Welcome to NexusFi: the best trading community on the planet, with over 150,000 members Sign Up Now for Free
Genuine reviews from real traders, not fake reviews from stealth vendors
Quality education from leading professional traders
We are a friendly, helpful, and positive community
We do not tolerate rude behavior, trolling, or vendors advertising in posts
We are here to help, just let us know what you need
You'll need to register in order to view the content of the threads and start contributing to our community. It's free for basic access, or support us by becoming an Elite Member -- see if you qualify for a discount below.
-- Big Mike, Site Administrator
(If you already have an account, login at the top of the page)
the idea behind the order flow grail is you can use a very small stop .. a few ticks not 4 or 5 points on the ES for example ... sure it can be done its possible. but it takes less time and pain to look for larger targets 12 to 20 points. you also suffer less slippage , commission , over trading , cost of the grail software and training ..ect. ect.
Given the very short time frame of the bars, and the fact that CL usually has a fairly thin book of resting orders, my guess is that a single institutional OTC order caused that sharp a move. It could be a speculator or hedger, but I’m pretty sure a fairly sizable institutional investor probably had an order to fill, was not overly sensitive to price (perhaps a position or risk limit was exceeded and they had to downsize their position), and the order caused the sizable, but very temporary price fluctuation. I usually trade Treasuries, which is a much thicker market than oil, and usually all the longer duration contracts (ZB, UB, TN, ZN) are in very tight correlation, but sometimes I will see one of them quickly move erratically , but quickly correct, and I can tell from the timing of it that it’s an institutional order from a customer that doesn’t really care that much if they leave ticks “on the table”.
A combination of factors. The chart pattern just before the break shows wedging / coiling / consolidating price action with a fairly well-defined trend support line. Once this support broke, it likely touched off several stop orders, which simultaneously hit at the market and took out a large number of resting bids. While this happened, market makers who got hit pulled their remaining bids and went offer (basic position and risk management), while range / volatility breakout traders and algos jumped in with their own sell orders just as liquidity was being pulled.
So it was not one large trader slamming the market for several hundred contracts all at once (no one trading that kind of size would do something that silly), it was several traders who were triggered for different reasons within a very short time-period. Actually quite similar to the flash crash, although on a much smaller scale. Happens all the time in crude - so much so that some traders build strategies around just this kind of a move (the range breakout traders alluded to above).
Look at a footprint chart for this move on a small range or reversal bar periodicity and you might glean some further insight.
At first I thought this was an old dead thread, but there will always be an interest in Order Flow Analysis, or what I call "Inventory Analysis" as a predictor of when Markets may end a trend, or pivot turn the trend, or hold support or resistance levels; all of which can be understood and well predicted when "Order Flow Analysis" is properly done.
The biggest issue is that "real Order Flow Analysis" is computationally, and memory-intensive. It is also "fractal" in the loose sense that Inventories are moving Net Long and Short dynamically, are are dependent upon the time width of the Analysis interval chosen.
First of all, I used to sell software which implemented this stuff; but I no longer sell it. I don't mind, however, describing the rough algorithmic basis for calculation of "useful Order Flow Analysis" in case some of you are smart enough, and skilled enough to implement it.
Yes, "Order Flow Analysis" is useful; and what does it tell you? It tells you whether Market Maker (considered on aggregate) is net Long or Short against the Retail population (both small and large retail institutional traders, who are NOT Market Makers) and that is an interesting piece of information. BUT... it is by no means that useful in predicting a pause, reversal or continuation of a Trend.
The real money is in a further refinement called "Inventory Risk Analysis" which, in a nutshell, is telling you whether the Inventory is Long or Short, sure, that's informative; but much more important is "whether Market Maker is in the money, or is losing money on the current Inventory, on the current Timeframe" since everything in trading is "timeframe dependent".
Consider that "major swings of the market" represent Accumulation and Distribution periods, during which Market Maker (MM for short) is doing "Accumulation" against the Retail players, and perhaps doing that with a steady down-trend; or whether in a "Distribution phase" usually associated with an up-trend in Pricing. Anyway, if it were the case, that major trend changes occurred over, say, 2 hour cycles, then you'd want to Evaluate your Inventory over roughly a 2 hour period in order to appreciate MM's Inventory situation, assuming that you wanted to trade also on roughly that timeframe.
COST BASIS. For MM's Inventory, which is determined over a calculation of "all trades" which occur over the interval of interest, we need to first establish an estimate of COST BASIS. If MM is "Long", meaning that more Retail Selling is occurring than Retail Buying so that MM is holding Inventory which she wants to Sell off at a Higher price, then we want to know the Cost Basis price for the "Net Long Inventory". That is done by starting with "the Lowest Price Retail Sale Trade (MM is a Buyer)" and searching for a "matching Retail Buy Trade (MM is the Seller)" which is at least 1 Tick higher. Inventory which was Bought by MM; let's say 1 Contract, to Sally the (Retail) Seller; can be matched with 1 Contract Sold by MM, say to Billy the (Retail) Buyer; at a minimum of 1 Tick or more profit.
The preceding operation "neutralizes the single contract which Sally sold to MM and which Billy bought from MM at least 1 Tick higher". So that single contract "disappears" from Total Inventory. It is a special case, most likely of "a spread trade" in which Market Maker (MM) is able to profit on the Bid/Ask spread, where there is no Retail opportunity for profit.
So, proceeding from the Lowest price MM-Buy trade, we search upwards for MM-Sell trade(s) which "neutralize" that volume. We eliminate the entire MM-Buy trade, and we deduct from the MM-Sell trades, the Contract volumes which are "matching". If an MM-Sell trade volume goes to zero, of course, that means that MM-Sell trade effectively disappears and has been "neutralized" or "removed from this Inventory evaluation" due to matching volumes.
TRYING TO STAY out of the weeds here; let's just say we do this until we have processed and eliminated MM-Buy trades (the ones generated from Retail Sell transactions) by finding matching MM-Sell volume at a higher price. BUT THEN WE FIND THAT we have a bunch of MM-Buy trades WHICH DO NOT YET HAVE matching MM-Sell trades. So, we have a bunch of "pure MM-Buy transactions" which have yet to find any matching MM-Sells at a higher Price. THIS IS MM'S NET "LONG" INVENTORY and the Volume Weighted Average of this collection of MM-Buy trades can be taken as the COST BASIS for MM's LONG position (against the Retail Market) and so then.....
...THEN WE KNOW THE COST BASIS PRICE which MM must achieve in order to Profit on this collection of LONG inventory. If the Current Market Price is below which COST BASIS, then I say that MM is "In Long Risk" or, in plain terms, "is losing money" on her LONG inventory, at this moment, over the timeframe of evaluation.
TO AVOID YOUR UNDERSTANDABLE REACTION "tldr;" (Too Long, Didn't Read) LOL ... This means that when MM is "in Long Risk" then she will be reluctant to deviate "too far below" her Cost Basis, which is a very Dynamic calculation.
Obviously, the inverse operation would be used in order to evaluate a situation in which MM had an excess "Short" Inventory, and when that Cost Basis is calculated, then any Current Market Price above that price level, would result in "Short Risk" resulting in a reluctance to raise the Price much further, as "Short Risk" (just like Long Risk) means that MM is actually "losing money" at that moment in time, on aggregate.
TO END THIS DISCUSSION, this process of extending Order Flow Analysis into yielding a Risk Evaluation; is the only way I know to predict when a Market Price will be "reluctant" to proceed further into Risk; and often signals a Market Reversal.
DISCLAIMER AND APPRECIATING THE SCALE OF MM'S ACTIVITIES. We can measure a particular market over time, by experience, and appreciate just how much Risk MM wishes to, or is willing to, take; before deciding to "reverse a trend direction" or proceed in the same direction. Knowing the Total Net Inventory in contracts, say, and its Cost Basis, we can then literally estimate (pseudo-calculate) Market Maker's total Risk in actual Units of Currency. In a Market such as ES, we might see Risk levels in the hundreds of thousands or even low millions of USD. These levels of Risk are, however, just "business as usual" for the Market Makers in such a highly liquid market.
MARKET MAKERS HAVE DEEP POCKETS. Markets are Dynamic and they show "sawtooth" behaviors on multiple "fractal-like" timeframes, precisely because MM's strategy is to use Retail Trade Volume in order to "take a Long position against Retail Players, and temporarily dip into Long Risk" and then reverse, so that she can then swing the other way and "take a Short position against Retail Players, and temporarily dip into Short Risk" as part of the game.
In my view, this is the best and only way to truly use "Order Flow Analysis" or what I call "Inventory and Risk Analysis" to predict when a Trend may come to an end, and show some pullback, or reversal, perhaps at some of the Support and Resistance levels. These, of course, are what I call "volume inflection points" so we can expect them to be violated. But WHETHER A SUPPORT LEVEL BREAKS is almost totally controlled by the RISK ENVIRONMENT which exists at that moment in time, over a significant Inventory Timeframe.
SO THAT IS MY WAY OF THINKING ABOUT, AND CALCULATING Order Flow, which means "Inventory and Risk Analysis" and is my "Commitment of Traders" (COT) evaluation method in real time. Comments welcome; but you can appreciate how complex and how dynamic "real" (in my view) Order Flow Analysis needs to deal with the Time and Sales tape; in order to be of significant value in trading.
[EDIT] An equally useful pairing with Inventory Analysis is Depth of Market (aka "Book") Analysis to determine Market Price direction. DOM Analysis and Inventory Analysis taken together can provide a day trader with useful information, so maybe we can discuss somewhere else how the DOM introduces concepts useful in prediction.
How are you categorizing what is a market maker's order or what isn't? Do you mean to say limit orders vs market orders? How do you match if someone say enters with two orders and exits eighth one? How do you match orders up at that point? Otherwise it would basically just be cumulative delta right?
By my definition a Market Maker is the entity which Buys the BID price, to which a Retail player Sells;
and Sells the ASK price, from which a Retail player Buys. The MM never "pays retail" but is the Market
Maker, while all other players (large or small) are Market Takers and are subject to the Bid/Ask price
penalty, that's all...
So a trade on the Time and Sales "tape" at the BID is a Market Maker "accumulation" from a Retail Seller,
while a trade at the ASK is a Market Maker "distribution" to a Retail Buyer.
Those are my criteria for categorizing events.
Regardless of whether a Retail player uses a Market Order or uses a Limit Order; to be "Retail" s/he is
suffering the Bid/Ask penalty. On some "unusual exchanges" so-called "retail" players are allowed
to place Bids and Offers on the Market Depth Book. For our purposes, that converts these guys
into "mini Market Makers" (in a very limited way),
but they are not the large entities to whom the term Market Maker largely applies.
Generally when a Retail player uses a Limit Order, that is converted to a Market order at the
moment of execution. S/he does not "make the bid-ask" but is forced to "pay the bid-ask" penalty.
[EDIT] Any trader or Market Maker who has genuine access to place orders on "the
Order Book" will also be in competition with other Market Makers who "jockey for position"
on the Order Book on the Price-specific FIFO queue. If a Market Maker "pulls" his Bid or
Offer which is placed on the Primary Book on which the Matching Engine operates,
then that MM is penalized, and loses her place on the Book in the FIFO queue as the
first one entitled to execution at that price level, against Retail players. This is what
makes the Depth of Market or "the Book" interesting as a possible predictor, but that's
another topic entirely. Unless I'm wrong, please let me know.
[EDIT] The term "Order Flow" is misleading; what I really mean is "Trade Flow" since
only Executions (Time & Sales) count toward the Inventory Analysis which I use.
Order Flow, in the
sense of Bids or Offers on the Order Book which DO NOT execute, have a different
role in giving us insight into where Market Maker intends to forcibly move the market
price. Hope that helps?...
Ok that makes sense and answers one part of my question. The second part had to deal with how do you match orders up. If you are just adding and subtracting all the orders that hit then you basically get cumulative delta. But it sounds like you're looking for the lot size to match up exactly? So if mm bought 10 at one level and sold two 5's at another level since the size didn't match up you'd still say there's 10 net long and 10 net short?
there have been some very good post on order flow concepts. however most traders me included would never be able to master the advanced concepts. give us some thing easier. when the market opens the market makers will have retail and institutional orders . depending on the size and where the orders come from they will short them or add to them with there own inventory . tell us how to tell with out a masters in math ..
