This thread hopes to discuss the participants of the Mini Nikkei futures contract, their intentions and the information value of volume for trades reported in their data feed. It could likely be generalized to any Futures market but my focus is currently on the Mini Nikkei, and by having a specific market we can break down the actual participants based on information available. I am approaching this from a blank slate of market micro-structure knowledge so will make some assumptions based on books read and terms used when describing the various participants. Any assumptions mare are open for discussion and hopefully we can filter the various opinions to establish a reasonable model of the market. What We Know There are two markets - J-NET and Regular. Mini Nikkei allows J-NET trading. This is described as an independent market for a number of different products, including the Mini Nikkei. Both J-NET and Regular volume is reported at the end of the session. J-NET volume appears to be minuscule compared to the regular volume (as reported here - 'Trading Volume by Participant' - http://www.jpx.co.jp/english/markets/derivatives/perticipant-volume/index.html) Morning and Night Session Volume Reported - Both have significant volume in the regular auction. Morning session is likely less influenced by US Markets. How is Volume by Participants Reported?? Yesterdays reported total volume was 524706 contracts. Participant volume for whole day is reported as Buy vs Sell with: Sell Total = 417243 Buy Total = 414918 Does this mean that participants absorbed a total of 417243 contracts on the Offer, and 414918 contracts on the Bid? Why is this total 832161 > 524706??? Anyone have a clue how this is reported?? Participant Generalizations for understanding Volume Flow In general it is my understanding that in a futures market we have: Market Makers (reported as above. Probably HFT algos) Institutional Traders / Investors (hedging stock portfolios etc) Speculators (the remaining volume) Market Makers are getting filled on the inside Bid and Offer and have incentives from the exchange to provide liquidity. They are mainly concerned with making their $$ from exchange incentives and the inside market spread. I assume they get offside sometimes when filled against institutional traders and have some way of offsetting this risk and / or play games in the market to get back onside. Probably no way of wrapping head around their strategies, but when reading volume intraday - they are on one side of the transaction so need to consider their intentions. Institutional volume is of longer term interest - hedging stock portfolios, economic risks etc. These guys are going to move large volume, and probably responsible for big shifts in the market while getting majorly net long / net short. They surely split their transactions - buy or sell throughout the trading session to build an overall net position. Speculators are everyone else playing in the market - and on most days are probably responsible for most of the volume getting moved around - getting played by market markets in the process. These are the patterns of puking / stop runs etc we see in the volume reported. Going to assume still that most volume traded is just a leg of a spread against another asset class (Calendar, index spread etc, individual stocks), since that appears to be how most proprietary traders operate - agree / disagree? That will do for now - lets see if someone can interpret the volume reported and has any more insights to add about who is playing in this market and their intentions. Then we will come up with some theories on price discovery.
The list you have is incomplete. If I buy 1 contract that you are offering, it's counted as 1 contract in the volume figures, not 2. AMRO, SG and ML are the major MM's and make up over half of the volume every day. Individual investors make up around 27% of the volume.
Thanks but can you please elaborate. The numbers I am showing are downloaded from the Osaka website. Can view here (http://www.jpx.co.jp/english/market.../20161107_volume_by_participant_whole_day.pdf). In referenced volume by participant doc - Amro reported as 172392 Sell Volume, 192805 Buy Volume. I would think this would mean that Amro's were responsible for a total of 172392 + 192805 = 365197 contracts of the whole day Volume. Correct? However if you add up each participant volume it totals 832161 contracts which is more than the Total volume reported across all sessions (morning, afternoon and night). The image I showed with total volume 524706 admittedly only includes the night and afternoon session. Total volume for whole day is 737319 ..... which is less than 832161. Why the discrepancy? How much volume is simply Market Maker vs Market Maker. Do you mean that AMRO vs SG vs ML are making up over half the volume just trading against eachother? Or that they provide over half the liquidity for all transactions? Thanks.
The list published by the exchange is just the top 20. As you can see in the attached, the blue column is the number of contracts sold at the bid, and red is bought at the ask. You can see that contracts are reported by ones, not twos. The volume in the far right column includes Monday's evening session. Yes. Edit: Ah, I see why you are confused. will explain later after market closes
>grahamglover The page you were looking at (http://www.jpx.co.jp/english/market...-att/20161107_volume_by_participant_night.pdf) shows both new and closing trades in each column. When the exchange receives a buy order it doesn't know if it's a new long trade or the closing of a short position. You may,or may not, find this page of interest: http://www.jpx.co.jp/markets/statis...0020m6d-att/Tousi_DV_W_201610_4_1024_1028.pdf
Ok.. I am feeling stupider by the minute here.. but each transaction (opening and closing) would count towards the total trade volume. You are saying that the Amro Sell Volume of 172392 is the Open Sell Volume (Someone lifted Amro on the offer) + Close Volume (Amro hit the bid) = 172392. Regardless that is 172392 traded volume. Add the Buy and sell columns up and huge discrepancy between that and total volume traded for the day (more volume reported by participants than total traded volume). Anyway we can clear this up quicker?? The other doc is huge help thanks.
No, the volume for the day, for example today's day session is 389184, is actually counting only one side of the trade. So if I buy 1 contract, it's counted as 1. The seller's action is not counted as only 1 contract changed hands. But for the broker's total buy volume posted by the exchange, both opening buys and buys to close short positions are included. The exchange can't tell which is which. Additionally, this will show the broker's position on balance, which some people use in their analysis.
So now have wrapped my head around that - am looking into the statistics of Trading by Type of Investors. I guess we have to make our own interpretation of this. By far the most volume for October was traded by Foreigners. I take it this can include foreign market markets, individual and institutional interest.
No matter how smart I try to be about who the participants are - most of the volume is traded against market markers, probably in any futures market. As @learner2007 has stated, over half volume accounted for is traded with AMRO, SG and ML alone. Now if I am a short term speculator in this market - it really makes sense to consider what these major market makers are up to and what we can look for in the volume footprint of this market to understand the game a bit better, and this is the case in all markets depending on the time frame you are interested in. At the micro level - with multiple market makers at play, those guys are probably competing against one another for queue position etc to gain the inside market edge, a game the short term speculator has not much chance at without the infrastructure. At the same time they are providing liquidity to speculators, and institutional investors. So by monitoring the volume vs price action intraday - the next assumption is, can you assume that market makers are going to get offside during price swings and that they will then need to use various tricks to manipulate prices to their advantage?? For example, if institutional money is constantly eating away at bid side liquidity during a down move, of which most liquidity is provided by different market markers, they will inevitably be building a net long position. How long do they keep this on their books and are they hedging it somehow? I take it this where you can use the Open Interest and Trading Volume by investor information to gauge the positions on balance. However some pretty clever people are probably employed to interpret this information way better than your average retail trader so is it worth the effort?