• Sharebar

For a while now I have been adamant about perusing a dual degree in statistics and economics, but after reading "The Black Swan: The Impact of the Highly Improbable," I've become flustered and lost in what to do. The arguments Taleb made about the Gaussian curve and using the hard sciences in the social science seemed logical and made me question the whole academic field of statistics and economics. I don't know, I guess it's a good thing that I'm not having this crisis when I'm actually working.

I do love finance and do love following current events, and I want to do something in finance after college. However, in college I wanted to major in something, at least slightly, technical because I know that I would never have that opportunity to do so again (or at least for a long time). That's why I wanted to major in statistics. But after reading the Black Swan, I think it might actually be detrimental if I do so.

Do you guys have any suggestions?

The WSO Advantage - Trading

Prop Trading Interview Guide

Video Library - 1 Year ($400 value). Learn More.

S&T Interview Prep Pack

30,000+ sold & REAL questions. Learn More.

Resume Help from Trading Pros.

Land More Trading Interviews. Learn More.

Find Your Trading Mentor Here

Realistic Trading Mock Interviews. Learn More.

Comments (22)

  • feenans's picture

    Taleb writes about very general concepts. In practice, statistics skills are invaluable. For example, even though asset prices are not normally distributed, you will still use black scholes as a pricing convention.

  • econ's picture

    If you find stats interesting, I'd say still go for it. If nothing else, it might help you break into finance to begin with.

  • chewingum's picture

    I'd really like to see this topic discussed in further detail. I'm in pretty much exactly the same position as the OP: came into college thinking quant shit was the be all end all, and only realized later that all the complicated math and stat doesn't really work. Right now, I'm wondering why I should even continue majoring in this crap when I can do something chill like Econ, rock the GPA and land a job. Then again, I might just be a pussy because stats is a pain in the ass to learn on your own when you have shitty teachers.

    (Might want to move this to the trading forum to get more relevant responses)

  • cplpayne's picture

    Don't let Taleb determine what you major in, but the one thing you should take away from his books is as he would say "don't be a sucker."

    "One should recognize reality even when one doesn't like it, indeed, especially when one doesn't like it." - Charlie Munger

  • GutShot's picture

    I understand your concern, having gone through the same mental debate.

    In the end, it's a debate that can be answered through empiric evidence. Look at HFT shops, look at Renaissance or DE Shaw. Their strats are quant based and work.

    That said, they can lose money too, and are hardly infallible. The value guys are rockstars too. Einhorn and the like. It's up to you what style you like most.

The WSO Advantage - Trading

Prop Trading Interview Guide

Video Library - 1 Year ($400 value). Learn More.

S&T Interview Prep Pack

30,000+ sold & REAL questions. Learn More.

Resume Help from Trading Pros.

Land More Trading Interviews. Learn More.

Find Your Trading Mentor Here

Realistic Trading Mock Interviews. Learn More.

  • LTV's picture

    I learnt stats as a part of physics and electrical engineering, and we never used stats to quantify human behavior or construct term structure models. For example, the normal distribution (that you say you've lost faith in) wasn't first developed to explain financial events, but rather to explain the movement of gas particles. The vast majority of quantitative models are actually applied physics equations massaged to explain financial instruments and markets (read Black-Scholes and Heat equation). Ito calculus is rocket science (quite literally). Have you ever wondered why the vast majority of quants from the 80s and 90s were physicists?

    To come to the point, financial math is not an exact science, while electrical engineering and physics are. The development of math in any field starts with some basic assumptions (like Newton's laws of motion or Maxwell's equations), and are built on that. These laws have stood the test of time to the extent that they're "exact" now. The assumptions of Financial Math, however, were formed relatively recently and were borrowed from physics and were calibrated to fit the markets. It's not "exact" yet, and hence the reason the math can't explain certain events.

  • In reply to LTV
    econ's picture

    LTV:
    The assumptions of Financial Math, however, were formed relatively recently and were borrowed from physics and were calibrated to fit the markets. It's not "exact" yet, and hence the reason the math can't explain certain events.

    I suspect it never will be exact. For a great coverage of these topics, see my post in the Econ Group (or, just read Richard Feynman's graduation speech "Cargo Cult Science" and F.A. Hayek's "The Pretense of Knowledge").

  • In reply to econ
    econ's picture

    econ:
    LTV:
    The assumptions of Financial Math, however, were formed relatively recently and were borrowed from physics and were calibrated to fit the markets. It's not "exact" yet, and hence the reason the math can't explain certain events.

    I suspect it never will be exact. For a great coverage of these topics, see my post in the Econ Group (or, just read Richard Feynman's graduation speech "Cargo Cult Science" and F.A. Hayek's "The Pretense of Knowledge").

    LOL -- that sounds really cocky and I didn't mean for it to. I meant for a great coverage of those topics via the links in my thread... Also, go on youtube and watch "Quants: The Alchemists of Finance" (or something like that).

  • In reply to LTV
    GekkotheGreat's picture

    LTV:
    I learnt stats as a part of physics and electrical engineering, and we never used stats to quantify human behavior or construct term structure models. For example, the normal distribution (that you say you've lost faith in) wasn't first developed to explain financial events, but rather to explain the movement of gas particles. The vast majority of quantitative models are actually applied physics equations massaged to explain financial instruments and markets (read Black-Scholes and Heat equation). Ito calculus is rocket science (quite literally). Have you ever wondered why the vast majority of quants from the 80s and 90s were physicists?

    To come to the point, financial math is not an exact science, while electrical engineering and physics are. The development of math in any field starts with some basic assumptions (like Newton's laws of motion or Maxwell's equations), and are built on that. These laws have stood the test of time to the extent that they're "exact" now. The assumptions of Financial Math, however, were formed relatively recently and were borrowed from physics and were calibrated to fit the markets. It's not "exact" yet, and hence the reason the math can't explain certain events.

    Math is not meant to explain specific events. It's a framework. Engineers and physicists grant specific meaning to abstract framework to do whateva they need to do, so are the financial engineers.
    "as far as the laws of mathematics refer to reality, they are not certain; and as far as they are certain, they do not refer to reality." ----Albert Einstein

  • In reply to GekkotheGreat
    LTV's picture

    GekkotheGreat:
    LTV:
    I learnt stats as a part of physics and electrical engineering, and we never used stats to quantify human behavior or construct term structure models. For example, the normal distribution (that you say you've lost faith in) wasn't first developed to explain financial events, but rather to explain the movement of gas particles. The vast majority of quantitative models are actually applied physics equations massaged to explain financial instruments and markets (read Black-Scholes and Heat equation). Ito calculus is rocket science (quite literally). Have you ever wondered why the vast majority of quants from the 80s and 90s were physicists?

    To come to the point, financial math is not an exact science, while electrical engineering and physics are. The development of math in any field starts with some basic assumptions (like Newton's laws of motion or Maxwell's equations), and are built on that. These laws have stood the test of time to the extent that they're "exact" now. The assumptions of Financial Math, however, were formed relatively recently and were borrowed from physics and were calibrated to fit the markets. It's not "exact" yet, and hence the reason the math can't explain certain events.

    Math is not meant to explain specific events. It's a framework. Engineers and physicists grant specific meaning to abstract framework to do whateva they need to do, so are the financial engineers.
    "as far as the laws of mathematics refer to reality, they are not certain; and as far as they are certain, they do not refer to reality." ----Albert Einstein

    I completely agree with you (and think you've done a better job of conveying my original point). The point of my post was that the framework for engineering and physics has been around for a long time and is well established (did not mean exact literally, hence the quotes). The Financial Engineering framework, however, is just a convolution of other framework and has room for improvement.

  • In reply to LTV
    GekkotheGreat's picture

    LTV:

    I completely agree with you (and think you've done a better job of conveying my original point). The point of my post was that the framework for engineering and physics has been around for a long time and is well established (did not mean exact literally, hence the quotes). The Financial Engineering framework, however, is just a convolution of other framework and has room for improvement.

    I agree. But no matter how much time they spend, they will never achieve what physicist has done. "Financial Mathematics" is a damn social science. And pretty much every social science assume, "People are always rational." Otherwise, nothing is predictable. But NO, they are not... Now, you think this assumption is a little absurd. Then you see the financial engineers try to fit a normal distribution on the future cumulative return on the market, given 20+ other assumptions.

    Back to the original post, IMHO, it's better to learn something and know the assumptions, than to remain ignorant just because now you think the assumptions are far-fetched. At least, you can always turn around and take advantage of those people who are brave enough to use the model without considering the implications of the assumptions and those who remain cautiously ignorant.

  • za3212's picture

    Combine a major in psychology with your stats major

  • trazer985's picture

    One thing I've come to realise from this line of works is the fallacies in the assumptions people make about data.

    Look at specifically what the facts are, and what the limitations of what they mean are. Humanity makes some very bizarre and often self harming decisions (she doesn't like me, you fancy her not me etc. etc.) because of this, based on very loose evidence.

    Interestingly, does anyone else think that Taleb uses random in the same way most people use the word God?

  • leveredarb's picture

    you misunderstood taleb, he doesn't say statistics are useless, he just argues they are misused by an aweful lot of people. Its still a worthwhile topic to study, just be aware of the limitations. Same applies for economics.

    The only issue is once you accept talebs framework of thinking and start looking for bias and fallacies in every data source you can never read a newspaper article again(this is more true of fooled by randomness I feel).

  • In reply to leveredarb
    Hamilton's picture

    leveredarb:

    The only issue is once you accept talebs framework of thinking and start looking for bias and fallacies in every data source you can never read a newspaper article again(this is more true of fooled by randomness I feel).

    God dam, you're right. I concentrate more on 'facts' now.

    "Have you ever tried to use a chain with 3 weak links? I have, and now I no longer own an arctic wolf."
    -Dwight Schrute

  • In reply to blastoise
    happypantsmcgee's picture

    blastoise:
    Read you bible and pray, every day.

    Says the guy who texts during a funeral

    If I had asked people what they wanted, they would have said faster horses - Henry Ford

  • In reply to Hamilton
    leveredarb's picture

    To unlock this content for free, please login / register below.

    Connecting helps us build a vibrant community. We'll never share your info without your permission. Sign up with email or if you are already a member, login here Bonus: Also get 6 free financial modeling lessons for free ($200+ value) when you register!
  • In reply to leveredarb
    Hamilton's picture

    "Have you ever tried to use a chain with 3 weak links? I have, and now I no longer own an arctic wolf."
    -Dwight Schrute

  • levelworm's picture