Buy Side Decision Science

In most investment firms, buy side decisions are increasingly being informed by the output of machine learning. This could be in automated decisions like in algorithmic trading to inform the traders decision to support value or growth investing.

Overview

Predict the future value of a given asset.

TARGET

Asset value

Challenge

Using machine learning to determine future asset prices is challenging for many reasons.  Outside of the difficulty of forecasting where the market is headed and its impact on the price of a given asset, the management and joining of all of the data that can influence this price is a major difficulty.  To start with the price history and trading volume of the asset (from Bloomberg for example) needs to be gathered and consolidated to understand both the current price and historical trends.  Next, the history and current macroeconomic environment needs to be included from someplace like FRED. For equities and corporate bonds, the SEC filings should be included.  Beyond these standard data sources, many investment decisions are beginning to be made with the help of alternative data.  These can include weather for agricultural commodities or satellite imagery to understand the state of the current agricultural crop. All this data needs to be cleaned, joined and transformed into valuable ML features before going into model training. This pre-modeling prep process can be frustrating and time consuming. We are here to help.


Modeling techniques and libraries

Asset price prediction

Build machine learning models using standard techniques (from linear regression to more advanced machine learning algorithms) to predict asset price. 


Package: 

  • Sklearn


Data features

No items found.

Related accelerators

No items found.

Get your data science on.