Machine learning, which is just an advanced form of artificial intelligence (AI), is changing how the world operate.
There are more use cases of machine learning in finance than ever before, a trend perpetuated by more accessible computing power and more accessible machine learning tools (such as Google’s Tensorflow).
What is Machine Learning?
An algorithm is just a procedure for solving a problem, based on conductiong a sequence of specified actions.
Machine learning is just a collection of algorithms that learn from data in different ways. These algorithms identify repeatable and persistant patterns in data. And the algorithms does not have to be told explicitly what kind of patterns and relationships to look for — the algorithms work that out for themselves.
In other words, machine learning provides computers the ability to automatically learn and improve from experience without being explicitly programmed to do so.
Uses of Machine Learning in Finance
Machine learning has a wide number of use cases in Finance. Some of they are:
- Modelling complex trading relationships that seek non-linear interactions between factors in order to provide an investment strategy that generates abnormal returns. The algorithm can even discover new factors based on financial past data and unlock new data sets.
- Fraud prevention because an algorithm is able to quickly weigh the transaction details against thousands of data points and make a determination whether or not the attempted activity is uncharacteristic of the account owner.
- Improve customer service. Banks and financial institutions can assess large data sets and understand what customers want and need using these algorithms. The solution, as provided by machine learning technology, is not to replace automated customer support systems, but to make them better.
Benefits of Machine Learning in Finance
The Financial Stability Board reported last November 2017 reported that some of the benefits would be:
- A more efficient processing of information, for example in credit decisions, financial markets, insurance contracts, and customer interaction, may contribute to a more efficient financial system.
- An improvement in regulatory compliance and increase supervisory effectiveness.
- Network effects and scalability of new technologies may in the future give rise to third-party dependencies. This could in turn lead to the emergence of new systemically important players that could fall outside the regulatory perimeter.
- Unexpected forms of interconnectedness between financial markets and institutions, for instance based on the use by various institutions of previously unrelated data sources.
Risks and Challenges of of Machine Learning in Finance
As with any new product or service, there are important issues around appropriate risk management and oversight.
It will be important to assess uses of AI and machine learning in view of their risks, including adherence to relevant protocols on data privacy, conduct risks, and cybersecurity.
Also, there are some challenges:
- Signal-to-noise Ratio: When using machine learning in finance, there is a lot of noise and a lack of clarity in the signals. We need to identify something that is barely perceptible due to the background noise.
- Non-Stationarity: Financial markets are in constant movement. They are dynamic. Machine learning works better with stationary data so models have to improve in order to deal with non-stationary data.
- Black-Box Aversion: Investors and Fund managers won’t invest all of their assets using something that they don’t understand. People don’t like to this approach.
How I know all of this?
I always like to be pending to the latest developments in technology, especially the ones that apply to the financial industry.
Dr Ledford is Man AHL’s Chief Scientist and Academic Liaison, based in the Man Research Laboratory in Oxford.
AHL has been researching machine learning techniques for around seven years, and trading machine learning components within its client programmes since early 2014.
He have us an introduction to machine learning techniques, and how they are increasingly being applied by quantitative investment managers.
Dr Ledford explained what machine learning is, how it can be distinguished from more traditional data analysis techniques and the opportunities and limitations of machine learning.
Other sources include: