News

How analysts can become citizen data scientists

machine learningAccessible machine-learning platforms enable investment analysts to take on the role of citizen data scientists in a short space of time, improving their fundamental research, says Charles Brecque, of Mind Foundry.

Today’s accessible machine learning platforms can be quickly and accurately tailored to unlock fundamental investment strategy insights in hours rather than months. This is enabling investment analysts to take on the role of citizen data scientist for the first time, augmenting their extensive industry knowledge with advanced machine learning capabilities.

Fundamental investment success involves building a thesis on global trends and future market directions, and identifying relevant investments that are aligned with this strategy. This second process is long and laborious.

Previously, applying machine learning to assist would require the use of expensive data scientists with little knowledge of investment issues, to help review financial data of potentially thousands of relevant businesses and identify under- or over-priced investments. Highly accessible machine-learning platforms are now changing all this, helping investment analysts advance through the machine-learning process in an intuitive manner to unlock greater insights from financial data.

Getting started: gathering data
It’s simple to get started. Take this example. An investment analyst would typically extract relevant financial data – anything from revenue and earnings per share, to operating margins – from a commercially available financial research platform such as Bloomberg. Once a suitably sized data set has been acquired, the investment analyst then awards scores of 1 to good investments, and 0 to inadvisable investments, for example based on whether the stock price increased significantly over the past year or not.

Data preparation is typically a tedious and time-consuming task for suitably qualified data scientists – with many stating in one survey that it comprises up to 80 per cent of their workload. With a humanised machine learning platform, powerful data preparation and manipulation capabilities are made accessible to employees of all skill levels – in this case the investment analyst – cutting the time, resources and expertise required to prepare financial data.

A key functionality of such platforms is providing actionable advice on how to best correct errors within data sets. In this investment context, take a data set containing missing values for share pay-out ratios. A humanised machine-learning platform will flag this issue to the analyst and provide several methods to correct it. The investment analyst may have contextualised knowledge of why this is the case – such as no dividends being paid out to investors – and take the opportunity to automatically fill in all missing pay-out ratio values with zero.

To save time on similar data preparation activities in the future, the investment analyst can save the data preparation workflow for auditing and reuse purposes by colleagues.

At this stage, it is possible to view a visualisation of how promising investment targets are distributed based on the various captured financial data points, such as cash flow. It’s a strong possibility that at this stage no clear pattern will be evident. However, this user-led data preparation is an invaluable tool in providing investment analysts with a unique opportunity to visualise and manipulate raw data prior to model selection and deployment.

Intuitive machine-learning platforms identity an effective model with a high level of accuracy to apply to data sets and provide justification for the suggestion. These models avoid overfitting – that is, fitting models to training data to the extent that they struggle with new and unseen data and fail to provide accurate predictions.

Using a Bayesian approach, which learns from each iteration what works and what doesn’t, the investment analyst can quickly identity the most effective model with a high forecasting accuracy. Selected models can be used within a platform or deployed through a specific application or Excel spreadsheet to make live forecasts.

Machine-learning models typically uncover intricate relationships between complex financial data points – drawing conclusions which would be missed by analysts without access to machine learning.

The investment analyst can go further and ask their machine-learning platform to cluster the companies. This can then be used for final risk analysis when constructing an investment portfolio.

Where next for machine learning in finance and investment?
Humanised machine learning platforms offer highly intuitive, user-centric interfaces to guide users step by step through the entire process, from data preparation to model deployment. The investment analyst can directly use machine learning to unlock detailed insights from financial data, at speed and scale, without specialist training or data science expertise.

The best way to get started is to pick a stage of the investment process which could benefit from experience gathering. This will enhance the analyst’s ability to perceive experience and as a result improve the performance of the portfolio by adding one more green or red flag in the decision-making process. At this stage, the investment analyst has truly become a citizen data scientist.

These capabilities also offer promising applications in the finance and investment sectors beyond informing investment analyst strategies. Strong use cases today include using machine learning to forecast company revenues to derive valuations, predict the failure of trades for trade settlement, predict and anticipate large drawdowns, forecast EPS beats and misses, and optimise marketing and sales operations for funds that rely on distribution strategies.

Charles Brecque is CX operations manager at Mind Foundry

©2019 funds europe