Order Flow Analytics' D.B. Vaello (President) - Ask Me Anything (AMA)

February 25th, 2013, 11:02 AM

PrymeTyme

PriceAction =
The Result in Price Change after Trades/Transactions are executed

OrderFlow =
What Happens Before PriceAction

Correct ? Right ?

please comment and add or critcic my POV.. TY

Plotting time and sales on a chart in rows and columns is just that - glorified time and sales - particularly if you have no idea what the data represents or how to use it. Sounds like you do so that's great.

I'm in agreement with you on market depth - I think there is little to be gained for the work that is put into it. Any claimed edge is subtle and should be treated as such.

Price action - check
Order Flow - sort of. I'd say it's what happens in order to result in price action.

Again - it relates back to having a process to make sense of the data and act on it. There are 100 companies that plot glorified time & sales. Most have no strategy for how to use that information.

DB

February 25th, 2013, 11:08 AM

FalseProphets

Isn't every software just a tool, a utility that bestows its benefits in varying degrees based on the user?
Sure, bid/ask analysis with an eye on immediate market depth can be accomplished by seasoned DOM traders. Do you mean to claim that you can watch the DOM for a few hours and know exactly when volume clusters ramp up in magnitude, sometimes a precursor to an immediate reversal or continuation, and you do that for multiple instruments simultaneously?? Even if you are blessed with photographic memory, why put oneself through that level of strain undertaking an activity that is stressful enough to begin with. Software, among other things, is supposed to make somethings we already do, easier.
It appears you went to great lengths to agree with most points on DB's explanation only to arrive at your main point that you don't need any software because you are really really great! Orderflow edge is not as it used to be since block orders are a thing of the past. But for scalpers, there is still plenty to exploit. By the way, any volume analysis tool that doesn't live and die by cumulative delta is already a cut above the rest.

P.S. I am not affiliated with DB nor am I a customer.

Thanks FalseProphets. My only objective with the software and training is to allow people who can't make sense of the info to understand and apply it logically. I really cannot stare at time & sales in a window and make sense of it. Nor can I gain much insight from endlessly staring at the DOM. But there is underlying logic that can be applied to start to make sense of it when arranged and explained. As stated earlier, CD doesn't make sense to me logically so that is why I dont use it or recommend it.

DB

February 25th, 2013, 11:17 AM

Titan

Didn't see this question asked unless I missed it but what instruments does OFA work with?

Sorry about that. The current version allows the user to input any instrument they want - so technically anything. That said, the original design was for futures, not stocks - so it's generally a disaster if you try to input ETFs or similar. There might be a few limitations on instruments with a non-standard scale. We had to do some modifications for notes and bonds - simply because we never planned to make this available to these markets and didn't plan for it.

OFA has about 20 formatted instruments:
ES/NQ/YM/TF
6E/6B/6J/6C/6F/6A
CL/GC/QM/Corn/Wheat
Dax/Bund/FESX
Bonds & Notes

And a custom symbol input field

DB

February 25th, 2013, 11:47 AM

RichardHK

OFA DB,

We hear much talk of needing to follow the large size institutions (despite some obviously losing like the majority), so lets call the winning guys 'smart money' for these questions. Everything relates to the ES market.

Q1. What are the most common footprints in the order flow that smart money leave for us to follow?

Smart money = the guy who didn't get stopped out. If you think that's limited to the institutions (which clearly you don't) then you're nuts. The best thing you can look for are the signs of accumulation or distribution at a concentration of prices (drawing in substantial trade without substantial price action) and the subsequent unwinding of the losing side being unable to find market depth to unwind into. We look for a pretty specific momentum pattern which is instrument specific - but in the ES we would expect to see sequential clusters of trade followed by a thin print in the target range of stops - let's say 8-10 ticks for the sake of simple conversation - and I'm sure some vendor-hater will blast me for that. But the point here is that you should have regular expectations of where losing volume will unwind - and when it's not met by depth trying to take advantage of the inventory, you will see very clear patterns in the print.

Quoting

Q2. How does your OFA logic method identify these footprints? Can it be done, albeit with much difficulty(?), using standard 'tape reading' tools like the T&S, DOM, and such?

I dont think I'd really be able to identify patterns of volume by reading the time & sales or DOM - which is why I built an application to do it the way I wanted to. There are posts here by others that indicate they know the secret to this magical process. I dont and cant claim that I do.

Quoting

Q3. How do we differentiate between smart money and dumb institutional money at work?

Start by realizing that your real competition is the retail trader - not the institutions. The retail trader is generally an under capitalized neurotic consumer of magic numbers. Take his money, it's easy.

Quoting

Q4. How do we handle HFT/other big guys action that flips aggressively between bid and offer heavy limit orders being placed and pulled (within seconds often)? Or do we just not trade at this time, despite practice getting very common?

I posted earlier about order flow really only providing usable feedback when there is directional movement. If I get into a trade with an expectation of momentum higher - and instead we churn across a small series of prices - I generally unwind my position because the expectation has failed to deliver. Imagine, you go long and price moves 4 ticks up and down for 2 hours... Is there directional movement? Will you get accurate feedback from the data? Would you hold that position? For me the answer is no to all 3. I try to limit my "reading of order flow" to specific criteria defined by my tactical approach to the chart that day. I'd lose my mind if all I did was watch volume prints all day. Probably not the answer you are looking for but at some point you have to agree that when a trade is not doing what you expected you need to go back to the drawing board rather than hope for it to work out.

Quoting

Q5. Have you compiled a list of typical 'games' that smart money plays to take others' money? If so, are these built into your OFA logic? Would you share same games list for others to comment on?

I would say the most common is just a very simple stop-run at ridiculously obvious places on the chart. People like to buy old lows and sell old highs - or buy/sell off trendlines, etc. And typical momentum traders like to bracket the same areas. They get beat up a lot and you can see it happening when you organize the data - but then again, I think a good trader can see this happening with a basic OHLC chart.

Hope that was helpful and gave my critics some good ammo - I hear it's in short supply these days!

DB

February 25th, 2013, 11:59 AM

artemiso

Which data visualization library are you using? Infragistics?

We are not using any library. We have our own graphics and drawing code.

DB

February 25th, 2013, 03:31 PM

I never got my 2 posts / questions answered,

scroll up please or pm me on the answers

thanks

February 25th, 2013, 04:14 PM

I have an OFA License that I do not use. I invested $3000 or more, but would like to sell the license. Can I do so?

David

February 25th, 2013, 06:12 PM

Sorry I was going to PM you earlier when I replied to the first batch but then the Euro started producing and I forgot about your other post...

kronie

A) do you teach your students to avoid the pitfalls, traps and tell-tales that you teach them to look for?, namely keep your:

A1) stops server side so as not to be spotted, and then only allow or preprogram the software to release the stops, when market is touched? or similar?

Totally depends on the client. We obviously teach them not to have their stops where we expect the bulk of stops to be triggered in directional moves. Outside of that I don't address how a small lot trader should try to hide his stops somewhere in the market depth. They have enough to worry about just managing a position. The stop is the price that invalidates your initial premise for taking the position. It might be modified slightly from the positioning of other nearby trade - but the song remains the same in the end.

Quoting

A2) using that premise, when the server side orders for stop and targets are managed in that manner namely hidden from the general order book, then how do you account for the change in or increase of bids/asks at the offer?

simply put, actually, there's no way to simply put this concept and question.
if you keep your stops/targets hidden and submit when market is touched, then they show up as recently added, and then contribute to the concept of increased participation (valid conclusion); increased demand (invalid conclusion as these are defensive orders that should have been placed upon order being filled)

Simply put, huh? I don't buy the premise so I don't hide my stops or worry about the things you're asking about. I put on the trade, manage it and then put on another. Maybe this is more relevant in thin markets? It could be I'm just ignorant on the subject.

Quoting

A3) since most of the conclusions derived from watching the at bid cross price and the orders and the initial verses most recent quantities at that fast moving level, then what can be derived from that analysis, other than sheer interest in market activity?

what can you do with that instant analysis?, place another order to further participate?, pull other existing orders?, confirm that markets are fat?, confirm that markets are thin?

which?

I don't think our conclusions really are derived from the same premise... Order flow or print reading or whatever you want to call it is really not that hard or complex. Like anything it just needs some explanation of what to look for and then practice implementing it with a simple strategy.

Which can be analyzed instantly?
1. place order
2. pull order
3. exit position
4. maintain position
5. add on to position
6. stay flat
7. go play golf / seek medical attention

I'm probably missing a bunch here but isn't that exactly what trading tools are supposed to assist you with? If they don't then throw them away.

We teach traders how to focus on the historical print, the current print and to have realistic expectations about the future based on what evolves in the print (History/Current/Future) all in equal balance when possible. With the historical you can find inflection points or invalidate areas that you'd think would have been important but weren't - and therefore shouldn't be part of your future analysis (note to Mike who hates chart patterns!). You can also learn a lot about what to expect in the future when these events happen. In the current print you can find patterns of momentum or exhaustion (or 2-sided trade) and act on them if they fit into your trading plan.

Will reading the print solve all of your trading woes? Nope. But it's real-time feedback and many of us see the value in that - whether through OFA or any of the numerous other tools that do similar things.

DB

February 25th, 2013, 06:16 PM

David_R

I have an OFA License that I do not use. I invested $3000 or more, but would like to sell the license. Can I do so?

David

Sure David. Just contact George with the new user's info and he'll transfer it with a small fee - that way the new user is actually in the system and can have access to upgrades, etc. I'd figure all that out before hand so there's no misunderstanding between the parties. Sorry we didn't give you what you needed.

DB

February 25th, 2013, 09:10 PM

OrderFlowAnalytics

We are not using any library. We have our own graphics and drawing code.

DB

Thanks. What I don't like about C# (I'm guessing?) is that the data visualization tools are rather weak. I saw some of the screenshots on your site and would recommend a graphing library. In the long run, $750-1500~/year is cheap compared to the time it takes to work with low-level, probably less performant/visually appealing/stable code. Back offices of .NET shops like GMO, JPM all use some kind of commercial library.

Order Flow Analytics' D.B. Vaello (President) - Ask Me Anything (AMA)

Discussion in Trading Reviews and Vendors

Order Flow Analytics' D.B. Vaello (President) - Ask Me Anything (AMA)