Welcome to NexusFi: the best trading community on the planet, with over 150,000 members Sign Up Now for Free
Genuine reviews from real traders, not fake reviews from stealth vendors
Quality education from leading professional traders
We are a friendly, helpful, and positive community
We do not tolerate rude behavior, trolling, or vendors advertising in posts
We are here to help, just let us know what you need
You'll need to register in order to view the content of the threads and start contributing to our community. It's free for basic access, or support us by becoming an Elite Member -- see if you qualify for a discount below.
-- Big Mike, Site Administrator
(If you already have an account, login at the top of the page)
Reminds me of the old Beach Boys track "Wouldn't it Be Nice". In my humble view, this is a starter for sure but we need to always be aware that the big boys (and girls?) are playing games with us and each other. Like chart patterns that stand out to 'everyone', tape patterns like this are equally well known so can be programmed to appear and trap unwary traders into poor positions.
Many HFT firms are operating like market makers, scalping the spread or a little more. They can be a big buyer one second, and just after we place our buy order, the HFT becomes a big seller 1-3 ticks up, taking their scalp and pushing the market back down again for a few ticks where it has its 'big buyer' orders waiting to cover the shorts. Rinse and repeat the process for as long as it works, or until another game sequence is initiated.
I often imagine that some of these big traders have their own secret competitions with each other where they are trading to meet objectives like getting the market back to the open by a certain time! Lots of scenarios to imagine but most of the time, I believe, enough big players will co-operate to add weight to each others' games for their mutual benefit, and our mutual loss.
How does anyone know an iceberg is broken? On the Jigsaw snapshot we have a historical record of what has been added or removed from level, but we do not know if there are any more orders outstanding from the iceberg (or icebergs more likely) to make further changes. Only when the inner bid/offer changes do we know the orders have been exhausted (in appropriate direction).
If some exchanges do not follow these rules then I guess they may automatically shift iceberg order up or down a tick. But what mathematical decision/calculation would make sense to do this? If I am Mr Big Trader I do not want my order being messed with in this way. If I am happy to buy 5,000 contracts at a price, they should stay there until gone. Oh what a dream.
Even Titanic could not break that iceberg, sadly.
Sorry for being so fussy. Been an engineer for far too long it seems.
Can you demonstrate these things with real life examples like a chart or video of live data or any other thing?That will be very helpful to less-knowledgeable people like me.
Hello.
Happy Holidays!
Thank you all for your answers, special thanks to Bloom for the reply and the video, it was very interesting to see.
I have two questions:
1. Why all use 576tik schedule? Why 576?
2. Can I use the Merket Replay in NinjaTrader? Or is there a replay data TS distorted?
Thank you.
I'm not trying to coin a phrase here but by "broken" we are discussing those times where:
- limit orders are being added to a level
- someone throws a big order at that level and price moves through it
- limit orders get refreshed at the original level, it pops back up there and the refreshing continues
If there's an iceberg on the inside bid and the price moves through it momentarily, there's no need to adjust the iceberg price at all. The moment you refresh limit orders at that level, as long as they outnumber the offers at that level, it becomes the inside bid again.
You never know if there are more orders going to be added at any moment but when they get refreshed, price pushes through and they get refreshed again at the original level, you know what it is you are seeing, right?
Remember - both platforms and exchanges provide functionality for refreshing orders and I really don't know which one's are used the most. I guess when I'm that big, I'll look into it.
I am no guru but simply watching for these strategies all the time while learning the tape. Will try and capture some video in future, as well as review a few videos posted here, but the ES is so complex with its high liquidity and action. For now, here is a description of a typical HFT strategy that would work against us going long when strength appears at the bid. Perhaps more common on a suitable stock but I stick to ES for now:
1. Trader A initiates an iceberg Bid order buying up 500 ES contracts at a time. Trader A is an institution who wishes to buy 5,000 contracts.
2 Trader B guesses an iceberg is on system and wants to make a few ticks by encouraging Trader A to raise her Bid. Trader B lifts the offer by buying 500 contracts on a market order, which for sake of simplicity here, is enough to raise the inner Bid/Offer each time.
3. As Inner Bid/Offer moves up, Trader A moves her iceberg order up to inner Bid.
4. Trader B buys another 500 contracts and lifts Offer, raises inner bid/off.
5. Trader A moves up iceberg.
6. This repeats for 3-4 ticks until Trader B sells all bought contracts for more than bought.
7. If Trader A does not want to play this game, Trader B can lose if Trader A does not move iceberg, and others come in and sell heavily into iceberg. No guarantees, even for big players, just probabilities.
So, for example, a summary tape might look like this: (Read from bottom up of course! Hope no typos.)
09:43:30 1429.75 300 | 20 1430.00 -- 'We' are losing and only been in market for much less than a minute.
09:43:28 1430.00 300 | 400 1430.25 -- Back to 1430.00 and 'We' wonder what happened to our sure buy.
09:43:28 1430.25 1300 | 200 1430.50 -- More sellers...
09:43:28 1430.50 800 | 0 1430.75 -- Sellers...
09:43:26 1430.75 2000 | 60 1431.00 -- Sellers follow...
09:43:26 1431.00 600 | 250 1431.25 -- Mkt ticks up on other buyers. Trader B sells all 2,000 contracts for profit*
09:43:16 1430.75 1000 | 500 1431.00 -- And again...
09:43:02 1430.50 800 | 500 1430.75 -- And again...
09:42:30 1430.25 500 | 500 1430.50 -- Trader B lifts again, Trader A follows up a tick...
09:42:09 1430.00 2500 | 500 1430.25 -- Trader A iceberg absorbs sellers. BUT Trader B lifts offer. 'We' buy!
09:42:08 1430.25 400 | 0 1430.50 -- Market moves down to level that Trader A's iceberg order starts.
Result: Trader A bought all he needed but had to pay higher bids. Trader B bought and sold 2,000 - 500 for 1 tick, 500x2t, 500x3t, and 500x4t = USD56,250 less very low commissions. Not bad for 2 minutes. HFT algos are not likely to get caught with this but less savvy institutions could. Once or twice a day is still good money. Add in the many other money-making strategies in use throughout the day and you soon get to a billion dollars.
The above is obviously simplified to focus on Trader A vs B action, and assumes Trader A allows Trader B to pull his iceberg up 4 ticks. The giveaway in this simplified example is the inequality between bid and offer traded volumes - a divergence overblown here for clarity.
Yes, that is the Beach Boy song. I was referring to the song title really, as lyrics are not relevant. Simply trying to say that "would'nt it be nice" if reading the tape was so easy.
And that video has 10,801 likes till now! Your 118 dislikes quote represents just 1% - just the tip of the likes iceberg.
Merry Christmas and Happy New Year everyone.
Taking a short break from the tape!
Richard
Hong Kong
Ahh... semantics. If limit orders are being added to a level that means an iceberg is still in operation to me. Broken suggests something not working but if orders are being added it is working fine.
If a big order moves to another price/level, then iceberg has finished in my mind.
To me, limit order refreshing at original level means the algo is creating orders to bring market back to level to allow a new iceberg to operate. This fluctuation is the way I believe these algos are programmed to disorientate other traders (algos and humans). But as I am a human I could be wrong of course.
I guess my thinking above applies to those big fellas/gals that can afford to be co-located within same building as exchange servers. Lots of program traders too who just jump in and out for arbitrage and other seemingly mindless trades. And plenty of poorly programmed HFTs messing up the tape too - they do not all win. Why we human discretionary types stick with the ES is strange given all this in-built stress.
When we both get bigger we can join forces and add some chaos of our own to the tape.
Thanks for your ongoing support here. Much appreciated. My counter-beliefs would not be possible without the extra insights afforded by that awsome toolset of yours.
Seasons greetings and best wishes for a fabulous New Year.
Richard
Hong Kong
I don't agree - fact is, you can only try to keep refreshing at a level, you can never guarantee price won't move thru you because of a large order. Just because price moves thru you momentarily does not mean you are done.
Fact is if you are refreshing to keep it at 600 and someone puts thru an order for 1000, it's going to push through you.
That is quite possible and indeed that does happen. In fact this is one of the setups in John Grady's "No BS Day Trading" ebook. The difference when this occurs is the time allowed for the inside level on the opposite side to fill up before whacking it. This is getting into the area of publicly revealing what's in his dead cheap ebook, so Skype me and we can discuss it but I'd rather not to it publicly as I learnt it myself from his book.
Anyway - when it moves thru a level being refreshed and then pops back up again, it's just the refreshing continuing in my opinion.
Best way I heard the ES described was "Crack for traders"
I like the discussion to be honest and it doesn't matter which tools people use, the discussion is the same.
It's good you are coming to your own conclusions and disagreeing with me on this. You've 'got it' now and you just have to turn the knowledge into action.
I think it's also worth asking you at this point, if you agree with the following:
1 - A newbie would have a hard time learning this by just "staring at the tape" with no idea of what to look for
2 - Once armed with basic knowledge of what to look for, you have to put in the hours yourself to move forward, no amount of teaching can replace those hours watching the tape
3 - Once you have gone through step number 2 above, the questions you ask about reading order flow change radically
Regarding the video, I thought you would have been happier if the number of dislikes were more..After all you believe that things are not supposed to be that easy("Wouldn't it Be Nice" to take trading easily!).and the song is just the opposite of that...just a thaught.
Ironically, you seem to feel happier with so many likes
BTW,waiting for your videos and Merry Christmas and Happy New year to you
I believe we are both right actually (that's nice) but each describing different kinds of iceberg orders.
The normal kind as discussed by me is the 'reserve order' where a hidden amount of orders is setup for auto refreshing as each part is hit. There may be 500 shown on the DOM, but as soon as that is taken, another 500 chunk is put up - until say all 5,000 orders are done. If a trader has such an iceberg order on the exchange, it will usually refill limit orders at same level this way until order is exhausted. Therefore, a larger than displayed market order hitting the bid for example, will just get swallowed up for as long as the hidden order still has room. Only when the iceberg has finished will the inner bid drop.
In your case, you are talking about the likely more common 'timed order' where iceberg chunks are delayed by a set time interval - or more likely to avoid detection, varying time intervals. This is also done by the exchange. Therefore, in this case a large order will take out level in between iceberg chunks. But as discussed earlier, the iceberg algo just has to pull market back into needed zone to continue operation. Subsequent iceberg chunks join the order queue when put up so that other traders (us) can still get their orders filled as normal, in time sequence.
Anyone attempting to use iceberg orders from their own non-exchange server/PC would have little chance to ever interrupt a standard iceberg order given the micro/nanoseconds between the exchange handling the orders. But obviously sometimes a rogue mkt order can slip in just based on exquisite timing and Murphy's Law.
I have that John Grady ebook and did not realise my example above was discussed there. Have read many articles and books on topic and it just comes out as common experience now. That strategy above is called 'towing the iceberg' in another book I read. Chinese Site Link discussing book - Use Google Translate! Nobody has the copyright on market games but I sure wish I could read the Goldman Sucks in-house games manual someday. For now, best to focus on how the HFTs operate and glean strategies from there.
1 - Agreed. I find it better to understand how the market works (what games are played) and then watch the tape to look for setups. Difficulty being that myriad games and transactions are being implemented in parallel of course. But like other fields, theoretical study together with hands-on practical experience is best way. No shortcuts.
2 - Yes, many hours. Does not seem to be something the brain can absorb quickly or easily. Too many distractions, spoofs, activity even watching one T&S. But hours offline studying the mechanics of the market pays off as you then learn to recognise tape movements more easily. I am still in this phase. So no amount of teaching can replace hours watching the tape, but I believe we still need both. Market strategies are always changing and as all markets are correlated, we need to know more than simple mechanics of the tape.
3 - Still working through my multi-hours learning but yes, the questions must change. But that does not mean my earlier statements are dumb! When I first looked at the DOM I was puzzled by how a single contract could move the price. Only after understanding the market auction process can you begin to start learning how the 'tape' behaves. Your question reminds me of basic education where you learn stuff (science/maths/etc) at one level, say pre-college, then later at college they tell you to forget that stuff in your head and learn entirely new ideas/concepts at the next level. And so on... Understanding comes from experience too, building on past knowledge and sometimes discarding 'truths' once held to be eternal.
