How to Predict Stock Market Crashes using Mathematical Models

preview_player
Показать описание
Some financial bubbles can be diagnosed before they burst. Let's see how.

0:00 Intro
0:33 Super-exponential growth
1:40 Log-Periodic Power Law (LPPL)
3:23 Bubble indicators
3:54 LPPL Singularity Confidence Indicator
5:03 Financial Crisis Observatory
5:23 Trading strategies
6:08 Herd behavior
6:32 Cryptocurrency bubbles
7:05 Deep learning experiments

References

Financial Crisis Observatory

Why Stock Markets Crash by Didier Sornette

A Stable and Robust Calibration Scheme of the Log-Periodic Power Law Model

Real-Time Prediction and Post-Mortem Analysis of the Shanghai 2015 Stock Market Bubble and Crash

Theses that backtest trading strategies based on LPPLS indicators

Mesmerising Mass Sheep Herding
Рекомендации по теме
Комментарии
Автор

Amazing explanation my friend. I do some research in this domain. I've found that stock markets from developing nations, commodities and stocks during stressful time obeys LPPL like anything. Herding being at it's peak ensures that.

bikramadityaghosh
Автор

awsome! Thank you so much! Would be nice if you do more videos of this topic

Aviszzs
Автор

Best channel that I have discovered in years. Keep it up!

andreizaika
Автор

Great information, especially the end with your own nn approach. Keep it up!

WhyGoThere
Автор

This was awesome. Thanks for explaining a topic that is otherwise way over my head :)

iKavt
Автор

A Norwegian quant team did some financial modelling for fun. Then the news got wind of them earning a 100 million on the stock market. They said "Oh we're just trying a few things out" and the media magically left them alone. Also there's this math professor in London. Yeah he just bought an apartment for £10, 000, 000.

kebman
Автор

awesome video! I am stuck in finding the value for the critical time variable, to compute Tc-T cheers

retrofuturst
Автор

Nice intro to the topic. Besides my lamenting the over constrained approach, I would assume the parameters are sector sensitive, thus train for each sector separately.

nd
Автор

Thanks for introducing mathematical approach. I'd like to read all of reference your provided

AIStockPrediction
Автор

excellent, this was a real video - thank you

BreezeTalk
Автор

Sayısal analiz ve matematiksel modellemeye ilgili bir mühendis adayı olarak, çok ilginç bir anlatı, teşekkürler! :)

fzycold
Автор

Çok güzel bir çalışma olmuş. Tebrikler sevgili oğlum..👍👏

mustafaiskdogan
Автор

Nice video, you do a great job explaining the model! You have some original ideas with regards to using the estimated params in a nn. No love for the BiT blog tho? :(

joshuanielsen
Автор

I use square of 9 to predict future dump or pump in crypto

luckychan
Автор

I think Sornette paper says P(t) = and NOT Ln{P(t)} = ?

klam
Автор

How to download the stock market data?

rakhibalasani
Автор

usta bir türkün başka bir türkün ingilizcesinden türk olup olmadığını anlaması max 5 saniye sürüyor bu mucizevi bir şey :) gayet güzel konuşuyorsunuz yanlış anlaşılmasın video için teşekkürler

applepie
Автор

Great video! Thanks for sharing, btw, is there any source code about this?

黃健寧
Автор

It's great bro and do you have a github account. I have some problems on coding this model.thx bro

heisenburg
Автор

Great Video!, I learned a lot. I am going to start working on my research as well including a Neural Network model, Utility theory including the SMA. There is no way that you can predict the Stock Market just with math. The Stock Market is moved by emotions and it's pretty irrational like the 97% of people that invest in the Stock Market... lol... do you have Twitter?

dkmk
visit shbcf.ru