AI Generated Quant Models

แชร์
ฝัง
  • เผยแพร่เมื่อ 14 พ.ย. 2024

ความคิดเห็น • 3

  • @xclordon_4696
    @xclordon_4696 12 วันที่ผ่านมา +1

    could you actually use this code to make money rn though

    • @CA219th
      @CA219th 12 วันที่ผ่านมา

      NO! I don't mean to be a jerk, but if it was actually this easy to make money, the quant hedge funds would've been out of business because everyone would be doing this. There's a clown on TH-cam called moondev or some bs that created a script that back-tests a million algorithms to "find one that works." This approach of creating so many algorithms WILL eventually yield one that looks like a money printer, however (!!!!): this is called overfitting. Models that show amazing-looking returns in their back-tests are most likely overfitted because they are trained on past data (meaning, among other things, they are trained on a fixed price behavior and number space, this is unrealistic because markets are dynamic as f*ck -> you need complementary models to run along your trading model to conduct sanity checks). Such "amazing" models almost always underperform when implemented because there is no understanding of the dynamics of the assets, market, or the model. A model is absolutely useless if you don't understand how the model uses mathematics or market structure to generate profits, so when it underperforms, you have no real metric to determine if the model is failing or succeeding (returns is not really a metric in this regard, if profits are highly negative, most naive 'quant' traders will liquidate their positions, so if they naively implemented a working algo, the algo never has a chance to shine.
      The real money-making algorithms and models are not public for a reason. It is possible to create something that generates profits, but it isn't as simple as asking AI because it'll just give you what you're asking for and it pulls information from the internet that others also have access to, meaning that many others are trying to do the same thing too, which in turn means decreased capacity for whatever model you're implementing. Please don't be fooled by these youtubers who are trying to make their money on views. Learn some math, learn some financial economics and some Python, and explore the field for yourself with your own original ideas, this is called Quantitative Research (think Point72, Citadel, Jane Street).

    • @jimmyhernandezzzzzz
      @jimmyhernandezzzzzz 9 วันที่ผ่านมา

      no