¿ªÔÆÌåÓý

ctrl + shift + ? for shortcuts
© 2025 Groups.io

Yikes smart exchange is not smart


 

I was mistakenly under the impression that if I create a SMART sell order that it would fill at the best price

Contract contract = new Contract();
contract.symbol(symbol);
contract.exchange("SMART");
Order order = new Order();
order.clientId(clientId);
order.orderType("LMT");
order.action("SELL");
?
but today in fact it did not find me the best price.
Now part was due to some lazy programming on my part. I received a quote of a bad, low price, and didn't double check it.

Regardless of that issue, I really thought that IB would fill my order at a reasonable price.? In fact though it took my order and sold it, at a large loss, and at a price whereby my price was the only one like it for that minute of trading.
Fortunately I am still using small share sizes while testing.

So does this mean that I cannot trust SMART, and I have to go find the exchange each time?
I tried to use some other smart options, but apparently they are only available in the TWS client, and are not valid for the API.



 

Why did you use LMT ? (shouldn't you specify a order.lmtPrice ?)
What was the instrument you dealt with ? (Liquidity can really be a nightmare)
Did you send your LMT during RTH ?
Did you look at the spread before setting your order ?
While conducting your test, open TWS Level II window on the instrument you test and look at price/exchange, that give you a feeling of what may happens.

I am guessing that you will feel you open a Pandora box.

IMHO I don't leave strategy to others unless they explain there algos. The bigger are your lots the more dangerous it is to leave other doing there cooking.
Assuming you are not an institution you may not need to hide your volumes/prices, this help finding simpler algo


simplest

and dig "Peg to Primary" (not great, but good starter if only using RTH) or MKT

or take some time and study


I never had opportunities to benchmark IB algo versus Jefferies , I suspect IB work better on simple condition because may execute faster (pure speculation)

Time invested in studying theses make sense if you need to manage high volumes on low liquidity stocks.

or
do your cooking using NBBO and Market Depth level II to decide yourself where and how you can optimize the split of your lots.
Also look at safer way to know the price you are dealing with.

While never found it very useful in RTH (tape B,C) as it's already past story and IB does deliver a very decent tick price in LI
or even better with LII

Personally (limited experience) I won't use SMART for an order (I am even surprised it worked)
And also a rule of thumb, you should prefer conid to identify an instrument rather than symbol,
SMART can be dangerous on an order, IB support a lot of market places, then if you don't specify currency you are taking a huge risk (IMHO as I don't use SMART for order)
conid+exchange is safer. exchange being one of contractDetails ValidExchanges (aside of SMART)


 

Thanks for the reply, I think.
I didn't put all the info on the order, no I don't need limit price.
This was premarket and I do agree that the liquidity was an issue.
But the bigger issue was my misunderstanding how SMART works.
I've done hundreds of orders just fine, but this exposed a nasty little issue I figured someone else should be aware.
But I am probably the only person doing small cap.


 

And FYI MKT orders can completely wipe your account, careful with that. Didn't happen to me, but I've heard and understand how you can lose everything with it.


 

Yes I see how you can loose a lot on SELL on low liquidity stocks. especially pre-market as some guys set fishing lines with 0.01$ bid, just in case somebody else made a mistake at time no 'normal' "ask" exist.

Anyway was more a suggestion of ramp-up on learning way to learn more elaborated method to control order.

?


 

"... But I am probably the only person doing small cap."

Most probably not, but may I suggest to use Level II analysis to educate lot prices/volumes and dispersion of buyers/salers (kind of muti-factors spread) just less an issue on big cap (but still)
For that level II is a very valuable tool.


 

"I was mistakenly under the impression that if I create a SMART sell order that it would fill at the best price"

Reg NMS is not required pre-market but it would be interesting what IB has to say about this.? I always assumed they'd route to the best price anyway.? Can? you provide the relevant data showing that they did not?