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

Financial market applications of LLMs

thegradient.pub

Post reply on HN
Print view View original

51–60 of 116 posts

Re: Financial market applications of LLMs

#51

Post by Onavo »

nostrademons wrote: post #32
btbuildem wrote: post #30

There were some developments using LLMs in the timeseries domain which caught my attention. I toyed with the Chronos forecasting toolkit [1], and the results were predictably off by wild margins [2] What really caught my eye though was the "feel" of the predicted timeseries -- this is the first time I've seen synthetic timeseries that look like the real thing. Stock charts have a certain quality to them, once you've…

I used to work in financial software, and when writing the charting UIs, I'd wire them up to a randomwalk to generate fake time series data. It was a relatively common occurrence for a VP or the company CEO to walk by, look at my screen, and say "What stock is that? Looks interesting." Unpopular opinion backed up by experience: a randomwalk is the most effective model for generating timeseries that have the "feel" of…

> Unpopular opinion backed up by experience: a randomwalk is the most effective model for generating timeseries that have the "feel" of real stock charts.

That's not an unpopular opinion. The BSM model is based on the assumption that stock prices are stochastic i.e. random walks. Monte Carlo simulations and binomial trees are the two common methods of deriving a solution to the BSM model.

Re: Financial market applications of LLMs

#52

Post by Bostonian »

nostrademons wrote: post #32
btbuildem wrote: post #30

There were some developments using LLMs in the timeseries domain which caught my attention. I toyed with the Chronos forecasting toolkit [1], and the results were predictably off by wild margins [2] What really caught my eye though was the "feel" of the predicted timeseries -- this is the first time I've seen synthetic timeseries that look like the real thing. Stock charts have a certain quality to them, once you've…

I used to work in financial software, and when writing the charting UIs, I'd wire them up to a randomwalk to generate fake time series data. It was a relatively common occurrence for a VP or the company CEO to walk by, look at my screen, and say "What stock is that? Looks interesting." Unpopular opinion backed up by experience: a randomwalk is the most effective model for generating timeseries that have the "feel" of…

Since volatility clustering does exist in returns, a GARCH model should produced more realistic-looking returns than a pure random walk.

Re: Financial market applications of LLMs

#53

Post by Rexxar »

bigyikes wrote: post #3
jsemrau wrote: post #2

A lot of words for not bringing much new content to the discussion. I think the most interesting application of LLMs in Finance are (1) synthetic data models for data cleansing, (2) journal management, (3) anomaly tracking, (4) critiquing investments All of this should be done by professionals and nothing is "retail" ready.

> All of this should be done by professionals and nothing is "retail" ready. Don’t worry, just train the LLM to always append “This is not financial advice.” to their responses. Boom, retail ready.

Or just append the string to output without asking the LLM to do it :-).

Re: Financial market applications of LLMs

#54

Post by Kon-Peki »

cpursley wrote: post #36
conorh wrote: post #26

We are working on a project for a client which functions as an analysis tool for stocks using LLMs. Ingesting 10ks, presentations, news, etc. and doing comparative analysis and other reports. It works great, but one of the things we have learned (and it makes sense) is that traceability of the information for financial professionals is very important - where did the facts and information come from in what the AI is p…

I assume you're ingesting PDFs. If so, how are you handling tables accurately?

If it was me, I would be ingesting the raw filings from SEC EDGAR and using the robust xml documentation to create very accurately annotated data tables that would be fed to my LLM

Re: Financial market applications of LLMs

#55

Post by neodypsis »

conorh wrote: post #26

We are working on a project for a client which functions as an analysis tool for stocks using LLMs. Ingesting 10ks, presentations, news, etc. and doing comparative analysis and other reports. It works great, but one of the things we have learned (and it makes sense) is that traceability of the information for financial professionals is very important - where did the facts and information come from in what the AI is p…

Could something like that proposed in "Training Language Models to Generate Text with Citations via Fine-grained Rewards" [0] work for you?

0. https://arxiv.org/abs/2402.04315

Re: Financial market applications of LLMs

#56

Post by __MatrixMan__ »

fragmede wrote: post #45
__MatrixMan__ wrote: post #14

Earlier quoted context omitted.

That should be the goal, right? Good ideas get the funding they need as if by magic, yet nobody is sitting on the sidelines collecting rent. The best thing that AI can do for finance is eliminate it.

that sounds really bad for everyone collecting rent

Not really, just do something else instead. It'll only actually happen if the AI-begotten efficiencies are real, and in that scenario there will more to go around re: supporting people whose current expertise is no longer relevant.

Re: Financial market applications of LLMs

#57

Post by mhh__ »

wuj wrote: post #39

HFTs exploit price inefficiencies that last only milliseconds. The time-series data mentioned in the article is on the scale of seconds. I wonder if its possible to get the time-series data on the scale of milliseconds, and how that would affect the training of the objective function in a LLM.

The data is reasonably easily acquired, for a price...

Re: Financial market applications of LLMs

#58

Post by mhh__ »

multicast wrote: post #50
wuj wrote: post #39

HFTs exploit price inefficiencies that last only milliseconds. The time-series data mentioned in the article is on the scale of seconds. I wonder if its possible to get the time-series data on the scale of milliseconds, and how that would affect the training of the objective function in a LLM.

Todays derivatives and their pricing are based on the premise that stock prices can not be predicted and behave like a Brownian motion system. If you take real time data from any stock and calculate in order how many times a stock went up in a row or down in a row you end up almost perfectly with a natural probability distribution. HFT's are involved in market making and arbitrage both of which already involves high…

Derivatives are priced under those assumptions because the aim is to calculate exposure/risk (where simple / assume you're wrong is desirable), the pricing is sort of an afterthought most of the time.

Re: Financial market applications of LLMs

#59

Post by throwaway14356 »

fragmede wrote: post #45
__MatrixMan__ wrote: post #14

Earlier quoted context omitted.

That should be the goal, right? Good ideas get the funding they need as if by magic, yet nobody is sitting on the sidelines collecting rent. The best thing that AI can do for finance is eliminate it.

that sounds really bad for everyone collecting rent

they can get jobs
Post reply on HN