Informing Investment Decisions Using Machine Learning and Artificial Intelligence

Informing Investment Decisions
Using Machine Learning and Artificial Intelligence

Why Data Matters

Quantitative techniques and new methods for analyzing big data have increasingly been adopted by market participants in recent years. This includes computerized trading, use of big data, and machine learning or artificial intelligence.

Big data and machine learning have the potential to profoundly change the investment landscape. As the quantity and the access to data available have grown, many investors continue to evaluate how they can leverage data analysis to make more informed investment decisions. Investment managers who are willing to learn and to adopt new technologies will likely have an edge.

Machine learning and artificial intelligence may unleash new insights; however, investors need to better understand the current landscape and applications for data analysis before making significant investment in the technology.

Big Data and Machine Learning ‘revolution’

A quantitative investor has access to real-time information, but organized data is not always readily available, and it needs to be analyzed to glean tradeable ideas.

The availability of new datasets, methods of analysis and more sophisticated computing has led to the growth in big data and the machine learning ‘revolution.’

3 trends that enabled the Big Data revolution

Exponential increase in amount of data available
Increase in computing power and data storage capacity, at reduced cost
Advancement in Machine Learning methods to analyze complex datasets

How will Big Data and Machine Learning
change the investment landscape?

The changes to the investment landscape will be profound. Big data will give an edge to quant managers who are willing to adapt and learn about new data and analysis methods. Machines have the ability to quickly analyze news feeds and tweets, process earnings statements, scrape websites, and trade on these instantaneously.

Big data and machine learning strategies are already eroding some of the advantage of fundamental analysts, equity long-short managers and macro investors, and systematic strategies will increasingly adopt machine learning tools and methods.

The transition won’t be without setbacks, though, as certain data may have no value and more complex techniques don’t always produce better forecasts.

Analyzing new data sources

Signals can now be found in data generated by:

INDIVIDUALS

(e.g. sentiment from news articles, Facebook, Twitter, etc.)

BUSINESS PROCESSES

(e.g. credit card use data)

SENSORS

(e.g. satellite imagery of mining sites and company parking lots)

New techniques using Machine Learning and AI

As data sets get larger and more complex, investors need to use sophisticated data analysis techniques. The tools used for these tasks include machine learning (drawn from traditional statistics) or deep learning (inspired by the working of the human brain).

These techniques can be used to analyze data and design trading strategies. Select a technique to learn more:

Supervised learning

Unsupervised learning

Deep learning

Using historical data points as training samples to infer a rule or ‘equation’ capable of predicting future outcomes.
For example: An algorithm that can evaluate the best momentum signal (derived from market performance over recent time periods) for predicting future market performance, or to try to predict how much the market will move if there is a sudden spike in inflation.

Aims to identify the common drivers behind the data points by identifying relationships between input variables.
For example: A successful algorithm could find that, at a certain point in time, the market is being driven by various factors, such as momentum factor, energy prices, level of US dollar, and liquidity.

Analyzes data via multiple iterations, or “layers of learning” – starting by learning simple concepts, and then combines these to formulate more complex concepts. This can be accomplished by passing the data through multiple layers of non-linear processing units in a manner similar to neurons within the human brain.
For example: Artificial intelligence that is capable of efficiently executing a given trade order with minimal market impact.

Key Takeaways

Investment managers who adopt and learn about new datasets and methods of analysis will likely have an edge.
Regardless of the timeline of these changes, analysts, portfolio managers, traders, and CIOs will need to be familiar with big data and machine learning developments and related trading strategies.
Both fundamental and quantitative investors will leverage big data and machine learning across asset classes.
Machine learning algorithms cannot entirely replace human intuition, or understand complex, long term investment trends.

Global Research Update

Marko Kolanovic
Global Head of Quantitative and Derivative Strategy

Clients can visit J.P. Morgan Markets to read the full report.