DeepTrade
Have fun building a customizable trading system that lets YOU set the return/risk ratio. Tailor your strategy, maximize gains.
Errungenschaften
Team
Prototyp
Nutzer
Partner
Umsätze
Investoren
Unser Angebot
A working stock trading system based on deep learning that beats Nasdaq.
Zielgruppe
Focus for now is not selling a product but investing using a trading company. One reason is that the market for financial products is strongly regulated.
Herausforderungen
Combining deep learning with deep reinforcement learning to increase returns and reduce risk.
Unsere Story
A beginners misconception of algorithmic trading is that you can predict stock prices and then just invest and get rich. But in reality there are only very few patterns in historic data you can learn. Having experimented with algorithmic trading for quite some time without success (success = better than some index) I got the breakthrough idea while reading Peter Thiel’s “Zero to One”.
I won’t disclose it here but it is a simple and beautiful concept on how to approach investing, which turned out can beat the market. It also allows you to trade risk against return, which is a default requirement in trading. With such a concept, the challenge was to build a system that takes it to a trading strategy and validate it with limited data (there is just about 30yr of stock prices with lower quality the further you go back in time).
The current challenge is to improve the performance by experimenting with all kinds of machine learning techniques. As a one percent improvement in yearly returns will result in a large compounding effect after 10 years, it is super exciting.
Algorithmic trading does not have much purpose embedded. For me it is a way to scale investments to eventually follow up with other topics as well. Those could be having easy access to trading for everyone or completely unrelated ideas.
Currently, I am working on that by myself with limited time. I am looking for one, at most two, other people to join.
You might ask why. There are two reasons, first my primary goal is to work on fun and challenging topics together with others and secondly, it makes progress faster. And I think small teams are the best.
If that resonates with you and you think you could implement a deep learning trading algorithm given the high level idea, would be great to connect. As mentioned, this is not only about solving technical problems but also about having a great time.
I won’t disclose it here but it is a simple and beautiful concept on how to approach investing, which turned out can beat the market. It also allows you to trade risk against return, which is a default requirement in trading. With such a concept, the challenge was to build a system that takes it to a trading strategy and validate it with limited data (there is just about 30yr of stock prices with lower quality the further you go back in time).
The current challenge is to improve the performance by experimenting with all kinds of machine learning techniques. As a one percent improvement in yearly returns will result in a large compounding effect after 10 years, it is super exciting.
Algorithmic trading does not have much purpose embedded. For me it is a way to scale investments to eventually follow up with other topics as well. Those could be having easy access to trading for everyone or completely unrelated ideas.
Currently, I am working on that by myself with limited time. I am looking for one, at most two, other people to join.
You might ask why. There are two reasons, first my primary goal is to work on fun and challenging topics together with others and secondly, it makes progress faster. And I think small teams are the best.
If that resonates with you and you think you could implement a deep learning trading algorithm given the high level idea, would be great to connect. As mentioned, this is not only about solving technical problems but also about having a great time.
Team
Hauptsitz
Weitere Startups
Diese Gründungsprojekte könnten auch interessant für Dich sein!
