sharetrader
Results 1 to 7 of 7

Thread: Neural Networks

  1. #1
    Junior Member
    Join Date
    Apr 2017
    Posts
    10

    Default Neural Networks

    I have been reading about neural networks lately namely how they can uncover hidden relationships in price data. Just wandering has anyone dabbled with them? Is it a waste of time? Should I just stick to traditional technical analysis and fundamentals?

  2. #2
    Senior Member
    Join Date
    Nov 2018
    Location
    Christchurch
    Posts
    1,063

    Default

    I recall the study of neural networks having a lot of quantitative math during my finance studies. They prove nothing. TA and fundamental analysis again only goes so far. If you subscribe to the likes of Buffet and Munger, their secret doesn't imply any of those analysis but rather, a more 'common sense' approach.

  3. #3
    IMO
    Join Date
    Aug 2010
    Location
    Floating Anchor Shoals
    Posts
    9,696

    Default

    Common sense approach being ?

  4. #4
    Reincarnated Panthera Snow Leopard's Avatar
    Join Date
    Jul 2004
    Location
    Private Universe
    Posts
    5,853

    Default

    I have been there and done that and now have a T-shirt that reads:

    "I tried Neural Networks and all I got was a Headache".
    om mani peme hum

  5. #5
    Senior Member
    Join Date
    Nov 2018
    Location
    Christchurch
    Posts
    1,063

    Default

    Quote Originally Posted by Joshuatree View Post
    Common sense approach being ?
    The common sense approach seeing the behaviour of consumers and how corporate responds to their demand.

    For several decades, Wall Street has employed the top mathematicians, the brightest graduates from Harvard, Yale, Stanford, etc in the area of Finance and Economics... to prove that they could do better investing than the 'market return'. They come out with all sorts of ideas like TA, your neural networks, fundamental analysis, and they crunch down the numbers. You know what? They ALL failed. So then they go out and try it again.. and they failed. You've got to wonder why there is such a huge industry of stock pickers (ie. actively managed funds) and they do worse than the market index return on a "CUMULATIVE BASIS" say over 10 years ????

  6. #6
    Junior Member
    Join Date
    Apr 2017
    Posts
    10

    Default

    I tend to agree that if neural networks gave you an edge hedge funds would be using them. So how do hedge funds make there trading decisions? Do you they use technical analysis or fundamental or a combination of both?

  7. #7
    Senior Member
    Join Date
    Nov 2018
    Location
    Christchurch
    Posts
    1,063

    Default

    Fund managers make all sorts of trades based on countless of factors. The fact is pretty much all of them can't do better than simply throwing darts on the board (where the board would have the names of the stocks they would want to invest in). It's been done in the past where top scholars in finance have done no better than monkeys choosing which stocks to buy. We're talking billions spent by Wall Street trying to 'consistently' beat the market. It just doesn't work. Yet we have a whole industry that supports investing into 'actively managed funds'.

    In the area of Finance, we've been taught that in order for TA to be relevant, we need to have a "strong form" market hypothesis (where markets are extremely efficient). That has never been the case because not all information is available (ie. we still have insider trading). As for fundamental #s, that only goes so far as certain variables can't be quantified in #s. ie Such as a decision by a high profile investor pulling out and the stock collapses, or geopolitical decisions which no math forecasting formula can ever predict.

Tags for this Thread

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •