What creates spikes?

[redbaron1981]

I'm sorry I have not fully read this thread as it seemed to get detracted quite quickly.

Spikes are caused in areas where there is an imbalance in the supply of liquidity and where there is an immediacy of demand required by traders.

If you want to learn more about this I recommend learning about market microstructure, liquidity and different types of orders. You might also want to research episodic volatility.

Hope this helps.

 

[k p]

I'm sorry I have not fully read this thread as it seemed to get detracted quite quickly.

Spikes are caused in areas where there is an imbalance in the supply of liquidity and where there is an immediacy of demand required by traders.

If you want to learn more about this I recommend learning about market microstructure, liquidity and different types of orders. You might also want to research episodic volatility.

Hope this helps.

Hey redbaron... yes the imbalance is something I see for sure. But I was thinking why this imbalance. What I was trying to get at was that in slow times, where its easier to manipulate the market, a huge firm that is used to trading several hundred contracts decides to pull all offers to buy at a critical level, knowing that this will cause a bunch of stops to be run, and since you know you caused this, you know that there is no reason why you couldn't buy them all up for a few points lower. I have no idea if this is how it works, but I think it certainly makes sense.

 

[redbaron1981]

A stop order is the same as a market order as in they both consume liquidity.

To trade size cheaply you need liquidity whether entering or exiting. If you don't have liquidity you will move the market in the direction of your intended trade thereby increasing or decreasing your average fill price. People who trade size can't just enter or exit where they feel like as there market impact can be substantial and therefore terrible fill prices will be costly for entering and exiting.

Now if you think of liquidity everyone thinks of the limit order book which is generally not that real. However as we know many retailers like to buy and sell in the same places and this also means placing stops just beyond S&R under swing low behind moving average etc etc

Because of these repeatable patterns of stop placement a concentration of stop orders will build up in certain positions on the chart. Now going back to our big trader if we know where a concentration of stop losses are then we know that we can fill our limit orders up with the liquidity provided by the stop orders of the retail traders and therefore reduce our cost of entering or exiting. It is this action that causes spikes pin bars and whatever other names you can think of.

This brings out the age old of argument of how effective is TA. Well if you take a note from above then you will gather that if your actions are repeatable then you are exploitable.

 

[k p]

Very interesting indeed. So basically, those who know what they are doing will more than likely have buy limit orders just below a support level, thinking that if this support level breaks, and people are desperate to get out, the smart guys will be there to offer liquidity and take the other side of the trade that the amateurs are desperate to close.

So now of course, how does one go about figuring out if those stops being run is just stops being run, and not the beginning of an actual down move. The general wisdom of course is that you wait for a breakout (of breakdown in this case since I am using the sample of a support level breaking), and after this break of a few points, we wait for price to rise back up to the support level, and see if this level will now be resistance. So the buying that happens to allow price to rise back up to this level, to see if we can go back above support, is now in a tricky situation because if there aren't enough buyers to go back above support, back into the range or what have you, but instead there is more selling, then those guys who were initially responsible for halting the spike down are now also going to want to get out. So if there a way to distinguish one scenario from the other? (when the spike down is just stops being run versus the spike down is indicative of more selling)

I mean its great to be the guy who can buy up all the cheap contracts once stops have been run, but if you're the only guy buying and there is no one left to buy, then its not just a case of stops being run, but a real selling wave.

 

[redbaron1981]

Exactly and the rest is all for you to fathom out but a little clue. It's got nothing to do with support resistance, moving average, candle patterns etc etc and I'm certain this will encourage some posts from the TA brigade.

But if you take at least one piece of information away from this it should be that if you are trading in popular areas like quite a few posters on this thread are then my friend your liquidity is exploitable by larger traders than most of the posters on ET.

 

[k p]

Oh god dammit... LOL... This is as bad as what the member monoid was sharing with me a few days ago. Since you didn't read it, I will reproduce the chart here. He basically said that at 9:08, we saw a signal, and at 9:12, he entered long. I put those two times on this 30 sec chart of ES that he trades. Now I don't study the ES, so I'm not sure where the trend is, where support and resistance lie, etc. But based on what I've been taught to look at, there was no indication to go long here.

Not this this matters I guess since you say its not a matter of S/R or anything else that I look at anyway. But this leaves me in the dark as to what to look at. I really don't know what else there could be. Feel free to PM me so that you don't give it away to the world and I will happily just keep it to myself! I just don't see how else to look at price action other than plotting price on a chart and visually seeing where price is, where price has been, where it has changed direction, etc.

 

[i960]

Interesting. So it looks like you are essentially buying what looks to be one tick above the bar that has the volume spike, is this correct? And if the trade hasn't taken off after 3 bars, you'd tried to get out BE +1 tick? What if you're already heavily down right after entry? I find that often, if right after entry, price goes against me and I barely ever see a profit for any amount of time, it is a trade that should be exited right away without even getting to the stop as these barely recover.

Alright so it's natural to see price go against you a certain amount of time but you shouldn't be seeing that 100% of the time. You also shouldn't be seeing a situation where it takes a long time to see profit, if any. You should probably be seeing atleast some profit even if the trade ultimately loses.

Think about it. If the trade goes against you it means it's going somewhere else. You want to be buying or selling at that else place to reduce your risk when you are wrong. This means seeing the most likely price point it will trade to (for entry) and putting on a limit order there beforehand (better to be earlier in this realization as well - as there's a queue).

Don't wait for confirmation with a higher high or lower low and then just hop on - wait for a pullback of some sort after that fact. You might already be doing this as well. I suggest posting some losing trades that were based off of sound ideas yet immediately went against you so we can see entries.

I find that a good fill is a patient fill.

 

[k p]

I suggest posting some losing trades that were based off of sound ideas yet immediately went against you so we can see entries.

I find that a good fill is a patient fill.

You're asking for too much if you want trades based on sound ideas! LOL... But just for a laugh, I went through some charts and here is one that I think shows my not so wise wisdom.

The green line in this chart was a previous day high. Overnight, price shot up at 7am through the high, and after this, price also came down to this same level and bounced off, so it was providing possible support now, having been resistance the day before.

You can see my first trade was a long at 35, which is 3 points above the level, far too late, but at the time, I guess I thought the bounce was good so it should be heading up. (One of the things I've really noticed about me is that I sit there and wait to see if I'm right first, to see if it is moving up, and then I get on board, which is clearly far too late at that point. That's why I'm really into this "no confirmation" idea now and all I really want is a good price)

This one isn't so bad mind you because a case could be made for support here, but the entry is far too late. Look at the exit though. My chart didn't even register the bar coming down to where the exit actually was... I guess I'm the only guy who sold at this level!

Then you see a short around 30, after price broke below this level, and I guess at the time I thought ok, we are dropping, so let me jump on board, but that trade is clearly not set up. You can see how I'm very much of a reaction trader. If anything, I now that often price penetrates, and if it comes back above the level, then its simply a running of stops, and there were lots of guys eager to buy as price penetrated, and they obviously bought from me since I sold short.

Then you see a couple of more longs after this, but tiny profits as I'm just trading scared at that point. Knowing what I know now, I would have played this much differently. Of course in hindsight it looks easy, but I would in no way jump into a spike down like that. If anything, I'd be looking for a reason to fade it.

Anyway since you asked, I just wanted to show you something.

 

[llIHeroic]

He basically said that at 9:08, we saw a signal, and at 9:12, he entered long.

That's impressive to be able to nail that move with such a high degree of precision. I don't really splice into bars past the 2m interval at the finest, and I'm sure that monoid and I don't use the same methods, but it's quite interesting to see other frameworks out there giving similar information to the paradigm I use.

Here were my annotations in RT for the day. Ignore the comments on the image, this is just a re-post of a chart I was sharing with someone else.

 

[k p]

That's impressive to be able to nail that move with such a high degree of precision. I don't really splice into bars past the 2m interval at the finest, and I'm sure that monoid and I don't use the same methods, but it's quite interesting to see other frameworks out there giving similar information to the paradigm I use.

Oh I didn't want to give the impression that this was his chart. I just took what he mentioned in his post about where he got his signal and where he entered, and added this info onto a chart of the ES so that I could see what he was looking at. He does say his charts aren't time/volume charts, more of a custom design, so I have no idea what he is actually seeing. Given his excellent example up above with the apples and the snakes and the sunglasses, he is more than likely seeing this completely differently. So disregard my chart really as I'm sure it does no justice to displaying how he is looking at price action.