It’s the Economy, Stupid.

GDP Weighted Government Bond Indices

In the last 17 years, Bond ETFs have seen a tremendous growth trajectory. According to a study from EY in 2017, the total assets under management for fixed income ETFs totaled to 3.5 trillion USD in 2016 and is expected to grow to 7.6 trillion USD in 2020  (EY, 2017). In 2002, iShares launched the first 4 bond ETFs, all concentrating on the USD Market  (Murphy, 2019).  With their offering in the USD Treasury Market and the USD Corporate Investment Grade Market, iShares for the first time enabled their clients an uncomplicated access to one of the largest capital markets in the world via a single exchange traded security.

Most bond indices follow a market cap weighting approach, meaning that the weight of a bond within a given index is determined by the product of a bond’s price and amount outstanding. If one wants to measure the performance of a given debt market, this weighting method is correct, as it reflects the true performance of the market  (Brown, 2002).

However, with the rise of investable market weighted security indices, the criticism of such indices and the assets tracking them has increased as well. Robert D. Arnott was among the first ones to voice his concerns about purely tracking market weighted security indices. He argues that indices, which don’t follow the pure market cap weighting scheme but other weighting approaches based on fundamental company data, are more mean-variance efficient  (Arnott, Hsu, & Moore, 2005).

In our white paper, we discovered the following key takeaways:

  • For developed markets, a GDP weighted Government Bond index outperforms a standard market cap weighted Government Bond index.
  • A GDP weighted Government Bond index invests in higher quality borrowers (better ratings) and exhibits a lower duration.
  • The main return driver for GDP weighted and market cap indices is the interest rate risk. Currency effects have a minor contribution to the performance.

To read the full white paper, please click on the link below: