[ 3 / biz / cgl / ck / diy / fa / ic / jp / lit / sci / vr / vt ] [ index / top / reports ] [ become a patron ] [ status ]

/biz/ - Business & Finance


View post   

File: 14 KB, 474x248, NNT.jpg [View same] [iqdb] [saucenao] [google]
10637527 No.10637527 [Reply] [Original]

brainlet here. wtf are fat tails and why are they important?

>> No.10637541

>>10637527

>he's never had a fat tail before

Never gonna make it

>> No.10637552

>>10637527

big fat cawks that you succ

>> No.10637594
File: 35 KB, 540x529, african-fat-tailed-gecko.jpg [View same] [iqdb] [saucenao] [google]
10637594

>>10637527

>> No.10638238

>>10637594

Leopard geckos are the patrician pet. Taleb approved

>> No.10638327

>>10637527

You never seen a white girl with a nice big delicious ass? That’s a fat tail. It’s important cause it makes your pee pee hard.

>> No.10638763
File: 522 KB, 836x596, George-Costanza.png [View same] [iqdb] [saucenao] [google]
10638763

>>10637527
fat tails are untradeable theoretical events and they are important for zerohedge articles

>> No.10638826

>>10637527
Basically it means low probability high impact events decide the course of things a lot more than the average day to day events. If you delve into this stuff deeply you start to get into philosophical discussions about probability, chaos, and our understanding of the universe itself. The market can be very thought provoking because of how utterly fucking baffling it can be at times. The overall takeaway you should have as a trader though is there are two ways to get rich...lose a little often, but make a ton occasionally and thus have a positive expectancy (hopefully) OR lose a ton rarely, but make a little often and thus have a positive expectancy (hopefully). Neither is necessarily better, because individual skill, risk management, and other external factors come into play. The former's problem is how do I maximize my winning opportunities, the latter's problem is how do I avoid or mitigate my catastrophes.

>> No.10639629

>>10637527
Fat tails mean the probability distribution isn’t a Gaussian bell curve. As a result any model you use for prediction is not valid.

A practical example, Long Term Capital Management went under to excessive of leverage during a financial collapse (irc something around 30x). According to the models the chance of collapse could have expected to happen once in a billion years (some bizarrely small number), however multiple market crashes would happen in succession day after day.

Gaussian bell curves are divorced from reality. How can these mathematicians say a crash could only happen once in a billion years when the instruments they’re trading have only existed for a few decades? The math is beautiful but deceptive.

It is important because you cannot use a mathematical model to predict stock prices if the underlying assumption is that gains / losses follow a Gaussian bell curve.