4 ms·

Spoilers: > Literally every single paper was either p-hacked, overfit, or a subsample of favourable data was selected (I guess ultimately they're all the same

Spoilers: > Literally every single paper was either p-hacked, overfit, or a subsample of favourable data was selected (I guess ultimately they're all the same thing but still) OR a few may have had a smidge of Alpha but as soon as you add transaction costs it all disappears. > I should caveat that I was a profitable trader at multiple Tier-1 US banks so I can say with confidence that I made a decent attempt of building whatever the author was trying to get at.

DennisP 7y ago
But also: > Almost every instrument is mean-reverting on short timelines and trending on longer timelines. i.e. he confirms the momentum factor, which isn't surprising since there's more solid evidence for it than anything else, going back hundreds of years. He doesn't say what fundamental factors he looked at, so it's possible that value, size, and profitability/quality would hold up as well. All those have been studied pretty extensively in academia, in papers going back decades. The author took only a fairly random sampling of recent papers.
- edoceo 7y ago
  All the finance folks I know use "fundamentals" to mean some class of attributes, that they all know what they are. Like when I say "GPL3" to someone who has a-priori knowledge. Not exactly sure but I think the "fundamentals" are top-line revenue, unit cost, unit margin, yoy-growth, EBTDA, cash-on-hand, default-alive - and likely the ratios those values produce.
  - DennisP 7y ago
    These sort of papers don't look at all the fundamentals at once. They'll isolate one factor, or a small related group, and see if it's predictive. Two simple factors that are well supported: on a risk-adjusted basis, stocks of small companies do better than big ones, and value stocks do better than growth stocks. ("Value" means the stock is cheap relative to some simple fundamental measure like the company's book value.)
  - MR4D 7y ago
    You’re pretty correct. A simpler way to think of it is any data that comes off the Income Statement, Balance Sheet, or Cash Flow Statement - or can be directly calculated using such data.
  - lootsauce 7y ago
    Fundamentals mean zilch for trading unless your'e talking earnings announcements or like news.
    - H8crilA 7y ago
      Please don't downvote this comment. There's a huge difference between short term trading and investing. And yes for short term trading (minutes/hours/few days) fundamentals matter rather litte - after all you can be a successful trader of Bitcoin, which has pretty much no fundamentals. Except on volatility events like macro data releases, earnings, FED decisions, surprise news (looking at you GE) etc.
      - Retric 7y ago
        > fundamentals matter rather litte Having a small impact is not the same a having zero impact. Traders operate over longer timeframe even if they are holding an individual stock for minutes there is a limited pool of stocks. Keep playing the game and longer term impacts add up.
        derefr 7y ago
        But do the fundamentals of an individual investment matter in the short term compared to the fundamentals of the market as a whole? I would suspect that the “longer term impacts add[ing] up” would just be market health. Day-trading randomly-picked stocks with random buys and sells is a Markov approximation of an index fund :)
        Retric 7y ago
        I think that depends on how stocks are chosen when you are day trading. Limiting things to high volatility stocks for example creates a bias in your index fund approximation.
  - derefr 7y ago
    I know as little about this as you do; my own gathering is that “fundamentals” are any signal you can get from a public company’s mandated quarterly reporting. So, every piece of data on the published balance sheet, plus maybe any reported board actions.
- bitL 7y ago
  Look up technical vs fundamental analysis. It's trading 101.
albertshin 7y ago
Discredits a whole field -- doesn't post any papers that he attempted to replicate.

DennisP 7y ago

But also: > Almost every instrument is mean-reverting on short timelines and trending on longer timelines. i.e. he confirms the momentum factor, which isn't surprising since there's more solid evidence for it than anything else, going back hundreds of years. He doesn't say what fundamental factors he looked at, so it's possible that value, size, and profitability/quality would hold up as well. All those have been studied pretty extensively in academia, in papers going back decades. The author took only a fairly random sampling of recent papers.

edoceo 7y ago
All the finance folks I know use "fundamentals" to mean some class of attributes, that they all know what they are. Like when I say "GPL3" to someone who has a-priori knowledge. Not exactly sure but I think the "fundamentals" are top-line revenue, unit cost, unit margin, yoy-growth, EBTDA, cash-on-hand, default-alive - and likely the ratios those values produce.
- DennisP 7y ago
  These sort of papers don't look at all the fundamentals at once. They'll isolate one factor, or a small related group, and see if it's predictive. Two simple factors that are well supported: on a risk-adjusted basis, stocks of small companies do better than big ones, and value stocks do better than growth stocks. ("Value" means the stock is cheap relative to some simple fundamental measure like the company's book value.)
- MR4D 7y ago
  You’re pretty correct. A simpler way to think of it is any data that comes off the Income Statement, Balance Sheet, or Cash Flow Statement - or can be directly calculated using such data.
- lootsauce 7y ago
  Fundamentals mean zilch for trading unless your'e talking earnings announcements or like news.
  - H8crilA 7y ago
    Please don't downvote this comment. There's a huge difference between short term trading and investing. And yes for short term trading (minutes/hours/few days) fundamentals matter rather litte - after all you can be a successful trader of Bitcoin, which has pretty much no fundamentals. Except on volatility events like macro data releases, earnings, FED decisions, surprise news (looking at you GE) etc.
    - Retric 7y ago
      > fundamentals matter rather litte Having a small impact is not the same a having zero impact. Traders operate over longer timeframe even if they are holding an individual stock for minutes there is a limited pool of stocks. Keep playing the game and longer term impacts add up.
      - derefr 7y ago
        But do the fundamentals of an individual investment matter in the short term compared to the fundamentals of the market as a whole? I would suspect that the “longer term impacts add[ing] up” would just be market health. Day-trading randomly-picked stocks with random buys and sells is a Markov approximation of an index fund :)
        Retric 7y ago
        I think that depends on how stocks are chosen when you are day trading. Limiting things to high volatility stocks for example creates a bias in your index fund approximation.
- derefr 7y ago
  I know as little about this as you do; my own gathering is that “fundamentals” are any signal you can get from a public company’s mandated quarterly reporting. So, every piece of data on the published balance sheet, plus maybe any reported board actions.
bitL 7y ago
Look up technical vs fundamental analysis. It's trading 101.

albertshin 7y ago

Discredits a whole field -- doesn't post any papers that he attempted to replicate.