Why do we mostly ignore half of the market?

Back to topics list To post a new topic, please log in or register
avatar
7
bigtent 2012.04.09 09:42
 

I know this is "convention", but all conventions need to be questioned, otherwise mankind would never have left the trees.

MetaTrader not only does not record ticks in it's history, but it only has the OHLC of the bid prices.

However, long trades are opened on the ask and closed on the bid, and short trades are opened on the bid and closed on the ask. So the ask is of as much importance as the bid.

Usually, the spread is variable so one does not "just follow the other". Also, if you watch tick charts, you will notice that there is often very different activity on the ask and bid ticks.

I have seen one platform that allows you to have charts based on the ask price (Can't remember which). This would actually be very handy when managing short trades, as the ask is what counts in this instance. In MetaTrader, you could build them recording current data, but you could not do any more than that as there is no historical data.

According to conventional wisdom, buying (ask) creates upward price pressure and selling (bid) creates downward pressure, and the difference in those pressures moves the market. So therefore it would make sense to glean what is happening at this fundamental level to help predict what what is likely to happen next, particularly when you consider the complaints about "lagging indicators" etc, the tick is the closest to immediate information we have.

I have tried searching, but to no avail, but has anyone done anything on tracking and analysis on ask and bid ticks to try and see any use in trading? (I know that we only get to see the ticks that our brokers provide, but they should be somewhat representative of the market in general).

For example, does a fluctuating ask and relatively still bid indicate more buying activity, therefore the price is likely to rise in the immediate future?

Any ideas? Anyone played with this kind of thing?



 
avatar
3849
ubzen 2012.04.09 10:32 #
 

We've been through this question countless of times on this forum.

analysis on ask and bid ticks to try and see any use in trading? Sorry, I don't see anything useful here. Ticks can be lost in Cyberspace. Hence the fact that they're un-reliable for back-testing..... then expecting to behave the same in real-trading.

(I know that we only get to see the ticks that our brokers provide, but they should be somewhat representative of the market in general). In my opinion, if you're concerned about differences between bid/ask for a particular broker, you should be equally concerned about differences between Broker-A vs Broker-B.

For example, does a fluctuating ask and relatively still bid indicate more buying activity, therefore the price is likely to rise in the immediate future? Open a real-account, built an indicator which logs the activity of bid vs ask, generate statistics on probable direction based on this information.

IMO, all of the limitations you've described above can be done with mql4, it'll just require far more effort than most people are willing or able to program. Example,

- Tick based back-testing, there's the tutorial for that here. *If you don't like the fact that it's based on particular broker spreads.

- You can always modify the spreads using something here. *If you still don't think its real because bid's not sitting static while ask is bouncing up and down..... then.

- You can create your own, historical data which logs the Ask as well. Search for File Functions. *Still not satisfied.

- Create a virtual order management. Use arrays which mimic the mql4 OrderInfo* properties. You can do anything you want now. From -Negative spreads. To slippage. To even Multi-Currency back-testing.

- There's always mql5, which is also bid only data. But makes allot of the above easier.

Conclusion: unless you're trying to trade some form of high-frequency-scalping system then it's just not worth it.... IMO.

 
avatar
18
rbhauer 2012.04.09 11:04 #
 
bigtent:

(snip) ... Any ideas? Anyone played with this kind of thing?

Short answer: You're frustrated and going through a phase. I've been there. The answer is not there. Let it go. Plenty of opportunities looking at other directions.

Long answer: Ratio between tick volatility vs spread is not proportional by significant order of magnitude. Risk adjusted return = (return-spread)/(risk+spread). If you need a lot of ticks to eventually generate a magnitude to overcome the (risk+spread) denominator, you'll no longer be working with tick scale/magnitude data set. It'll just be a complete waste of time and resource instead of working straight off the larger timeframe and use M1 as needed.

 
avatar
7
bigtent 2012.04.10 07:23 #
 
ubzen:

We've been through this question countless of times on this forum.

analysis on ask and bid ticks to try and see any use in trading? Sorry, I don't see anything useful here. Ticks can be lost in Cyberspace. Hence the fact that they're un-reliable for back-testing..... then expecting to behave the same in real-trading.


Conclusion: unless you're trying to trade some form of high-frequency-scalping system then it's just not worth it.... IMO.

Thanks for your reply Ubzen, but mostly, you are missing my point. OK, my fault, for not expressing myself clearly.

As someone who built an ISP company from the ground up and ran it for 12 years, I am very aware of internet packet loss; something we all have to live with, even though it is far better than it used to be say 15 years ago, regardless of what you are using the internet to achieve. It may occur, it may not occur. just as you can lose connection to the internet altogether (which, speaking from experience, is really annoying when you have just entered a trade). The inherent data loss problems of the internet is not relevant to what I am talking about, it's just an unpredictable pain the arse with any real-time application (latency is another in the same category) and they are just things we have to put up with. Things such as unscheduled press releases are also similar, in that they can move the market suddenly. (An IMF release last week caused a sudden drop of 4o pips in AU - killing a trade I had just placed, for example.

These kinds of things we have to just bear as externalities, and no matter what you are doing, all you can work with is the data which we you DO get.

I am also not talking about different spreads at different brokers ( have live accounts with different brokers), nor am I unaware of the methods of overcoming metatrader's shortomings in this regard.

So, since there is also another response below, I'll try to better explain myself in a new post, that you both can consider.

Then, maybe you can tell me if it has been discussed before, and maybe direct me to it (because I have NOT been able to find anything on it).




 
avatar
7
bigtent 2012.04.10 07:28 #
 
rbhauer:

Short answer: You're frustrated and going through a phase. I've been there. The answer is not there. Let it go. Plenty of opportunities looking at other directions.

Long answer: Ratio between tick volatility vs spread is not proportional by significant order of magnitude. Risk adjusted return = (return-spread)/(risk+spread). If you need a lot of ticks to eventually generate a magnitude to overcome the (risk+spread) denominator, you'll no longer be working with tick scale/magnitude data set. It'll just be a complete waste of time and resource instead of working straight off the larger timeframe and use M1 as needed.


Thanks for the reply!

No, I'm not frustrated at all, just curious about something that has made me ponder a lot over the last few years, when looking at tick charts. (I am a very curious individual, but seeing I am not feline, that has not proved dangerous over the last 50 odd years)

I am more looking at Ask volatility vs Bid volatility (which is quite different to what you are looking at), and this could be plotted against time, or it could be plotted vertically against price.

More below....

 
avatar
7
bigtent 2012.04.10 07:50 #
 

OK, so I'll try to explain it more here....

I have noticed that, quite obvious on the tick chart, ask and bid do move independently and quite often. (I have watched ticks many times when trying to nail a good entry in manual trading). This has made me curious over the years, as to whether anything could be gleaned from it. I have also studied many indicators and am aware of the ideas and assumptions made with them, e.g. Stochastic is effectively just the position of moving average in a Donchian channel, etc.

The author of this article was dismissing/questioning many "mistakes" of TA because of lack of fundamental understanding of the market, etc.

In my response to that article I said:

I do not believe there will ever be holy grail science around it. Rather, I see that intuition, unquantifiable and undefinable like a surfer riding waves, will always be a big part of trading. Technical analysis and indicators can help traders "read the market", and since we all see things differently, no system, no indicator, whatever, will provide a "one size fits all" solution. Many of these "tools" may well be scientifically nonsense, but some people will be able to use them to create profit, while others, trying to do exactly the same thing fail, just like one surfer somehow manages to consistently make it through the tube while others consistently get wiped out!

We can get fooled looking at charts and fancy looking indicators that make it look like we are part of a NASA program, but in the end, that is just a display, and it is, in reality, people driving the markets, and I repeat, as Elders said, (and it has been scientifically quantified) people a far more emotional than rational when it comes to decision making.

So maybe it is more that we need to recognize technical analysis for what it is, rather than expecting there ever to be an intrinsic understanding of the market to build a science around, because the latter is the real unachievable "holy grail".

After reading it revived the interest in something I'd been thinking about for some time, and since I've never found any discussion on it (I have searched many times), I thought I'd throw this topic into the forum.

What I am wondering is:

(Caveat: I'd love to know exactly how, technically, the networked distributed forex market network actually arbitrates price, but for now we will go with accepted theory)

Theoretically, ticks are the fundamental building block of the market. a) Sell orders put downward pressure on price, and b) Buy orders put upward pressure on price.

Now, there are two prices, bid and ask. Generally, we ignore the Ask for all intents and purposes, except when coding an order function.

But I propose that, based on conventional theory, the following may be true:

  • Buy orders are at the Ask price. Therefore, an upward movement of the Ask indicates the filling of a buy order.
  • Sell orders are at the Bid price. Therefore a downward movement of the Bid, indicates the filling of a sell order.
  • Ask and Bid are somewhat "tied together with a rubber band", so that if there is enough upward volume on the Ask, it will drag the Bid up, and if there is enough downward movement on the Bid, it will drag the Ask down.

Counting/measuring/collating of these movements with respect to the time that they happen, and/or the price levels they happen at, (producing something similar to OANDA's Order Book) may give some extra information to traders. (Note: obvioulsy, with MetaTrader's limitations, this could only be live collected data, unless logged or downloaded data is added)

Now it may be useless, but I still feel it is worth looking at, so, before I go and reinvent the wheel, has anyone done or is anyone aware of any analysis done along these lines? Because I'd really like to have a look and see what has been done!

 
avatar
3849
ubzen 2012.04.10 11:08 #
 

Let me began by saying; I'm not here to judge anyone nor his/her methods. I just present my opinions based upon my experiences. Your approach is unique in regard to details. "Counting/measuring/collating of these movements with respect to the time that they happen, and/or the price levels they happen at, (producing something similar to OANDA's Order Book)". It'll be quite difficult to find someone who'd done this the way you're thinking of doing it. However allot of people including myself have done something similar.

This exercise reminds me of three approaches, Tick-Volume, Spread, and Tick-Charts.

Tick-Volume is the easiest to dis-count. This one most newbies including myself starts out investigating. I wanted a scalping system and it seemed logical to start with m1-charts. The problem is I mis-toke Tick-Volume for Market-Depth. Once I realized what Tick-Volume was it became obvious it was not what I was interested in. See I wanted to know Size-of-the-Long-Positions, not the Number-of-Transactions/ Broker's Out-Cry on Ask/Bid within 1-minute.

Spreads: Here's a question, (Bid stays static) but (Ask move down 1-point) does this indicate a change in spreads or a change in buying pressure or both? To me, changes in spreads could easily explain the differences between Ask vs Bid. This will create inconsistency in my system form broker-to-broker example Fixed-Spread Broker vs Variable-Spread Broker. It'll also create inconsistency between 4-digit brokers vs 5-digit brokers.

Tick-Charts: I've been through threads with people asking for 1-second charts, 5-seconds charts etc. It'll be nice if mt4 had all these charting options but to what end? Tick-Charts are nice in that, if you care about support and resistance on inter-minute time-frames it gives you a flat-plane view like Renko/Box Charts. It also high-lights the Bid vs Ask nicely. But using this information to gain Market-Depth can be trick-business to say the least.

Well the above are the ones which I've tried and those are my conclusions. If this thread could be summarized as a search for Market-Depth then IMO it'll be very difficult to figure out where Big money is using Ask/Bid changes. As I'm sure you know, Forex is de-centralized. This might come off arrogant but I believe even if all the retail brokers sent you the positions and size of all their retail clients it'll still not be enough to give you a good sense of the market depth of the inter-banks.

Similar to your Order_Book, you can try looking at the Market-Depth and Trader's Sentiment Here. I would encourage you to complete your case study. And hope you can keep us posted on your findings.

 
avatar
7
bigtent 2012.04.10 12:55 #
 
ubzen:

Let me began by saying; I'm not here to judge anyone nor his/her methods. I just present my opinions based upon my experiences. Your approach is unique in regard to details. "Counting/measuring/collating of these movements with respect to the time that they happen, and/or the price levels they happen at, (producing something similar to OANDA's Order Book)". It'll be quite difficult to find someone who'd done this the way you're thinking of doing it. However allot of people including myself have done something similar.

This exercise reminds me of three approaches, Tick-Volume, Spread, and Tick-Charts.

Tick-Volume is the easiest to dis-count. This one most newbies including myself starts out investigating. I wanted a scalping system and it seemed logical to start with m1-charts. The problem is I mis-toke Tick-Volume for Market-Depth. Once I realized what Tick-Volume was it became obvious it was not what I was interested in. See I wanted to know Size-of-the-Long-Positions, not the Number-of-Transactions/ Broker's Out-Cry on Ask/Bid within 1-minute.

Spreads: Here's a question, (Bid stays static) but (Ask move down 1-point) does this indicate a change in spreads or a change in buying pressure or both? To me, changes in spreads could easily explain the differences between Ask vs Bid. This will create inconsistency in my system form broker-to-broker example Fixed-Spread Broker vs Variable-Spread Broker. It'll also create inconsistency between 4-digit brokers vs 5-digit brokers.

Tick-Charts: I've been through threads with people asking for 1-second charts, 5-seconds charts etc. It'll be nice if mt4 had all these charting options but to what end? Tick-Charts are nice in that, if you care about support and resistance on inter-minute time-frames it gives you a flat-plane view like Renko/Box Charts. It also high-lights the Bid vs Ask nicely. But using this information to gain Market-Depth can be trick-business to say the least.

Well the above are the ones which I've tried and those are my conclusions. If this thread could be summarized as a search for Market-Depth then IMO it'll be very difficult to figure out where Big money is using Ask/Bid changes. As I'm sure you know, Forex is de-centralized. This might come off arrogant but I believe even if all the retail brokers sent you the positions and size of all their retail clients it'll still not be enough to give you a good sense of the market depth of the inter-banks.

Similar to your Order_Book, you can try looking at the Market-Depth and Trader's Sentiment Here. I would encourage you to complete your case study. And hope you can keep us posted on your findings.

Thanks again, ubzen. No you are not arrogant, you express your opinion and that is great and thanks for the link.

On the decentralized market, as someone who has had a lot to do with servers and getting them to "play nice together", I'd really love to know how they arbitrate the price on with the market so fractured, even among the big banks!

I will be back with anything I come up with...


 
avatar
1
waforex 2015.03.19 02:58 #
 
Hi bigtent, did you get any further on this subject. I have been attempting to analyse the same thing as a means of measuring buying or selling pressure. Is there a way i can contact you please?
Back to topics list  

To add comments, please log in or register