Brandon Leon

Toolkit · Free template

An IB-grade LBO & M&A workbook. Free.

17 tabs, every formula auditable: sources & uses, year-by-year debt schedule with covenant tracking, returns attribution (EBITDA growth / multiple expansion / debt paydown), reverse-LBO take-private analysis, peer-blended comps, accretion/dilution under three financing mixes, sensitivity tables. Built around the same cycle-aware methodology as the DCF template — with two worked Micron examples shipped together so you can see how the through-cycle and AI-cycle scorecards disagree on whether the name is cheap. The Python pipeline that auto-populates this for any ticker is available on request.

Get the workbooks

Two 17-tab Excel workbooks — MU through-cycle and MU AI-cycle — so you can see methodology disagreement in action on a real cyclical name. Drop your email, the downloads appear here. ~36 KB each, no follow-up unless you opt in.

Just the files by default. Disclaimer.

Thanks — here are the workbooks.

Workbook — MU Through-cycle (.xlsx, ~36 KB) Workbook — MU AI-cycle (.xlsx, ~38 KB) Tear sheet — MU Through-cycle (.pdf, ~80 KB) Tear sheet — MU AI-cycle (.pdf, ~78 KB)

Run the workbooks side by side. The methodology disagreement IS the recommendation. The tear sheets are one-page PDF summaries — IC-ready.

Want the Python pipeline that auto-populates this for any ticker? Scroll down ↓

Want the full Python pipeline?

The workbooks above are static templates. The Python + Jupyter pipeline I built around them pulls live data from yfinance, runs the full end-to-end build (loader → sector scorecard → 5y forecast → DCF → LBO → M&A → comps → reverse analytics → Excel) and produces fully populated workbooks for any ticker in ~60 seconds. I send the source on request to analysts, researchers, and students with a real use case — tell me what you're working on and I'll send it over.

I read every request and reply personally. Usually within 24 hours.

Got it — I'll be in touch within a day.

What's in the template

How to use it

Two paths, depending on which tier you have.

Workbook tier (manual fill-in)

  1. Download both workbooks

    Drop your email at the top to reveal the downloads. You get MU_through_cycle.xlsx and MU_ai_cycle.xlsx — same target, two scorecards. Open in Excel or LibreOffice.

  2. Read both side by side

    The point of shipping two workbooks is to show methodology disagreement. The through-cycle scorecard winsorizes the FY23 trough out and lands at $90/sh implied. The AI-cycle scorecard uses peak-half margins and lands at $134/sh. The spread between them is the option value of your structural call.

  3. Replace MU with your own company

    Overwrite the yellow-highlighted Drivers cells: ticker, shares, beta, debt, revenue / margin / capex assumptions, entry & exit multiples, leverage, hold period. Live formulas recalculate automatically. Source the data manually (yfinance, SEC filings, FactSet, Bloomberg) — or get the pipeline tier and skip the manual sourcing entirely.

Pipeline tier (request access above ↑)

  1. Email me a use case

    Use the request form near the top. Tell me what you're working on. I send the source over after a quick read.

  2. Run setup

    Mac/Linux: ./setup.sh  ·  Windows: double-click setup.bat. Installs Python dependencies (yfinance, openpyxl, xlsxwriter, pandas, jupyter, matplotlib).

  3. Edit Cell 1 of the notebook (or use the CLI)

    Plug in your ticker, peer set, and scorecard. jupyter notebook IB_Model_Template.ipynb → Run All. Or use python run_model.py MU --peers WDC,STX,NVDA --preset semiconductors. Pipeline produces a fully populated 17-tab workbook in ~60 seconds.

What you need

FAQ

Why ship two workbooks instead of one?
For cyclicals especially, the methodology choice is the recommendation. Shipping the through-cycle scorecard (winsorized — drops the FY23 trough and the FY21 peak) alongside the AI-cycle scorecard (peak-half — assumes HBM long-term agreements put a floor under memory margins) makes the bull/bear framing explicit instead of hiding it inside one preset. If both presets agree, you have conviction. If they disagree (as is currently the case for memory), the spread between them tells you how much your view depends on the structural-vs.-cyclical call. This is the same framework I'd present to an IC.
What's the difference between the workbook and the pipeline?
The workbooks are the two 17-tab Excel files (through-cycle + AI-cycle). They come pre-populated with Micron as a worked example so you can see what a finished IC analysis looks like. To run a different company, you replace MU's inputs manually with your own data. The pipeline is the Python + Jupyter system that automates the data sourcing — it pulls financials from yfinance, builds the peer set, runs the full build, and produces fully populated workbooks for any ticker in ~60 seconds. The workbooks are a free direct download; the pipeline is sent on request to people with a real use case.
What's the Reverse-LBO tab?
It solves backward. Given preset exit multiple and leverage, what's the maximum entry multiple at which a sponsor still hits 20% IRR? Convert that entry multiple to a $/share take-private price. Compare to current market. If take-private > current price, the public market is paying LESS than a sponsor would — possibly attractive. If take-private < current, the public is paying MORE than patient capital would; sponsor wouldn't bid. Useful framework even for names that aren't actual buyout candidates — tells you where private capital would step in.
What's the Reverse-DCF tab?
Same logic, applied to the public-market valuation. Decomposes the current EV into the perpetual-growth-rate the market price implies, given preset margin assumptions. Often surfaces "the market is implying NEGATIVE long-term growth" or "the market is implying 8% perpetual growth — above global GDP" — either of which is a real signal. Includes a joint-sensitivity grid: every (margin × growth) cell shows the implied EV ÷ current EV ratio. Cells near 1.0 are where the market sits.
Is this really free?
Yes — both tiers. The workbooks are a free download. The pipeline is also free for analysts, researchers, and students with a real use case — I just want to know what you're working on before I send the source. Built on open-source libraries (yfinance, openpyxl, xlsxwriter, pandas, jupyter). No paid data feeds required.
Why gate the pipeline behind a request form?
Two reasons. (1) The pipeline is a real piece of work and I'd rather it goes to people with a use case I can engage with. (2) Every request is a conversation — if you tell me you're building a thesis on a name, I'll often have something useful to add. The form isn't a filter to keep people out; it's a filter to start better conversations.
Does this work for any ticker?
The workbooks work for any public company — you just have to source the data manually. The pipeline (request access above) auto-pulls from yfinance, which covers most US-listed equities and major international tickers (NYSE, NASDAQ, LSE, TSE). For names yfinance doesn't track, both versions handle missing data gracefully (visible "N/A" in the Notes tab instead of silent magic numbers).
What sectors are calibrated?
Six scorecards ship with the template: semiconductors, semiconductors_ai_cycle, energy, materials, industrials, generic. Each is a calibrated set of through-cycle factors — mid-cycle EBITDA margin method (winsorize / median / peak-half), capex intensity, NWC %, target leverage, entry/exit multiples. Cyclicals (semis/energy/materials/industrials) use winsorized mid-cycle by default to avoid being dragged by a single trough. Memory in the current cycle uses peak-half (semiconductors_ai_cycle). Add your own scorecards by editing one Python dict.
Will this give me the same answer as a top-tier sell-side desk?
It uses the same methodology (Hamada re-levering for WACC, mid-year discount convention, blended Gordon + exit-multiple terminal value, peer-blended multiples with percentile bands, returns attribution). The answer depends on your scorecard choice and any overrides — growth path, margin assumption, entry/exit multiples. The template is dynamic — change one Drivers cell, the entire model recalculates. The Sanity Checks tab will tell you when you've drifted into territory the model doesn't trust (TV > 85% of EV, terminal g above GDP, leverage covenant breach).
I'm not technical. Is this hard to use?
The workbook tier is just Excel — if you can open a .xlsx and overwrite cells, you can use it. The pipeline tier requires Python (free; install from python.org) and double-clicking a setup script — the README walks through every step and most users have it working in 5 minutes.
Can I customize it?
Yes — every formula in the workbook is editable. If you have the pipeline tier, every Python module is open too: want a 7-year forecast? Edit financials.py. Want to add a new scorecard? Edit presets.py. Want a new sensitivity dimension? Edit sensitivity.py. Every module is independently testable.
Does it work on Mac?
Yes. Tested on macOS, Windows, and Linux.
How is this different from a free LBO spreadsheet on the internet?
Most free templates are static one-pagers with a hardcoded company. This is a 17-tab IB-grade workbook with a full sources & uses, year-by-year debt schedule with covenant tracking, returns attribution decomposition, reverse-LBO take-private analysis, reverse-DCF, peer-blended comps with percentile bands, accretion/dilution under three financing mixes, sanity-check auto-flags, and a full 3-statement projection. The pipeline tier auto-populates everything for any ticker via yfinance. Methodology shares a backbone with the DCF template — same audit standards, same sell-side IC discipline.

Two ways in.

Workbooks are a free download. Pipeline is on request.

Get the workbooks ↑ Request pipeline access ↑

Universal Edition · May 2026 · License: MIT-style for personal, academic, and research use; please credit if you publish off this work. For the matching DCF template, see DCF Template. For other free templates, see the toolkit.