@annanay The market has decided that price discovery is an important enough job to have that kind of incentive. Finance is the brain of the economy and price discovery is its fundamental tool
Neural networks are pretty magical, they can create features that are predictive of price by virtue of some high-dimensional pattern recognition.
BUT, it's not easy to create good neural networks in trading, and I've read a lot of papers on neural networks in trading and boy are they practically impossible to parse. I want to keep my articles on neural networks practical, and you should walk away from this thinking: "Huh, that's all there is to it, doesn't seem too difficult to try."
Here's how to implement a convolutional neural network that can predict short-term price direction from limit order book data with > 80% accuracy.
As usual, happy to share articles with anyone who comments and retweets on THIS post.
The Problems With PCA
We've all had pipedreams of extracting predictive features from PCA, but they've mostly fell flat. Why? I think in part it's because the factors have no stable meaning. PCA extracts whatever linear combination explains the most variance.
Secondly, the stocks themselves have no stable identity either. With this, it rewards memorizing idiosyncratic patterns specific to each stock. You have enough flexibility to fit noise rather than signal.
Today's article is about a different form of PCA that overcomes this and actually is able to extract predictive features!
You guys know the shtick, comment and retweet for a draw of the free article.
There are signals that just "makes sense", and when you find them, you need to do almost no kind of backtesting and fitting.
This is one such example. Once you understand this signal, you will go, "That totally makes sense, why didn't I think of that?!"
Most people don't understand what it means to be "creative". Every signal must be momentum or mean reversion or *shudders* pairs trading. I hope to share with everyone that good signals can be found by thinking deeply about behavioral effects and existing algorithmic signals.
Random comments and retweets will get this article for free.
You've read it somewhere, run PCA for "statistical factor analysis"; but material on this is either so shallow that it's meaningless (run pca and the eigenvectors are factors), or so dense that you'll need a PhD in Statistics to parse it.
This is the most information dense article on why PCA can actually extract factors, and how to reason about it.
Happy to share the article with some people who comment and retweets!
You've heard of crypto trend signals.
Have you wondered how they were created?
Today, I actually show you guys how to create a performant, scalable trend signal. Hint: It's going to be a smart ensemble of weak trend estimators.
Randomly giving out free articles for retweets!
If you've used machine learning to solve problems, one of the common questions you will have is how to do feature selection?
One possibility is the Sequential Forward Selection (SFS), which builds a feature subset one feature at a time. However, it is not feasible when working with an enormous set of features AND has issues with path dependency!
In today's article we cover a strictly superior variant of the Sequential Forward Selection!
Find it at the bad place, or retweet for a chance at a free article!
Karl Marx’s “das Kapital” has to be in the world’s best fiction. I have little more respect for flat-earthers than for communists.
The flat-earther stands alone on the lunatic fringe, mocked by virtually by everyone else, by even a casual observer with a smartphone or a glimpse of the horizon. Their error and serious disjunction from reality is considered obvious, their humiliation immediate, considered deserving, and universal. There is a brutal honesty in their isolation.
Communists, by contrast, enjoy a strange, perennial resurrection. Every generation produces a fresh harvest of well-meaning sophists who declare with perfect confidence, immense moral smugness, and zero embarrassment that “this time”, it will be different. This time, the human nature glitches will be patched, they promised. But as someone said, you can ignore reality but you can’t escape the consequences of ignoring reality. The corpses of a hundred million don’t register to the communists as data to cause any modicum of embarrassment but they’re instead boldly dismissed as “not real communism” or unfortunate growing pains.
The flat-earther’s fantasy harms no one but his own reputation. The communist’s fantasy has repeatedly produced mass graves, secret police, engineered famines, labor camps, firing squads at dawn, and societies so economically broken that people risked machine-gun towers and minefields just to taste basic bread.
Flat earthers and Communists : One is a clown in the corner of the internet. The other keeps getting invited back to run countries.
The communist is the most dangerous kind of fool that exists: the one who is perpetually convinced that next time his ideology will finally be implemented by sufficiently virtuous people. Therefore, never under any circumstances, a communist should be allowed anywhere near the levers of power.
Your signal with the HIGHEST SHARPE might be your biggest drag on live performance because trading it costs more than it returns.
If you want to understand how PMs think about managing fast moving signals in real portfolios, check out this article below!
Random 20 retweets + comments get free access to the article. Cheers everyone!
👇
@DearthOfSid Why is it that all your criticisms of libertarianism are generic sentences with no actual substance? Basically you can replace libertarianism in that sentence with anything which you dislike.