./mdx_sources/StructuralBreaks/structural_breaks.mdx

Structural breaks, like the transition from one market regime to another, is one example of such a confluence that is of particular interest.

What's the distinction between organized and unstructured data? We will learn how to work with unstructured financial data and then transform it into a structured dataset that can be used by $mathrmML$ algorithms. In general, you should avoid consuming someone else's processed dataset because you will most likely find what someone else already knows or will figure out shortly. Ideally, your starting point will be a collection of unstructured, raw data that you will analyse in order to generate relevant characteristics.

Structural Breaks

./mdx_sources/PortfolioConstruction/portfolio_construction.mdx

In practice, mean-variance optimal solutions tend to be concentrated and unstable. How to deal with the instability caused by the noise contained in the covariance matrix?

Portfolio Construction

./mdx_sources/DangersOfBacktesting/dangers_of_backtesting.mdx

A common misunderstanding is to think of backtesting as a research tool. Researching and backtesting is like drinking and driving.

Backtesting Dangers

./mdx_sources/MicrostructuralFeatures/microstructural_features.mdx

The level of detail contained in FIX messages provides researchers with the ability to understand how market participants conceal and reveal their intentions.

Microstructural Features

./mdx_sources/FeatureImportance/feature_importance.mdx

Why repeating of backtest may fail? and how to prevent it. Why repeating a test over and over on the same data will likely lead to a  discovery?

One of the most common errors in financial research is taking some data, running it through an ML algorithm, backtesting the predictions, and repeating the process until a nice-looking backtest appears. Such pseudo-discoveries abound in academic journals, and even significant hedge funds are prone to falling into this trap. It makes no difference if the backtest is an out-of-sample walk-forward. The fact that we are repeating a test on the same data will almost certainly result in a  discovery. This methodological error is so well-known among statisticians that the American Statistical Association warns against it in its ethical guidelines (American Statistical Association [2016], Discussion #4). It usually takes around 20 iterations to find a () investment strategy with a standard significance level ( positive rate) of 5%.

Feature Importance

./mdx_sources/EntropyFeatures/entropy_features.mdx

Entropy Features

When markets are not perfect, prices are formed with partial information, and as some agents know more than others.

Sat Mar 05 2022

Entropy Features

TODO