What is AI/ Machine Learning?

AI/Machine Learning is a field at the intersection of Math and Computer Science.

The field is very broad but the gist is to teach machines to interact with the world the way that humans do. At the infancy of the field this was done explicitly, but over time we learned how to develop machines that can learn to recognize patterns in data on its own. This is Machine Learning.

Machine Learning seeks to approximate the process that was responsible for creating the data. You can think of it as reverse engineering in a sense. We have data. We want to understand how it was created. So we can make assumptions about the process that created it that allow the machine to learn and produce more data. If the data our machine produces is a good match to the data we started with, then we know a little more about how that data was created and our machine learned as well.

Do you offer trading signals or algorithms?

No. Let us explain. If we offered to sell you a signal or algorithm that would allow you to make a $1,000,000 for a small fee of $1000. Should you take this deal?

We don’t think so.

Spoiler Alert: If you took this deal, the signal is to find 1000 people just like you and sell them the same signal.

Professionals that develop real signals and algorithms don’t sell them. Have you ever heard of any quantitative hedge fund offering to sell the signals or algorithms that sustain their profitability? We haven’t.

If a signal or algorithm provider can sell you a signal or algorithm that annotates when you should buy and when you should sell or claims to trade for you, they could put their own capital in it and make significantly more than the fee they’re charging you.

Most signal/algorithm providers are not quants, researchers, or engineers. They are marketers. They are good at automation and drawing images on charts in hindsight and coming up with buzz words to drive engagement. AI, algorithmic, liquidity, bot, etc. are the latest. Humans are visual and have a proclivity for pattern recognition, making them susceptible to cognitive biases that pay marketers.

We offer models not signals or trading algorithms. Trading is all about decision making. We conduct proprietary research that mimics the process of research and development in investment banks and hedge funds, to produce models as a decision support tool.

See our Disclaimer.

What is a Quant?

The term Quant is an abbreviation of Quantitative.

This term refers to someone who builds mathematical models. Quants typically have formal backgrounds in Math, Computer Science, Physics, Economics, or some other engineering field.

The connotation of the term has evolved since its inception. Originally, the term was used as a pejorative for non-traditional members of Wallstreet. Over time the term has taken a more positive view as markets have become more quantitative.

There is a lot of misinformation online. In short, quant means math. It doesn’t mean Python or any programming language. It doesn’t mean automation or algorithms. It is impossible to be a quant in the absence of mathematical models.

What are Models?

Models are function approximations.

To clarify, we’ll first explain what models are not. Models are not automation coded in Python or some other programming language. They are not indicators that draw zones and levels on charts. They are not trading bots of which is not AI or Machine Learning, but simply automation.

A model is math. Things in the real world can be “modeled” or represented mathematically. These math equations explain how and why certain events happen. These are models as they approximate real world functions, albeit, mathematically.