I am not trying to talk you out of anything I just have seen a large number of people who are not comfortable taking risk. To decide on starting capital, obviously the methods and stats are important considerations but I think that the 10,000 foot view is a good place to start.

Mike Cautillo
Senior Member
Senior Member
Posts: 27
Joined: Thu Mar 15, 2012 2:59 pm

Post by Mike Cautillo » Thu Apr 19, 2012 9:30 am

Toosday....I can certainly appreciate your feedback and concern. I totally understand that in this business-its all about managing your risk and that is the delicate trade off for what you want out of the markets.

The funny thing is I tend to lean more towards more risky side then conservative. I am not afraid of risk but again I know that we must control and understand our risk.

I just want to make sure when I am ready, I can at least have the proper start to give this thing a fighting chance...do you really think 10,000 is enough to start trading a TF system.

Thanks for looking out....appreciate it.

fab1usa1
Roundtable Knight
Roundtable Knight
Posts: 383
Joined: Sat May 21, 2011 8:30 pm
Location: New York

Post by fab1usa1 » Thu Apr 19, 2012 9:53 am

My experience is that $10,000 at a low-cost broker is sufficient for stocks. Not so for futures unless you plan to trade one contract at a time.

Mike Cautillo
Senior Member
Senior Member
Posts: 27
Joined: Thu Mar 15, 2012 2:59 pm

Post by Mike Cautillo » Thu Apr 19, 2012 10:26 am

How about for futures....that brings me to another point-have you ever looked at the performance of trading one future at a time and how successful this would be ...if you did have limited captial.

Aaron01
Roundtable Knight
Roundtable Knight
Posts: 113
Joined: Thu Jun 02, 2011 1:21 pm

Post by Aaron01 » Thu Apr 19, 2012 3:13 pm

Mike Cautillo wrote:How about for futures....that brings me to another point-have you ever looked at the performance of trading one future at a time and how successful this would be ...if you did have limited captial.
Are you talking about taking 1 trade and then rejecting all of the signals that are generated afterwards until the trade is completed or trading many singals, all in a size of 1 contract?

If the former, it would, perhaps, be best to include some sort of filter to sort which signals you're taking. Just taking the first trade that is generated has an element of randomness to it. It may be the 15R home run trade we all dream of or it may be a trade of much less consequence, that will also cause you to miss out on said home run.

rhc
Roundtable Knight
Roundtable Knight
Posts: 464
Joined: Tue Nov 18, 2008 8:46 pm
Location: Oz

Post by rhc » Thu Apr 19, 2012 10:12 pm

Mike Cautillo wrote:How about for futures....etc etc ...if you did have limited captial.
Mike, these links may prove helpful to those that are considering trading futures on a limited budget.

viewtopic.php?t=2359&highlight=

viewtopic.php?t=8842&highlight=

viewtopic.php?t=9367&highlight=

Mike Cautillo
Senior Member
Senior Member
Posts: 27
Joined: Thu Mar 15, 2012 2:59 pm

Post by Mike Cautillo » Fri Apr 20, 2012 8:01 am

Thanks RHC....great help.

If I was following lets say 25-30 markets.....LTTF .....is 250K sufficient to give this a good go.

Toosday
Roundtable Fellow
Roundtable Fellow
Posts: 70
Joined: Fri Jun 03, 2011 11:21 am
Location: Austin, TX

Post by Toosday » Fri Apr 20, 2012 2:23 pm

Mike,

As on outside observer a couple of things strike me about your last few posts.

1) The jump from a $10K account to a $250K account.
2) The desire to trade futures with a $10K account.

With respect to "trend" trading futures on a 10K account. I would say that this is virtually mathematically impossible. You data fees per annum will be about 8-10% of your account value. A huge amount to make up for using LTTF. If you were a great scalper, then I would assume a couple of markets would be workable but tough.

I view a $250K investment as a pretty decent chunk to start LTTF but it may not be enough depending on your goals or your system. Sticking to mini contracts may help this but I do not know if you will find 25 to 30 tradable mini markets. Margin requirements and drawdowns will sneak up on the $250K account size.

The last thing is the difference in diligence I would do on a 10K account vs a 250K account. Mistakes and leakage from a 250K account can amount to real money over time. I would make sure you have the time and inclination to manage the 250K account. All of this is not to mention the 100s of different ways I would look at my system for a 250K account. Many more reports and tests to verify that my investment has a decent chance of working.

All of these comments are meant to be helpful, please let me know if any of this is unclear.

Mike Cautillo
Senior Member
Senior Member
Posts: 27
Joined: Thu Mar 15, 2012 2:59 pm

Post by Mike Cautillo » Fri Apr 20, 2012 3:14 pm

Hey Toosday,

I do not have any desire to start trading with only 10K nor did I express this, I only mentioned this as a response to your post that you suggested maybe this amount would make a good starting point after I had asked-how is everyone calculating start up capital required to trade their system...though I fully understand what you are saying and you make valid points.

I do think with the type of system I am looking at and the amount of markets....250K is ample.

Post Reply