At industry gatherings such as the Fixed Income Leaders Summit, electronification, automation, AI and smarter execution remain central themes.

And rightly so. Electronic pricing, RFQs, portfolio trading, automated workflows and straight-through processing have delivered genuine improvements.

But I think we're overlooking a more fundamental problem.

Are we transforming fixed-income trading, or simply automating the complexity we've created?

Consider the role of trading venues.

Their networks of liquidity providers and buy-side participants are enormously valuable. So are their trading protocols, price discovery and ability to bring buyers and sellers together.

But somewhere along the way, many venues have also become increasingly elaborate workflow platforms.

Order staging, execution management, analytics, automation and increasingly sophisticated interfaces.

And much of that functionality is replicated across competing venues, each with its own workflows, protocols and integration requirements.

Meanwhile, the buy-side invests in OMSs and EMSs, sometimes at considerable expense, just to connect and interact with these fragmented ecosystems.

The result?

Multiple layers of technology, overlapping functionality, expensive integrations and traders still navigating between systems to get things done.

We've become very good at automating individual pieces of a complicated operating model.

But who's questioning whether all those pieces need to exist?

Perhaps we should revisit what we actually need from a trading venue.

Make access to liquidity as frictionless as possible. Provide reliable, actionable market information. Support efficient price discovery and execution, with the necessary controls and transparency.

And importantly, improve the quality of the signals being provided.

Axe quality remains a concern for the buy-side. So do stale indications, inconsistent liquidity signals and the gap between what appears available and what can actually be executed.

More automation won't solve poor-quality information. It may simply propagate it faster.

The buy-side doesn't fundamentally trade orders. It implements investment mandates and strategies.

Orders are one mechanism for achieving those objectives.

The intelligence to evaluate market conditions, select execution strategies, navigate liquidity and determine how to interact with venues should increasingly sit with the buy-side, not be fragmented across every venue it connects to.

That doesn't make venues less important.

It makes their core value proposition more important.

The objective shouldn't be to automate every step we've inherited. It should be to question which steps need to exist, who should own them and where they add genuine value.

Perhaps the next breakthrough in fixed-income electronification isn't another layer of automation.

It's removing the unnecessary layers we've already built.