Also, its not only about liquidity, my question was more about if the open price is tradeable? As one poster said, only 100 shares might trade at open price - and it holds true for liquid stocks also. This is very worrying, because it questions the usefulness of the opening price for any analytical system development.
once again-imo-useless,unless it's a relatively liqid stock. at least >1M avg.vol. S&M- http://en.wikipedia.org/wiki/Sadomasochism again it might work on nasdaq-IF stock you are in is really liquid and your size is relatively small. otherwise you are asking to be raped. "they" will pull off bids and offers , and execute your order at whatever price "they" feel like. because when you choose MOC-it mean for YOU that you will accept ANY PRICE. something like this might happens to you too- http://www.youtube.com/verify_age?n...utube.com/watch?v=0qz4L1O84_I&feature=related
You can trade in the opening and closing auctions with Market-on-Open and Market-on-Close orders. Your biggest problem is that Google Finance data is wildly inaccurate, particularly for the opens. Often Google will show opening prices nowhere near the actual opening auction price and sometimes outside the day's range as well. Yahoo Finance is even worse. So your backtesting will have little relevance.
Hi gmst, From my experience with TradeStation, Market-on-Open gave me horrible fills. Sometimes the fill was at a price that did not even appear on the chart! (As above the high for the day on a long RTM strategy). And it was for scarcely a 100 shares on liquid stocks. I just chalked it up to market makers trying to make a quick buck on some newbie. I would say look at the opening 1 minute charts, to compare with the daily, and if you see a big difference between the high and low, especially on low volume, you wonât be able to get the opening price. For me the opening price is not that significant anyway because I base my buy price off the previous bar close using a limit order. Closing prices are much more reliable as volume is usually heavier as institutions often place their orders at the close. I actually wait 1 minute after the market opens when placing orders to avoid price spikes that would cause my strategy to execute on spurious signals. You are basically at the mercy of the market with Market-on-Open orders which is not a good feeling.
Thanks Kevin, this is useful information. Seems I am better taking price data from some more reliable source.
RoI, Thanks for the info. This is a very practical suggestion to wait for 1 minute after market opens before start trading. Your experience has more or less confirmed my worry that open orders might not be the best deal. However, hearing what you and others had to say, it seems MoC orders are useful. Thanks.
Opens are not ideal for trading, but if your strategy depends on it, you can certainly try. To do so, you need to specify that the order is "on-open" and that it is to be routed to the primary exchange. Places like Yahoo report the "consolidated" open price, which is almost always the same as the opening Nasdaq auction price (for Nasdaq stocks), but is almost never the same as the NYSE open price (for NYSE stocks). The reason is that most NYSE stocks will trade very small on some ECN right at 9:30, but the NYSE specialist will not open the stock until 9:31-9:35 in many cases. So the "open" you see on Yahoo is really just the first 100 shares that traded anywhere after 9:30. The Nasdaq opening auction occurs right at 9:30, so it is usually the same as the consolidated published open price. SO, bottom line is that with Nasdaq stocks, you can count on your backtested open price data at least being accurate (though will not include any possible price impact of your order of course). For NYSE stocks, it going to have a lot of error in it, which may or may not make the strategy invalid. You can grab actual NYSE open prices from Bloomberg, but nowhere that I know of for free. As for market impact, you need to do a study of how many shares typically trade on the primary open in a given stock, and plan to be no more than x% of that volume. x is difficult to value, as you cannot easily empirically test market impact, so you will have to make an assumption that if I am more than x% of the open volume, I am going to move the price too much. Be careful with ETFs, esp. leveraged ones. They tend to trade less than you would think on the open (small pct. of overall daily volume, compared to most stocks).
One more thing: firing off a market order in the first minute of trading is pretty dangerous since market makers can be quite wide at that time and markets can be moving violently. As for closing orders, I think MOC orders are fine for most liquid stocks. But no reason not out an order in at 3:55 if the liquidity is there.
Thanks for very insightful experience. I did not know that google/yahoo data points are consolidated prices and not NYSE open prices. I assume even if I am looking at Tradestation intraday data, the 1st minute bar data for a security might also be consolidated prices. So, for any analysis it should be safe to only consider data post 930 EST, I guess. Since by then all stocks would have opened on NYSE and will be liquid enough to trade 100-1000 shares each, depending on stock's daily volume. Can someone please confirm this. Thanks