BlowFish

Dunno what filtro is? By the way the indicator works on volume traded not ticks. Feel free to change it to suit your needs but I think it is correct (with the caveats earlier in the thread) as it is. Of course I have been known to be wrong

Hey UB, do you know if they plan to offer historical best bid / best ask data (or at least flag trades that occurred at BB/BA)?

Jim. There are many many different types of participants. They each have different motivations and quite different methods of operation. It is a vast vast business ranging from second and sub second market making to hedgers protecting against a variety of changes in physical costs. Participants range from governments to little 'prop' shops. Earlier in the thread I pointed out that it is naive (and will never result in a good understanding of how markets really work) to lump everything together as 'institutional'. I hope you don't think I am having a dig Jim but this post illustrates that. I mentioned a reference earlier if you want to know who trades, why and how.

Indeed. However it a does nswer his question pretty fully. I can't think of a better source describing how large orders are worked, Incidentally it is not just his successes that hold important lessons.

Have a read of "Reminiscences of a Stock Operator" by Edwin Lefevre. It is a 'fictional biography' of Jesse Livermore. Apart from answering your question in some detail it is jam packed full of trading wisdom and insight. Truly a classic.

There is a series of threads if you work through them it is all described.

Hey ES, Reminiscences is a fantastic book, it is no exaggeration to say that I have read it a dozen times or more over the years. I like it on many different levels if I could only choose one trading book it would be a strong candidate. I think we have got as far as we are going to get on stops and such. I am sure people must be getting bored by now Having said that I just wanted to clarify that the interest is not in stops at all. it is in whether an order is a limit order vs. everything else. A limit order provides liquidity pretty much everything else takes liquidity. A participant using a limit order is passive/patient they may get a fill they may not they certainly do not demand immediacy. Market orders (however they have arrived at the market) are from aggressive/impatient participants, they demand a fill come what may. Anyway glad that you have found what resonates with you, I'd agree tape reading is a pretty 'pure' in so far as you are as close to the market as you can be. I wish people talked more about the practicalities of it. I guess it is such a dynamic thing that it does not really lend itself to static media. Still if you ever did a thread I'd be right there.

Actually I don't believe I said anything of the sort and I guess that is where the main problems lies. You are disagreeing with what you think I said. I simply described how orders are matched on nearly all centralised electronic exchanges and pointed out that stops are never traded against directly, they are converted to market orders. If you then treat limit orders that are submitted within there limit prices as market orders too (as the exchanges do) you only have two types of order to contend with, limit orders and market orders. There are data feed providers that provide a full data feed, DTN's IQFeed is the one that is commonly recommended. You might want to search for Fulcrum Traders posts on the subject. As for infrastructure I see you are in the UK, are you out in the sticks? ADSL is more than up to he task in fact I have used Zenfire with good results (though some people have issues with odd bits of data missing) it is a 'full' service though it uses UDP so if your infrastructure is poor you can loose the odd chunk. My 'main' broker is IB but there data is not good enough for this work. I'd be interested in any book recommendations you have on tape reading, one of the big problems is a distinct lack of them. I agree it is as much art as science but there really is not an awful lot written about it. There are classics like "Studies in Tape Reading" (Wyckoff under the nom de plume Rollo Tape) and "Tape Reading and Market Tactics" (Neil) . There are a couple of more recent modern ones Like Techniques of tape reading (Gaifer) and No BS trading (ebook). I have a few obscure titles too but nothing that really compares to those, any you would recommend? You are using logical fallacies to argue your case and attributing things to me that I did not say and clearly did not mean. Of course traders should not fuss over T&S or if an order was a limit or a stop. What I said was that traders should know the rules of the game they are playing or they will fail, plain and simple. This includes but is not restricted to exactly how different order types work, where they are held (most important), under what circumstances they are elected etc. They need to know how the orders they are submitting are handled. There is a whole bunch of other stuff they should know too if they want to avoid being on the wrong side of a market that opens limit down 3 days in a row too.

What I find it mind boggling is that people trade without an even rudimentary understanding of order types, exchange rules and just a basic understanding of how the environment that they plan to operate in actually works. It is hardly surprising that so many fail. These things are absolutes they are unequivocal, they are the rules of the game there is no reason not to know this basic stuff (except perhaps laziness?). In just about any other endeavor you just would not expect success without knowing the rules of the game. You may think it academic I think it is fundamental and mandatory. Whether it is useful in determining market direction is academic for the reason above however it certainly can be helpful in that regard. All of these new wave of studies based on 'market delta' that attempt to detect order flow, resting inventory, buying pressure etc. rely on measuring volume@bid vs. volume@ask. I too was skeptical if there was any real value in this as some of the core premises I felt where flawed (to do with how different participants behave under different conditions). Still, traders that I respect use this information profitably so I decided to investigate properly. There have been numerous studies been conducted (by academics granted ) on the efficacy of this approach in detecting 'order flow' long story short it is (72% - 80%+ depending on the study). That gave me some confidence that conducting my own research would not be time wasted. It was not. Of course others mileage might vary.

You can determine that they where not stop orders as stop orders do not trade directly. It is not 'academic', that part is wrong I would not have bothered chiming in it but it is impossible to have a complete and proper understanding of how orders are matched without understanding how stops are held and matched with limit orders. Of course for simplicity you can ignore stops (as they simply become market orders when elected) but since you mentioned them I thought it important to understand properly how they work. It is the basis of all sorts of cool stuff too. It is also important to realise (which is why I am re-iterating it) that limit orders below the market are to buy and stop orders below the market are to sell. Limits above the market are to sell and stops above the market are to buy. The post I commented on clearly showed best bid and best ask no where limiting things to just T&S is moving the goal posts! Anyway that information is available through any credible platform. If a trade is at the best bid it is a market order to sell. It could be one of the rare cases I mentioned but did not describe, a limit order to sell that 'crosses the market'. To all intents and purpose (including liquidity rebates, exchange fees, order matching etc.) that limit order is as a market order as the market is already at the limit or better. Anyway hope estrader is getting something out of this.

Why do you want to buy them? There are great indicators in the public domain that cover pretty much all you could want. Btw indicators won't make you a trader (many would argue they will delay the process). The fact that you are asking for 'great indicators' suggests you might not be aware of this fact.

This is all true except stop orders are converted to market orders when they are elected (there trigger price is traded at). Stop orders are invisible on the book (usually) though in some markets some participants can see them (like a NYSE specialist). So, the bid you see are limit orders to buy at that price. There are also stop orders to sell at this price (they do not show on the book). When elected these will be immediately converted to market orders and so match against the current limit orders at that price. If there is not enough volume to satisfy them at that level (in the form of limit orders) they will be matched to the next level as they are now market orders. This is one of the reasons you see price spike on high volume at S/R levels or break outs it is also why you can get slippage on stop orders. Short version - There is enough info to determine if a stop or limit order was filled as stop orders don't get filled directly, they get elected to market orders. There are a couple of special cases but that's likely to detract from the bread and butter working of centralised electronic exchanges.

I should add that there is a point to my pedantic bickering and that is this.....it is not a trivial matter to find an exploitable edge based on market A leading market B. What does seem to happen is that you spot something that appears extremely reliable for days or even a week or two and then bang the relationship just disappears or even reverses. I am always sceptical when people say the YM leads the S&P or the S&P leads bonds (inversely) or the US leads the EU during the NY morning. I think you need to be pretty fleet of foot and quite sophisticated to utilize these phenomena.

Indeed, however originally you suggested one could look at volume to determine the 'leader'. Arbitrage is a complicated business with several variables. Again being devils advocate I would argue that it is just as likely that the market that has moved has moved away from 'value' and will return to the others. An arbitrager would not care which scenario occurs as they would be short the the market that had risen and long the markets that had not. They only loose if something fundamentally has changed and the divergence continues. As an it is interesting to plot the relative strength of a pair of markets (like the ES and YM for example) its interesting how far and how long out of line things can get.

Exactly. PTJ had a right hand man that did the analysis, they would agree they wanted to be long x DM and then PTJ would get busy 'working' the order. If you are a buy side trader working a large order your job is to get filled at the best possible average price. It's pretty obvious. Is it an 'oh shoot' emotional response or is it a planned entry where you are fully aware of the risk parameters throughout the process.

If it was that simple trading would be trivially simple. My observations actually suggest the reverse, illiquid correlated instruments tend to move faster and further than their heavy weight brothers, a bit like flies on an elephant. The "arbitrage relationship between the S&P and all other world stock indicies" is not great despite them being correlated. Convergence is likely to be slow, if it occurs at all (one of the bigger risks of arbitrage). Also the arbitrage relationship is non stationary for example it is not hard to see that changes in underlying currency exchange rates are going to have a large effect on any arbitrage opportunity,

Pretty sure there are a couple of occasions in the famous PTJ video where he is buying into a market that is moving against him. Of course I could be completely mistaken. Wasn't there a segment where he is buying a falling Deutschmark all day? Finally stops out with a fairly substantial loss. Dunno it's a while since I watched it. I guess it depends on your definition of 'loser' if price is still in your 'buy zone' (even if lower) then perhaps it is not a loser.

There are a couple of variables you can use or you can hold one constant and use the other as a measure. Of course how you sample (slice & dice) the data will largely depend on the reversals that you are interested in analysing. Let's say you are interested in reversals of approximately 10 minutes, you might choose to use 3 minute bars and look at 3 bar reversals. It is not hard to conceive that a 10 point reversal (i.e price travels 10 points from the swing high 3 bars ago) is stronger than 3.25 points in a 3 bars reversal. Price is the obvious 'indicator'. If price moves '23 ticks' away from an extreme that is stronger than if it moves '22 ticks' away. Conversely you might hold price constant and see how long it takes for a move away from an extreme. In this case you might conclude that a 10 point move away from an extreme that takes 3 minutes is stronger than a 10 point move that takes 5 minutes. There is really no need to use fancy derivatives (indicators) of price or time when you have......price and time. There are other variables that you might want to sample against e.g. volume, ticks, market delta.

It also raises an interesting question 'do markets rise because of buyers wading in or sellers pulling out'? Of course it's a bit of both but actually I think it is often predominantly the latter.

That was where I was going next you beat me to the punch With TSSuport you will need to keep right on there case if you want it implemented properly, nice bunch of guys but techies rather than traders.

If you can hit an order before they can pull it the market will trade there. They are actual resting orders which can of course be pulled. BTW, When I traded on the Nasdaq many moons ago you quite often would have the best bid equal the best ask (a spread of zero if you like). In fact the best bid could be higher than the best ask (a 'crossed market'). This was because the Nasdaq is not a centralised marketplace, it is a bunch of different 'networks' each with there own queues and matching algorithms. Of course this situation would get quickly arbed out.

Sure it is though you need a robust system and preferably to diversify (that will prevent many a wipe out). Funnily enough simple systems with few variables/parameters tend to be more robust. If it only works with tweaked parameters chances are it is garbage. You need to run it on out of sample data or it will fail going forwards. The only real way of doing this is seeing live results from a real account. You could run tests yourself but you will never know if they optimized on data you are testing against. Take a look at LS Chads video sequence here to see how easy it is to put together profitable systems.......until you run it on out of sample data. I just read Curtis Faiths Turtle book on holidays, good read if you are interested in evaluating systems.

Best to offer a choice, most would prefer entry stops/limits to be held client side other wise technical issues can leave unmanaged orders at the exchange.

Does kinetic provide full inside bid inside ask data like IQFeed (30 days worth I believe)?

Right! When I was a kid it counted for something nowadays my cleaner is one. Incidentally in UK English John Thomas is slang for *cough* a penis, I'd be wary.