I know this varies a LOT depending on the security but can anyone take a guess at what percent of volume of the index of your choice is institutional? Same question for block trades. Anyone know of a site or software that will sms me if institutional volume or block trades suddenly drop?
Honestly, I think you're looking for ideas in the wrong place. Watching volumes went out the window last century ! I mean come on, its a known fact that HFT accounts for something like 70–80% of volume. Technically HFT would come under the auspices of "institutional", and yet you would be wasting your time looking at the noise and useless trades they generate. So it means you're then left with the remaining 20–30% of "real" institutional trades. insitutions can trade for so many reasons, that you're basically wasting your time there too. How are you going to differentiate between a pension fund, nominee account, investment trusts etc. etc. Then what about all the guys who trade in dark pools ? Watching block trades is also a waste of time. Bloomberg did an interesting research piece on that .... they found that block trades had no lasting impact on the price of stocks.
The instrument of my my choice is yes made up of institutions but I would also add professionals. I define pro's as being on the right side of the market and institutions not always. Any increase in volume will be institution or pro activity and should never be treated in isolation to price. I'm not sure if your aware of this but an important aspect of order flow is not all volume is open buying and selling.
Personally I wouldn't bother trying to 'detect' large individual trades/traders and here is why. I used to trade energy at a firm. Block trades happen all the time off exchange you can't track them and they don't move the underlying market in my experience as they are off exchange. If a larger player wants to say get long 10,000 ES contracts they are not going to bang a 500 lot 20 times. They will likely have the broker work their order for them to try and achieve a decent average price either via manual or algo fill. If I was going to track volume I would only track collective volume via volume profile, for example you can study how price reacts around high and low volume areas. GL.
Exactly right. I went over the original question to see what he is asking, and the key points were institutional volume and block trades. Well, institutional volume is very tricky. Its not like in addition to getting Time and Sales, the exchange tells you who bought and who sold. (ie. 1000 contracts bought by Goldman Sachs and sold to 1000 day traders in their basement doing 1 lots) LOL... how awesome would this info be? When Buffet says he sold all of his Walmart stock.... who did he sell it to? I doubt it was Joe Blow... so clearly its another institution. Is an ETF fund even an institution? They are obviously buying and selling to balance their funds, but maybe these funds are owned more so by retail... who knows. So really, institutional volume is almost useless since you don't know who is buying, for what purpose, and if they are buying for themselves or their clients. (heck, it might even be as a hedge where they don't care if it goes up or down). Now in terms of block trades, its very similar. Clearly there might be a different depending on what instrument we are talking about, but from my understanding when it comes to futures, trades can easily come in as a serious of 1 lots spread by just a micro second. In addition to this, I think I even saw some rules about how exchanges can group a bunch of trades together and lump them as 1 trade, to make it look like a large block, but really, it could just be made up of lots of little trades. I've seen lots of videos where traders try and guess institutional volume by having their platform only display trades above a certain size, like 47 or what have you in the ES, and they constantly change this number as if they supposedly cracked the code of how many units the institutions are trading in that day. Who knows, maybe they are onto something, but with so many variables, its really hard to tell. So back to the OP, the question you ask is very difficult to answer, and there might not even be an answer because you will never know.
Not only that, but by the time word gets out that Buffet has sold all of his stock, its old news, infact its not old, its virtually prehistoric ! Buffet would have spent the last X months quietly selling off his stock, small chunks here and there in trades that are carefully designed to hide in the general daily noise and fluctuations of the market.