The bond markets are a massive network of players and protocols that trade trillions of dollars a day. A high-stakes game is continuously being played by buyers and sellers seeking to buy low and sell high.
Stories are told to explain the status quo and make sense of what’s changing. Here are ten statements I’ve heard that are unraveling with the test of time:
- Bonds are too complex to price automatically
Traders used to take comfort in the complexity of the various issuances, the sheer number of CUSIPS 1M+, and the lack of observable trades which would make it difficult for algorithms to price. Alas, taking quantitative inputs into a model and outputting a price, without the “gut feel” adjustment of a human, has been accelerating with most desks having an “algo” and automated pricing already responding to 30% 1+ of the IG market and new algos coming online in the muni market. While transactions at a certain scale and complexity require skilled coordination, the goalposts keep shifting as confidence in automation grows. - Clients depend on dealers for liquidity
A prominent club of dealers was once the center of the bond universe, setting prices for clients and taking the other sides of trades. Over the last decade, we have witnessed the rise of electronic trading, the decline of dealer risk-taking, and the entrance of non-bank liquidity providers. Most importantly, new electronic protocols have gained traction, allowing clients to respond to clients for the first time. Over one-third 2of the volume on MarketAxess is done on the all-to-all network, implying a rising share flows away from traditional dealers towards a longer tail, and between clients themselves. - Retail doesn’t want to own bonds
Retail is often too narrowly defined as a self-directed individual acting like a portfolio manager. The reality is much wider in scope: retail flows are driven by self-directed investors, brokerage products, advisors, and even separately managed accounts by large asset managers. As electronic trading grows, costs to trade small lots have come down, and retail demand has never been higher. Recent Fed data shows significant household holdings in direct debt securities of 6.3T, higher than fund holdings of 5.6T3 . Products continue to evolve to improve end-investor access to portfolios. - The blockchain will reduce costs
The blockchain has shown the financial world that technology can enable fast, cheap, and always-available atomic trading/settlement, and the financial world has responded with efforts to reduce settlement times (T+1), increase trading hours, and reduce costs. Defecting from the netting benefits of settling through the traditional rails at clearinghouses would be very challenging for any new system to displace. Nor does the market need the crucial value prop of blockchain, which is decentralization. The traditional financial system will continue to adapt and even subsume the most relevant parts of blockchain assets, such as swallowing BTC into centralized managed funds. Tokenization will serve more as a metaphor for digital improvement rather than a movement for a shadow financial system. - Dealer markets should be separate from client markets
Dealers used to be able to transact with one another through interdealer channels that were obfuscated away from clients, bifurcating liquidity into dealer pools and client pools. Traditional boundaries continue to bleed away and now clients can participate directly or indirectly in previously out-of-reach channels such as the traditional wholesale interdealer brokers or order books. The resistance sought to protect margins and rents, but the march is in full force towards eliminating boundaries in an increasingly interconnected market. - Dependence on bond algos and ETFs will exacerbate a liquidity crisis
The speculation of how poor liquidity will be in a crisis is often pondered, though it’s a bit of a circular conjecture. A crisis, by definition, is a surge in volatility leading to increased trading costs and worse liquidity. There is always a price for a trade, and in free markets the best-suited participants step in to capitalize on opportunities. Electronic markets and ETFs make the sell-off experiences more visceral since prices are continuously observable. More liquidity providers of different types have resulted in decreased spreads, and both the frequency and magnitude of crises continue to decrease, as seen in the long-view chart of credit spreads. - Portfolio trading is the wave of the future
The portfolio trading protocol has ballooned in the last few years, with participants benefiting from the ease of executing many line items in batch. Platforms tout the benefits of their new protocol, and savvy dealers can intermediate blocks of risk immediately. Are execution costs actually superior to competing individual line items? Are desks subsidizing their liquidity provision in PTs to compete for share, distorting outcomes? Or is this a clever way to shift accounting costs between bonds? There’s been fairly light analysis or theory as to why a bespoke basket should be traded all or none bilaterally, and is unlikely here to stay as the buy side is able to save on both workflow and costs with automated competitive line item execution. - Block trading saves cost
Preferring to transact bonds in large blocks is still the norm, but the market is gradually shifting towards slicing up risk into smaller transactions. Block trades in equities are now the exception as risk is often parcelled out and routed by execution algos over time. Bond execution algos are nascent today, but they will grow in relevance as the market matures past initial stages of being electronic. The average corporate bond trade size is down 35% over the decade 4and will continue. As algo liquidity provision continues to grow, the buy side will respond with algos as well, and the robots will trade with the robots. - Bond issuance must be done at scale
The primary process is still a concerted effort across stakeholders to price and distribute billion-dollar deals. Ultimately, the primary market will follow the secondary market to prefer more issuance at smaller sizes, leading to an even greater proliferation of CUSIPs. The algos won’t complain when they price smaller quanta of risk since their marginal cost is nil. Portfolios will benefit from a greater continuum of securities, allowing further customization. Issuers will be more deeply integrated into the secondary market to run open competitive issuance and buy-back processes. - Technology is an enabler, not a replacer
This refrain is often repeated at conferences discussing electronic trading and AI to mitigate natural fears with changes, but does not adequately acknowledge the pivots that need to happen for practitioners to stay productive in a changing world. The industry is ruthless in its search for efficiency, and you can juxtapose the pictures of older trading floors with today to appreciate living through a transforming construct of market interactions. The curious and avid might find new leverage to supercharge their goals, and the resistors and slow to adapt will be vulnerable. The pace of change also keeps rising and is harder to intuit.
- https://www.greenwich.com/press-release/bond-dealers-increasingly-automate-trading ↩︎
- https://investor.marketaxess.com/news/news-details/2025/MarketAxess-Announces-Trading-Volume-Statistics-for-December-and-Fourth-Quarter-2024/default.aspx ↩︎
- https://fred.stlouisfed.org/series/BOGZ1FL154022375Q, https://fred.stlouisfed.org/series/BOGZ1FL154022075Q ↩︎
- https://www.greenwich.com/market-structure-technology/equitization-bond-market-real#:~:text=Bond%20traders%20don’t%20like,bond%20market’s%20equitization%20a%20reality. ↩︎
