Skip to content
All library documents

Using PMFG Centrality to Compare Peripheral and Central Stock Portfolios

Article Hudson & Thames
Read the original

Summary

This article explains how a Planar Maximally Filtered Graph (PMFG) represents similarities among assets while preserving more network structure than a Minimum Spanning Tree. It ranks nodes by a combination of graph centrality measures, then compares portfolios of peripheral and central stocks. The portfolio selection uses daily prices for 92 stocks and indexes before the Covid-19 market decline; performance is examined from just before the decline through September 2020.

In that sample, the peripheral portfolio recovered to 90.7% of its pre-decline value, compared with 76% for the central portfolio. It also had a smaller trough decline and consistently lower rolling volatility, while rolling Sharpe ratios diverged in its favor later in the period. These observations suggest that peripheral stocks fared better in this particular crash and recovery window. The result is a historical comparison, not a general guarantee: the article uses one sample, a chosen centrality ranking, and a crisis period, and does not establish that the pattern will persist in other markets or periods.

Key ideas

  • A PMFG adds high-similarity links while preserving planarity and contains the Minimum Spanning Tree as a subgraph.
  • Centrality measures can rank assets from peripheral to central within the network.
  • The article compares portfolios selected before the Covid-19 decline and evaluated through September 2020.
  • In this sample, the peripheral portfolio recovered more and showed lower rolling volatility than the central portfolio.
  • The observed advantage is limited to the studied assets and historical period.

Tags

EquitiesPortfolio constructionRisk managementUS markets

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.