When FMZ encounters ChatGPT, remember an attempt to use AI to assist in learning quantitative transactions

When FMZ encounters ChatGPT, remember an attempt to use AI to assist in learning quantitative transactions

ChatGPT

Recently, ChatGPT has become very popular, and there are various applications on the web for video in real-life scenarios. A brief introduction of what ChatGPT is, ChatGPT is a language model developed by OpenAI. Unlike previous artificial intelligence, ChatGPT has a very powerful language understanding ability.

When FMZ encounters Chat GPT

In the field of quantitative trading learning, beginners have always been very difficult to get started with. Since most beginners are zero base, they will find it very difficult to learn because they lack both grammatical knowledge and logical thinking.

OK, the next article is not a short story I wrote, but we'll show you a little appetizer.

part1

When FMZ encounters ChatGPT, remember one attempt to use AI to assist in learning quantitative transactions.

As a beginner, I was looking for effective ways to learn quantitative trading. After searching online, I found FMZ and ChatGPT. So I decided to try using them to deepen my understanding of quantitative trading.

First, I learned on FMZ how to write trading strategies using PINE scripts. On the platform, I found templates for various strategies and tested their effectiveness by adjusting parameters. In addition, I learned how to evaluate the returns of strategies using feedback functions.

Next, I used ChatGPT to get more help. On this platform, I found an AI assistant called Assistant, which was able to answer all my questions. By interacting with Assistant, I learned a lot of useful knowledge and got a lot of valuable advice.

In the process of using FMZ and ChatGPT, I have reaped a lot. Not only have I learned how to write a trading strategy, but I have also gained more knowledge about quantifying trading. I believe that as I continue to learn and practice, I will be able to improve my trading skills.

part2

I will be able to further improve my trading skills and get better returns in the future.

While FMZ and ChatGPT can help a lot, I also realize that quantifying transactions is a continuous learning process. I still have a lot to learn, and I will continue to work hard and improve myself.

Finally, I hope that my attempt will be helpful to other beginners. If you also want to learn how to quantify trading, why not try using FMZ and ChatGPT, two platforms that may be a great help in your learning.

part3

Of course, quantifying transactions is not an easy task and requires constant learning and practice in order to improve skills. Therefore, I also recommend that you participate in various training courses and community activities, exchange trading experiences with professionals, and always remain curious and curious.

In addition to this, it is very important to pay attention to risk management. There are many uncertainties in quantitative trading, so we must set up a reasonable stop-loss strategy and carefully choose the trade varieties and trading cycles.

part4

In summary, I believe that using FMZ and ChatGPT is the best way to learn.

Learning quantitative trading is a good option. They provide us with a wealth of learning resources and help to make it easy for us to master knowledge and practice. However, we must also remember that learning quantitative trading is an ongoing process and we need to keep working to be more successful in this area.

Ok, I started writing again, can you believe it! the above part 1 - part 4 was written by ChatGPT, let me give you some proof:

We're back on topic again.

Let the AI help me write a strategy.

If I were an entry-level learner of PINE Language 0 on the FMZ platform, I could make a simple request to ChatGPT to see if the AI could help me write code:

The demands:

Can you help me write a binary equator strategy using the PINE script? The equator indicator for different cycles: gold forks open multiple, dead forks open empty, 100 stop losses. And give me the complete PINE script strategy. Use the version control code://@version=4 to start.

// @version=4
strategy("双均线策略", overlay=true)

longTerm = input(title="长期均线周期", type=input.integer, defval=200)
shortTerm = input(title="短期均线周期", type=input.integer, defval=50)
stopLoss = input(title="止损点", type=input.integer, defval=100)

longMA = sma(close, longTerm)
shortMA = sma(close, shortTerm)

if (shortMA > longMA)
    strategy.entry("做多", strategy.long)
if (shortMA < longMA)
    strategy.entry("做空", strategy.short)

strategy.exit("止损", "做多", stop=strategy.position_avg_price - stopLoss)
strategy.exit("止损", "做空", stop=strategy.position_avg_price + stopLoss)

It's really written, so let me run it through the FMZ platform's feedback system!

The PINE script is up!

Let's try using ChatGPT to help us explain the strategy so that we can also learn how to quantify transactions.

The question is:

strategy("test", overlay = true)

varip a = na
varip highPrice = na
varip isTrade = false 
varip offset = 30

if not barstate.ishistory and not isTrade
    strategy.entry("test 1", strategy.long, 1)
    strategy.exit("exit 1", "test 1", 1, trail_price=close+offset, trail_offset=offset)
    a := close + offset
    runtime.log("每点价格为：", syminfo.mintick, "，当前close:", close)
    isTrade := true 

if close > a and not barstate.ishistory
    highPrice := na(highPrice) ? close : highPrice
    highPrice := close > highPrice ? close : highPrice

plot(a, "trail_price 触发线")    
plot(strategy.position_size>0 ? highPrice : na, "当前最高价")
plot(strategy.position_size>0 ? highPrice-syminfo.mintick*offset : na, "移动止损触发线")

可以解释一下这个脚本的详细逻辑吗？

The screenshot:

I have a little doubt that there is a quantitative trading researcher behind this page answering my questions...

The imperfections

Of course, ChatGPT is not particularly perfect at the moment, for example, when communicating with it in Chinese, it speaks frequently and needs to be interrupted to continue the conversation. For some problems that it does not particularly understand, it is going to be a bit confused and you will be blindsided, we need to combine the search engine to determine the answers given, and can not believe it. However, ChatGPT is still very helpful for learning quantitative trading on FMZ, it can help us understand the hard code, at least it can give us some information, let us know what to do.

I hope that in the future, there will be no more "start to quit" in quantitative trading learning!