The answer to whether execution quality is improving, based on discussions at the Fixed Income Leaders Summit, is, ‘it depends’. For liquid, standardised instruments traded in modest size, execution quality has improved markedly over the past decade. For large block trades in less liquid credit, the picture is more complicated – and the tools that have transformed the former have not yet fully reached the latter.
The improvement in liquid markets is real and widely acknowledged. The effect of the exchange-traded fund (ETFs) ecosystem on continuous pricing has been marked. Pre-trade analytics are richer. Pricing feeds are more reliable. Portfolio trading has provided an effective mechanism for managing fund flows. The growth of all-to-all trading protocols has expanded the pool of potential counterparties. One experienced buy-side trader noted that the early months of portfolio trading required considerable comfort with new pricing models and execution analytics, but that data quality had improved to the point where algo-based pricing – inconceivable a few years earlier – is now operationally viable.
Yet the same speaker was candid about the limits. “I don’t know that you need more information to execute your more liquid business,” he observed.
In high-yield credit and other less liquid segments, the art of trading – the network of relationships, the ability to source a block, the judgment call about when to pick up the phone and when to just press a button to execute a block remains a personal choice, based on workload.
“It depends how lonely you are,” joked one trader.
Participants described a market in which high-touch and low-touch activity can be clearly separated, with human capital directed toward the former and technology handling the latter, to optimise efficiency of execution and liquidity management.
The proliferation of execution venues and protocols has also introduced a new source of friction, through protocol opimisation. With request for quotes (RFQs), portfolio trading, direct dealer connectivity, all-to-all venues and streaming liquidity all available simultaneously, choosing the right protocol for a given trade has become a meaningful decision in its own right.
Several panellists pointed to protocol selection as one of the most promising near-term applications for AI — a layer of intelligence that can assess current market conditions, available liquidity and bond characteristics to recommend an approach, without requiring the trader to process all of that information manually.
On axe data specifically, speakers reached a nuanced conclusion. Real-time inventory information from dealers works well for smaller trades where the data is reliable and actionable. For block liquidity, it is less dependable — dealers managing stretched resources find it difficult to keep inventory data live and accurate across multiple client feeds, and stale information erodes trust quickly. While the market is improving, but it has not yet reached the point where streaming liquidity can replace the phone call for a meaningful portion of credit flow.
However, broadly speaking, liquidity cost has fallen and depth has increased, just not evely across all fixed income instruments.
©Markets Media Europe 2025
