Will exchange-traded funds shape the future of bond dealing?

“Just-in-time” could come to the bond market

In the 1970s, Toyota adopted a new system to build its cars—one that would become well-known and widely adopted across industries: just-in-time (JIT) production. The essence of the JIT approach is that manufacturers build a product, such as a car, only as they need it. Historically, car makers avoided the risk of a delay in operations by stocking car inventories at warehouses. After the adoption of JIT, car makers were able to keep a much smaller inventory on hand.

Now, a similar JIT approach could be coming to the business of bond dealing. Traditionally, dealers keep bonds in their inventory, anticipating that investors will buy them but not knowing exactly which ones. Like maintaining a complete stock of finished cars, carrying large inventories is costly, even for bond dealers. JIT may be seen as a welcome change.

The rise of exchange-traded funds (ETFs) makes JIT possible in bond markets. ETFs are securities traded on an exchange, just like stocks, that entitle the bearer to a share in a pool of assets (such as stocks or bonds). For example, a fixed-income ETF might entitle its bearers to a share of a pool of 100 bonds. We find that a dealer can use bond ETFs as a warehouse to meet investor demand to buy and sell bonds. Similar to a car maker using JIT production, the dealer can reduce its inventories of “parts” and order them from its “suppliers” through a JIT approach.

Admittedly, ETFs make up only a small share of the Canadian asset management industry. Our analysis shows that ETFs in Canada are not yet used as warehouses to a large extent. But the practice is growing in the United States. These changes to the way dealers handle bonds can transform the market by:

  • improving prices
  • reducing the costs of large trades
  • making it easier for issuers themselves to borrow funds

ETFs enable dealers to warehouse bond inventory

ETF warehousing is when dealers use bond ETFs to deposit and withdraw—or push and pull—bonds instead of using inventory. Figure 1 compares the ETF warehousing model with the traditional bond dealer model:

  • Typically, a dealer distributes bonds by keeping them in inventory until it finds a client that wants them. If the dealer does not hold bond inventory, it cannot fulfill client orders promptly. As a result, it might lose a trade to another dealer.
  • The ETF warehousing model works differently. Instead of holding individual bonds in inventory, a dealer relies on a pool of bonds held within an ETF—an outside warehouse. Using a JIT approach, the dealer could pull bonds from the ETF to fulfill client orders or push bonds acquired from clients to the ETF.

Figure 1: Exchange-traded fund warehousing is a new way to deal bonds

Note: ETF is exchange-traded fund.

ETF warehousing is most valuable to bond dealers when custom exchanges are performed. This occurs when an ETF agrees to create or redeem ETF shares for a group of specific bonds that a dealer has negotiated with the fund. If an ETF manages a large pool of bonds, it can be flexible about what it buys and sells because the pool is large enough to tolerate small deviations from the fund’s index or strategy. As compensation, ETFs can charge a fee for custom baskets, or they can tactically buy or sell bonds at favourable prices.

The emergence of the ETF warehousing model goes hand in hand with the recent growth of portfolio trades. In bond markets, portfolio trades happen when a basket of bonds is exchanged in a single transaction. The rise of ETFs, which now hold thousands of bonds, has created more flexibility in how both dealers and clients can execute portfolio trades. For example, if a client can put together a group of bonds that matches the needs of an ETF, the dealer can facilitate the trade in a single transaction at a lower cost to the client.

We give a proxy for the magnitude of ETF warehousing in the United States based on the volume of portfolio trades taking place. Chart 1 shows that portfolio trading is small but growing rapidly. In 2019, portfolio trading accounted for 3.8 percent of weekly trading volumes in US fixed-income markets. The 2019 US Securities and Exchange Commission ETF rule, which allows all ETFs to conduct custom exchanges, could make ETF warehousing more common. In addition, bond ETFs may become more willing to engage in custom exchanges as they grow their assets under management

Chart 1: Exchange-traded fund warehousing, proxied by portfolio trading, is small but growing rapidly in the United States

Note: The analysis for identifying potential bond warehousing transactions is inspired by an approach outlined in "Wall Street's New Balance Sheet is an ETF" by Lisa Abramowicz (May 25, 2016).
Sources: Bloomberg, FINRA and Bank of Canada calculationsLast observation: 2019

Our estimate of the size of portfolio trading is consistent with other studies. Morgan Stanley estimated the size of portfolio trading at US$88 billion in 2019 (Surane and Leising 2019). Citigroup, which focused on the investment-grade bond market, estimated it at roughly US$66 billion (Banerjee and King 2019).

ETF warehousing could bring efficiencies and drawbacks

The growth of ETF warehousing could bring benefits to the bond market, though we also foresee some vulnerabilities. In Table 1, we present the potential advantages and disadvantages of ETF warehousing for bond markets.

Improved efficiencies in bond distribution

ETFs can hold inventory at a much lower cost than dealers, so dealers can save money by passing bonds on to ETFs to hold. Bond ETFs agree to be flexible about the bonds they hold and, in return, receive favourable prices, collect a fee or obtain bonds they would rather have. The cost savings to dealers are amplified by recent changes to banking regulations, which increase dealers’ costs of holding bonds in inventory. This may further drive dealers to use ETFs for warehousing.

Reduced segmentation

Segmentation occurs when markets for similar goods trade apart from each other and at different prices. ETFs could reduce the large degree of segmentation that exists among the many bonds that have only minor differences. ETFs are tolerant of small differences in bonds, such as maturity dates or coupons, because a large portfolio of corporate bonds can have the same risks and returns as the index. Since ETFs are more willing than other participants to trade different bonds as substitutes, they could help bring markets for these bonds closer together.

Increased price discovery in bond markets

When a bond is placed in an ETF, dealers can offer quotations for that bond. A corporate bond that was issued a few years ago tends to be stale, and many dealers do not advertise prices at which they would like to buy and sell the bond. With the rise of bond ETFs, fund managers can elicit quotes for the bonds in the fund. This means more traders reporting more views on the same group of bonds, which benefits price discovery.

Lowered transaction costs for portfolio trades

Placing more bonds in transparent and public places could also make the execution and pricing of large portfolio trades more efficient. A portfolio trade occurs when multiple bonds are traded in a single transaction. ETF warehousing makes these trades easier because dealers can move a group of bonds into and out of an ETF rather than having to find a place for each bond individually. Going forward, this greater liquidity might even increase the volume and frequency of portfolio trades.

More complex bond market

ETF warehousing introduces more individuals into the bond dealing process. Whereas before dealers simply interacted with each other, now they communicate with ETF managers and intermediaries (for more on the structure of ETFs, see Foucher and Gray 2014). This introduces new connections, and thus more opportunities for error—operationally and otherwise. For example, if a cyber incident leads to an outage at the ETF, its pool of bonds could become cut off from the rest of the market. The question remains whether the ETF warehousing model of bond distribution is as resilient as the traditional warehousing model that it could replace. This could be a subject for future research.

Reduced access to individual bonds

We also see potential drawbacks to efficiency. In a scenario where most bonds are held in ETFs, it could become harder for clients to buy and sell individual bonds because ETFs typically do not create or redeem shares for single bonds. Still, individual bonds might remain accessible if ETFs accommodate requests to create or redeem ETF shares for relatively few or single bonds. Insights gathered through market intelligence indicate that ETFs, at least those in the United States, have become more accommodating.

ETF warehousing is small but could be transformative

ETF warehousing remains small in Canada but could gain more traction due to its cost savings. It could move bonds out of individual dealers’ inventories and into places that are transparent and more public. Dealers could then source the bonds they want, when they need them—a just-in-time approach. The cost-saving benefits are also being amplified by the recent regulatory changes noted above. Through competition among dealers, some of these savings could be passed on to investors and, by improving liquidity, would help issuers borrow funds at lower costs.

In sum, ETF warehousing could be an effective way to manage bond dealing. But the net effects on bond liquidity and market stability are still unknown, so the Bank of Canada will continue to monitor its development. As we have noted, ETF warehousing offers both advantages and disadvantages. While it does bring efficiencies, it could create difficulties in buying and selling specific bonds. One particular vulnerability of interest is the new connections ETF warehousing creates, with new intermediaries becoming pivotal to bond dealing. Further examination and research is warranted.


  1. Abramowicz, L. 2016. “Wall Street’s New Balance Sheet Is an ETF.” Bloomberg Opinion, May 25, 2016.
  2. Banerjee, A. and M. King. 2019. “The Coming Revolution in Credit Portfolio Trading.” Credit Research Citigroup, November 2019.
  3. Foucher, I. and K. Gray. 2014. “Exchange-Traded Funds: Evolution of Benefits, Vulnerabilities and Risks.” Bank of Canada Financial System Review (December): 37–46.
  4. Securities and Exchange Commission (SEC). 2019. “SEC Adopts New Rule to Modernize Regulation of Exchange-Traded Funds.” Press release (September 26).
  5. Surane, J. and M. Leising. 2019. “Bond Trade That’s Gone from Zero to $88 Billion in Two Years.” Bloomberg Future Finance, November 18, 2019.


We thank Jean-Philippe Dion, Virginie Traclet and Jun Yang for helpful comments and suggestions. We are also thankful to Manu Pandey for excellent research assistance. Finally, we are grateful to Meredith Fraser-Ohman and Nicole van de Wolfshaar for editorial assistance.


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.

DOI: https://doi.org/10.34989/san-2020-16

On this page
Table of contents