Post New

A trading experiment uses five retail brokers in 2022, namely, E*Trade (acquired by Morgan Stanley), Fidelity, IBKR, Robinhood, and TD Ameritrade (acquired by Schwab). https://onlinelibrary.wiley.com/doi/full/10.1111/j...

The broker sample covers 87% of the retail brokers who report DAT numbers. TD Ameritrade has by far the most active retail investor base. Across these five brokers, total trading volume is large, averaging around 14 million trades per day. It assumed a typical stock retail trade size of $8,000.

About 53% of total trading volume was on public exchanges and 9% on ATSs, and another 38% was executed by off-exchange market-makers.

The prevailing best quote, defined as the NBBO.It does not include many of the best prices available on exchanges, such as odd lots25 and nondisplayed orders. Indeed, a significant amount of activity transacts within the NBBO (even on exchanges). Fidelity, for example, reports that it generally executes trades at effective spreads that are around 25% of dollar quoted spreads. Fidelity receives very little for payment for order flow; Ameritrade's (Schwab bought) had a lot.

The average round-trip cost ranged from 0.07% to 0.46% for the same trades at different brokerages.

TD Ameritrade was best, followed by Fidelity.

I use IBPro. I set a limit order at the mid point between the quoted bid and ask and it's usually filled immediately. I also use Vanguard. Recently I noticed that Vanguard has a wider spread on the bid/ask than IB. However, Vanguard usually beats my limit orders which are usually placed towards the one third point.

In the study I think they were placing market orders. Were they getting better than midpoint?

Aussi

Getting data for such a study is hard; I know only the words of the paper on the study's conclusion of net. Size of market share determines spreads so it would have varied considerably based on the type of company. Other than reading the paper and its limits I don't know what to conclude. It would be interesting to have included Schwab since it purchased Ameritrade. Fidelity was decent.
If you can always get your trades executed at the midrange of the spread, good for you; the risk is missing trades when you don't. For longer term traders it shouldn't matter for big stocks; I trade short term so missing out on a trade altogether is more of a risk for me. E.g. I bought TPR a few months ago and it is up 62%: if I had missed it due to a misjudgment on my limit order I would have missed that gain.

If I haven't got the execution within about a minute I move halfway between my limit (usually, but not always, my limit is one side of the spread), I move my limit to the new half way point. Very rarely do I see the other side move away from my new limit. Generally I am trading NAS100 stocks.

Aussi

Interesting study, but bear in mind that it is entirely about at-market orders. So it is primarily a measure of how much and where the market pools were taking advantage of fools.

Use limits, or you're placing a large wager on someone you don't know being nice to you. It is a nice test of who is (was) nicest in that situation, but it's not a situation that is relevant to me.

Jim

Ans if you use a limit and it is not executed and you miss a massive quick run-up it can cost you much more than you saved by using a limit order. On large stocks the spread is so tiny the risk is not worth it.

Ans if you use a limit and it is not executed and you miss a massive quick run-up it can cost you much more than you saved by using a limit order.

That is extremely unlikely and therefore hypothetical, unless you get greedy with your limits. I place me limit stock buy orders at the ask and sell orders at the bid, only to protect against being fleeced, not to shave a few pennies. With options I go deep inside the spread because it's more of a price negotiation with the market maker.

Elan

place me limit stock buy orders at the ask and sell orders at the bid, only to protect against being fleeced, not to shave a few pennies.

For things that are passably liquid I'm a hair more aggressive. I often start with a limit at the last trade price if it's within the bid/ask range.

When I'm doing a lot of trades at once, I set the limit each one to a bit more aggressive even than the midpoint. e.g., sell at .7%ask + .3*bid and submit the orders. A few will fill, others won't.
Then I go back to the first trade and shift the limits of all the still-unexecuted ones to the midpoint.
Then I go back to the first trade and shift the limits of all the still-unexecuted ones to the far side of the midpoint, e.g., sell at .3%ask + .7*bid
If there are any left unexecuted, I do as you suggested, hit the far side of the spread.

I'm happy to pay the far side of the spread, it's the cost of doing business. But I'm too cheap to pay that if someone is willing to do be more generous : )
It's five minutes well spent, if you're a miser.

Jim

This has occurred to me often; it occurred last week with APP. It occurs for me about once a month when I try to save by using a limit order. APP went from 363 to 380 the next morning so yes it can cost you if you miss a purchase by using a limit order that doesn't execute.

Sometimes the risk is big. I am up 62% in TPR in two months; if I had missed that by using a limit order trying to save 1/2%? You can see the risk.

This has occurred to me often; it occurred last week with APP. It occurs for me about once a month when I try to save by using a limit order. APP went from 363 to 380 the next morning so yes it can cost you if you miss a purchase by using a limit order that doesn't execute.


Sometimes, it goes down after you miss buying it too :)

I've not been trading that actively recently, as I used to with FolioFN.
At IB now, I usually use BasketTrader and their MidPrice algorithm, which tries to execute at the midpoint and follows it around.
Typically, most orders get executed in 30 minutes, a few stragglers take up to the rest of the day, and now and then they only get partially filled.
I don't have time to monitor them, so I take what I can get. Executions have seemed good though.

Mark

yes it can cost you if you miss a purchase by using a limit order that doesn't execute.

FWIW, I never let a given limit sit for more than a minute before moving it, and rarely even that long, so that's not really an issue. I'm just trying to see where the bot on the other side will trade. It's usually a bot, whether it has a bid in already or not.

On rare occasion you can outsmart them, for very thinly traded stuff. You put in a bid, the ask moves away from you. You move your bid up, the ask moves up the same amount. This happens a few times and you s see the pattern--now your bid is above its original ask! So, you cancel your bid. Wait a short while, and generally the original low ask comes back. You pounce on that price (still a limit order). Often you'll get the fill there. Nobody will get rich this way, most of the bots are too smart now, but it's entertaining.

Jim

This has occurred to me often; it occurred last week with APP. It occurs for me about once a month when I try to save by using a limit order. APP went from 363 to 380 the next morning so yes it can cost you if you miss a purchase by using a limit order that doesn't execute.

Sometimes the risk is big. I am up 62% in TPR in two months; if I had missed that by using a limit order trying to save 1/2%? You can see the risk.

A limit order is not a catholic wedding. Yes, maybe in 10% of trades the price will move away from you by a few cents. I always check immediately after placing my order, and I adjust the limit to get the execution. I never place an order and wait till the next day to find out that it didn’t execute.

Another 10% of the time you actually get an execution that is a few cents better than your limit.

Elan

Post New

US Policy FAQ

---

Contact Shrewd'm
Contact the developer of these message boards.