This is why the trading mean reverting conditions have been tight in range bound for at least the last 30 days according to these metrics. As you can see in these Instagram posts, the Hurst Components has been less than 0.5 which is mean reverting! This is for most time periods I bpick which is the passed argument in the Pyton script.
Simplest mean reversion is moving averages but how effective is it and which timeframes work best for crypto? You can pick 5 best long and 5 worst outlook for pair trading in crypto but Correlated relationships don’t last too long due to crazy volatile moves.
Qunatopian has some interesting discussion on other mean reverting trading ideas
https://www.quantopian.com/posts/search?q=Mean%20reverting
RSI2 and Long stocks at minus two times Average True Range(ATR) looks interesting? As for moving average, you need to find the fight parameters for moving average against which timeframe?
https://www.johndeoresearch.com/mean-reversion-trading-strategies/
This could be extra complicated with overlaying linear regression over mean reversion?
https://www.marketcalls.in/amibroker/linear-regression-based-mean-reversion-strategy.html
Is this useful with typical mean reversion?
https://dailypriceaction.com/blog/mean-reversion-guide-to-market-timing/
Hurst Exponent is most reliable to find when a time series is random walk (Geometric Brownian Motion), mean revert, and trend follow/momentum.
https://www.quantstart.com/articles/Basics-of-Statistical-Mean-Reversion-Testing/
decent entry/exit description with basic mean reverting strategies
https://blog.quantinsti.com/mean-reversion-strategies-introduction-building-blocks/
Another easy Quantopian Python logic example
https://www.quantopian.com/posts/simple-mean-reversion-strategy
Z-Score can be highly reliable to measure cointegration
https://www.quantstart.com/articles/Backtesting-An-Intraday-Mean-Reversion-Pairs-Strategy-Between-SPY-And-IWM/
Some say it is possible to beat random walk but I doubt that
https://steemit.com/mathematics/@sicilian/could-it-be-possible-to-profitably-trade-a-random-walk
I am no expert here from a math point of view but I believe in this headline which means avoid random walk
https://quant.stackexchange.com/questions/1710/proof-that-you-cannot-beat-a-random-walk
TIP to Find decent articles with DuckDuckGo by searching:
Find mean reverting strategy ideas
NOTE I now post myTRADING ALERTS into my personal FACEBOOK ACCOUNT and TWITTER. Don't worry as I don't post stupid cat videos or what I eat!
I re-activated an old trading bot that focuses on the Ether market vs Bticoin. All the big crypto coin like BTC, ETH, BCH, LTC are all pointing negative. I must remember the Ether bot seems to be getting performance than the old Bitcoin. I always forgot about this and wish I remembered is earlier if you want better profit and loss.. It would have got these signals up and running faster.
This is the current profit and loss running after the first hour:
total Profit and Loss: 6.000000000000037e-06 win/loss ratio:100.0 total Rate of Change: :0.7001166861143567
The bot has been transferred to my cloud server finally. This means All system GO! It will go through an initial 24 hour before fully being operational. Exciting times ahead. Signal generation will follow soon courtesy of of this service!
As I get closer to moving bot code to a remove server since it its doing well, I am looking at making a available a custom virtual machine so you can received trading signals. I can deliver by Telegram messenger or Email. What do you prefer? I find Telegram is much easier. This VM will be prebuilt with all the necessary to received these signals and process the trades into a preferred exchange like Binance. This will all be part of my monthly Quant Analytics service. Would you be interested in this type of service? Details here: https://quantlabs.shop/products/quant-analytics
I get so many articles sent over to me on linear regression. I would suggest reading this article if you are into learning how to apply this technique with Python’s Scikit Learn package for machine language application.
So, I believe everyone who is passionate about machine learning should have acquired a strong foundation of Logistic Regression and theories behind the code on Scikit Learn. All right… Let’s start uncovering this mystery of Regression (the transformation from Simple Linear Regression to Logistic Regression)!
I did not even know this existed. The first question is will it replace tradng data scientists humans. The answer is no but we heard that before a this is fairly new technology. If I was in this field, I would be afraid of this technology. This article also covered the results of each cloud solution of Google vs Microsoft Azure. I am sure there will be growth here in the near future.
I remember when I first heard about Fourier Transformations with periodic functions. This article was posted on my Facebook group so people responded to it. Here are some highlights of the complexity of it:
The
Sine and Cosine functions are arguably the most important periodic functions in
several cases:
The
periodic functions of how displacement, velocity, and acceleration change
with time in SHM oscillators are sinusoidal functions.
Every
particle has a wave nature and vice versa. This is de Broglie’s
Wave-Particle duality. Waves are always sinusoidal functions of some
physical quantity (such as Electric Field for EM Waves, and Pressure for
Sound Waves).
There is even a section described for artificial
intelligence as well.
