By clicking “Accept All Cookies”, you agree to the storing of cookies on your device to enhance site navigation, analyze site usage, and assist in our marketing efforts. View our Privacy Policy for more information.
Creating a Custom Factor Based on Dividend Yield
Kiski's Quantitative Research team developed a custom factor centered around dividend yield. This blog explores the process and significance of creating this custom factor.
Client-Driven Innovation

The creation of this custom factor was driven by a specific request from a client, a long-only fund with almost $2bn in assets. The client sought to boost their performance by examining the impact of dividend yields on their portfolio. This factor would enable the investment team to assess exposure and examine return correlations, gaining insights into their portfolio that they did not have at that point. Our Quant team delivered a tailored solution, integrating it on the custom client portal and focusing on the ease of access to the requested infrastructure.

Factor Models and Dividend Yield

The Fama-French factor model initially included market risk, size, and value to explain stock returns, and has since been expanded to include profitability and investment patterns. These factors capture the nuanced behavior of different stocks, leading to more accurate predictions and better-informed investment decisions.  

Dividend yield, the ratio of a company's annual dividend to its share price, is a crucial indicator of a stock's value. By developing a custom factor based on this ratio, we aimed to identify the client’s portfolio returns' connections, exposures, and correlations to dividend yield.

The Process of Developing the Custom Factor

Data Collection and Preparation: Our already robust data infrastructure allowed us to efficiently handle and process extensive information, including high-level indices data and granular ticker-level data. As data quality was critical for the success and development of this feature, we employed rigorous cleaning and validation processes.

Factor Construction: We computed dividend yields for each company and standardized these values for comparability across sectors and market conditions. This involved normalizing the data and adjusting for outliers and anomalies.

Integration into the Model: Integrating the custom dividend yield factor into our existing multi-factor model was done after extensive testing and multiple iterations. We assessed the factor's predictive power and its correlation with other established factors.

Implementation and Monitoring: Once validated, the custom factor was integrated into our daily calculations and portfolio analytics, provided through direct email reports as well as live dashboards on our portal. Continuous monitoring and dynamic adjustments ensure that the factor remains relevant and effective and responsive to changing market dynamics.

Results and insights

Kiski Platform Solutions was envisioned with flexibility as a core feature – our customization capabilities allow our analysts and developers to deliver these tailored solutions quickly and efficiently. Clients receive precise, data-driven insights that align with their objectives and strategies, helping them achieve their goals without putting an additional burden on internal investment staff. By providing a swift response to client requirements, we make sure they maintain a competitive edge while staying focused on the markets.

Explore our solutions and start getting answers.
Book a call
About the author
Janko Sikošek
linkedin logo

Janko Sikošek is a Quantitative Analyst with a strong background in finance and analytics. He currently applies his expertise in quantitative research to enhance investment strategies. Janko's academic credentials include a Bachelor's degree in Economics from the Faculty of Economics in Belgrade, alongside extensive experience in various internships within finance and sales. His skill set is complemented by a strong interest in economics, trading, and strategy.

More Blog Posts
Read all Blog Posts
->