Spreadsheets and models

Most investment models still live in a spreadsheet. The tools here make that work faster: the spreadsheets themselves, which now include AI help for formulas and analysis on some plans, add-ins that pull market data and financial statements into cells, and services that extract figures from filings into structured data.

AI help in spreadsheets is good at writing formulas, explaining someone else's model and summarizing a table. It is weaker at knowing whether the data is right. Check any formula it writes on a few rows by hand, and keep the inputs traceable to a source.

Data add-ins save copying and pasting, but they bring the same questions as any data provider: where the numbers come from, how often they update, how they treat adjustments and what the license allows. A model that refreshes automatically can change without you noticing, so record the date of each pull.

Professional teams use services that extract figures from filings and presentations straight into their models, with links back to the source page. That traceability is the feature to look for, since it lets anyone check a number in seconds.

Structure the model so it can be checked. Keep inputs on one sheet, calculations on another and outputs on a third, and color input cells so anyone can see what is typed in and what is calculated. AI assistants work better on a clear layout too, since they can find the inputs and explain the steps.