The following content is for reference only and is not investment advice. Always do your own research and consult a financial professional before making any investment decisions. The strategies and observations discussed in this article are based on historical data and market patterns and are no guarantee of future results. The cryptocurrency market is extremely volatile and trading carries significant risks, including the potential loss of your funds. Please ensure that you fully understand the associated risks before participating in any trading activities.
Preface
We all know that Binance is the largest centralized cryptocurrency exchange, so its announcements can serve as important market indicators and may form the basis for developing a trading strategy.
The idea of this strategy is based on the fact that every crypto project hopes to be listed on Binance because Binance provides the highest liquidity in the market. For a project, listing on Binance can be said to be the best thing. Suppose we think about it the other way around, what will happen if the token is delisted by Binance? Let's start from this perspective.
Research Methods
When we say “delisted from Binance,” we are referring to the removal of spot trading. Why spot and not contract? Based on the patterns we have observed, Binance typically delists contracts first and then delists spot trading. When spot trading is removed, it means all liquidity for the token is withdrawn from Binance.
Therefore, our research focuses on spot trading, specifically trading pairs whose titles on Binance’s announcement page are often “Binance will delist…”.
我們的樣本代幣有：[CVPUSDT, EPXUSDT, FORUSDT, LOOMUSDT, REEFUSDT, VGXUSDT, BONDUSDT, DOCKUSDT, MDXUSDT, POLSUSDT, OMGUSDT, WAVESUSDT, WNXMUSDT, XEMUSDT, DREPUSDT, MOBUSDT, PNTUSDT, ANTUSDT, MULTIUSDT, XMRUSDT, BTSUSDT, PERLUSDT, TORNBUSD, WTCUSDT, SNMBUSD, SRMBUSD, YFIIUSDT, NEBLUSDT, AUTOUSDT, HNTBUSD, AIONUSDT, MIRBUSD, ANCBUSD, BEAMUSDT, MITHUSDT, TRIBEBUSD, REPUSDT, BTCSTBUSD, DNTUSDT, NBSUSDT, BTGUSDT, TCTUSDT, HEGICBUSD]。
Observation period:
Study Pattern: 1 Minute Chart
Start time: Start observing the price 1 minute before the announcement.
Ending time: Observation will end 7 days after the announcement. Why seven days instead of six or eight days? Because after the announcement, Binance usually gives users 8 days to process their positions. After this period, users will no longer be able to trade the delisted trading pair.
Define the first K line:
start_bar: The K-line of the minute immediately following the announcement time is defined as "start_bar" for subsequent convenience. After that, it is called the "first K-bar".
Decline of the first K bar = (opening price of start_bar - closing price of start_bar) / opening price of start_bar
Hypothetical entry point:
second_bar (second K-bar): The opening price of the next K-line (second K-bar) formed after start_bar is regarded as the hypothetical entry point. We use the second candlestick instead of the first one because the crawling process takes some time (less than 15 seconds) and in the worst case we may enter the trade on the second candlestick, thus taking the worst case scenario As a hypothetical entry point, the stop loss point is set 1% above the opening price of start_bar.
Measure the decline:
The maximum drawdown is measured from the opening price of second_bar.
During the 7-day observation period, track the maximum percentage decline, and the price during this maximum decline cannot be higher than the opening price of start_bar, because this is our stop loss point.
Research objectives
Calculate the drop of start_bar physical K bar
= (start_bar opening price - start_bar closing price) / start_bar opening price
Why does the drop in start_bar matter? Because it represents our risk. But why do we set the stop loss at 1% above the opening price of start_bar? Because the market we assume is that after the announcement, the target will continue to fall and fall without looking back. Therefore, the physical K of start_bar is strong The range represents our risk!
Calculate the maximum decline within 7 days after the entry point (second_bar's opening price), and the price cannot be higher than the start_bar's opening price during the process.
Explore whether there is any correlation between the above two items, and if not why?
Can corresponding methods be developed to operate?
Case Study 1 – CVPUSDT
CVPUSDT is an asset that has been removed from the shelves. Let us put the terms defined above into the chart to view. In addition, it is currently not possible to view charts on Binance or TradingView, but the data can be obtained through Binance’s API.
As we mentioned, the observation period is 7 days and we measure the maximum percentage decline starting from the opening price of the second candle.
In the CVPUSDT case, the "first K-stick drop" is 0.739%, and the "maximum drop" is 66.538%.
Case Study 2 – REEFUSDT
Next, the case of REEFUSDT is quite special. Let’s look at the results first. In the case of REEFUSDT, the “drop of the first K stick” was 26.027% and the “maximum drop” was 7.284%. We can see that the first K bar of REEFUSDT has a large drop, but the maximum drop is actually very small.
From this, we can deduce a hypothesis: when the first K bar drops significantly, it will squeeze the space for the maximum drop to a certain extent. To put it simply, the first one will fall completely, and the subsequent decline will be relatively small.
It can also be inferred from this that if the follow-up strategy fails, a similar situation will generally occur. Participants rush into the market at the same time and perform the same behavior, causing the price to slide downwards significantly, causing a very large drop. K bar, the relative maximum decline will also be compressed, which is a sign that I think the strategy may fail in the future.
final result
There are 43 tokens in the sample list. After applying the same process to all samples, the results are shown below:
The chart above (blue) shows the largest drawdowns
The chart below (green) shows the decline of the first K bar
Visually, when the first K bar's drop (lower chart) is larger, the maximum drop (upper chart) tends to be smaller. This is the situation of REEFUSDT mentioned above. If the room for decline in a short period of time is limited, then the decline of the first K bar will compress the room for the maximum decline. Additionally, the chart is sorted by time, with 2024 on the left and 2022 on the right. It can be seen that although it is not obvious, the trend of the maximum decline decreasing over time still exists. This is also a sign that one of the strategies may fail. This data must be tracked regularly. If the maximum decline slowly decreases (profit) in the future, it also means that The strategy may gradually become ineffective.
But when we perform correlation analysis on these two sets of data (the drop and the maximum drop of the first K bar), the correlation is -0.214. The data shows that the two are weakly correlated, but tend to have no correlation. Why? So? You can take a look at it first and then continue reading.
Explore whether there is any correlation between "the decline of the first K bar" and "the maximum decline". If not, what is the reason?
The following ideas are personal subjective opinions for reference only. The current interpretation is that this strategy has not failed or there are not enough samples of transaction failure (stop loss), so the number of money-losing transactions in the only samples is not enough. Make them very relevant. However, it is inferred that in the future, as more and more people know about this strategy, its maximum decline may slowly decrease, and the decline of start_bar may also increase relatively. The possible reasons are as mentioned in the REEFUSDT case. When multiple participants perform the same behavior at the same time, the drop in the first leg will increase, and potential profits may also decrease.
Can corresponding methods be developed to operate?
First of all, if the entry point and stop loss point continue with the above assumptions, then the profit stop point and filter remain to be considered. As mentioned earlier, if the price of the first K-stick drops too much, it will relatively compress our profits. Therefore, we need a filter to filter out some not-so-good opportunities. When the price of the first K-stick exceeds a certain value, We won’t enter the market when the time comes, so I call this parameter Start_bar limit.
Next, we also need an appropriate stop profit point, which I call Tp_point. So I ran the optimization to find out its parameter plateau. First, the x-axis, y-axis, and z-axis of this optimization model are Start_bar limit, Tp_point and PPF in order. The formula of this PPF (Pessimistic Profit Factor) as follows:
PPF is a measure of robustness that is calculated by adjusting for the square root of the number of wins and losses (note the influence of the sign). Such a design would make small transactions with smaller sample sizes more susceptible to adverse effects. Since the sample size is extremely small, we choose to use a model that is disadvantageous for small samples as the optimization metric. However, since this optimization model is not scientific enough, it is for reference only.
Assume that the number of profits and losses for data A is 100 times, and the number of profits and losses for data B is 10 times. Assume that the average profits and losses of the two are the same:
Win count: 100-sqrt(100) ; Lose count: 100+sqrt(100)
Win count: 10-sqrt(10) ; Lose count: 10+sqrt(10)
When there are few samples, the win count will be deducted in a larger proportion, and the lose count will be added in a higher proportion. The calculation formula is as follows:
sqrt(100) /100 = 0.1
sqrt(10)/10 = 3.162/10 = 0.3162
Why is it said that this optimization model is not scientific enough?
Subjectivity of parameter selection and data distribution properties
When selecting parameters, due to the high degree of subjectivity, we must pay attention to the fact that there is no significant correlation between the two variables "the drop of the first K bar" and the "maximum drop". Therefore, the parameters of this model may be inconsistent in practical operations. Invalid. However, considering the future adaptability of the strategy, it is still necessary to use the "drop of the first K bar" and the "maximum drop" as the x- and y-axis parameters. In particular, the "drop of the first K-bar" can not only be used as a screening condition, but also reflects potential risks to a large extent.
How to interpret charts
Caution is required when interpreting the graphs, especially given the relatively small number of failures so far. The fragmented parameter distribution means that the stability of the results may be affected, so when analyzing the model, it is important to note that these variables are not absolutely influential. Charts primarily provide a reference for evaluating the possible impact of parameters under different scenarios, rather than making final decisions.
It can be seen that when the decline of start_bar is set at 0.04 (4%) and 0.02 (2%), the model performs relatively well. However, considering that the start_bar decline may increase and the maximum decline may decrease in the future, I will set the limit range of start_bar between 0.04 and 0.05, and set the Tp_point between 0.3 and 0.375. The above are the reasons why I chose these parameters, but it should be emphasized that because the number of failed samples is limited, this model is only my personal subjective judgment, please do not adopt it directly.
In addition, some readers may want to ask:
How to short the spot?
First, you need to go to the Binance official website to familiarize yourself with the Relevant Rules for Spot and Margin Lending, especially pay attention to the Risk Rate rules.
Here is a brief description of its main processes:
Transfer USDT to Isolated Margin or Cross Margin Margin Account
Lend the token
Just go to the isolated margin or cross margin margin account to sell
Repay when closing the position and buying it back
It is very simple to say, but there are many details and rules in the middle. What is the risk rate (maintenance rate) of the margin to be maintained? How to borrow it? How to pay it back? Will I not be able to borrow the borrowed currency? How much can I borrow? Is the liquidity sufficient when closing a position? These are all factors that need to be considered, but you can only slowly read the rules one by one. If leveraged borrowing is too troublesome, another method is to go to other exchanges to see if there are contracts, which is a simpler and more direct approach.
in conclusion
This article explores the impact of Binance’s delisting announcement on cryptocurrency prices and develops potential trading strategies based on this. The research method focuses on Binance’s spot trading delisting announcement and tracks the largest price drop with a seven-day observation period. The research process defined nouns, such as the first K bar (the current K line announced) and the second K bar (the hypothetical entry point), and calculated the drop of the first K bar and the maximum drop after entry. to explore the correlation between the two.
Key findings include:
1. When the first K bar drops significantly, the subsequent maximum drop space is relatively compressed.
2. After conducting correlation analysis on 43 samples, it was found that there was almost no correlation between the two, showing a weak correlation.
Based on the above research, a preliminary framework of trading strategy is proposed, considering setting start_bar limit and tp_point as parameter filtering conditions. The 3D model optimized for parameters shows that the strategy works better under specific start_bar and tp_point combinations.
Regarding how to conduct short selling operations in the spot market, he briefly introduced the process of Binance's leveraged borrowing and reminded that the risks and rules need to be understood. This strategy still needs more samples and verification.
Social media:
Twitter: @UncoverTrading will not take the initiative to send you private messages!!!
Twitter: @wwwp407 will not take the initiative to send you private messages!!!
TG telegram link: https://t.me/+L-zWOjhkyz5jMzk1
If you don’t have a Binance account yet, here are my referral links:
Binance：https://www.binance.com/zh-TC/activity/referral/offers/claim?ref=CPA_001AVXVMLR