Limit orders

Quote from jimrockford:

This is very disturbing, especially coming from a credible knowledgeable source like you. I would like to see some explanation from IB.

Can you please provide some specific data for a specific trade: execution prices, times, bids and asks at the various market centers, and make sure you include enough details to prove a Reg NMS violation?
More...

Hi Rockfish.

I'm not trying to prove regulatory non-compliance by IB...
That is the job of their NASD examiners...
And IB is always pushing the envelope on any and all regulations...
So their compliance is always borderline at best...
And clearly illegal at worst.

You see...
I don't have a problem...
As long as I get the NBBO from a reputable source like Thomson...
And enter it manually or automatically into TWS...
For 100% of all orders...
And then I get NBBO 100% of time (subject to latency).

I'm just trying to understand what IB is doing...
To the idiots that enter market orders based on phony IB data...
Or even bigger idiots that enter NBBO+x market orders.

These clowns have no idea they are being chiseled on 20-30% of all executions...
Often by MUCH more than $0.01

How about starting with a simple question...
That I am not getting answered by Google:

Is there an "official" NBBO for stocks?
Meaning is there an "official" list of liquidity centers that comprise the NBBO?
If so... a link to the regulation or algorithm?

Or does every major player...
Like a quote vendor or exchange or broker...
Pick 5 or 10 market centers and create their own NBBO...
A process that is WIDE OPEN to abuse.
 
ARCA, BATS, BRUT, CBSX, CHX, DATA, EDGA, EDGX, FLOW, INET, NSX, NYSE, PHLX, SM, TRAC. This is unofficial and the market makers(i.e. GSCO) are NOT in there.

Don
 
Quote from Don Bright:

ARCA, BATS, BRUT, CBSX, CHX, DATA, EDGA, EDGX, FLOW, INET, NSX, NYSE, PHLX, SM, TRAC. This is unofficial and the market makers(i.e. GSCO) are NOT in there.

Don
More...

Thanks.

Apparently...
The "official" definition is...
"all exchanges and market makers registered to trade in that security".

But as far as I can determine...
There is NO central entity creating a real-time NBBO quote.

What this means in real life...
For less liquid NYSE stocks (the bottom third in average volume):

I want to buy 500 shares of XYZ

(1) I have a very comprehensive NBBO quote from a high end vendor like Thomson.
eg XYZ 10 20.01 x 20.04 20

(2) I enter buy order into TWS...
And IB uses an ** internally manufactured ** quote as a default into my order...
eg XYZ 10 20.01 x 20.08 5

So IB enters the default price 20.08 into my buy order...
EVEN THOUGH a major player in the Thomson NBBO... is offering 4 cents lower.

If I send the IB default...
I may pay 20.04 or 20.06 or 20.08...
It's a total crapshoot... anything can happen.

If I manually or auto enter buy at 20.04 NBBO...
That is the ONLY way I am guaranteed to get real NBBO... subject to latency.

This is very clear and indisputable in less liquid stocks...
Because the market is fairly stable.
I can even buy stock a 2 different prices depending on how I enter my order into SMART.

But for hi volume stocks... it would be impossible to detect.

It's not illegal...
Because "best execution" has such a fuzzy definition.

But the point is...
Trusting a broker's or clearing firm's Routing System to get you the NBBO...
Is unprofessional.
 
I've not used IB nor Thomson financial, but on our Goldman Sachs Redi monitors we see all the bids and offers at each tier. The only exception we have is, if for example, the market is 20.01 x 20.04, and I try to pay 20.04, but there is a 20.02 "hidden pool of liquidity" within the Goldman family, they fill me at 20.02.

With depth of New York Open Book (NYOB), and a full Level 2 screen, I would think the NBBO would have to show up.

I assume IB offers Level2 quotes, and direct ECN access, so you can default your "providing liquidity" bids and offers to ARCA, etc. for collecting liquidity rebates?

It seem funny to me that IB would make it necessary for you to have to pay an outside source for market data??

I'll try to snag a screen shot.

Don.
 
Redi3.jpg


I realize this is a thick stock, but this is what I'm trying to show, depth of book with NBBO exchanges/ECN's.

Don
 
Quote from speculatus:

This makes sense if your strategy is to accumulate your line on the way down. Otherwise, good luck getting fills on limits orders on the way up :)
More...

This is a good Point Speculatus. I should have written to place limit orders, if the time expected to get filled is x minutes / hours in the future AND the price is < currentprice
 
Quote from Don Bright:

At the risk of being part of the "other half" of people on the board who disagree with your market order comment, let me explain this once again (for NYSE, Naz makes little difference what you do).

Marketable limit orders are executed immediately at the best price ( so send in 5 cents above offer for buys, below bid price for sells), market orders are held and batched or matched by the Specialist himself, and this sometimes takes time and costs money.

Speculatus has a bit of a point, trying to over-engineer trading generally just takes time away from the time you could actually be trading.

Develop skills and engage in profitable strategies manually, then get into automating them, if that's what you're trying to get to.

FWIW,

Don
More...

At the risk of being part of the "other half" >>> LOL

I definitely agree w/ your comment about sending marketable limit orders vs pure market orders.

However, I was trying to state that there is a time when you should look to go long on the ASK and times when you should go long on the BID. Do you agree with my conclusions, that this is based up how much time in the future you expect your order to be filled. If instantly then send a marketable limit order, if > some threshhold of time then send a buy limit order on the bid?

I am fairly new at this game, and definitely no-where near Bright-caliber, but I have collected some amount of data to make an actual conclusion vs rather than just saying things that are out of "feel"

"Develop skills and engage in profitable strategies manually, then get into automating them, if that's what you're trying to get to."

I've never known how people develop skills manually, What works for me is to backtest everything, and if it works, then that's your strategy with specific rules. I really see no difference between you trading it vs the computer.
 
Quote from ProgrammerGuy:



I definitely agree w/ your comment about sending marketable limit orders vs pure market orders.

I've never known how people develop skills manually, What works for me is to backtest everything, and if it works, then that's your strategy with specific rules. I really see no difference between you trading it vs the computer.
More...

(Don't hate me for this)...but backtesting, if of any value whatsoever, is going to tell what doesn't work, not what will work. Too many variable like VIX, FV Prem/Disc, peers, pairs, dividends, outside news sources, etc. All this is vital to what "is" and you cannot duplicate it for what "was."

My comment about "manually" is really quite simple. I've been doing opening only orders manually for decades. Now, with automated programs of various flavors, we have traders making a $half million per year doing that working strategy. My brother, after trading Pairs for years manually, can now trade a couple of hundred pairs with the better programming that we have.

Don
 
Quote from Don Bright:

(Don't hate me for this)...but backtesting, if of any value whatsoever, is going to tell what doesn't work, not what will work. Too many variable like VIX, FV Prem/Disc, peers, pairs, dividends, outside news sources, etc. All this is vital to what "is" and you cannot duplicate it for what "was."

My comment about "manually" is really quite simple. I've been doing opening only orders manually for decades. Now, with automated programs of various flavors, we have traders making a $half million per year doing that working strategy. My brother, after trading Pairs for years manually, can now trade a couple of hundred pairs with the better programming that we have.

Don
More...

That's interesting Don. Do you know of any traders that have been able to make the backtesting route work... these guys from California that I know swear by this.
 
Quote from Don Bright:

(Don't hate me for this)...but backtesting, if of any value whatsoever, is going to tell what doesn't work, not what will work. Too many variable like VIX, FV Prem/Disc, peers, pairs, dividends, outside news sources, etc. All this is vital to what "is" and you cannot duplicate it for what "was."
More...

Well, you'll see the net change of those variables on the tape right away. Just watch for price and volume...

I know you going to say that the tape reading is all encompassing :)
 
You must log in or register to reply here.
Back
Top