A spreadsheet template for tracking investment decisions, and the three ways it fails

A spreadsheet works. Below is a free template that does the job properly using Google Sheets: one row per decision, a scorecard that grades each one against a benchmark at six horizons, and a tab that computes hit rate and average alpha. What kills tracking spreadsheets is not the design. It is the upkeep. Recording decisions in the middle of earnings season, adjusting for splits, adding dividends, handling ticker changes, etc. One busy quarter later the updating has stopped. Two limits survive even perfect upkeep: you can always edit the past, and nothing in a file you own carries a timestamp you do not control. That makes it a good private decision journal but cannot graduate to credibly prove your skills to someone else.

Why a spreadsheet can work just fine

One row per call, entered the day you make it, with the thesis you had before you knew how it went. Buy, Add, Hold, Trim, Sell. The Scorecard measures each call against a benchmark at six horizons, from six months to five years. The Analysis tab turns that into hit rate, average alpha, and slugging. Your decisions compound into a record you can learn from.

Calls expire, which is the part most templates get wrong. A decision covers twelve months. Repeat the view with a "hold" on the same name and the clock restarts, so a five-year grade exists only if you kept renewing. An opposite call closes it immediately. Without that rule every call you make stays open, and hindsight picks which date you decide to record as the close to analyze that decision.

Two conventions carry the rest: entries go in the same day, and rows are never edited. Corrections are new rows. The date cell warns you when the date is not today, which is as far as a spreadsheet goes. It can warn but cannot refuse.

Keep those and you have a real record. A disciplined paper portfolio in a workbook beats no record at all. Where it fails is the rest of this page.

Failure one: the upkeep stops

A call you make today creates work on six future dates. Six months out, then a year, then eighteen months, and on to five years, that row wants a closing price and a benchmark price. Make twenty calls this year and you have booked 120 dated chores across the next five years, none of them urgent, all of them yours.

A live feed helps less than it looks. It fetches the horizon prices, which is real work saved. It cannot make the entry, because nobody but you knows you decided something today, so that row gets typed by hand or it does not exist. And splits make it worse: the feed restates history afterwards while your typed entry price stays as you typed it, so a decision you entered two years ago needs to be adjusted. Dividends are also left unaccounted for.

None of this is hard. It is just endless, and it competes with your day to day work. Entries get late, then sporadic, then you skip a quarter and backfill it from memory. What you are left with is not what you decided. It is what you remember deciding, and memory can become selective to favor good outcomes. Spreadsheets work but require discipline and for most they slowly go quiet.

Failure two: the past stays editable

This one is not about cheating. It is about tidiness. An entry price that looks wrong now gets fixed. A call you would rather not count gets deleted. A thesis gets one small rewrite so it reads better next to what happened. All of it is hindsight with edit rights.

The append-only convention fights this and it is worth keeping, but nothing in the file enforces it. The sheet can't lock a row against its own owner, and protection comes off for whoever put it on. A record you can revise is partly a record of your revisions, and you cannot tell which part is which. That is also why reviewing old decisions without hindsight bias is a problem in its own right.

Failure three: it proves nothing to a stranger

Every date in the sheet is a date you typed, and the file's own timestamps move every time you save. Six honest years and one long editing night look identical from the outside. A spreadsheet is testimony, not evidence.

Screenshots inherit the flaw. Version history in Sheets looks like an answer and is not: copy the file and the history starts clean, with nothing to show an earlier one existed. Emailing yourself the file gets a server timestamp, which proves a file existed that day, not that the record inside it is complete or unchanged.

Discipline holds off the first two failures. It cannot touch this one. Proof means the record sits somewhere its author cannot reach back into, and a file you own is the opposite of that. The sheet can teach you. It cannot make you credible to a stranger, and strangers can be your next recruiting audience: the interviewer, the PM you cold-messaged, the allocator.

Where Investor Strides fits

Equity research has run on spreadsheets far longer than it should have, and this is one more case of it. Software does this job better. Investor Strides records decisions at the official market close on the day they are made, in a model portfolio it prices for you, and nothing can be backdated or edited after the fact. The methodology page documents how records are written and why they cannot change. The record belongs to you and leaves with you. It only counts forward.

If a private journal is all you want, and you believe you can be the outlier in the industry with enough discipline to maintain it, then take the spreadsheet. Otherwise, make it easy on yourself and let Investor Strides handle the maintenance while adding a credibility layer to help your career.