The models, and the deck at the end of them
Leveraged buyouts, accretion and dilution, comparables and the presentation the work exists to produce. Built to refuse a degenerate input rather than return a confident number from one.
Free plan. No card required.
What tools does investment banking analysis need?
A leveraged buyout model to test what a sponsor could pay, a merger model to test what an acquirer can afford, a trading comparables set to anchor both, and a way to turn all three into materials. The models themselves are not complicated. What separates a usable tool from a dangerous one is what it does when an input is missing or absurd.
The specifics
Leveraged buyout model
Sources and uses, a debt schedule across tranches, and returns to the sponsor, with the internal rate of return solved rather than approximated.
Merger and accretion analysis
Cash, stock and mixed consideration, with the share issuance derived from a real acquirer price and the earnings impact computed from it.
Trading comparables
A peer set with the multiples that are actually computable per peer and the count behind every median stated.
Discounted cash flow
The anchor valuation, loaded from filed fundamentals, with the terminal value share made explicit.
More on thisPresentation output
Content generated separately from layout, so every deck is visually consistent and no model ever asks a language model where to put something.
Document workspace
For the diligence half of a process, with everything questionable traceable to a page.
More on thisThe comparison in full
| Input | What most tools return | What this returns |
|---|---|---|
| Acquirer share price of zero on a stock deal | Zero shares issued, so the deal looks strongly accretive | A refusal naming the input |
| A negative acquisition premium | A negative offer price and a flattering result | A refusal naming the input |
| Zero earnings before interest and tax in an LBO | An internal rate of return of several hundred per cent | A refusal, because there is no root to solve for |
| A non-integer number of holding years | A schedule silently one year short | A refusal naming the input |
| A missing net debt figure | Debt treated as zero | The valuation withheld, with the reason stated |
Every row is a real failure mode. A model that returns a plausible number from nonsense is more dangerous than one that will not run.
This is not a full banking stack
There is no deal database, no league tables, no capital markets calendar and no messaging network. It is the modelling and the materials, which is the part where a small team is at the biggest disadvantage. If your work depends on the network, that is a different purchase.
Frequently asked questions
Does Khan Terminal do LBO modelling?
Yes. Sources and uses, a multi-tranche debt schedule and sponsor returns, with the internal rate of return solved numerically. Degenerate inputs are refused with a reason rather than producing a number, which matters more than it sounds: an all-zero cash flow set will return a several-hundred-per-cent return from a naive solver.
Can it model an acquisition?
Yes, with cash, stock or mixed consideration, deriving share issuance from the acquirer's price and reporting the earnings impact. Where accretion is genuinely undefined, for instance when the acquirer earns nothing, it is displayed as undefined rather than as a dilutive verdict.
Can I export a pitch deck?
Yes. Content and layout are generated separately: the analytical content comes from the models and the language layer, and a deterministic layout engine decides how it looks. That is what makes decks consistent, and it is why no model is ever asked to choose a slide layout.
Which plan includes the banking tools?
The Premium tier, which also covers the due diligence document workspace and the research library search. There is no trial on that tier; the recommended route is the Pro trial to see whether the workflow suits you, then moving up.
Model a deal you know the answer to
Run a transaction that already happened and see whether the output matches what was announced.