B3's Commitment to a Resilient CSD
B3, Brazil's leading market operator, recently executed a maneuver that typically goes unnoticed: moving the "registrar core" of its system without stopping market operations. On March 12, 2026, the second phase of its central securities depository (CSD) transformation was launched, utilizing VeriSafe® technology from Vermiculus, a cloud-native, microservices-based system. This development comes with a noteworthy detail, which, for any CEO or CIO, is more valuable than empty promises of innovation: this marks the second consecutive year of significant deployments, following the first phase in December 2024.
In market infrastructure, the epic narratives often miss the mark. It is not the one who announces the "new" that wins; it is the one who can change a structural component without disrupting settlement, ownership, and reconciliations. B3 and Vermiculus assert they have accomplished this through a gradual and "soft-launch" approach, operating in parallel with the legacy system to ensure stability. Viviane Basso, COO of B3 for issuers, deposits, and OTC, summarized this with language that intrigues me for what it omits: a “frictionless” transition and “immediate” benefits for participants, without dramatizing the change as if it were magic.
The second phase expands functional capabilities: settlement transaction processing, participant transfers, deposits, withdrawals, IPO support, and creation and redemption operations for Brazilian securities, including ETFs, deposit receipts, and fractional shares. In other words, the CSD transitions from a scalable pilot to a fully operational machine.
The CSD as the Backbone of the Market, Not an IT Project
A CSD is not just "another system." It is the registry that determines ownership, when ownership changes, and under what conditions an asset can move. If I illustrate it as a blueprint, it serves as a slab of transfer: receiving inputs from trading, clearing, custody, and corporate events, and distributing them to participants, issuers, and regulators. When that slab cracks, it does not merely cause market operations to "slow down"; it transforms into a more risky, expensive, and less trustworthy environment.
B3 is treating this element as it is: critical infrastructure. Phase two designates VeriSafe® as the “golden source” of depositor data, implying a master truth point for positions, movements, and records. In high-complexity trades, that detail determines whether reconciliation becomes an overnight process fraught with friction or a nearly continuous flow with limited exceptions.
The communiqué emphasizes elastic scalability and microservices architecture. I translate that into mechanics: instead of a monolithic machine where improvements require halting part of the engine, a set of interchangeable modules allows for scaling capacity in real-time. Vermiculus claims VeriSafe® handles hundreds of millions of accounts, investors, and daily transactions. While we lack volume figures for B3 in this announcement, the structural point is clear: the design is justified when user, asset, and operational growth stops being linear and begins to leap.
There is also an operational governance aspect. The phased deployment with parallel operation serves as a means of transferring risk from a "big cut day" to a controlled sequence of validations. Taraneh Derayati, CEO of Vermiculus, describes this as phased replacement combining stability with value delivery from the outset, precisely because it runs parallel to the legacy system.
The Less Glamorous Decision Was the Migration Channel, Not the Cloud
Many people read “cloud” and assume automatic modernization. I focus on the transition channel. A CSD is not replaced like a CRM system. Its launch is tied to settlement events, operational windows, integrations with trading systems, and regulatory tolerances. The announcement leaves two significant operational clues.
First: the “soft-launch” approach and parallelism with the existing system. That is a risk-reduction strategy by design. In architecture, it’s about reinforcement while under load, not after. It avoids the classic scenario where a total migration exposes edge defects that did not appear in tests.
Second: the first phase already yielded specific functional benefits. According to sources, it enabled a tenfold increase in capacity, extended opening hours, generated new reports for participants, and accelerated the rollout of new features. There are no figures regarding costs or returns, but there is something tangible: the system started paying its rent before it was “finished.” This is a critical distinction between a program that drains the budget and one that begins to reduce operational friction early.
The announcement also mentions integration with B3's trading system, BSX, for processing operations, settlements, and ownership changes, based on context gathered in the briefing. In terms of load, this integration poses a typical failure point. If the connection between trading and the depository is not robust, errors do not present as “bugs”; they surface as ownership discrepancies, rejections, and manual processes.
Henrik Rouet-Leduc, project manager of CSD at Vermiculus, claims that the project is “setting a new standard” through speed and efficiency while maintaining the timeline and delivering continuous value. As an analyst, I perceive that statement as a hypothesis to monitor rather than a verdict: standards are not set by the communiqué; they are defined by operational incidents, exception rates, and stability on peak days.
Applied Atomization in Infrastructure, Fewer Promises, and More Specific Loads
The mistake in many large transformations is attempting to please all stakeholders with a universal narrative. Here, I see a well-applied atomization, though it’s not named as such: B3 isn’t saying “we’re going to reinvent the market.” It lists specific loads that the new CSD must handle.
Phase two targets settlement transactions, participant transfers, deposits and withdrawals, IPOs, and the creation and redemption of instruments such as ETFs, deposit receipts, and fractions. That list reveals a strategy: first covering flows that concentrate operational risk and volume while simultaneously enabling products where elasticity matters. Fractions and ETFs often amplify counts of positions, events, and movements, pushing systems to their limits.
In parallel, the microservices architecture promises “fit” between function and capacity. If a specific module requires more resources due to an operation surge, that component gets scaled instead of inflating the entire platform. Rodrigo Nardoni, B3's CIO, framed the technical reason in 2023: flexibility and scalability for Brazilian demand and its future. That statement is sober and accurate: the market changes due to volume and variety of instruments, and a rigid CSD becomes a bottleneck.
There’s a second angle of atomization: the participant. In phase one, new reports for participants were mentioned. In a depository, reports are not “nice to have”; they reduce manual reconciliation and operational litigations while lowering transaction costs in the ecosystem's back office. It’s distributed value: not only does B3 operate better, but its members also lower friction and risk.
The Financial Mechanics Behind the Milestone, Fixed Costs Under Pressure
There are no details on contracts, total costs, or ROI in the sources. This compels analysis to be conducted without speculation. Still, the design hints at a clear financial intention: to transform a segment of market operation costs from rigid fixed expenses to a more elastic structure.
A legacy CSD tends to penalize with rising marginal costs: more volume requires more hardware, longer night windows, larger support teams, and more manual processes to address exceptions. The cloud-native and microservices approach, in theory, enables capacity scaling without proportionally replicating the entire operation. If this holds true, the benefit is not merely “modernity”; it’s the expansion of operating margins per transaction, especially on peak days.
There’s also an accounting and risk benefit that is rarely mentioned: speed of delivery. Vermiculus emphasizes agile methodology, frequent releases, and rapid time-to-market. For a stock exchange, this affects two lines: the ability to list and support new products with less delay and a lower backlog of pending work that later explodes into giant projects.
The primary risk of this type of transformation typically resides in integration and coexistence. Running in parallel reduces the risk of discontinuity, but it necessitates reconciling two sources of truth until the replacement is complete. If the new system becomes the “golden source” too quickly without the edges being sealed, discrepancies arise. If the legacy shutdown lingers too long, the cost of double operation hits hard. The announcement suggests the phased design aims to navigate between those two walls.
For Vermiculus, the value is reputational and commercial. Executing one of the largest CSD transformation projects, as described in the sources, serves as a calling card for markets looking to modernize without exposing themselves to public failures. In critical software, that reputation proves more defensible than any marketing campaign.
The Lesson for Leaders: Replace the Engine Without Turning Off the Lights
The milestone for B3 should not be seen as a victory for the provider or as a cloud case study. It serves as an operational reminder: in complex markets, transformations that endure are those designed as a bridge in use, with spans that change without shutting down traffic.
Phase two in production suggests that B3 opted for engineering discipline over grand gestures: phased deployments, parallel operations, early delivery of capacity and functions, and a modular design promising scalability without inflating the entire structure. There is public information lacking on costs, incident metrics, and the schedule for total legacy shutdown, and this absence prevents celebrating a total victory. What can be asserted based on what has been published is that the project is avoiding the most common pitfall of modernization: turning a necessary change into a binary gamble.
Companies do not fail due to a lack of ideas; they fail when the pieces of their operational model do not fit together to produce measurable value and sustainable cash flow.
