DEBT MARKET DATA SCIENCE

#DEBT-MARKET-DATA-SCIENCE

Goal: To cluster leaders vs followers in entry to new debt markets

#Goal:-To-cluster-leaders-vs-followers-in-entry-to-new-debt-markets

Contents

#Contents

Insight 1: The data show that the dynamics of the most popular markets changed significantly after 2010

#Insight-1:-The-data-show-that-the-dynamics-of-the-most-popular-markets-changed-significantly-after-2010
  • Did the nature of the market change after 2010, or did the nature of the data collection change?
  • Note to only use post-2010 data for modelling
  • Note that reporting period dates used (some latency)

Insight 2: The data show that the dynamics of less popular markets are consistent before and after 2010

#Insight-2:-The-data-show-that-the-dynamics-of-less-popular-markets-are-consistent-before-and-after-2010

Question 1: Are we interested in identifying leaders issuing debt in new markets with some measure of success?

#Question-1:-Are-we-interested-in-identifying-leaders-issuing-debt-in-new-markets-with-some-measure-of-success?
  • What would constitute success in terms of the financials?
  • Are successful leaders only determined by numbers of followers?
  • Are successful markets indicated by the rate of increase of debt issuances?
  • Do we want to identify leaders who venture successfully into new markets (rate of success)?
  • Is currency the only available market identifier? (EUR - Greece is quite different from EUR-Germany)

Question 2: Timelines and risk profiles?

#Question-2:-Timelines-and-risk-profiles?
  • Is reporting period too coarse a time frame?
  • Would target 'follower' customers like to venture into new markets weeks, months or years after leaders?
  • Is there a time period where followers would like to see some excellent performance by leaders before jumping into new markets?
  • Do followers and leader have different risk profiles and should this be part of the modelling?

Summary

#Summary
  • Very interesting data

  • Some initial insights and views of the data found

  • Wary of black box modelling approaches

  • Need more understanding of the debt market landscape before modelling

Import libraries and initialise plotly

#Import-libraries-and-initialise-plotly
Loading output library...

Exploring market entry points with debt_balance

#Exploring-market-entry-points-with-debt_balance
Loading output library...

Find all the rows where reporting age is 1 as we are interested in earliest market entry points

#Find-all-the-rows-where-reporting-age-is-1-as-we-are-interested-in-earliest-market-entry-points
Loading output library...

5 most popular markets

#5-most-popular-markets
Loading output library...

Interestingly many of the median popularity markets do not exhibit the pre and post 2010 split in popularity dynamics that is evident in the more popular markets

#Interestingly-many-of-the-median-popularity-markets-do-not-exhibit-the-pre-and-post-2010-split-in-popularity-dynamics-that-is-evident-in-the-more-popular-markets
Loading output library...

Lots of one hit (flop?) wonders in the debt market space

#Lots-of-one-hit-(flop?)-wonders-in-the-debt-market-space
Loading output library...

Upper quartile of popular markets show split in dynamics pre and post 2010

#Upper-quartile-of-popular-markets-show-split-in-dynamics-pre-and-post-2010
Loading output library...

Enter any currency market to view trajectory e.g. EUR, AUD, USD, SYP

#Enter-any-currency-market-to-view-trajectory-e.g.-EUR,-AUD,-USD,-SYP
Loading output library...

Use Debt_balance to filter the other files

#Use-Debt_balance-to-filter-the-other-files

filter large financial data files to extract only the rows relating to companies with debt issuances

#filter-large-financial-data-files-to-extract-only-the-rows-relating-to-companies-with-debt-issuances

Viewing files

#Viewing-files
Loading output library...
Loading output library...
Loading output library...
Loading output library...
Loading output library...