For some ETFs with lower trading volumes the spread between the bid and ask can be as high as 1% or 2%. That makes it impossible to make a profit with any system that has a low average profit per trade. From Interactive Brokers I read this about buy at the open orders: A Market-On-Open (MOO) order is an order to be executed at the day's opening price. Market-On-Open (MOO) orders can only be executed when the market opens or very shortly thereafter, but must provide the first printed price of the day. If I understand that correctly, at the open, the buy price will be the same as the sell price. There isn't a spread between the bid and ask. Am I right about this?
On the open and on the close orders only match with other on the open and on the close orders. That means, if you are buying, you could potentially get a better price than the ask price or a worse price than the ask price. As with any market order, there is no guarantee what price you get. A limit on the open would protect you from getting a "bad fill". In general, most lower volume ETFs have little or no on the open or on the close activity. Which means, you are far more likely to get a "bad fill" than a good one. If you broker provides you time and sales data, it is relatively simple to see the on the open and on the close activity for a specific stock from previous days. Is there a specific ETF you are interested in?
Just a followup to my previous post... There is no spread, but I think you are thinking about it in the wrong way. Just like every trade, the buy price is always equal to the sell price. Let me give you a scenario for a low volume ETF with little or no on the open activity: You enter a buy market on the open order at 9:00 AM. There currently are no sell on the open orders to match your buy order. The NYSE sends this "imbalance" information to market participants. A market participant sees this and enters a sell limit on the open order 2% above the NAV for your particular ETF. At 9:30 AM, your orders match and you get the opening price. You didn't pay a "spread" but you still paid 2% above the fair value. Here is some info from the NYSE about opening and closing orders: https://www.nyse.com/publicdocs/nyse/markets/nyse/NYSE_Opening_and_Closing_Auctions_Fact_Sheet.pdf On the open and on the close orders are mostly used with individual stocks, even thinly traded ones, not ETFS though.
Thank you so much FroggerMan for the information. The ETF that I'm interested in is PALL for physical palladium. The bid ask spread is typically very high, above 1%. My trading system for it only averages around 3% per trade. The only way I can see to avoid this large spread is to switch to trading palladium futures, which still runs a fairly high bid ask spread of at least 0.5%
Date Bid Ask Open 04/03 201.42 207.76 207.65 04/02 191.07 203.03 200.12 04/01 211.00 215.55 212.31 03/31 219.21 225.00 221.00 03/30 211.11 215.00 211.11 I would give it a try, but would recommend a limit on open instead. One thing to keep in mind, you don't always get an instant fill notification with on the open orders. Meaning, you could be left waiting a few minutes to know if your order got filled but I wouldn't say that is typical. Have you considered just placing a regular limit order instead and just adjusting the price as the spread changes? I am go on a tangent here, because this reminds me of my first ever trading strategy back when I was in college. The strategy isn't important, but it was dependent on getting the opening price for various sector ETFs about 1 minute after the open, and closing the position at the end of the day. I back tested it over ~10 years and it showed to make consistent profits every year. I knew I wouldn't get the opening price each day, but I figured it would average out, some days better some days worse. Once I started trading, and losing money, it took me two weeks to figure out what was going on. Unfortunately, my assumption that it would average out was wrong. All the profits from my strategy were coming from the wonky opening price, not from how the ETF traded the remainder of the day. And unfortunately, for my strategy, you had to know the opening price of the ETFs before knowing what ones to buy. I lost some money but learned a valuable lesson. Back testing and simulated trading is the right place to start, but until you are trading in live markets you never know how well your trading strategy will work.
For an experiment, I bought the ETF SLV and sold it in a different account, both using a market order on the open. So it would seem, at least for this relatively high volume ETF, that I could buy and sell it at the same price. If I wanted to do the same thing during the trading day, I would have to sell at the bid(13.68 as of now), and buy at the ask (13.69 as of now) and so my cost would not only be the commissions, but the spread as well. Of course I could send a limit order to buy at 13.68 or limit order to sell at 13.69, but the trade wouldn't be instant and I might miss the buy or the sell.
there is no benefit to OPG or MOC when the security is thinly traded. I was trading a strategy in thinly traded CEFs a couple of years back. Backtest looked great but out of sample was very different. Turns out I 'became' the market to some degree.
There is no delay, it's always instant notification. But auction don't happen at the same time on NYSE.
I did another experiment this morning where I placed a simultaneous buy and sell order for SLV on the open for two different accounts, except this time it was for limit orders instead of market orders. The orders were both executed at the same open price: