Years from now, you'll feel like this is the most valuable article in your investment career, figuring out the source of the return and the source of the risk.

Many years later, the aunt who had grown up as a box of cabbage, remembering herself as a frog in her childhood, he forgot about the gold fork death, forgot about the boom and the mentality, even forgot about the woman who was with him that year, but he must have rarely remembered, having seen the fifth issue of the children's story Quantum.

……

He is also very concerned about the situation in the country, because what is being said in this issue makes him aware of the importance of figuring out the source of the income and the source of the risk.

In order to make it easier for people without a background in programming to benefit from monkey talk quantum mechanics, I'll try to avoid code as much as possible in my column, and to implement functions as much as possible with pseudo-code.

Everyone who invests has been subjected to a torture of the soul from a family friend.

What do you make your money on?

Or to expand on it:

You're not going to lose money.

Most people only come back when they have stumbled upon their own, and they turn their lips against each other.

In other words, what is your source of income?

Most people are faced with this question, very simply saying, "I buy at a low price, sell at a high price, it doesn't hurt?"

Oh, you say you can just feel it? Blind. I give you a K-line, you tell me the high is the low.

Are you saying it's high or low now?

Oh, you said you wanted to see the brown line?

You say you've broken through the pressure point, the low point is undoubted?

I've been thinking about this for a long time, but I'm not sure I've ever thought about it before.

You say you don't count, let me repeat?Come on, two breaches of the pressure point, and you're saying you're going to fall?

After learning the lesson, this author is sure to tease me. So you say, fall, you fall:

I'm sorry, I was just going upstairs.

So, what do you say, is it high or low?

I don't dare to say it. Next is a waterfall.

This is the point of SB waveform theory. The current waveform does not conform, he would say this is only temporary, the next waveform definitely conforms, the next one does not conform, he would say the next one definitely conforms. It's like having a boy or a girl, and if you guess wrong, he says your next baby is definitely a boy/girl.

Like taking a drug, all drugs are marketed on the assumption that it is ineffective, it must prove that it is effective before it can be marketed. (Except for the drug, the drug is ineffective unless it kills a bunch of people, and can't be suppressed when it is reported, otherwise the death can only say bad things, can't say that the drug is toxic.

You have to assume that your current idea is wrong, and prove it works by using logical fallacies or statistical fallacies.

You can say he works.

It's only through proven sources of income that you can be a credible source of income.

So, the next question is very simple. How do you prove it works with a logical puzzle?

It's very simple, for example I know an exchange, his raw_Kline_info generation, is generated through price and depth information of binance, tokens, etc. So it must have a lag relative to his peer exchange, right?

So as soon as I can find his peer exchange and fit the calculations to his K-line price, then I can get the information from his next K before he draws K. Through this price information, we know the price trend information for a given exchange within a few hundred milliseconds.

See here, the impatient Cinnamon is happy to open up and ready to write. The patient Cinnamon is still taking notes.

I'm not going to go into detail about what you're doing here, because you can't get any profit from it.

For a few hundred milliseconds, the price of the exchange varies by about 0.5 parts of a million. And at this point, you also have to consider whether you are a taker or a maker, because logically, the taker, that is, eating other people's pending orders, is obviously faster and more stable to catch the trend. But the taker's transaction fees at the exchange are often much higher than the usual.

If you can't find the problem? Even if you can determine the trend, you can't necessarily make money. Then your source of income here, in addition to the fact that you can determine the trend, is an additional information bar that you have previously ignored and can catch the trend.

In addition to the above, there are also two ways to capture trends: 1⁄2, can grab the ticket. 2, the profit from the falling trend can cover the cost of ordering (procedural fees).

These add up to your source of profit and risk.

I'm going to read it again: 1, find an exchange D for the price of A, B, C 2 Algorithms that match the D exchange to the other exchanges to draw K 3⁄4 Determine the short-term trend of a trading pair on exchange D based on the algorithm used 4, change positions by hanging or eating trends according to well-determined trends 5 Guarantee that the cost of changing your position is less than the benefits of your trends.

Okay, these five points make up the source of revenue and the source of risk for your strategy. Next, lay out the first task to extract the source of revenue and the source of risk from these five points.

What happened after the extraction?

Answer your own two questions.

First, can your assumptions about profitability be proven logically or statistically? 2 - Can you solve your risk sources?

Whenever you look at your trading system, first torture yourself with these two souls, and then you can honestly answer why you make money and why you don't.

And if you don't like homework enough, please talk about quantification in class.Quantitatively speakingLee analyzes the source of the benefits and the source of the risks of each teaching strategy.

If you see a student here who might ask: "So you just said logical proof, statistical proof?" If it's a digital currency, use FMZ to measure:https://www.fmz.cn/sign-up/1974419I've been trying to get a hold of you for a while, but I'm not sure if I'll be able to get a hold of you. In statistical terms, it is generally a re-test of data over several different cycles, and then the re-test proves to be valid, then the simulation disk, the real disk, actually runs for a while, that is, the practice is the only criterion for the truth of the statistical method.

Strategic pseudo-code (here suppose you have seen the children's talk about quantification, written the middle layer, here only do the logic layer pseudo-code, do not fill the content, tolerate errors, optimize, perfect the strategy you have to do it yourself;;):

'''
class high_freq():
    def __init__(self,mid_class):
    '''
    这个用来初始化各项数据，自己根据需要做
    '''
        pass
        
    def refreash_data(self):
    '''
    这个用来刷新行情，深度，账户信息
    '''
        pass
        
    def refreash_target_data(self):
    '''
    这个用来刷新对标用交易所数据
    '''
        pass
        
    def make_price_condition(self):
    '''
    这个用来处理价格信息
    '''
        pass
        
    def make_amount_condition(self):
    '''
    这个用来处理数量信息
    '''
        pass
        
    def make_deal_condition(self):
    '''
    根据价格信息和数量信息，给出判断交易条件，是做bids，还是asks，还是等待
    '''
        pass
    
    def make_trade_dict(self):
    '''
    根据交易条件和深度，生成需要交易的订单簿
    '''
        pass
        
    def do_trade_and_cancel(self):
    '''
    根据订单簿信息，取消老单，范围过远单，并且填充新的挂单价格，并挂单
    '''
        pass
        
    def check_deal(self):
    '''
    检查挂单情况，是否成功挂单，是否有网络问题遗漏单，仓位风险
    '''
        pass

    def lower_risk(self):
    '''
    根据自己的设置，降低仓位风险。比如倾向于持币，则平时多进微小买单
    倾向于空仓，则平时多卖出微小单。
    这个很容易理解吧，上行期设置倾向持币，下行期设置倾向空仓
    '''
        pass
    
    def trade_controller(self):
    '''
    处理和交易相关的逻辑,整合到一个函数里
    '''
        pass
        
    def clear_info_controller(self):
    '''
    处理和清理线程相关的逻辑，整合到一起
    '''
        pass
    
    def target_controller(self, target_class):
    '''
    处理和对标用交易所信息相关的逻辑，整合到一起
    '''
        pass
    
def main():
    raw_base_class = mid_class(exchanges[0])
    base_class = high_freq(base_class)
    
    raw_target_class =  mid_class(exchanges[1])
    target_class =  high_freq(target_class)
    
    While True:
        Sleep(100)
        
        base_class.refreash_data()
        target_class.refreash_target_data()
        
        base_class.target_controller(target_class)
        base_class.clear_info_controller()
        base_class.trade_controller()

So, that's the end of this article. I'm going to share in the next issue how to ignore the small and medium-sized exchanges that give you a zero-fee trial period so that you can actually manipulate the strategy of this example. I didn't share the code here, but this market-making strategy I'm currently running out of myself, depending on the exchange and the trade, about 1 to 5 thousandth of a profit per day.