BlowFish

I have it on order, should arrive in Europe this week I guess. Can't promise a review. What interested me was the patient buyers patient sellers paradigm. It's a slight twist on looking at things from the point of view of aggressive buyers sellers, which seems the 'norm' (I guess EMC2 would call them anxious) . Uses similar metrics as far as I can tell (Utick vs Dtick rather than V@Bid vs V@Ask, same deal really). I probably drink way too much mediocre tequila, strong beer and fine wine, all it will take is a couple of interesting pages to make it worthwhile.

Sure there will but the will be hugely less likely. Look at RoR all the answers are there, the risk of ruin diminishes hugely as the % rises. Who would not want a smoother equity curve?

yeah by assuming too much risk Which is how he lost similar amounts in a tenth of the time! Still the late great Livermore was prone to ....how shall we say..... over extend himself, so he is in good company I guess.

That's just one way of doing it. The most important metric to focus on is risk of ruin RoR. Funnily enough higher percentage approaches that have smaller R:R (or even fractional) have much smoother equity curves with smaller strings of losers. The type of approach (lower % higher RR) you describe is hard to trade as you will have to deal with 10+ losers in a row. Oh and pigs get slaughtered don't they?

Indeed, to get the fantastic results they publish the systems are usually hugely curve fitted to historical data. There are lots of 'tricks' that the less scrupulous can use to produce fantastical results on historic data sadly things will unravel going forwards.

Thanks for your detailed reply. I don't quite understand how you compare bid ask and last for historical data. If over 1 second you get 3000 ticks (lets say) of bid ask and last data there is no way to tell the order exact order they came in as bid ask and last are held in separate databases? The sequence information is not preserved for those 3000 ticks. You know the order of all the bids and you know the order of all the asks and of all the trades. There is no way from MC's database of knowing the sequence across the three databases? Are you somehow reading straight from the IQ history and by passing MC database? maybe using Qcollector or something? I hate saying impossible (nothing is impossible right ) but I feel pretty safe saying that MC wont sequence historical data from it own database correctly, that's why I wonder if you are constructing the history directly from Qfeed somehow?

Just looked at the charts more carefully. They look quite close in 'shape' (though not quite the same) but the absolute numbers seem quite different (maybe an artifact of how things are reported). I notice that the bands (presumably VWAP & SD Bands) look slightly different. There are a couple of discussions on those calculations on Jperls trading with market statistics threads. I wonder if you are re weighting all the old bars when a new bar completes? Just thought I'd mention it though I am sure it's good enough.

I would argue that TT is perhaps not a great benchmark! Arguably DTN.IQ is currently the one to judge from. The thing is it can all change again within 3 months! If one is using delta to gauge order flow or resting inventory it is not a precise measure anyway. It's a question of whether things are good enough. Of course that is no reason not to strive for perfection and to pressurize the companies that make this software to offer greater precision. Out of interest how do you deal with the fact that MC only has 1 second precision for historical data? This kind of make the 30 day history form DTN academic dosen't it? I guess you can only run your indicators live? This is actually my biggest criticism of both MC and NT as it make any sort of qualitative testing on historical data ....well.....impossible really (without writing your own database and data handling).

Always find your posts interesting Steve apart from one thing....this whole 'institutional' vs 'retail' thing. 'Institutional' traders have all sorts of motivations and modus operandi. Some of the very largest are not even 'profit motivated' in the traditional sense. Just as well really, as an ongoing study of the CoT will show 'institutions' to be on the 'wrong' side of the market just as often as retail traders!! It rather smack's of Tradeguiders 'smart money' rhetoric. Your analysis stands on it's own merit, there is no need to try to add extra weight. I only mention it as I feel it distracts from otherwise excellent posts. It is not necessary to know who is trading, why and how. If one is interested then a proper study into market microstructure would pay dividends (Harris Trading and Exchanges: Market Microstructure for Practitioners is a decent starting point). There are many types of 'institutional' trader that trade in vastly different ways with quite different objectives (which to be fare you acknowledge at the start). I guess what I am saying is, a better question is what would an informed trader be doing vs an uninformed trader (assuming that they where both speculating on the short term direction of the market). To answer the original question one would need to know a lot more about the motivation of this 'institutional' participant. BTW have your company raised your forum embargo or have you changed jobs?

Resting inventories will be the problem. This requires accurate order book data and charting that has tick precision across multiple data streams. MC does not have the internal precision and is prone to sequence issues and race conditions across multiple data streams (like bid ask and last). In fact race conditions can occur with a single data stream. Profiles of actual trades are pretty straight forward and if you look around you will find public domain snippets for both MC & Ninja. Have a look for Gomi's stuff for ninja.

But there is no point 4 on the image? I think what Waveslider is saying is that a WW requires points 1 2 3 & 4 to be 'in' (described by price action that has already occurred). In your image price has not reached point 1 yet and does not have points 4 & 5 drawn at all. A truly remarkable feet if it pans out . Could you draw in a 1-4 line and your final target projected from the intersection of 1-4 line and point 5?

The participants, me, you, whoever. Anyone that places a limit order through their platform.

Both parties can open, both can close, or one can open and one can close. In the example you posted C is closing by selling the options they bought on Jan 2nd, E is a brand new buyer. Lets say the brand new ES contract has just started trading. None have traded yet. I can sell you 5 (by going short). I am short 5 you are long 5 and open interest is 5. Clearly no one is closing. In fact if someone had to close, no contracts would ever trade (as no one has a position in the new contract). You do fully understand the concept of being short?

Re OI, when one party is opening and one party closing OI remains the same. C is simply transferring their holding to E. On the other issue, I guess it depends what you mean by 'market maker'. If I place a limit order to sell 100 ES @ 1200, I am essentially 'making a market' at that price. If I simply place a market order to sell I am taking liquidity from someone 'making a market', that could be Kiwi sitting with his limit buy order at 1199.75.

Keep at it ...

It might be worth noting that the adage holds true for most markets. 'Zero sum' is an expression that often crops up around the same time. Take equities for example, if you want to sell you need a buyer however if you want to sell short you need to find someone that will loan you the stock. Stocks also have 'floats' that further restrict things (there may be willing buyers and willing sellers that can not trade with each other) You might want to google 'open interest' too

That is one of the perennial issues imvho. When is a climax the climax. Quite often you will get what looks like a climax in an area where you might anticipate demand, only for price to move lower and do the same at the next area, on even higher volume.

Worthy of it's own thread Had planned to answer the OP but most has been said. Forums are well .....forums, they haven't changed much from BBS's (Bulletin Board Systems) that pre date the WWW by an age and even the commercial availability of the internet (that is waaaay older than the WWW). People hang out that share an interest, that's about as far as it goes.

What's to say really? The trading products and service business is choc full of all sorts of dubious operators ranging from the out and out con men to the mildly deluded. You have issued your warning, what purpose is to be served in a public slanging match?

Enough to fully understand and calculate risk (and hence position sizing). That is basic + - * /. A real beginners understanding of probabilities would do no harm too. This is simply to understand that trading is about consistently applying an 'edge' rather than individual trades. Not much, not much at all.

But trading isn't rocket science. In fact it is pretty darn simple, it's just not easy. As an example I have just read Curtis Faiths books on the Turtles. The first batch where taught to trade in 2 weeks, the second some time later they cut to 1 week. The approach was very simple though some found it hard to follow (I guess with 70% odd loosing trades not surprising!) Anyway... I would say if you have the slightest doubt, pass. If you search around you will find a couple of threads on what to expect from a mentor, what a good one will likely expect of you and ideas of questions you should ask them. Try here and ET. If my memory serves a poster called 'NihabaAshi' was a contributor, that should help you narrow it down.

Hi Bill, Finalg is one of payment. Ninja licensing is a bit odd. It's free if you have a live account with one of there brokers (Mirus & Amp being the most common futures one). Certain things (that your probably wont need) cost extra (one off) license fees. Multibroker support springs to mind. I am not sure if there is some sort of minimum required account activity, I do not believe so. Many people just sign up for a free demo account with the aforementioned brokers when they expire they re-sign. From that point of view it is completely 'free'. I think nowadays Mirus prevent you signing from the same email address within a month.

I disagree on part of this:). Sure it needs some re-architecting (it was a mistake to copy TS's limitations). Historical data needs proper sequencing information and/or millisecond or better time stamps, that's simply a design limitation (as you say) they will have to fix it eventually to remain competitive. I have had some indication in emails they will but who knows when? Arguably real time data handling is 'bugged'. It is possible for race conditions to exist, they seem to have fixed a lot of the issues there but it is possible to have variables (like close) change between the beginning and ending of a function. That should plain just not happen. This is not really a design decision, it is that the internal architecture is simply not 'robust'. (Probably some multicore issue). I would go further, if you want to do any work where precision is important MC perhaps is not the best choice. It's a shame really as it's a decent product that I have a lot of time invested in (since version one if memory serves).

A while back there was a bit of discussion on Cutis Faith's books. I just thought I would report back and say I am a good portion through the first and am pretty impressed. For a 'system' trader I might even be tempted to use the rather dangerous "must read" expression. Also very impressed with the latest kindle (three). Shame it is not water proof so you can read it in the pool.

Keep us posted how you get on