I've been running workloads on both AWS and Azure for the past four years. Not as a neutral evaluation exercise, but as a practical reality: we ended up on both because different parts of the business made different decisions at different times, and now I maintain infrastructure on both clouds.
I want to share what I've actually learned, which is different from most cloud comparison articles because it comes from genuinely operating both, not from benchmarks or feature lists.
First: the cloud wars narrative is exhausting
Every few months there is a new round of "AWS vs Azure vs GCP" content that treats choosing a cloud provider as if it is the most important infrastructure decision a company can make. In my experience it is not even close to the most important decision. It is roughly as important as choosing which brand of server rack to use.
Both AWS and Azure are mature, reliable, and capable of running almost anything you'd want to run in financial services. Both have had significant outages. Both have extensive compliance certifications. Both will negotiate enterprise agreements that include security terms your legal team needs.
The decision of which cloud you are on matters much less than how well you use whichever one you are on.
That said, there are real differences, and they matter in specific contexts.
Where Azure actually wins for us
We are an investment management firm. We use a lot of Microsoft software. We have Active Directory, we use Excel everywhere (I know, I know), and we have several products in the Microsoft ecosystem that matter to the business.
Azure's integration with all of that is genuinely better than AWS's. When we need our cloud workloads to authenticate against the same Active Directory that our on-premise environment uses, Azure makes this almost trivially simple. AWS can do it - there's AWS Managed Microsoft AD and various federation options - but it is more work and more complexity.
For regulated workloads that need to sit within specific boundaries and use enterprise identity management, Azure tends to be a cleaner fit if you are already a Microsoft shop.
Where AWS wins for us
When I need to build something new and move fast, I almost always reach for AWS. The breadth of managed services is larger. The documentation is better. The community of people who have solved the specific weird problem I'm dealing with is larger. Stack Overflow has more AWS answers than Azure answers by a significant margin, and while that might sound like a trivial thing, at 11pm debugging something obscure it is not trivial at all.
AWS Lambda (serverless functions - small bits of code that run on demand without you having to manage a server) is also meaningfully more mature than Azure Functions. This matters to us because we have a handful of event-driven processes - things that need to run when a file arrives, or when a market close happens - and Lambda handles them cleanly.
The thing nobody talks about enough: egress costs
Both providers charge you for data coming into the cloud (ingress) cheaply or free, and charge you for data leaving the cloud (egress) at rates that can genuinely surprise you.
In financial services you move a lot of data. Market data comes in from multiple sources. Reports go out to clients. Data moves between systems. If you don't model your egress costs properly when you're designing workloads, you will get a bill that looks like a mistake but isn't.
We have one workload that generates about 4TB of report data per month that gets sent to a client-facing system. The egress cost on that alone is several thousand dollars per month. It doesn't make the workload uneconomical, but it is a number I track, and it was not in the original cost model.
Model your egress before you deploy, not after.
The honest answer to "which should we use"
If you are starting from scratch with no existing cloud presence: use whichever cloud your team is most familiar with. The learning curve is the biggest real cost, and familiarity compresses it.
If you are already deep in the Microsoft ecosystem: Azure for things that need to integrate closely with that ecosystem, AWS for everything else if you want more flexibility.
If you are in financial services specifically: both have the compliance certifications you need. The decision is operational, not regulatory.
The cloud is infrastructure. Good infrastructure is invisible. The goal is to make it invisible as quickly as possible and get back to building the things that actually differentiate your business.