Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Multi-Agents LLM Financial Trading Framework

github.com

Post reply on HN
Print view View original

51–60 of 90 posts

Re: Multi-Agents LLM Financial Trading Framework

#51

Post by fg137 »

XenophileJKO wrote: post #32
dakolli wrote: post #30

Earlier quoted context omitted.

Lmao, you llm people have some crazy delusions. You realize markets are zero sum, and if you're using a public model that everyone else also has access too, you llm psychos will destory eachothers "agentic" edge (not that there ever was one). Not to mention all the other obvious flaws with llms, lime having an effective memory of ~200k words and no ability to judge whats actually going on in the real world.

You really have no idea what you are talking about. I have been running an intermittent experiment with a multi agent "investment firm" for over a year now across model releases. They certainly can beat indexes, BUT.. the model families have some biases that you have to design around. The stop loss that bit the parent is certainly one. The models like to create rules. Often rules, one of those is making all kinds of…

"just trust me bro"

Re: Multi-Agents LLM Financial Trading Framework

#52

Post by jeanmichelselli »

This would make sense only if the LLMs would be constantly updated with a new data set and training phase every day. In that case, I'd could see this approach as having some sense. But, otherwise, these are just stochastic machines trained on static (outdated) data and I don't see how their predictions should be better than any other method around or even better than a human guessing.

Re: Multi-Agents LLM Financial Trading Framework

#53

Post by lordnacho »

looofooo0 wrote: post #12
hacker_9 wrote: post #8

Having worked in hedge funds for the last decade, this seems to miss the mark. Firstly we often reward skillstacking ie a technical person later becoming a trader. The more one person knows the better. These people are rare though hence the reason there is still many seperate job functions, so a person can specialize. But an AI agent? They all have the same brain, so why nerf them by specialising. Secondly, browsing…

Given what a know about the 2008 financial crisis, wouldn't an AI analysis in the years before that crisis of the real state funds helped to understand the risk of them better and avoid the big exposure.

I'd say it was more to do with how incentives lined up among different actors.

There's always risk somewhere in a financial system.

Re: Multi-Agents LLM Financial Trading Framework

#54

Post by harrouet »

whazor wrote: post #20
_pdp_ wrote: post #18

If this worked it wouldn't have been open source? Anyway, I have been running my own trading experiment and so far it has lost a bit of money. That being said I have not tried to optimise anything - just let it do whatever it wants. The losses are small and it might be able to recover later this year. Who knows. The agent writes a blog about its progress here https://trades.chatbotkit.space/ I am thinking to output a…

It would be more interesting to compare trading agents with index tracking ETFs. The better version of an ETF could maybe be a model where you zoom in on the companies and add/remove to your portfolio on the company related news, but keeping a broader portfolio. Maybe agentic trading still performs worse than ETFs. But alternatively, if it were meaningfully better then it would be okay to opensource, similarly how ET…

The issue at stake is that financial markets are order-2 chaotic system, i.e. acting on them can change their outcome.

Put simply, if you open source your magic recipe, the behavioral change will affect the prices and you recipe will not work anymore.

Re: Multi-Agents LLM Financial Trading Framework

#55

Post by chasd00 »

I vibe coded a little stock market sim game then i wrote an agent who’s job is to win the game. A couple friends and family members play the game too. If it works I’ll just follow along and hold the same portfolio the bot does. The bot, named stonker, makes its first trades in about an hour actually. Assuming it works I mean hah

https://stonks.jettdigital.app

edit: i just talked to my little sister ("Ginger" on the leaderboad) who is doing a good job beating the sp500. She said she just asks her ai what she should invest in and then executes those trades.

Re: Multi-Agents LLM Financial Trading Framework

#57

Post by csomar »

ppalata wrote: post #47
_pdp_ wrote: post #18

If this worked it wouldn't have been open source? Anyway, I have been running my own trading experiment and so far it has lost a bit of money. That being said I have not tried to optimise anything - just let it do whatever it wants. The losses are small and it might be able to recover later this year. Who knows. The agent writes a blog about its progress here https://trades.chatbotkit.space/ I am thinking to output a…

The "blog" says that it's just holding cash or am I misunderstanding something? Judging by the "buying power" number, you allow it to use 4x leverage against the cash it has? That seems dangerous unless you are fine with losing 380k+.

A lot of people don't realize that when a leveraged position goes against you (especially with a regulated broker), you get liquidated once your equity runs out, but you can still owe the shortfall on top of that. And the broker can come after your assets to collect it. So your real exposure here is 380K, not 98K.

That being said, I think the parent AI is using paper money. Though who knows, this is the brave new world of AI.

Re: Multi-Agents LLM Financial Trading Framework

#59

Post by mshafir »

Instead of going after money printing schemes like this (which I feel are unlikely to actually work and carry a ton of risk). I (thoughtfully) vibed up tooling and skills to just help LLMs manage my investments long-term. While not flashy, tax loss harvesting and direct indexing are areas where LLMs can reliably save you on fees that you'd otherwise have to pay. It's been very helpful, the LLM automatically deploys new cash based on my priorities, implements tax harvesting for me, and provides customized exposure guidance while keeping my goals in mind (ESG/sustainability focus, no FF rules) . Way better than any financial planner has ever done for me. https://github.com/mshafir/investbot
Post reply on HN