range vs trend and how to predict it

I think that the idea of being able to predict a trend day versus a range day is enticing, however, I personally have never made much headway in that regard. Aside from the broad observation that a string of trend days is generally followed by a period of consolidation, I have never been able to get much more specific than that from a practical standpoint. I never have advance notice of when a day's worth of trend movement is about to start or end. I guess we all have to live with our own weaknesses.

Therefore, I trade a method that works a bit like an all-season radial tire. In theory it should adapt to the environment, but from time to time it is found wanting. Even so, from my own experience, I found this to be a better approach than trying to time the change-over from trend to range tires. At least for me. Perhaps I am not sufficiently creative, but I find that the idea of trying to determine in advance whether to play the whole day trend or range seems to work better in theory than in practice. At least for now. Perhaps one day I will see the light. Or not.
 
Quote from travis:

Right now I am kind of tired of all this testing, but I will keep it in mind for the future. Thank you.
More...

I know travis, I know.

But actually, you're work has just begun.

You've discovered something new and it's excitiing and different, and the $$$ ramifications are enormous.

Now track your results every week, no exceptions, none ... then:

1) start playing with position sizing and profit taking setups which are designed to maximize the potential of the type of day you are trading, and;

2) sit down and start applying the process to markets which have little or no correlation to the market you are currently trading.

When it's all said-and-done, you'lll never have to work again (but at this point trading will be like a full-time job), which is what it should be.

Best Regards,

Jimmy :)
 
Quote from Thunderdog:

I think that the idea of being able to predict a trend day versus a range day is enticing, however, I personally have never made much headway in that regard. Aside from the broad observation that a string of trend days is generally followed by a period of consolidation, I have never been able to get much more specific than that from a practical standpoint. I never have advance notice of when a day's worth of trend movement is about to start or end. I guess we all have to live with our own weaknesses.

Therefore, I trade a method that works a bit like an all-season radial tire. In theory it should adapt to the environment, but from time to time it is found wanting. Even so, from my own experience, I found this to be a better approach than trying to time the change-over from trend to range tires. At least for me. Perhaps I am not sufficiently creative, but I find that the idea of trying to determine in advance whether to play the whole day trend or range seems to work better in theory than in practice. At least for now. Perhaps one day I will see the light. Or not.
More...


i think what you have summarized reflects my opinion too.

i had gone down this path and analyzed the stats on range, volume and other types of data.

in some cases, i had thought i saw some relationships which looked like they could be profited from.

in the end, after spending some time on this, i had to put all of my data though basic tests for random normal distribution, and at first i was shocked that i had been trying to squeeze something out of totally random data - next, though, i was glad i did because it has saved me from spending more hours on mining through random data.

i'd like to see if Travis comes up with something that's not random - and perhaps there is something but i was just not able to find it in the range / volume on ES (of course there may be a good chance my analysis wasn't good enough :) )

best of luck.
 
Newbie2006,

May I ask how many stocks and how many days/months/years etc you sampled on to get the below result. Also, did you take into account of those trades that occurred between bid and ask?
This is because I ever did something like this quite a while ago for an extremely small sample and although I did not specifically notice for the 1st 30 minutes, my impression was that generally the delta will be positive and not zero regardless of whether its a trend day or range day. Thanks.


Quote from newbie2006:

Although I am still testing my automated system but I thought to put in my 2 cents.

in my experience to differentiate between trending day and range day you need to watch the market delta.

I usually wait for the first 30 minutes (I am talking about stocks) to pass then I start counting the market delta, in range days you will usually find the mean of the delta to be round 0 (zero) and for trending days you will find otherwise.

I hope this help.
More...
 
What do I consider maximum drawdown?

I have started trading my system with real money, after depositing a margin of 2300 dollars on my trading account in early August. As soon as they'll get below 2000, that will be the end of my trading, because the broker stops me from trading with a lower margin. Until then, my plan may sound reckless but I don't plan any money management - that is I will reinvest all profits. Why? Because I don't have time to waste, and because all I stand to lose is 300 dollars based on my initial capital.

Also, for the same reason, no extensive and thorough tests (which are impossible to make anyway, for the amount of intraday data I would need) are preceding my real trading. With so little money invested and only 300 dollars to lose, I think I can worry less about testing. and more about practical and immediate action.

So far, the margin is still there. I am up less than a hundred dollars compared to two months ago, but at least I am gaining some practical experience. Besides, with the about 400 dollars paid in commissions, it must mean the system, which I continue to modify every day, makes some sense.

Yes, replying to another question, my system is not solely based on the correlations we talked about - they just help it. The mentioned correlations - daily volume with daily range, pre-open volume with daily volume, and pre-open range with daily range - account for just 25% of my trading system.

Basically they are very important because they decide whether the system will trade in a range mode (accepting only trades within support and resistance) or in a trend mode (accepting only trades beyond support or beyond resistance), but then it's up to my system to decide all the other things - support and resistance levels, and the other events triggering a trade.

I hope this added 25% will bring my system from break-even as it seems to be now, to being profitable. Yes, I am not talking about anything incredibly precious, as for the past 2 months it just broke even. I guess that's why I am not too secretive about it.

It is an intraday system based on a US stock index future and it makes about 2 trades per day. This is to say that it reacts pretty quickly to the changes I make, and I am able to see if I made a good change from my real money income, rather than from tests on the past. If it loses money for a whole week, certainly something is wrong.

To answer another question, by fader (who was wondering whether these ideas really worked), it may be insignificant, and a small test period, but this week, the "pre-open/daily" method has predicted correctly that monday to wednesday would be trend, and thursday and friday would be range, and I show in the picture I attached in my earlier post what I mean by trend and range.
 
Quote from travis:

Here I would like to sum up all the ideas I talked about in the preceding threads, and come up with a conclusion.

The system will weigh pre-open range, pre-open volume and weekday, estimate whether today is a range or a trend day, and then trade accordingly - within support and resistance on a range day, and beyond such levels on a trend day.

Look at the attached picture and tell me if it makes sense to you that on a given day an intraday system should be permitted either trend or range trading and never both.

According to this concept, this week would have seen trades within support and resistance on thursday and friday, and trades beyond S & R on the first three days of the week. (That is, it would have done so provided that the "pre-open weighing" method worked and predicted correctly which were trend and range days).

On the picture, you see daily candlesticks of the Dow Jones for this week. You can also see "micro-seasonality" at work. Just as August and December, also Thursdays and Fridays tend to be range days, and they were on this occasion.

By the way, it can't be seen in the picture, but this week confirms my assertions and tests that say that the larger is the range on a day, the larger is the volume (it happens on monday, tuesday, wednesday). Also, I have verified once more that pre-open volume is directly proportional to the final daily volume, and that the pre-open range is directly proportional to the final daily range. In other words, the "pre-open weighing" method predicted correctly the whole week.
More...

Travis,
Your investigation sounds quite interesting.
Where are you getting the pre-open range and pre-open volume figures from?
 
Quote from travis:

I am trading a future, so I simply did some tests on the data I bought from here (see previous threads for detailed discussion on tests):

http://disktrading.is99.com/disktrading/
More...

Jeez,

Since you're being so nicey nice about sharing, there are several brokerages you can goto for much better margin and performance bond rates.

Global Futures will allow you to go down to $500 before they pull-the-plug on ya ... just something to know, seeing as how you're (currently) conducting all of these high-tech experiments out of your garage and all. :)

Best Regards,

Jimmy Jam
 
billp,
let me elaborate more.

First let me clarify how I count market delta:

1- I define a variable "currentMidLine" its simply (bidPrice+AskPrice)/2 , of course this value changes whenever level1 quote changes, ie whenever the bid or the ask changes.

2- I watch for all the prints (all ticks), if the print is above the currentMidLine I increment the marketDelta with the volume of that print and if its below the currentMidLine I decrement the marketDelta with the volume of that print

3- I don't reset this marketDelta variable ever for the day

From the above you can conclude that if the print occurs exactly midway between the bid and ask I just discard it as I have no way of knowing whether it was aggressive buyer or seller, but when a print occurs between the bid and the ask but tends to one side I count it on that side.

This marketDelta variable represents the net difference between aggressive buyers and sellers (in terms of volume) from the start of the day up until now.

so now we have 3 cases:

1- the market is trending up, this means that there are more aggressive buyers than sellers so the market delta should be a positive number and this number increases as time goes on as long as the market is still trending in this direction

2- the market is trending down, this means that there are more aggressive sellers than buyers so the market delta should be a negative number and this number increases as time goes on as long as the market is still trending in this direction

3- the market is going side ways, this means neither buyers nor sellers have upper hand and this put the net marketDelta in the zero region

Notes:
1- Saying the marketDelta is zero doesn't mean exactly zero, it means its within a range and this range is different from a stock to another, for me I consider it zero if its in a range of only 30 multiples of the average volume needed to move the stock one cent

2- For the 30 minutes I mentioned, usually the first 30 minutes are trending to define the initial range of the day, I don't consider this a part of the real trend, usually in a trending day you will find extension to this range beyond those 30 minutes, in range days you find the price fluctuating within this range.. Almost.

3- I have tested this for few volatile stocks like AAPL

This is my humble explanation and I am very interested to know your input regarding the above whether you agree or not
 
You must log in or register to reply here.
Back
Top