Concentration in the market of authorized participants of US fixed-income exchange-traded funds

Introduction

The US market for fixed-income exchange-traded funds (FI‑ETFs) has grown rapidly over the past 10 years—by 30 percent annually, reaching US$990 billion in net assets in September 2020. FI‑ETFs make it easy for investors to invest in the bond market because, unlike bonds, FI‑ETF shares can be bought and sold on exchanges throughout the day and tend to be highly liquid (Johnson 2019). Hundreds of FI‑ETFs provide cheap and immediate access to a variety of government and corporate bond portfolios.

We focus on authorized participants (APs) in US-listed FI‑ETFs. APs are critical to the primary market because they have the exclusive right to create and redeem ETF shares.1 Data from the US Securities and Exchange Commission show a high level of AP concentration among FI‑ETFs (SEC 2019):

  • three APs performed 82 percent of gross creations and redemptions of FI‑ETF shares in 2019
  • these same three APs were prevalent across almost all FI‑ETFs

In contrast, we find a more diverse group of active APs in the equity ETF market. In 2019, eight APs accounted for 80 percent of gross creations and redemptions of equity ETF shares in the United States, and the largest AP had a market share of less than 20 percent. The high concentration of APs in the FI‑ETF market may simply reflect the fact that intermediation in the fixed-income market is dominated by a few players (Di Maggio, Kermani and Song 2017; Li and Schürhoff 2018).

The high level of AP concentration in the FI‑ETF market raises questions about the resilience of market liquidity in times of stress. How much do liquidity conditions of FI‑ETFs rely on a well-functioning primary market—where APs create and redeem ETF shares—to keep share prices close to the funds’ net asset value? What would happen in an extreme scenario where the few active APs stop creating and redeeming FI‑ETF shares?

The COVID‑19 outbreak sheds light on this question. In March 2020, the few active APs created and redeemed fewer FI‑ETF shares. As a result, most FI‑ETFs behaved like close-ended funds. However, we find that the trading volume of FI‑ETF shares surged during this period. FI‑ETF investors continued experiencing high levels of liquidity in the secondary market despite the reduced AP activity.

Authorized participants across equity and fixed-income ETFs

We use data from the US Securities and Exchange Commission’s Form N‑CEN for the dollar value of ETFs created and redeemed by APs (SEC 2019). Our sample includes 1,655 equity and fixed-income ETFs in 2019. These funds represent about 87 percent of US-listed ETFs and 99 percent of total ETF market value (Table 1).

Table 1: Equity and fixed-income exchange-traded fund market coverage from our sample in 2019

US equity ETFs
Sample US equity ETF market Sample coverage (percent)
Number of ETFs 1,312 1,512 87
Value of ETFs (US$ billions) 3,420 3,432 99
Fixed-income ETFs
Sample US FI-ETF market Sample coverage (percent)
Number of ETFs 343 381 90
Value of ETFs (US$ billions) 803 809 99

Note: ETF is exchange-traded fund; FI is fixed-income.
Sources: US Securities and Exchange Commission Form N-CEN and Morningstar Direct

Chart 1 shows the share of total gross creations and redemptions of equity ETFs and FI‑ETFs in 2019 by each AP. We find that the group of equity ETF APs is diverse:

  • 36 APs created and redeemed shares
  • the largest AP had less than 20 percent market share
  • 8 APs accounted for 80 percent of gross creations and redemptions of equity ETF shares

These findings about the equity ETF market are consistent with recent research on the ETF ecosystem in the United States. (BlackRock 2019).

However, in the fixed-income market (left panel), APs are much more concentrated, with three APs accounting for 82 percent of total gross creations and redemptions in 2019.

Chart 1: Market of authorized participants is much more concentrated in fixed-income (FI) exchange-traded funds (ETFs) than in equity ETFs

Chart 1: Market of authorized participants is much more concentrated in fixed-income (FI) exchange-traded funds (ETFs) than in equity ETFs

Source: Form N-CEN Filings Data—Securities and Exchange Commission  Last observation: Fiscal year-end 2019

Chart 2 shows that these three APs were prevalent in most FI‑ETFs in our sample. Of the 331 FI‑ETFs that had creations or redemptions, 298 had 50 percent or more of their gross creations and redemptions performed by these top three APs (light blue bars). Furthermore, we find that these APs had the most activity regardless of the investment style, age or sponsor of the fund.

Chart 2: Top three authorized participants are the most active in 90 percent of fixed-income exchange-traded funds (FI-ETFs)

Source: Form N-CEN Filings Data—Securities and Exchange Commission  Last observation: Fiscal year-end 2019

Chart 3 shows that most of the 51 APs registered in the FI‑ETF market are inactive. The blue line indicates the number of FI‑ETFs for which each AP is authorized to create or redeem shares, while the yellow bars reflect the total level of activity in dollar value for each AP. Often, FI‑ETFs have many APs listed (182 FI-ETFs, on average), but most APs were inactive, implying that only a small number of these APs actually created and redeemed shares.

Chart 3: Most authorized participants in the fixed-income exchange-traded fund (FI-ETF) market are inactive

Source: Form N-CEN Filings Data—Securities and Exchange Commission  Last observation: Fiscal year-end 2019

Drivers of authorized participant concentration

Several reasons could explain the high level of concentration of active APs in the FI‑ETF market. First, the characteristics of the fixed-income market favour a high concentration of intermediaries. In particular, trading in the fixed-income market requires specialized infrastructure and expertise. Di Maggio, Kermani and Song (2017) find that the top 50 dealers (out of thousands) account for almost 80 percent of trading activities in the US corporate bond market. Li and Schürhoff (2018) find similar evidence in the US municipal bond market.

A second and related reason could be that a small number of APs act as intermediaries to create and redeem shares on behalf of numerous market participants (Arora et al. 2020). For example, a principal trading firm that acts as a market maker in the secondary FI‑ETF market could use an AP to create or redeem shares and fulfill client orders. Insights gathered through market intelligence confirm that this practice is common in the United States.

Finally, it may also be that some bank APs have chosen to reduce their footprint in the bond market, partly due to changes to banking regulations after the global financial crisis.

Implications of authorized participant concentration for fixed-income exchange-traded fund market liquidity—lessons from COVID‑19

The high level of AP concentration of the FI‑ETF market can raise questions about the resilience of FI‑ETF liquidity in the primary and secondary markets in times of stress. It is estimated that about 80 percent of total FI‑ETF trading volume comes directly from the exchange (ICI 2018). It is unknown how much market liquidity conditions of FI‑ETFs rely on a well-functioning primary market where APs are creating and redeeming ETF shares to keep share prices close to the net asset value of the funds. What would happen to FI‑ETF market liquidity in an extreme scenario where the few active APs stop creating and redeeming ETF shares?

The COVID‑19 outbreak provides an opportunity to shed light on this question. In March 2020, many FI‑ETFs traded at large discounts relative to their net asset value. Chart 4 shows that these discounts were:

  • significant, around hundreds of basis points
  • persistent for several weeks
  • widespread across government, municipal and corporate credit ETFs

Chart 4: Fixed-income exchange-traded funds traded at large discounts in March 2020

Note: Annoucement corresponds to the Primary Market Corporate Credit Facility (PMCCF) and Secondary Market Corporate Credit Facility (SMCCF) Federal Reserve intentions to purchase investment-grade bonds and FI-ETFs. Implementation represents the first operation date.
Sources: Bloomberg and Bank of Canada calculations Last observation: September 1, 2020

The large and persistent discounts can be attributed in part to a decline in AP activity. When bond market liquidity was impaired in March 2020:

  • costs of carrying a large bond inventory rose
  • net asset values were increasingly affected by stale bond prices

As a result of both effects, APs became less willing to add bonds to their balance sheets and execute the arbitrage despite the large FI‑ETF discounts. The discounts thus remained large for several days, and FI‑ETFs behaved like closed-end mutual funds.

Despite the discounts, we find that liquidity in the FI‑ETF secondary market surged during this period. Chart 5 shows that, compared with February, the trading volume on exchanges for FI-ETFs that track municipal bonds surged by 245 percent in March, whereas it increased by 67 percent for FI‑ETFs that track corporate bonds. These results suggest that market liquidity conditions were resilient in the FI‑ETF market throughout the crisis. Moreover, the results suggest that FI‑ETF prices continued to provide a real-time view of the value of the underlying bonds during the crisis, as explained by Aramonte and Avalos (2020). In contrast, the net asset value of FI‑ETFs with less liquid holdings provided only a lagged indication of their “true” value due to poor bond trading activity.

Chart 5: In March, on-exchange trading volume for fixed-income ETFs spiked

Sources: Bloomberg and Bank of Canada calculations Last observation: September 1, 2020

We note that the announcement by the US Federal Reserve on March 23, 2020, to buy investment-grade FI‑ETF shares also played a role in stabilizing bond markets. FI‑ETF prices rebounded sharply following the announcement. Corporate bond ETFs even traded at premium after the announcement, suggesting that the Federal Reserve’s prompt actions after the onset of the COVID‑19 crisis helped improve market functioning and liquidity conditions in bond markets.

Conclusion

We document a high level of concentration among APs of FI‑ETFs listed in the United States. In 2019, three APs performed 82 percent of gross creations and redemptions of FI‑ETF shares, and their prevalence was present across a wide range of FI‑ETFs. These results contrast with equity ETFs, where AP activity is more diverse.

We also study the implications of AP concentration on the resilience of FI‑ETF market liquidity in the context of COVID‑19. Our analysis reveals that reduced AP activity led FI‑ETFs to behave like closed-end funds. Nevertheless, we find that volumes in the secondary market surged at the peak of the crisis, which suggests that:

  • liquidity conditions remained resilient
  • FI‑ETFs continued to facilitate price discovery during this period of market turbulence.

Appendix

Data

We collected new data from the US Securities and Exchange Commission’s Form N‑CEN (SEC 2019), which became available in 2018. Our sample of funds includes 1,655 of the 1,893 ETFs listed in the United States in 2019. The other 238 funds are not included in our sample because they had not submitted their Form N‑CEN before we collected the data at the end of March 2020.

  1. 1. See Box 1 in Foucher and Gray (2014) for more details on the creation and redemption process.[]

References

  1. Aramonte, S. and F. Avalos. 2020. “The Recent Distress in Corporate Bond Markets: Cues from ETFs.” Bank of International Settlement Bulletin No. 6.
  2. Arora, R., J.-S. Fontaine, C. Garriott and G. Ouellet Leblanc. 2020. “Will Exchange-Traded Funds Shape the Future of Bond Dealing?” Bank of Canada Staff Analytical Note No. 2020-16.
  3. BlackRock. 2019. “iShares Investigates: The ETF Ecosystem Part 2: Authorized Participants by the Numbers.”
  4. Di Maggio, M., A. Kermani and Z. Song. 2017. “The Value of Trading Relations in Turbulent Times.” Journal of Financial Economics 124 (2): 266–284.
  5. Foucher, I. and K. Gray. 2014. “Exchange-Traded Funds: Evolution of Benefits, Vulnerabilities and Risks.” Bank of Canada Financial System Review (December): 37–46.
  6. Investment Company Institute (ICI). 2018. 2018 Investment Company Fact Book: A Review of Trends and Activities in the Investment Company Industry, 58th ed.
  7. Johnson, B. 2019. “ETFs Add Liquidity to the Bond Market, Not Risk.” Morningstar. October 4.
  8. Li , D. and N. Schürhoff. 2018. “Dealer Networks.” Journal of Finance 74 (1): 91–144.
  9. US Securities and Exchange Commission Subcommittee on ETFs and Bond Funds (SEC). 2019. “Report on the Design of Exchange-Traded Funds and Bond Funds—Implications for Fund Investors and Underlying Security Markets Under Stressful Conditions.”

Acknowledgments

We thank Jean-Philippe Dion, Jean-Sébastien Fontaine, Virginie Traclet and Jun Yang for helpful comments and suggestions. We are thankful to Narahari Phatak and James McLoughlin for helpful discussions about the data. Finally, we are grateful to Nicole van de Wolfshaar and Meredith Fraser-Ohman for editorial assistance.

Disclaimer

Bank of Canada staff analytical notes are short articles that focus on topical issues relevant to the current economic and financial context, produced independently from the Bank’s Governing Council. This work may support or challenge prevailing policy orthodoxy. Therefore, the views expressed in this note are solely those of the authors and may differ from official Bank of Canada views. No responsibility for them should be attributed to the Bank.