Fins in the glass - trying to understand what happened by the ticks
Actually Si-6.19 instrument 03.04.2019 13:00
I was confused by the spread in the tester - over 60 pips, so went to look at the tick history.
Here's the statement
| <DATE> | <TIME> | <BID> | <ASK> | <LAST> | <VOLUME> | VolSumm |
| 03.04.2019 | 13:00:11.950 | 65811 | ||||
| 03.04.2019 | 13:00:11.979 | 65875 | ||||
| 03.04.2019 | 13:00:11.979 | 65811 | 5 | |||
| 03.04.2019 | 13:00:11.979 | * | * | 12471 | ||
| 03.04.2019 | 13:00:11.979 | 65875 | 55 | |||
| 03.04.2019 | 13:00:12.007 | 65821 | 3 | |||
| 03.04.2019 | 13:00:12.007 | 65822 | 4 | |||
| 03.04.2019 | 13:00:12.007 | 65822 | 3 | |||
| 03.04.2019 | 13:00:12.008 | 65814 | 65819 | |||
| 03.04.2019 | 13:00:12.009 | 65819 | 3 |
The asterisk omits the purchase volumes.
Here is a graphical representation of the tick
And I don't understand, what kind of situation was this? Suppose someone bought 12471 lots at once (doubtful), but how he bought lots cheaper than in the initial part of the market, namely 65822, if he bought 65875 on this tick? Or is it a chronology recording error? If it's a mistake, then the next ask is already at 65819!
In the section of this tick I have a stop loss, and it is clear that the tester closes at the maximum price, but in fact the price should be lower, for the broker would not have had time to drive the trade into this movement, and the result is a difference of 30 points between ticks and OHLC.
Attached is a minute with ticks.
The Si-6.19 tool itself 03.04.2019 13:00
...
Attached is a minute with ticks.
Look carefully at the file attached by you. For 2019.04.03 13:00:11.979 there are hundreds of ticks with different volumes. The order of ticks in the list provided is not guaranteed. So the price of 2019.04.03 13:00:11.979 65875 was not really the first.
Judging by the ticksheet, indeed someone threw a bid for 12471 lots and thereby ate up the whole cup. It can be seen by the number of bidders who have traded this bid - there are more than a hundred of them, for sure.
In general, the situation is not typical for Sishka, because even suckers do not trade with such volumes on this instrument with market bids.
What's not clear? Somebody decided to sell on the market, so he collected all the limiters.
Two things that are not clear are the consistency and the strong price bounce, i.e. the appearance of the pending orders almost in the same place after the market buyback.
Look carefully at the file attached by you. On 2019.04.03 13:00:11.979 there are hundreds of ticks with different volumes.
The tick is just one, i.e. it is one movement on the cup, since there is no Ask and Bid between them. Actually, the price has jumped, and in the interval those volumes that were eaten. And note the timing of the return tail - it does not coincide with the main movement, which also raises questions...
The order of ticks in the list provided is not guaranteed.
Ticks are price-deterministic, not time-deterministic. A tick is a database entry that brings together a buyer and a seller. Entries can be made simultaneously, in batches, one tick from a batch has no time advantage over another tick from the same batch, even if their prices are different. That is, they are actually created simultaneously.
The tick is exactly the same...
There is only one tick. You can see on the tape that there are hundreds of them:
2019.04.03 13:00:11.979 65811 5.00000000 2019.04.03 13:00:11.979 65812 17.00000000 2019.04.03 13:00:11.979 65812 3.00000000 2019.04.03 13:00:11.979 65812 4.00000000 2019.04.03 13:00:11.979 65812 4.00000000 2019.04.03 13:00:11.979 65813 2.00000000 2019.04.03 13:00:11.979 65813 1.00000000 2019.04.03 13:00:11.979 65813 17.00000000 2019.04.03 13:00:11.979 65813 1.00000000 2019.04.03 13:00:11.979 65813 4.00000000 2019.04.03 13:00:11.979 65813 1.00000000 2019.04.03 13:00:11.979 65813 5.00000000 2019.04.03 13:00:11.979 65813 1.00000000 2019.04.03 13:00:11.979 65813 1.00000000 2019.04.03 13:00:11.979 65814 1.00000000 2019.04.03 13:00:11.979 65814 3.00000000 2019.04.03 13:00:11.979 65814 2.00000000 2019.04.03 13:00:11.979 65814 17.00000000 2019.04.03 13:00:11.979 65814 10.00000000 2019.04.03 13:00:11.979 65814 1.00000000 2019.04.03 13:00:11.979 65814 25.00000000 ... 2019.04.03 13:00:11.979 65875 5.00000000 2019.04.03 13:00:11.979 65875 10.00000000 2019.04.03 13:00:11.979 65875 1.00000000 2019.04.03 13:00:11.979 65875 20.00000000 2019.04.03 13:00:11.979 65875 1.00000000 2019.04.03 13:00:11.979 65875 55.00000000
Two things that are not clear are the consistency and the strong price bounce, i.e. the appearance of the pending orders in almost the same place after a market buyout.
Competitive market. The flow of limit orders is very powerful: dozens of orders per second can go in. After all the liquidity is gobbled up, the glass is immediately filled with new limit orders, maintaining liquidity. No one would simply move the price by 50 rubles, even by a single large bid. In illiquid, non-competitive markets, yes, it is possible; in markets like Si, it takes some pro-trading before the price changes.
The famous flash cross happened due to the fact that the limit order flow was not big enough. The market order flow turned out to be much stronger, everybody rushed to sell on the market and price-givers did not have enough time to place their orders. Having analysed the situation, the exchanges introduced a restriction on the placing of market orders that work in certain cases - the idea is to give liquidity providers time (a few seconds) to place limit orders, protecting the market from a failure.
Ticks are price-deterministic, not time-deterministic. A tick is a database entry that brings together a buyer and a seller. Entries can be made simultaneously, in batches, one tick from a batch has no time advantage over another tick from the same batch, even if their prices are different. That is, they are actually created simultaneously.
What about one. You can see on the tape that there are hundreds of them:
OK, let's put it this way - 1 trade can contain more than one tick. After each trade we see asks and bids in the file - do you agree?
Here was one trade, because in the gap there are no other asks and bids - the buy and sell prices, because they are updated after the end of the trade.
Then how is it that the deal was split into time slots? How did the price and time returns occur later than the main deal.
Competitive market. The flow of limit orders is very powerful: there can be dozens of orders per second. After all the liquidity is gobbled up, the glass is immediately filled with new limit orders, maintaining the liquidity. No one would simply move the price by 50 rubles, even by a single large bid. In illiquid, non-competitive markets, yes, it is possible; in markets like Si, you need to trade through before the price can change.
I can't agree with that - at strong moves just the limits move after the price, and if the whole glass is eaten, only MM can fill it immediately with the same prices - why do other participants need it? As the result we can assume that MM processed the order and put limits in the glass at the same frame. I.e., he did it not after the deal closure, but immediately in the same frame, i.e., he either receives information before all market participants, or it is a conspiracy.
...
I can't agree with it - at strong movements just limits move after the price, and if the whole glass is eaten, only MM can fill it immediately with the same prices - why do other participants need it? As the result we can assume that MM processed the order and put limits in the glass at the same frame. It means he did it not after the deal closing, but at once in the same frame, i.e. he gets the information before all market participants or it is a conspiracy.
This is your speculation with a conspiracy theory. There is no MM on Si.
OK, let's put it this way - 1 trade can contain more than one tick. After each trade we see asks and bids in the file - agree?
...
One trade = one tick. As many trades, so many ticks. There is no difference between a "trade" and a "tick".
This is your conspiracy-theory speculation. There is no MM on Si.
Forum on trading, automated trading systems and strategy testing
Market maker, sharks and stock plankton
Aleksey Vyazmikin, 2019.03.05 22:19
Looked at the Si futures there and didn't understand, it turns out that 3 companies are now acting as MM at once? As they have it listed that the contract is renewed at the end of 2018.
Here is the list of MMs:
"Joint Stock Commercial Bank Derzhava Public Joint Stock Company"
Limited Liability Company 'Brokercreditservice Company
Limited Liability Company "INTRUST Financial Company
If the first two organizations are known and profiled, the third one I have doubts about, the fact is that it is registered (its founder is)ZAO "Asmk", with an average number of employees 2 people, and the main activity is the " Publication of books, pamphlets, brochures and similar publications, including printed dictionaries and encyclopedias, including for the blind. In other words, the firm is clearly a risk minimiser. At the same time,Asmk has14 founders (shareholders) :)
Ranking of market makers - leaders of the Ranking Programme
based on March 2019 results
| Programme option | 1 place | 2 position | 3rd position | 4th position | 5thplace |
|---|---|---|---|---|---|
| Futures on the MosExchange Index, MosExchange Index (mini) | RIKOM-TRUST INC. | Renaissance Broker Ltd | OOO BKS Company | INTRASTRASE TRADING COMPANY LLC | |
| Futures contracts on currency pairsSi,Eu,ED | Renaissance Broker Ltd | LLC "Kompaniya BKS" - Futures contracts on currency pairs Si, Eu, ED | - | PJSC JSCB Metallinvestbank | LTD "INTRASST" FX |
| Futures contracts on sharesGAZR, SBRF, VTBR, LKOH,ROSN,GMKR | ALOR + LTD. | INTRUST FINANCIAL COMPANY | I.T. Invest investment company JSC | PJSC "Best Eforts Bank |
- www.moex.com
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
You agree to website policy and terms of use
Actually Si-6.19 instrument 03.04.2019 13:00
I was confused by the spread in the tester - over 60 pips, so went to look at the tick history.
Here's the statement
The asterisk omits the purchase volumes.
Here is a graphical representation of the tick
And I don't understand, what kind of situation was this? Suppose someone bought 12471 lots at once (doubtful), but how he bought lots cheaper than in the initial part of the market, namely 65822, if he bought 65875 on this tick? Or is it a chronology recording error? If it's a mistake, then the next ask is already at 65819!
In the section of this tick I have a stop loss, and it is clear that the tester closes at the maximum price, but in fact the price should be lower, for the broker would not have had time to drive the trade into this movement, and the result is a difference of 30 points between ticks and OHLC.
Attached is a minute with ticks.