20170705-三星证券-Pair_Trading_Strategy_22页_1mb

Summary of Quantitative Issue: Pair Trading Strategy for Holding Companies and Subsidiaries

Core Content

This document outlines a pair trading strategy focused on holding companies and their subsidiaries, particularly those with a single major subsidiary. The strategy leverages the stationary correlation between the share prices of holding companies and their subsidiaries, which is crucial for identifying mean reversion opportunities. The report highlights that Korea has seen a rise in the number of holding companies, creating more pair trading opportunities.

Main Points

1. Holding Companies and Subsidiaries

• Holding companies in Korea have increased from 67 (end-Sep 2015) to 76 (end-Sep 2016).
• Companies like Hyundai Heavy Industries, Maeil Dairy Industry, BGF Retail, and SK Chemicals are transitioning to holding company structures.
• Holding companies that rely on a single listed subsidiary for most of their NAV are ideal for pair trading due to their stable share-price ratio.

2. Pair Trading Strategy

• The strategy involves long/short positions based on the share-price ratio between a holding company and its subsidiary.
• Buy signal: When the ratio falls below -1 sigma.
• Sell signal: When the ratio exceeds +1 sigma.
• Unwinding signal: When the ratio reverts to ±0.5 sigma.
• Ceiling on investment period: Positions are unwound after 14 days.
• Ceiling on maximum loss: Positions are automatically closed if the combined return falls below 20%.

3. Pair Trading Universe

• The report identifies nine pairs, divided into three large-cap pairs and six mid-cap pairs.
• Each pair has a specific start date for trading after the share-price ratio stabilizes.
• Large-cap pairs include:
  • AmoreG/Amorepacific
  • Hankook Tire Worldwide/Hankook Tire
  • Hanjin KAL/Korean Air Lines
• Mid-cap pairs include:
  • Youngone Holdings/Youngone
  • Halla Holdings/Mando
  • Nongshim Holdings/Nong Shim
  • Nexen/Nexen Tire
  • Hansae Yes24 Holdings/Hansae
  • Poongsan Holdings/Poongsan

4. Backtesting Results

• Large-cap pairs generated annual after-tax returns of 15.9% with an information ratio of 1.2.
• AmoreG/Amorepacific had total after-tax returns of 182% over 2582 business days.
• Hankook Tire Worldwide/Hankook Tire had total after-tax returns of 56% over 941 business days.
• Hanjin KAL/KAL had total after-tax returns of 55% over 739 business days.
• Mid-cap pairs showed annual after-tax returns ranging from 19.0% to 41.6%, with the average return of 22.5% and standard deviation of 8.7%.

5. Sensitivity Test

• The strategy was tested for various parameters:
  • Opening position: Adjusting the sigma values (0.90 to 1.10) resulted in information ratios ranging from 2.28 to 2.58.
  • Closing position: Adjusting the sigma values (0.40 to 0.60) also showed information ratios from 2.44 to 2.58.
  • Maximum investment period: Testing from 12 to 16 days led to information ratios from 2.31 to 2.54.
  • Maximum loss: A 20% loss ceiling was implemented to control risk.

Key Information

• Value of holding company is calculated as:
  $$
  \text{Holding company market cap} = \sum (\text{subsidiary market cap} \times \text{ownership rate} \times \text{holding company discount rate}) + \text{other value}
  $$
• If the holding company only has a single major subsidiary, its share price can be approximated as:
  $$
  \text{Holding company share price} = \text{subsidiary share price} \times C
  $$
  where C is a constant reflecting the discount rate.
• Transaction costs are 0.8% (0.4% for tax and 0.4% for commissions), and after-tax returns are calculated by subtracting this from pre-tax returns.
• The strategy is event-driven and short-term, with volatility in share prices of both holding companies and subsidiaries influencing the timing of trades.
• Exclusions:
  • Holding companies with multiple subsidiaries (e.g., LG Corp, SK Corp).
  • Pharmaceutical holding companies due to high volatility and unstable share-price ratios.

Conclusion

The pair trading strategy for holding companies and their subsidiaries in Korea is profitable and efficient, especially for those with single major subsidiaries. The backtesting results show positive returns and stable performance, while the sensitivity test confirms the robustness of the strategy. However, regulatory changes may slow the transition to holding company structures, and frequent trading is necessary due to transaction taxes.