Trading NQ via Price Action

[k p]

What I'm saying is nothing new. You've heard it before, from Db and from others, here and elsewhere. It is just pure logic. You do not get as many opportunities with e.g. an 1 hr chart as with an 1 min chart, therefore you will be able to keep yourself calm easier, you do not have to be 100% concentrated for several hours a stretch, and it is easier to take a short time-out so you do not go fubar.

Ah.. this clears it up beautifully. I thought that you might have been saying that there was a way to catch the better moves. But now I see that you just mean that by trading less, this takes care of having to be focused constantly and sweat through many entries and exits.

If you want huge moves go to huge(er) charts.

This makes perfect sense.

On the other hand, if I recall correctly, I think you answered one of Db's questions earlier that you wanted to day trade. So. Make up your mind.

Ah yes.... this conversation a couple of months left me a bit confused, not unlike where I am right now!

Its this constant battle between the macro and micro view. The macro view guides you, but then you need to get into a trade by looking at the micro view. The idea is that I want to trade less, but unless the entry is in the perfect spot and price goes instantly in my favor, micro-managing it can't be avoided.

Today for example, it certainly looked like once we dropped below 30 and fell below any swing lows that were on the hourly chart, that price would just continue down. But here we are, having essentially gone right back into this trading range of the past few days. (Coincidentally, where we turned around at 4119, this is precisely the top we hit back on Sep 19 that ended up being the start of the huge down move. I for sure wasn't looking that far back, but once it was obvious we weren't heading lower, I zoomed out and I see this level matches perfectly).

Anyway, so although my main goal is to make money in the least stressful way that fits my personality, I think that for the moment, I will for sure have to day trade. I don't want to scalp, I want to be able to capture 20 or 30 points, but if the market says that trend is changing on the 1 minute chart, I don't want to hold so that price comes all the way back to my entry. AMT is a big part of this of course, and since we are still below the upper channel line, holding a short here for long term wouldn't be wise. (It would be nice to catch a move such as the long off the 3793 level on Oct 20 that you pointed out... over 300 points. I still remember when Db asked in my journal why I didn't take the test of the level that happened just a few minutes after the open that launched the huge move up)

Anyway... as you can see... lots of indecision on my part. Its more important now to do things right so that I don't lose more money, but this is of course quite the corner to back myself into since focusing on the money is the worst way to trade.

 

[Roffe]

Ah yes.... this conversation a couple of months left me a bit confused, not unlike where I am right now!

You must be more confused than you give yourself credit for. This was not only a couple of months ago, but last time was approx. 2 weeks ago: http://www.elitetrader.com/et/index...-via-price-action.281995/page-94#post-4040133

The idea is that I want to trade less, but unless the entry is in the perfect spot and price goes instantly in my favor, micro-managing it can't be avoided.

"Instantly" for a 1 hr chart will probably not be the same thing as "instantly" on a 1 min chart for you. Check for yourself.

 

[k p]

You must be more confused than you give yourself credit for. This was not only a couple of months ago, but last time was approx. 2 weeks ago: http://www.elitetrader.com/et/index...-via-price-action.281995/page-94#post-4040133

"Instantly" for a 1 hr chart will probably not be the same thing as "instantly" on a 1 min chart for you. Check for yourself.

LOL.... Aye... so much "stuff" has happened since then that I didn't realize it was only 2 weeks ago that Db posted this.

Well you know what... I have to just throw my hands up right about now! Although you pointed out that the line I drew in on that chart is more of a trend line as opposed to a SL for you, with me calling that an SLA trade but Db correcting me, its obvious that I will never "get" the finer points of SLA. Furthermore, if by focusing on the 1 minute chart, it takes me away from the macro view, but then when I look at the macro view, I cannot find any places to get in or to hold, then this confusion between the "1 minute/ 1 hour" chart isn't on its way to being solved.

Hmmmm.... what to do....

 

[dbphoenix]

I guess for me, when I mention the first SLA trade, it would be based on this SL breaking. Granted we are very much in a range, but once price rejected 42 and came up and broke whatever SL we could have drawn around the open, then this first RET I would think qualifies as an SLA trade. (I guess I'm wrong though )

Now if what you say is in reference to using the hourly charts, then this is something I haven't really explored yet as the stops just seems so wide. (although your PDF of course mostly uses the hourly charts). Its true that I don't want to scalp. I want to get into a move and then just do less, but this isn't always possible I guess until you get the best entry... so perhaps I'm thinking a few scratches and some re-entries are called for until you're in and price takes off in your direction.

I guess that ever since I read a reply you made a couple of weeks back with reference to using 5 minute bar charts, and saying that this gets you into the move late, which I can of course see as well, I figured that focusing on the 1 minute chart was still the way to go, but of course making sure to be looking for trades at the extreme.

There is no "granted we are very much in a range, but". If you're in a range, the SLA won't work, as I've said many times. It won't work because you have to switch to a reversal strategy, and you have no experience with that. Yesterday you said "I'm not making firm progress with testing rules for a trading plan that will work in real time." You can't formulate rules, much less test them, until you have a base of metrics to work from (and Donna explained in detail how to go about building this), and you can't build a base of metrics until you've at least begun backtesting. But for reasons that escape me, you refuse to do any of this, insisting instead on trading.

I'm not trying to be discouraging, but until you've done the backtesting, built up your metrics and formulated your rules, you're trading feelings. The SLA has nothing to do with any of this. There's probably no point in my repeating this yet again, but you must stop trading until you've completed the preliminaries. If you can't stop, and it appears that you can't, then this will go on for another thousand posts and another thousand after that. Is that really what you want?

 

[k p]

I guess in my head, there is this disconnect between evaluating the imbalances between demand and supply, and testing setups/patterns. I guess that if what it means to trade behavior can be put on paper in terms of "if this... then do this", then this should be easy to automate, and I even recall you saying that this can't be done.

Donna's help has been excellent, but when it came down to it, I just couldn't get started. How I drew my 5 minute trendlines even needed to be tested. Do I use a swing high from the close of yesterday... or just a more recent one? At what point do I fan the line, or do I keep the original and just call an attempt to exit a FBO? etc... So I got stumped even before I started. I even had trouble getting my head around the different variables such as MFE and MAE... I just couldn't figure out at what point to "take my measurement". You had also replied that often, when using a 5 min trendline, you're into the move already lates. So in a way, this also leads to my confusion because if I'm too late into a move by trading a 5 min chart... how the heck am I going to get into a move on an hourly chart if I'm taking into account the general trend and context? I of course do understand that in one time frame we could be trending, and in another time frame we are in a range.... etc.

I do have a way of twisting and turning everything in such a way that makes one guess what the heck I'm doing, but given so many different ways that I can look at something, I'm trying to find a way for something that is more of an absolute. Rejections I like because they are obvious on a 1 minute chart, and could even be seen on an hourly or daily if the rejection is prominent enough. Therefore, a test of this rejection might prove to be a juicy trade, so I've been paying special attention to this. Then once a move is under way, the 50% level seems like a good way to gauge if a pullback is just a pull back or serious trouble for the trend.

I figured that if I stick with SLA as a backbone, there would be less to test, the backbone being to buy after a RET once a change in trend has occurred, or price broke out of something, like a hinge or a range. I figured that if I could use whatever ability I had to read "price action", that this might keep me out of some bad trades, such as trading in a range, or at the very least just get out when its not working and not be stuck like a deer in headlights. But to hear there there weren't even any SLA trades today leaves me quite confused because, at least on a 1 minute chart, there seems to be lots there to trade. (the hourly chart does look more like a range though).

Anyway, so as stupid as it sounds, I wasn't even sure what I would be backtesting. ND's 5 min BOPB is great when it presents itself, but since these trendlines cause me problems the way I draw them, looking for stuff that is actually in the market as you would say, was I think the better way to go (this stuff being the rejections, the higher lows or higher highs, etc.... these are in the market)

I guess I have been continuing with my write-up this way for the past few weeks to see if what I'm at least looking at is the right way to look at things. There is nothing worse than putting so much time into backtesting only to find out that what you have is a random occurrence or not statistically positive enough to warrant actually trading it. Since I have so much trouble with backtesting, I thought I was in some way forward testing, multiple things at the same time mind you, but by doing this enough, I would start to see what works and what doesn't.

Rejections at key levels when you get them right are a solid trade, so going long or short above the bar that hits this level seems like a solid trade (and if the entry doesn't trigger, there is your confirmation), and at least your stop can be just on the other side of the rejection. If the inclination is to wait for the RET after this REJ, this is possible too, but then you're getting a much worse price perhaps. This is what I see over and over again, and although I guess this could be backtested on a static chart, when you see this in real time, and perhaps even have a tick chart or 15 sec chart to also watch, the behavior of this jumps out better. So I really thought I was going somewhere with this. If I only look at static charts, I feel as if I'm only backtesting patterns and not behavior.

 

[Roffe]

I guess in my head, there is this disconnect between evaluating the imbalances between demand and supply, and testing setups/patterns. I guess that if what it means to trade behavior can be put on paper in terms of "if this... then do this", then this should be easy to automate, and I even recall you saying that this can't be done.

I don't know how everyone defines "trading behaviour", but if you are unsure then it might help if you view it as you're actually trading other traders. You know what other traders are seeing, what they are thinking, and when. You know what they expect to happen; when, where and how. And therefore you also know how they might react when this doesn't happen, or perhaps when it does happen, but at an somewhat other location in time and/or price. Etc.

 

[k p]

I don't know how everyone defines "trading behaviour", but if you are unsure then it might help if you view it as you're actually trading other traders. You know what other traders are seeing, what they are thinking, and when. You know what they expect to happen; when, where and how. And therefore you also know how they might react when this doesn't happen, or perhaps when it does happen, but at an somewhat other location in time and/or price. Etc.

Hmmm... I technically don't know what they are thinking. I know that if someone buys, they do so because they must think price is going up and they can sell for higher, or they are buying to get out of their short, but this would also mean that they no longer think price is going lower and hence want to get out of their trade for whatever profit they already made, so they too think its going higher.

If I knew whatever everyone else was thinking.. it would be easy! LOL... If I knew everyone wanted to buy here, I would rush my buy order in. But alas, what if only 100 people want to buy, so price goes up a little bit, and nobody is left to buy? What if the next 1000 people waiting to buy wont buy until price is 10 points lower and me, having just bought, have to watch it drop???

Don't get me wrong, thinking this through seems to be what is necessary, and yet all this talk about backtesting just means you're looking for patterns. I figure Google would have already come up with an algorithm, and this is in fact what the HFT firms do in a way I guess, but since we still have traders out there that are able to consistently make money, human traders, putting on just a trade or two a day, then there is more to exploit than what is just on a static chart.

But its true that all we have to go on is what happened in the past, and this is what you get when you look at an EOD chart. So I guess if I have to come up with a "if I see price bounce off this line here and an entry one point above this bar leads to a profit of at least x amount of points y percent of the time"... then this is what I gotta do.

This way I don't even need to watch price move... I just have to wait to see the bar at the end of the minute. If my backtesting only shows me complete bars, then whatever I come up with will only work on fully formed bars. In a way, this would be nice because I get anxious watching that bar move if price isn't going in my direction... so perhaps I should just look away or 55 seconds, and only come back to the chart once the bar is fully formed. I can of course have a sell already in place if price goes below some previous bar if that is my exit (hence my exit can be in the middle of a bar to get out)... but by focusing on backtesting, I am only trading bars. I might as well just look for a pattern where if I see 3 up bars, what are the chances that the next bar will be down.

 

[dbphoenix]

Donna's help has been excellent, but when it came down to it, I just couldn't get started. How I drew my 5 minute trendlines even needed to be tested. Do I use a swing high from the close of yesterday... or just a more recent one?

Find a hundred instances and plot both, then determine which one has the higher probability of success.

At what point do I fan the line, or do I keep the original and just call an attempt to exit a FBO? etc...

Find a hundred instances and plot both, then determine which one has the higher probability of success.

And so on.

This is how it's done, not by trading what you feel and what you think and what you guess.

 

[k p]

Find a hundred instances and plot both, then determine which one has the higher probability of success.

Find a hundred instances and plot both, then determine which one has the higher probability of success.

I see a repetitive theme here! Ok... let me see what I can come up with.

 

[bh_prop]

There is nothing worse than putting so much time into backtesting only to find out that what you have is a random occurrence or not statistically positive enough to warrant actually trading it.

Except maybe: not doing the work first, having a smaller account balance to show for it, and months later zero backtested results because you have bounced all over the place with ideas and no structure.

Imo forget about trying to learn and monitor all these different setups (hinges, 50% retracements, range breakouts, etc) right now. Pick ONE setup on ONE timeframe and start to backtest it. Mastering that alone could easily take weeks. Then and only then move on to the next setup you want to include in your arsenal.