I have been trading the financial markets for multiple years now and am now switching to swing trading stocks since i want to spent less time in front of my computer. The strategy i have developed is doing very well. I do open my positions at the close using MOC orders, i also close 95% of my positions at the close with MOC orders. I do understand how MOC orders work yet i feel there is still some stuff i need more knowledge about, mainly the volume of MOC orders and the influence on the closing price. Other people have warned me about MOC orders and having slippage. Yet in my personal experience so far the slippage of the closing price versus the last traded price is very minimal, most of the time just a few cents. Even when the market on close volume is around 3 million of which 2,6 million where sell orders and the total daily volume was 19 million, with 0 slippage on October 11th. I am not trading a large size yet, however i am planning on scaling up quickly since the system is doing so well, before i do i would like to understand more about the impact of MOC orders and the relation to the closing price. A Google search gives me little information so i am hoping to get some more information here. Why is it that other people have told me to not go over a few percentage of the daily traded volume while there are daily examples of huge market on close orders with huge imbalances which represents a huge part of the daily volume, without having an impact on the closing price versus the last traded price? Taking ABEV as an example again, how is it possible that the market on close was 3 million of which 2,6 million where sell orders and closing price was exactly the same as the last traded price, not even 1 tick down while the market on close orders where of such significant size with huge imbalance. Do MOC orders get special treatment of some kind and so they don't influence the price as much as volume during regular trading hours?
https://docplayer.net/418102-Trading-around-the-close.html here is a doc about it that I have replicated. The last trade price is the wrong benchmark to measure the market impact from. The market impact of MOC orders occurs when the imbalance message is released. For nasdaq that occurs at 15:55 and will see the far majority of the market impact realized at that time since new orders can not increase or switch the side of the imbalance, where as on nyse the imbalance begins being published at 15:50 and since d-quotes are able to change the direction and magnitude of the imbalance the market impact is realized over time as those orders come in. Ideally you'd get imbalance data from nasdaq/nyse and some tick data and then create a market impact model in order to determine the cost of trading the close, it's the approach I've taken. Short of that realize that as you've been told, the larger the % of adv you try trading the larger the market impact and it can be very significant. I have found it is largely a function of % of adv as opposed to % of the close auction volume.
You need to define what the last price is. MOC price is based on auction that sets the OFFICIAL closing price. So if the stock is listed on NYSE, the auction will create the closing price for the stock which you will get. However, the stock can still trade on other exchanges at any price. What dinn13 is saying is that your big order is likely to move the official price itself.
Thanks for your reply, i will read the document tomorrow. I am using the imbalance data from Market Chameleon and am looking at the IB DOM for the price and volume information. The last 15 minutes of Friday's trading session in ABEV we traded between 4,57 and 4,55. 4,55 being the last traded price and also the close price. It also looks like the previous numbers from above (from memory) where a little bit off: the total market on close was 3,248,468 and according to Market Chameleon 3,055,634 was sell volume. Shouldn't we see a larger move then? The total daily volume that day was around 19 million, so the market on close was 17% of the average daily volume. I have been trying to find extreme closing examples for multiple months now but never found any, can you maybe provide me with some examples? A few cents difference doesn't bother me at all, i am just curious how so much volume can be traded with so little impact on the market.
Why does the last price matter? I am trying to get the official close price and hold my position for multiple days/weeks, this is how i created and tested my strategy, so what difference does the price after the official close makes? Do you mean after market trading? The last order in my DOM was 1 minute and 50 seconds after the close at 4,54 by ISLAND exchange.
The ABEV imbalance you're seeing is at 15:50 which includes only the MOC/LOC interest. At 15:55 the imbalance begins to also include d-quotes and at that time the ABEV imbalance is a buy imbalance of only 278731. Nyse imbalances are harder to model since the imbalance isn't firm. For example even though d-quotes are included after 15:55 they can still be cancelled thus once again changing the imbalance. Of note dquotes can be sent and cancelled up until 3:59:50 at which point the imbalance is finally firm. https://www.nyse.com/publicdocs/nyse/markets/nyse/NYSE_Opening_and_Closing_Auctions_Fact_Sheet.pdf I'll come back later with some examples of imbalances that caused obvious market impact
Can you please tell me where you got your data for the market on close from friday 10th of October? The only data i have for market imbalances is from Market Chameleon (free) so i can only see the imbalance delayed, however i do not use the data for my trades so it doesn't really matter, i was just wondering why this large volume and large imbalance had so little impact on the market, not that i am trading anywhere near that size, i do want to understand how and why these things are happening. So you are saying at 15:55 on October 11th there was a buy imbalance of 278731? Tis means 3,2 million sell volume was added by d-quotes in the last few minutes?
Just started reading this document and now i realize what you mentioned earlier, i have to look to the price at 15:50 for NASDAQ and 15:45 for NYSE to see the market impact of the imbalance. Stupid i didn't realized this earlier . It turns out that on average for an imbalance that equals to 1% of adv, the price impact is about 0,04%. This probably explains to why i didn't noticed it before since i was looking for much larger impact.
The times have changed now, that doc is fairly old now. Nasdaq is 15:55, for nyse the imbalance is first published at 15:50 which only includes MOC/LOC (and why the ABEV imbalance didn't immediately have a large impact), at 15:55 it also includes dquotes (of note dquotes could have very well been there since hours ago, they simply aren't included in the imbalance prior to this time), so in the ABEV example there either could have been 3.5 M buy dquotes prior to 3:50 or they could have come in after the first imbalance, no way to know from the data. I use pcap data from the live multicast feed and prior to doing the pcap I bought the historical data from nyse and nasdaq.