In the previous session, we introduced the concept of continuous auctioning, where buyers and sellers continuously send their trading intentions. We discussed how the subsequent prices are formed, why there are price disparities, and how fluctuations and trends are created. Today, we will delve into further details about the aforementioned concepts and phenomena.

To have continuous auctioning, we must first establish a state of call auctioning. This is similar to an auction where there must be an initial bid price and a clear minimum bidding increment before further bidding can take place.

Thus, we need to use the results of the call auction to create the premise for continuous auctioning.

Advertisement

What is a call auction?

At the start of trading, a specific time period is set during which buyers and sellers can submit their genuine trading intentions. When you submit your intentions, it doesn't mean they are immediately matched. There is a time frame within which you can submit and, in some cases, withdraw your bids. In other time frames, you can still submit bids but cannot withdraw them.

What is a trading intention?

For a given product, do you want to buy or sell? At what price? In what quantity? After you submit, once a certain time point is reached, you cannot withdraw, and then the call auction matching can proceed.

What are the principles of matching?

After the submission period ends, the system will match the bids and offers based on the highest bid price and the lowest ask price. The goal is to achieve the maximum number of matches at the best possible price for both buyers and sellers. This process ensures that the market opens with a fair and representative price, which then serves as the starting point for continuous auctioning.If there can be an intersection between the lowest asking price and the highest bidding price, meaning a transaction is feasible, then I would set an optimal price. What is the optimal price? The price that can generate the maximum trading volume is called the optimal price, and it is found using an iterative trial calculation algorithm.

Let me give you an example.

The first column of this table is the asking price, the second column is the quantity, and the third column is the direction. You can see that the first four rows are all sell orders from sellers. The first person is selling two lots at 10 yuan, the second person is selling five lots at 9.9 yuan, the third person is selling three lots at 9.8 yuan, and the fourth person is selling ten lots at 9.5 yuan. The one with the most to sell and the cheapest asking price is 9.5 yuan.

When sellers are quoting, buyers do not know. Both parties base their decisions on historical data, especially the results formed at the close of the previous day, or the settlement price or closing price formed after the close, as well as the historical price trends and information derived from technical analysis over a certain period, and the macro news and specific news of the subject that we believe affect the price trend. I make a comprehensive judgment on today's price and the price I am willing to bear.

Then the buyers also start quoting. The buyers do not know how much the sellers have bid. The first person is bidding two lots at 9.6 yuan, the second person is bidding five lots at 9.9 yuan, and the third person, who is buying the most, is bidding seven lots at 11 yuan.

At this time, everyone first looks at whether there is an intersection between the buying and selling prices. The cheapest selling price is 9.5 yuan, and the cheapest buying price is 9.6 yuan, which can be traded. The most expensive buying price is 11 yuan, which can certainly be traded. So, what price should be traded?

For example, we can set the transaction at a price of 9.55, which is in the middle of 9.5 and 9.6.

At this time, everyone should note that because 9.55 is higher than 9.5, the ten lots at 9.5 will not be sold. However, the prices above 9.5 can be sold. But everyone should note that this buyer is at 9.6, and 9.6 is willing to buy above 9.55. But those at 9.8, 9.9, and other prices think 9.55 is too cheap to sell, so they won't sell. So at this time, the ten lots at 9.5 can be sold, but the sell orders above 9.5 cannot be sold.

For the buy orders, I can use 9.6 to buy the ten lots at 9.5, and 9.9 will also buy the ten lots at 9.5, with the final price set at 9.55. However, the ten lots of the sell order at 9.5 can be absorbed by the two, five, and seven lots at 9.6, 9.9, and 11, respectively. Since the price of eleven is higher, it should first be traded to the seven lots at 11 yuan, and the remaining ten lots minus seven is three lots, which are traded at 9.9.

The final actual transaction is that I traded at 9.55. The person who was willing to sell at 9.5, because it was lower than 9.55, sold his ten lots at a transaction price of 9.55, and he sold ten lots for 95.5 yuan.In this scenario, there are two prices involved: 11 yuan and 9.9 yuan. The 11 yuan is fully traded, with the final transaction price actually being 9.55 yuan, with priority given to the transaction, resulting in seven transactions. Out of the five transactions at 9.9 yuan, only three were partially executed, all at the price of 9.55 yuan. At this point, it is determined that there are ten transactions.

However, is ten the optimal number of transactions?

We can continue to raise the price, for example, to 9.8 yuan.

When the price is raised to 9.8 yuan, the three transactions at 9.8 yuan can also be sold. Of course, the sellers at 9.5 yuan are willing to sell at a higher price. From the buyer's perspective, both 9.9 yuan and 11 yuan are higher than the selling price of 9.8 yuan, so I am willing to buy, and this adds up to twelve buy orders. Therefore, at this time, transactions can still occur at 9.8 yuan.

The transactions at 11 yuan are fully executed, taking seven out of the ten buy orders at 9.5 yuan. Then, from the five buy orders at 9.9 yuan, two more are taken from the previous three, leaving two more that can be taken from the sell orders at 9.8 yuan.

So when the price is set at 9.8 yuan, the transactions increase to twelve.

Can it go higher? Of course, it can be set to 9.9 yuan.

At this point, all twelve buy orders at 9.9 and 11 yuan are fully executed. For selling, ten transactions are made at 9.5 yuan, two at 9.8 yuan, but the 9.6 yuan cannot be sold, so more is sold at 9.9 yuan. However, the buying side remains at twelve buy orders, with no new buy orders, so the optimal quantity is twelve.

Therefore, it is more appropriate to set the optimal price at 9.8 yuan.

Thus, because they are willing to sell at 9.8 yuan, but if we consider fairness, we can take a mid-price between the cheapest buy order at 9.9 yuan and the sell order at 9.8 yuan, and sell at 9.85 yuan.If I sell at 9.85 yuan, for both the buyer and the seller, the transaction volume would be twelve lots, the maximum transaction volume. However, it is a bit more advantageous for the seller. In this way, through a trial calculation method, the transaction price with the maximum transaction volume is found.

If you don't understand this example, that's okay. The main point here is to give everyone a sense of how it works. At the beginning, when we talked about the opening call auction, after you submit these orders, how does it find the transaction price that can match the maximum transaction volume through an algorithmic approach?

At this time, I hope you can either not understand the algorithm or forget the algorithm, but what you need to remember is that after a period of collective bidding, a result is formed, which is the result after the matching is completed. At this point, the orders that have been matched are traded, and this price is called the opening price. The opening transaction volume is the twelve lots I just matched.

What about the unexecuted orders?

Note that the unexecuted ones are the source of continuous bidding. They form a series of orders with the intention to trade during the call auction period, but they were not able to be traded during the call auction. So, I just leave them there, and they will be continuously matched as the continuous bidding proceeds.

Now we have four concepts:

The first is order flow;

Order flow sounds very fashionable, but in reality, it is quite simple. It is the new buying and selling intentions at a given time point, which we call a timestamp. When you place an order, I place an order, and anyone places an order, it is called the order flow. It converges into the exchange.

This order flow has a timestamp that indicates when the order was placed, indicating the buying and selling intentions, how much money? How much quantity? This is called the order flow.

The order flow is like water, continuously flowing into the exchange. After flowing in, one order at a time is collected, and I match it with the existing pool of buying and selling intentions formed by the call auction just now.The second order book;

At a specific timestamp, the accumulated willingness to buy and sell is called the order book.

This order book represents the accumulated buying and selling intentions at a given point in time. When my order flow comes in, it is matched with my existing order book. If the match is unsuccessful, my order flow is then placed on the order book, updating the status of the order book.

Of course, some order instructions are special, such as all-or-none or cancel, or fill-or-kill. If it cannot be filled, it is canceled. However, these special instructions are not considered here. After my order comes in, if it cannot be filled, it is hung there, updating the order book.

If it can be filled, I will reduce it. The matching of buy and sell orders follows the rules of trade matching, such as time priority and price priority, etc. After the matching is completed, the status of the order book is updated, forming a new order book.

So, the order book has a concept of time; it is a kind of buying and selling intention that is accumulated at a specific timestamp.

The third trade record;

Correspondingly, if the order flow can be successfully matched with this accumulated order book, the result of this match is called the trade record.

That is to say, the buying and selling intentions I accumulated before have finally been matched by the newly added order flow. For the person who newly sent this, my buying and selling intentions have been matched by a previous stock of buying and selling intentions. Because time is discrete, it is like a continuous dynamic matching process.

If a trade has been completed, whether it is the person who previously formed the order book with the wrong buying and selling intentions or the person who newly sent the buying and selling intentions to form the order flow, they both receive a trade record: the same time, the same price, the same quantity, just with different buying and selling directions, this is the trade record.What is the fourth thing called the market trend?

If you understand the process of continuously matching transactions just like buying Chinese cabbage, then the market trend is a sampling display of transaction records and order books.

Due to some rules of data management and the performance of the exchange's information system, it does not mean that all data is displayed to you. So the market trend we see is a result of sampling.

Generally speaking, it can be considered that 500 milliseconds is a relatively small limit for displaying market trends. I summarize and display the transaction records every 500 milliseconds, and we usually call this Tick data.

Please pay attention to this chart.

It gives a 5-minute intraday Tick chart of the continuous contract of rab on August 22, 2016. You can see that within a very short period of time, the price remains unchanged. So this Tick is a continuous horizontal line within a specific period. This is the most refined display of intraday Tick that we can obtain relatively.

What is the market trend? The market trend is a sampling display of transaction records.

Understanding the mechanism of these data generation, when we see this plate in trading. We not only know what it is, but also why it is.

So here I give a brief introduction to stock and futures market trends:

Stock market trendLevel 1: Every 500 milliseconds, summarize the transaction records once, serving as pseudo-tick data, and update the order book every 6 seconds on average, displaying five levels of market quotes.

Level 2: Display transaction data at the 10-millisecond level, and update the order book every 3 seconds on average, displaying ten levels of market quotes.

Although the Level 2 market data is continuously evolving with the purification of the securities trading group. For example, the Shenzhen Stock Exchange has begun to provide what is called a holographic order book, showing you all levels of quotes in the order book, and even all the queues above each level are displayed. As for transaction data, it tries to provide you with high-precision timestamped data as much as possible.

Futures Market

On average, summarize transaction records every 500 milliseconds as pseudo-tick data, and update the order book, displaying one level of market quotes.

For the futures market, after forming a transaction every 500 milliseconds, you can see in the transaction details that approximately two entries are updated per second. Due to some fluctuation in the time it takes for the data to reach the local server, generally speaking, on the same second timestamp, sometimes there are two market quotes, and sometimes there are three.