AgeKay

Same here. High/low volume areas as support or resistance make a lot more sense to me than 1 standard deviation or some other arbitrary range or prices level. Thank you for the examples which I found very helpful to understand how you can trade these high/low volume areas. Let me us know when you notice something else after signing up for their software again.

Thanks for the picture of depth. Can you confirm that this market is largely retail driven, i.e. changes in the depth are small at a time?

Thanks for the charts, but I still don't get the point. To me, both market profile and the volume distributions of both contracts look very different. Yes, the general shape looks kinda similar but if you look closely you'll see that there are big differences at the individual prices.

I am not going to look more than 1 year back and this topic is not about historical prices or price patterns, but market profile or volume distributions where the volume at a specific price level is important even if that means that you have to adjust for the gap.

alleyb, so what are you saying? It does not really matter? Or should you use only the front-month contract despite lack of volume when it was a back-month contract since you would see the same volume distribution?

Thanking very much for sharing your experience with their software. It is much appreciated! Especially since you pointed the weak and strong points of their software/methodology. Do you know whether the 180-day volume histogram is based on the continuous contract or the front-month contract only? What you pointed out regarding the volume histogram taking precedence over MP is because MP is just an approximation of the volume histogram. I think MP was only used as a proxy to determine the volume traded based on constant trading activity over time (which everybody knows is not necessarily true) because they did not have the real time data we have today back then. I don't see why anybody would still use MP when they can just use the real volume histogram. The concepts probably still apply, but the volume histogram just offers you more accurate data for that.

What exactly are those shaded regions? Value Area Low/High? POCs?

hahaha, zdo, that is very true. I can see that in myself. I had a big problem in the beginning accepting that the markets are dynamic and you can't just apply some indicator or combination thereof that "works". I've gone from fully automatic system (wannabe) trader to fully discretionary. Yours is probably the best overlooked advice in this thread.

I don't think you need to worry about the money that much when you're at his level. Also I don't even think it makes economic sense. You might get 2 extra man hours for every hour you pay, but it will still take him 2 more than your professional developer (so that's 5 vs 1 hour). You cannot believe the difference in productivity between different developers. Here is an example. You should not forget maintenance of the application which is very important. If you're super talented student gets his master and expects to be paid as much as a professional or isn't interested in your program anymore, you can probably throw away all the code, because it's unmaintainable (once he is knows what he's doing he'll probably even tell you that and he'll want to rewrite it from scratch).

Thanks for the links. Sorry, I was tired and lazy last night. The round turn commission for MHI on IB is HKD 34.00 (4.35 USD), so the commission is huge compared to tick size, but as somebody pointed out it might be ok if the daily range is also huge. Can anybody give me an idea of the daily range?

Darth, it really depends on how much experience the computer science student has in developing an actual APPLICATION, not just a sample app or a function that does something specific. From my experience, most computer science students are completely overwhelmed developing a workable application. I have even found that some high school students that have been programming for a while at home just for fun are more qualified than most computer science students. I'd go with a Pro since you're trying to make money with your trading software. Imagine there is some stupid little bug that causes you to lose money. The money you lose will over time be more than what you saved employing a computer science student vs. a professional developer. As always, it all depends on the programmer. Make sure to let them show you a complete working application that they have developed before. I might be biased here since I am a professional developer myself, but there is a big big difference in skill, quality and efficiency between hobbyists and professional programmers just because professional developers have a lot more experience and had to work on some real challenging problems, not the silly problems you work on in school.

The comment above suggests that a scalp is the same as in another markets. What is the daily range of this market? I am not familar with IBs website. Can't somebody just tell me the commission if they already know?

Hey Soultrader, first of all thanks for the intro. Sounds like a great market for scalpers that use limit orders if the tick value is so large and bid/ask depth is low (depends on the commission of course). I have a few questions: 1) What is the commission with the local Korean broker 'NHFutures' that you mentioned? 2) Are the KOSPI 200 Index Futures and the U.S. stock index markets correlated? 3) How do you know it's largely retail driven? Because of low sized trades?

What's the round turn commission on the Mini Hang Seng Index Futures? I assume it can't be much if traders are scalping it for a few ticks since the tick value is so small.

Why does everybody quote Steenbarger. What does he know? He is not a professional trader. He even admits that he's not terribly successful at trading. He's just a psychologist that found an interest in trading and tries to relate everything about trading to psychology or things he found interesting about his clients. I do not disagree with Steenbarger in this case since what he says about edge is very true, but what he is saying most of the time is not going to help anybody trade better (it's more likely you become a psychologist).

I would like to know if anybody knows about a good source (e.g. study based on exchange data) that tells you the current average holding period in the U.S. stock indexes and U.S. Treasuries. I've found a study for the European Futures markets (Dow Jones Euro Stoxx, DAX, Bund, Boble, Schatz) for the years 1999-2002 which I found very interesting. It's called "How Short-termed is the Trading Behavior in German Futures Markets?" by Gregor Dorfleitner and it shows that the average holding period the stock index futures decreased by a factor of about 3 in those 3 years while the debt futures have a generally a very short holding period. Here is a table to give you an idea of what it shows: FESX: from ~21 days to ~7 days FDAX: from ~15 days to ~5 days FGBL: from ~2.25 days to ~2 days FGBM: from ~3.5 days to ~3 days FGBS: from ~3.75 to ~3 days This suggests that there are many day traders in these markets. I'd like to see a more current version of this study, but I am especially interested in the U.S. markets. I think this information could be useful to give you an idea on how much data to look back. Do you think it make sense to look at price and volume that is several months old when the average holding period is shorter than a week?

Thanks brownsfan. Sorry, did not look at the date. This thread popped up on the front page of Trader's Laboratory, so I assumed that it was current. Well maybe, Soultrader, if you read this, would you mind updating this thread with your current "tools of the trade".

Thank you Soultrader. Very interesting. I was always wondering what kind of tools you use and what you use them for. It would be very cool if more traders share their "tools of the trade". Do you use the same tools and layout for every market? I had the impression you were also trading Japanese or Chinese markets.

Thank you sevensa. Yes, that's a very good point. I have not thought about it that way.

I was aware of this, and I guess this is what troubles me. Does it make more sense to include the volume of expired contracts which is potentially irrelevant (since nobody has any open positions/interest in it anymore) or to use the volume of the front month contract only where there isn't much volume before it became the front month contract. Intuitively, I am thinking the later, because those traders that wanted to keep their positions had to roll it over to the next front month contract anyway when their contract expired (even if this aren't that many), so the current front month contract should include all the volume that is relevant. If that's a lot less volume then so be it. What do you think? Does my reasoning go wrong here somewhere or am I missing something? Why should we include the volume of a contract that has expired?

Sure. I would of course trade only the front-month or almost front-month contract, but my question was with regards to which contract to use to calculate the longer term VWAP. Wouldn't it make sense to use only the volume data of the current front-month contract going back 6 months or longer since traders don't have any open positions in the expired contracts anymore? I thought the whole idea of looking at the volume distribution and VWAP was to get an idea of the commitment of traders - at which prices traders have traded the most or least (correlating this with traders having open positions). If you use a continuous contract, wouldn't that include positions that have already been closed and therefore don't matter anymore? Longer term traders would have rolled over in the front-month contract anyway as you have described. Sorry, I don't wanted to take over this thread with my questions, I just wanted to fully understand the reasoning since I am still a little confused in that regard.

Jerry, thanks for your reply. Is there are reason why you use a continuous contract consisting of 3 months front month contracts instead of the current front-month contract that might have been trading for 7 months already (mini Russel 2000 also trades for up to 9 months)? I know this is not really possible for a 1 year VWAP, but in case you wanted to calculate the 6 month VWAP, wouldn't the later approach be more accurate since you would only include the trades in the calculation that were actually traded in that contract? Does it make a difference or am I making things more complicated than they should be?

I did not say that it's a bad thing and it has nothing to do with being too optimistic. It's just incorrect. For example, The Estimated Position In Queue (EPIQ) that X_Trader calculates is actually the worst-case scenario since they calculate the worst position that you could possibly have. Your position is most of the times better since contracts in front of you might have been canceled. There is no need to add some randomness for not executing your order like NinjaTrader does. I am just saying that it's unnecessary and probably hurts more than it helps since you can't get a realistic feel for getting executed (there are trading strategies that are based on this fact, so I think it's important).

Yeah, you know, talk is cheap. Why don't they just post their algorithm, so we can see for our self if it's realistic or not. I personally tried NinjaTrader's Sim and it was totally unrealistic (but not necessarily in a bad way). I'll give you an example of what I saw. The are 100 contracts offered. I join the offer (that means, I also put a limit order at the offered price). That means 100 contracts are in front of my contract since the markets I trade use the FIFO algorithm for order matching. So when 100 contracts trade at that price, my contract should execute next. There is no way around that, that are the rules of the exchange. But in NinjaTrader there might come 200 contracts after my contract, 100 contracts trade (so I should be next), but then another 100 trade and you are still not executed. No matter what they do, it's just not correct. You don't need some fancy algorithm (which obviously does not work correctly) to determine whether to fill a limit order or not, just use the exchange rules, but I guess the NinjaTrader developers don't exactly know how that works.

Yeah, it's very important to understand how order matching works. That's why I said, the only realistic simulator I have seen is X_Trader because they do it correctly. NinjaTrader does not do it right either.