The tracking performance of equity exchange traded funds: a consideration of fund replication strategy, fund domicile and crisis period.

Abstract

An Exchange Traded Fund (ETF) is an investment vehicle that issues securities that are essentially claims on an underlying pool of assets. Tracking error measures, the ability of traditional passive ETFs to replicate the returns of their respective underlying index accurately. This measure is commonly reported for all funds with a mandate to replicate some benchmark index. Despite the primarily passive nature of ETFs, fund managers can apply active investment management techniques to them. The application of active management to these funds may include the respective index holding an actively selected basket of securities or entering derivative contracts that deliver the performance of an index, or some mixture of the two. The importance of looking at the passive and active characteristics of funds corresponds to the replication strategies followed by ETFs. Here replication refers to the concept of mirroring the returns of a benchmark index with the returns of an ETF. Bloomberg Professional’s categorisation of replication strategies shows that ETFs replicate their benchmark indices using the following strategies: full physical, stratified sampling, optimization, synthetic and leveraged replication. This study analyses the tracking performance of 52 equity-backed ETFs, focusing on replication strategies, fund domicile, and crisis period. Four methods of tracking error estimation are applied to the ETF sample which have an inception date before 1 January 2006 or 1 January 2012 for the fund replication and domicile analyses due to the observed lack of ETFs following certain replication strategies and domiciled in emerging markets with an inception before 2006. For the crisis analysis the research period spans 18 years to account for the documented price impacts the 2008/2009 Global Financial Crisis (GFC) and the COVID-19 pandemic had on various indices and their replicating funds. We find that overall partial physically replicated ETFs provide superior tracking performance. Full physically replicated ETFs exhibit the highest level of tracking error. Synthetic ETFs demonstrate superior tracking performance in comparison to full physical ETFs. Considering the same underlying benchmark index, leveraged ETFs with lower leverage multipliers exhibit lower levels of tracking errors than their counterparts. ETFs domiciled in developed markets limit tracking errors to a greater extent than emerging market ETFs and synthetic ETFs show superior tracking performance when tracking emerging market indices. All fund replication strategies (noting that leveraged ETFs are excluded in this section of the analysis) for both emerging and developed market ETFs show increases in tracking error during the GFC and the COVID-19 pandemic. Optimized ETFs exhibited the highest increase in tracking error during the GFC while full physically replicated ETFs exhibited the highest increase during the pandemic. Synthetic ETFs showed the most resilience to the effects of the pandemic. Emerging ETFs exhibited higher increases in tracking error during both crises in comparison to those in developed markets. This study provides both institutional and individual investors with valuable knowledge on the consideration of fund replication strategy, fund domicile and the performance effects of documented crisis periods when selecting an appropriate ETF. Investors and portfolio managers are provided with relevant insights on which type of ETF replication to follow in countries with different development levels and during volatile market periods. Partial physical replication provides superior tracking performance when focusing solely on replication strategy. Synthetic ETFs are recommended when investing in emerging market indices and aiming to minimize exposure to volatile markets marked by crises.