Artificial Intelligence (AI) and Machine Learning (ML) are quietly revolutionizing nearly all areas of our lives. Did you know the latest trading algorithms are using these technologies extensively?

And, believe it or not, Machine Learning hedge funds already vastly outperform generalized hedge funds as well as traditional quant funds according to a report by
ValueWalk.

While hedge fund managers may hesitate outsourcing investment and risk assessment decisions to robots - ML and AI systems can be incredibly helpful to humans during the decision-making process.

This article offers a few examples of how Machine Learning can
be applied to trading strategies but the possible applications are vast.

AI/Machine Learning Hedge Fund Index vs. quants and traditional hedge funds. Source: Eurekahedge

Below are AI applications.

Arizona Financial Text system (AZFinText)

Imagine a system that can monitor stock prices in real time and predict stock price movements based on the news stream. That’s precisely what AZFinText does. This article recounts an experiment that used Support Vector Machine (SVM) to trade S&P-500 and yielded excellent results. Below is the table that shows how it performed relative to the top 10 quantitative mutual funds in the world:

Experimental trading strategy using Google Trends

There are a plethora of articles on the use of Google Trends as a sentiment indicator of a market.

For example,the experiment in this paper tracked changes in the search volume of a set of 98 search terms (some of them related to the stock market). The term "debt" turned out to be the strongest, most reliable indicator awhen predicting price movements in the DJIA.

Below is a cumulative performance chart. The red line depicts a "buy and hold" strategy. Google Trends strategy (blue line) massively outperformed with a return of 326%.

Performance of Google Trends strategy

Advent of robo-advisory in wealth management

The vast majority of online advisory programs are just algorithmic software or passive portofios. They meet a need in the market, but are relatively blunt instruments.

In the future, AI-powered portfolio management and online advisory systems can play a significant role in helping people with relatively small savings (less than $250k) who cannot currently afford professional consulting services. By bulking together small accounts that are unprofitable on their own, an investment firm can amass significant assets that may be profitable when managed on an aggregate basis.

Of course, some worry that this management could potentially undermine to some degree the position of traditional wealth advisors and lead to personnel reductions. But here’s the reality: the more experienced advisors have nothing to worry about. Human input will always play an important role in investment decision-making process. AI portfolio management systems will serve as a useful tool for investment professionals and may be used for managing lower risk, onventional investments. In other words, AI will allows the wealth advisor to provide clients with added value.

An example of this in action today can be seen with US firm Wealthfront, which offered clients a service that harvests tax losses that was previously only available to portfolios of $5 million and above. It claims the service can add extra 1.55% a year to portfolio return.

The frontier of AI is called Deep Learning (DL) and it is expected to constitute a core wealth management and planning platforms of the future.

Beyond trading and market trend prediction.

Aside from market prediction, Artificial Intelligence can empower all kinds of automation needed in an effective organization. All the tasks normally performed by human analysts (think text processing, document classification, finding not-obvious patterns, text selection based on certain criterion, etc.) can be done by AI, freeing up people to focus on other areas of the business, such as client relations.

Context-aware Natural Language Processing (NLP) can be used for many of the above activities. It enables the machine to extract the meaning from text, analyze its sentiment (whether the writer has positive or negative attitude towards something), and automatically categorize the content into very specific groups etc.

Interested in business applications of Natural Language Processing? Click here.

Add to this the continuous monitoring of the internet (social media and other sites) and you will get a machine that empowers your analysts.

Based on the outcome of content analysis, the system can automatically fetch additional information if they are needed to make a more accurate decision.

Can I learn ML myself?

Applying Machine Learning to trading is a vast and complicated topis that takes the time to master. But if you're interested, as a starting point we recommend:

Once you're familiar with these materials, there is alo a popular Udacity course on hot to apply the basis of Machine Learning to market trading.

If you want to speed the learning process up, you can hire a consultant. Do make sure to ask tough questions before starting a project .

Or, you can schedule a short call with us to explore what can be done.

How is Deep Learning different and why I should care?

Deep Learning is on the forefront, on the bleeding edge of technology.
In contrast to a standard algorithmic trading or Machine Learning, Deep Learning-driven system does not need specific rules, indicators or features (feature engineering process). It's able to invent them itself saving analysts huge amount of time and, it performs well when the scale of the data increases (often, using DL requires a larger volume of data comparing to classical ML methods).

DL system is an Artificial Neural Network with many (often hundreds) of hidden layers. After it's constructed, it needs to be trained.

Training is a process of feeding the network with data and right answers (in this case - historical market data), and "penalizing" it for every incorrect prediction in a process called back-propagation. It often requires significant GPU unit(s) (for heavy computation), but nowadays they are accessible on the Cloud (e.g., Amazon P2). The outcome of the training is a computed matrix of weights defining the connections in the network.

See a short introduction on how to train a Tensorflow network on Amazon P2

More often than not, Deep Learning provides an End-To-End solution for a given business/engineering problem. Using pure Machine Learning, on the hand, may require breaking the pipeline into more distinct steps and phases.

The image below presents a simple artificial neural network with two hidden layers:

neural network

The next illustration presents the GoogLeNet - a network with over 100 layers that was the winner of ILSVRC 2014 (ImageNet Large Scale Visual Recognition Challenge). It's successor - Microsoft ResNet - outperforms human intelligence in object recognition and visual tasks.

googlelenet

Another interesting difference is interpretability.

Deep Learning methods, despite often being supreme in performance to ML, do not give a clear explanation or a reason why an accurate prediction is generated (and what particular neurons are modeling); it's hard to interpret the results.

I need more specific examples applicable in my industry.

This paper describes how Deep Neural Networks (DNN) were used to predict 43 different Commodity and FX future mid-prices .

Another experiment describes trading on Istanbul Stock Exchange with NN and Support Vector Machine (SVM) .

Interestingly enough, this paper presents how genetic algorithms support vector machine (GASVM) was used to predict market movements .

Still hungry for more details or practical applications?

Let's spend 15 minutes together to explore your idea or project and find out if our team of data scientists and engineers can help!

Summary

Implementing a successful Machine Learning investment strategy is hard. You will need extraordinary talented people with experience in trading and Data Science. By using ML, your portfolio can yield a greater alpha.