I have been working on a couple of EOD strategies for US equities that I am looking to start forward testing and are backtested using Market orders to execute trades. Question I have is am I better off in the long run keeping it simple and sticking with MOO and MOC orders or trying to actively trade the open/close myself with limit orders? I plan on running both methods side by side, but would like to hear from people's experience with this. Would this possibly depend on the liquidity of the stocks I am looking to trade? For the record, min avg daily volume that I trade is 1 mil, but I am looking to test with stocks down to 500,000 / day any opinions or advice on the matter would be greatly appreciated
Some stocks trade between 9:30am (Eastern) and the official open, and most data sources report the first trade after 9:30 as the open, rather than the MOO price. I think a significant number of NYSE stocks trade MOC, and NYSE MOC orders must be committed at least 20 minutes before the close. These factors make the published open/close prices somewhat fictitious. <!--[if !mso]><object classid="clsid:38481807-CA0E-42D2-BF39-B33AF135CC4D" id=ieooui></object> <style> st1\:*{behavior:url(#ieooui) } </style> <![endif]--><!--[if gte mso 10]> <style> /* Style Definitions */ table.MsoNormalTable {mso-style-name:"Table Normal"; mso-tstyle-rowband-size:0; mso-tstyle-colband-size:0; mso-style-noshow:yes; mso-style-parent:""; mso-padding-alt:0in 5.4pt 0in 5.4pt; mso-para-margin:0in; mso-para-margin-bottom:.0001pt; mso-pagination:widow-orphan; font-size:10.0pt; font-family:"Times New Roman";} </style> <![endif]-->
I have been running a couple of ATS for the past 5 months that involve getting in at MOO and out on MOC, and here is my take: You can send MOO/MOC orders up until 9:28am afterwords they will be rejected as the opening auction process is comenced. after 9:28 you cannot cancel them. you will get filled at 9:30 on all your NSDQ orders.. NYSE can take up to a couple of minutes after trading has commenced for your order to be filled.. usually its at around the same price it's trading.. nothing really substantial. the stocks I trade go from 50k volume up to Facebook volume. What i did was look at the volume traded at the opening print for various stocks and i make sure that my orders aren't bigger than 5% of the volume traded on the open print (on average), since i don't want to affect the opening print too much on those low liquid stocks... if you're trading stocks with 500k volume and over dont even worry about that you wont affect the opening price at all. If getting into the position is important then i highly recommend getting in MOO, otherwise if you try to get in with a limit order you'll probably only get fills on the stocks that are moving against you.. while not getting fills on the ones going your way. Last, actually alot of the data providers offer the official open/close print as EOD data. yahoo for example gives the official open print as their "Open" price and not the first trade after 9:30.. ofcourse it all depends if your system is automated, how many symbols you're gonna be trading.. etc.. so limit orders could be good as well.. test them both if you can. good luck.
and for MOC you have to send them before 3:45pm.. after that you are not allowed to send MOC orders at all. BUT you can still send limit on close orders only if there is an imbalance on the stock and in the direction opposed to the imbalance... so just make sure to send the them before 3:45
A lot of stuff is inaccurate here. For one, Yahoo open prices are actually the first trades >= 9:30 am ET. They are *not* the opening auction price for NYSE stocks. This is a significant issue for backtesting using open prices. As for avoiding impact on the open...I wouldn't look at the avg. open auction volume as much as I would the distribution of volumes over time. If you're having an impact 20% of the time, that may not show up in avg. volume data.
ohh allways thought it was the official print... thanks for the correction. hmm and for the auction volume it's very subjective.. for exampl some 50k volume stocks do about 2k volume on the open print... 5% of that on a 5$ stock is 500$... so with a 500$ order i start impacting the print... now it's either i size down and make less profit.. or i don't mind my effect on the print and I make a smalller % profit on a larger size trade... lower volume stocks have allways shown me very good results on backtests where i enter MOO and exit MOC.. but once you run the system you will see that actual results are lower because as an extra participant in the auction process even small orders will have an impact on the open/close print... i havent thought to look at the distribution of the volume instead of the average.. i will take a look at it. thanks.
Unfortunately I didn't save it, but there's a paper out there that studies the impact of MOC orders (don't recall if MOO was included). If I remember correctly, the main takeaway was that an imbalance will cause the closing print to move equivalent to the bid/ask spread at the time of the submission deadline. Interestingly, the amount of "slippage" was independent of the imbalance size. Again, take all that with a grain a salt since it's been a while (2 years?) since I read it, but you should be able to find a copy of the paper with enough googling.
When holding equities longer than a day, the close is a better liquidity point than the open. It is harder for conflicted intermediaries to play muppet unfriendly games with the close. Doing so exposes them to overnight inventory risk. Also the settlement process requires real cash to change hands and if a big boy gets hurt in the process they will draw unwanted attention.