Scroll through any finance forum and you'll find the same argument on repeat:
"Excel is dead, learn Python." "No bank runs on Python, master Excel."
Both camps are half right — which means both are half wrong. In real banking and risk teams, Excel and Python aren't rivals. They are two tools for two different jobs, and the professionals who get hired fastest are comfortable in both.
What Excel is genuinely great at
Excel is still the common language of finance, for good reasons:
- Transparency. Every number has a cell, every formula can be clicked and traced. A reviewer, auditor or credit committee can follow the logic without writing code.
- Speed for small models. A loan amortisation schedule, a quick sensitivity table or a ratio analysis is faster to build in Excel than to code.
- Communication. Managers, regulators and clients open spreadsheets, not notebooks.
- Prototyping. Many bank models are first built in Excel to agree the logic, then automated.
Where Excel struggles: millions of rows, thousands of simulations, repeatable pipelines, version control, and anything that needs to run automatically every day.
What Python is genuinely great at
- Scale. Loan books with millions of accounts, years of daily market data — Python handles them comfortably.
- Simulation. Monte Carlo VaR, stochastic interest-rate paths or credit portfolio simulations need thousands of runs.
- Statistics and machine learning. Logistic regression for PD models, gradient boosting for default prediction, time-series forecasting.
- Automation. A model that runs every morning, writes a report and flags anomalies — without anyone clicking anything.
- Reproducibility. The same code gives the same answer, and changes can be tracked.
Where Python struggles: explaining the model to a non-technical audience, and quick one-off what-if analysis.
Side by side
| Task | Better in Excel | Better in Python |
|---|---|---|
| Loan amortisation or a simple DCF | ✅ | |
| Presenting a model to a credit committee | ✅ | |
| Quick sensitivity analysis | ✅ | |
| PD model on 1 million loans | ✅ | |
| 10,000-path Monte Carlo VaR | ✅ | |
| Machine learning for default prediction | ✅ | |
| Daily automated risk reporting | ✅ | |
| Understanding a model step by step | ✅ | |
| Model validation & back-testing at scale | ✅ |
Don't forget the supporting cast
- SQL – almost all bank data lives in databases. Pulling the right data is step one of every model.
- Power BI (or similar) – turning model outputs into dashboards management actually reads.
- SAS – still widely used in many banks' credit risk teams, especially for older models.
- AI coding tools – increasingly used to write, explain and debug code faster — but only useful if you understand what the code should do.
The right order to learn
Here's the approach that works best for most finance learners:
- Understand the concept – What is a PD? What is VaR? Why does it matter?
- Build it in Excel – See every step. Make mistakes you can actually see.
- Rebuild it in Python – Now that you know what the answer should be, code it, scale it and automate it.
- Compare the two – If Excel and Python give different numbers, you've found a bug or a hidden assumption. That's where real learning happens.
- Present it – Turn the output into a chart or a two-slide summary for a non-technical audience.
This is exactly how models move from prototype to production inside banks — and why employers love candidates who can do both.
How Peaks2Tails teaches it
The Peaks2Tails approach is built around this Excel-then-Python method. Its programs promise production-style Excel modelling, Python workflows with interpreted outputs and visualisations that build intuition for complex maths, with refreshers in maths, statistics and coding for anyone starting from scratch.
In CPRF (for students and freshers), Semester 3 is dedicated to Excel & Coding:
- Advanced Excel and Power BI
- Financial modelling and equity research
- Python basics
- SQL and SAS basics
- AI tools for coding
…and those skills are then applied in Semester 4's credit, market, treasury and operational risk modelling.
In the specialised programs, every major model is built in both tools:
| Program | Tools | Example builds |
|---|---|---|
| Mastering Credit Risk Modelling (225+ hrs) | Excel + Python | PD/LGD/EAD models, scorecards, IFRS 9 ECL |
| Market Risk & Counterparty Credit Risk (180+ hrs) | Python + Excel | Monte Carlo VaR engine, FRTB capital simulator |
| Asset Liability Management (100+ hrs) | Excel + Python | Liquidity statements, IRRBB, FTP |
| Deep Quant Finance (200+ hrs) | Excel + Python | Derivative pricing, portfolio optimisation |
| Risk and AI (155+ hrs) | Excel + Python | Default prediction, anomaly detection |
Peaks2Tails' internship even includes converting Excel models into Python and SAS — one of the most common real-world tasks in bank risk teams.
The verdict
Stop asking "Excel or Python?" Ask "Can I build it in Excel, scale it in Python and explain it to anyone?" That combination is what gets you hired.
- 🎓 Explore CPRF for a full Excel + Python + SQL foundation
- 📊 Browse specialised programs on the Peaks2Tails short courses page
- 💬 Book a free 1-to-1 counselling call on WhatsApp
Peaks2Tails — turning risk into opportunity.
FAQs
Should I learn Excel or Python first for finance? Start with strong Excel to understand models step by step, then move to Python to scale and automate them.
Do banks really use Python? Yes, especially for large datasets, simulations, machine learning and automation. Excel remains common for review and communication.
Is SQL necessary for risk jobs? Very often. Most risk data sits in databases, and pulling it correctly is the first step in any model.
I have never coded. Can I still learn Python for finance? Yes. Programs with coding refreshers start from the basics and use finance examples throughout.