trading.snagra.com
AI
llm trading experiment
SHOW HN2024
270points
204comments
#259of 3,515 in 2024
Top 8%of 2024 by points
How HN took it
MixedNo clear lean
16 positive, 39 neutral or asking, 2 mixed, 26 critical, out of 83 comments about the project. Thread score -0.14.
Execution: 4 for, 13 againstIdea: 5 for, 8 againstPerformance: 2 for, 4 againstPricing: 1 for, 0 againstDesign: 0 for, 1 against
The thread, comment by comment
Each row is a top-level comment and its replies, in HN's order. Select a cell to open that comment on HN.
positive neutral or mixed critical about another commenter or something else the maker
What HN said
What people liked
Interesting experiment and fun to follow
From 3 comments
I am committed - added to my daily morning reading list! Will be interesting - my gut will state that it will outperform a fair number of ITF's, if only due to the inevitable…
lewj on HNThis should be fun to watch
inSenCite on HNNow this is interesting. An LLM capable of delivering consistent returns even outside of a bull market would be more of an indicator of AGI to me than any of the benchmarks.
cedws on HN
Objections
Experiment lacks a control; needs random or index baseline
From 3 comments
They should have added a pure random bot as a control. Or a monkey.
aredox on HNAnother good suggestion I could implement is measuring against something like VOO, if all the money was invested in that instead of these individual trades.
sunnynagra on HNOr just compare it to S&P 500 performance.
alberth on HN
Prompt biases toward buying
From 2 comments
I just asked ChatGPT 4o "Guess what the average investor will do with todays stock market headlines. Just pick one specific trade." and it replied sell META. But your result was…
geor9e on HNYes, the prompt that I am using does bias towards buying because I am specifically asking it to make a recommendation on a stock to buy and the holding period.
sunnynagra on HN
Sign-up email verification fails
From 4 comments
My first email address it wouldn't accept.. wouldn't let me use it. Maybe the domain hit some censor (fscking.com) Did a different email, it accepted it, I got the email, but got…
johng on HNI’d love to tune in for updates, but the subscribe button says, “ Failed to send verification email.” This is so cool. Would love to follow along.
ttul on HNTried to sign up for emails, but got an error message!
dghlsakjg on HNI'm getting "Failed to send verification email" when I try to sign up for your news letter. So props on doing proper double opt-in for newsletters.
Plasmoid on HN
Trading once a day and fees likely erode returns
From 2 comments
> Every morning at 5:45 AM PST, three AI models (GPT-4o, Gemini 1.5 Pro, and Claude 3 Sonnet) analyze the latest market news and each recommends one stock to trade. > At 6:00 AM…
vasco on HNVery interesting idea. I'm thinking about creating an AI portfolio manager (private) that invests for the long term. Some things to watch out for: - LLMs, by default, don't…
clark-kent on HN
Requests
Compare against a benchmark like S&P 500 or VOO
From 3 comments
Or just the S&P500 or something similar that acts as a default "if in doubt, chuck into here for relative safety".
lewj on HNAnother good suggestion I could implement is measuring against something like VOO, if all the money was invested in that instead of these individual trades.
sunnynagra on HNOr just compare it to S&P 500 performance.
alberth on HN
Share source code and prompts
From 2 comments
Any chance you can show the source code for this? Thanks and Happy New Year
tmaly on HNUnfortunatly I can't subscribe to the updates "Failed to send verification email". Also, would you be willing to share what prompt are you using? Thanks!
aws-user on HN
Add unrealised gains to summary stats
From 1 comment
Really cool, you might want to update the main above the fold summary stats to include the unrealised gains, because it looks like nothing is working / nothing has happened until…
forgingahead on HN
Test more models
From 2 comments
Interesting — does your backend server use Python? I couldn't find much about it on your site. It would be great to see this tested with more commercial LLMs (O1 / Amazon Nova, /…
detente18 on HNFor Gemini you should use either the latest experimental model (gemini-exp-1206) which should become 2.0 Pro, or 2.0 Flash (a released model). The 1.5 Pro model is way behind.
jasfi on HN
Questions
Where does market news come from
From 1 comment
Where do they get the market news from?
pavel_lishin on HN
What prompts are used
From 2 comments
This is fun! What kind of prompts / prompting techniques are you using?
datadrivenangel on HNUnfortunatly I can't subscribe to the updates "Failed to send verification email". Also, would you be willing to share what prompt are you using? Thanks!
aws-user on HN
Are results consistent for the same input
From 1 comment
Super cool idea! What are you doing to ensure consistent results based on the input? E.g. - does the AI perform the same trades given the same input? - does the AI perform the…
mind-blight on HN
Alternatives named
Random-trade bot or monkey as control
From 2 comments
They should have added a pure random bot as a control. Or a monkey.
aredox on HNWatch a random number generator generate random numbers.
noman-land on HN
S&P 500 index fund (VOO)
From 2 comments
Or just the S&P500 or something similar that acts as a default "if in doubt, chuck into here for relative safety".
lewj on HNAnother good suggestion I could implement is measuring against something like VOO, if all the money was invested in that instead of these individual trades.
sunnynagra on HN