For no reason at all other than for my own amusement, I like to produce short video clips documenting those rare occasions when the price ladder seems to make (some) meaningful sense (to me). Taken from last week, NASDAQ - NQ Futures; my market of interest for the moment. The videos and enclosed comments should speak for themselves. Second opinions welcome, let's try and keep it civil (heh...) Third one is my personal favourite. Will upload more when I get some time. Was a bit silly and should have included a live clock. In future I will do.
What is this? A DOM for ants?! Anyways, I'm of the opinion the vast majority of any large/outlier sizes seen have other non-obvious motives at play. I always assume it's a fake out, even if it gets filled, as unless someone wants to get in or out of something in a real hurry for a rare reason, there's really no reason for them to do it this way. In fact, we might just see "spoofing" and reverse signaling take different forms as HFTs become more aware of what is obviously fake vs what isn't. The parties doing this may be relying on the fact that the more aggressive they are the more the market is going to move away from them but in reality they comprise some amount of the other side of the market in the first place.
Why not rely on computers to do this faster and with much greater accuracy? Unless this is some type of archaic exercise like those guys who hunt deer with spears and rocks, i dont get the point. surf
Would you have been able to spot and trade these real time or do they happen too quick to allow time for decision making etc?
1) I dont know how to tell a computer to watch the DOM and front run in the event that an order looks like it's real. Or to fade it if it looks like the market isn't impressed by the size 2) Yes, you can spot and trade real time. I'll include a live clock in future so you can get a better feel (as the prior videos were in parts sped-up). Mind you, there isn't much time to think about it. But that's scalping isn't it?
You dont need to tell it, Harry. There are multiple off the shelf softwares that tape and DOM read. Its so easy even a caveman can do it. surf
So I finally got around to looking into these and will share some thoughts on why I think every single one of them would have been a bad signal to follow from the "obvious" side. (for note, my timezone is Pacific time, which is a -8h delta from you) In the above video, you see a large "seller" chasing an offer down and consider this as a sign to go short. The problem is, that seller probably knows that and is leading the sheep to the slaughter. The large order is pushing price down as people front-run it while those in the know are having their buy order liquidity filled. Even if that 200 lot offer was filled, they may still be net long afterward (consider how many people hit the bid on seeing that 350 lot and who knows, maybe they're +300 long by then). In the last candle, the invisible hand (and probably others) help to push price up into the stop zone igniting a serious breakout via cascading buy stops and chasers. In the above video, you see a large "buyer" wipe out the offers and leave a 750 lot on the bid and consider this as a sign to go long. One still would have had a shitty entry by following that signal and in reality it may have simply been someone covering their shorts from the large run down. Price consolidates afterward and drops again. Whether the 1000 lot buyer had any idea or not is of no consequence, buying would have gotten you nowhere good. In the above video you see a large "seller" chasing the price down - seemingly worried about getting out ASAP (you described it as a "large mess" for somebody). This implies price is going to be dropping quickly and someone wants out now. Part of it gets filled (200 lot or so), and everyone who took that as a sell signal got completely screwed (by design). This is basically the same situation as the first video: induce fear, get liquidity filled on the buy side in the process, rip the price up and profit. Consider that getting multiple 100s of lots filled may even have more power than straight up spoofing because other traders saw it - they now believe that order (and intent) is/was real whereas it's easy to spot spoofing and use it as a reverse signal. It may even be possible that other traders "in the know" with similar size used those large orders against the original trader, but I kind of doubt that. Consider too that if these signals were valid then why didn't buyers or sellers yank their orders on the opposing side quickly? Who knows, but you don't have to get caught up in this mess by following bogus signals and helping to provide liquidity to these entities. What's the consistent picture here? What you see isn't always as it appears and most obvious signals are a trap by design. They're not going to show their hand unless it's an extreme event - and if it is they'll just go market and be done.
Great and thorough response. I'm not yet convinced these were false signals, in so far as I believe you might have been able to snatch several ticks in each case. I agree that it makes little sense for anyone to show their hand in this way and therefore safer to assume the larger move is going to take place in the opposite direction. I have in fact been recording every day this past week and have skimmed through NQ... I didn't notice a single instance these last five days of a bid or offer persistently sticking out like a sore thumb. Actually the one exception is an Algo that pops up periodically flashing 250 on the bid or offer, but seems to be programmed to remain four or five ticks away from the inside market at all times. I didn't discern that the market paid these any mind but will keep an eye open. Hopefully more videos to come from next week though.
What's the consistent picture here? What you see isn't always as it appears and most obvious signals are a trap by design. They're not going to show their hand unless it's an extreme event - and if it is they'll just go market and be done. So do you find any use in these orderflow ideas?