The DIY investor's tool stack: tracker, journal, screener, or one workbench?

Here’s a thought experiment. Imagine you hired four very competent assistants to run your portfolio, and then, as a sort of cruel social experiment, forbade them from ever speaking to each other. One lists everything you own, one records why you bought it, one hunts for new ideas, and the long-suffering fourth sits in the middle frantically copying notes between the other three so the whole thing doesn’t fall apart.

That fourth assistant is your spreadsheet. The reason it’s so tired is that you’ve built exactly this office, probably without ever deciding to.

The four assistants you accidentally hired

Nobody sits down and architects a tool stack. It accretes, like plaque. You start with a tracker, because your brokerage dashboard cheerfully shows you one account and pretends the others don’t exist. Its whole job is the question where do I actually stand, which you feel you should be able to answer instantly and somehow can’t, especially once your book is scattered across several brokerage accounts.

Then a journal, because you keep making the same mistake with the cheerful regularity of a man stepping on the same rake. It answers was that a good decision: plan before, outcome after, win rate and expectancy from records instead of from your brain’s flattering highlight reel.

Then a screener, because you need new ideas and you’d rather they came from somewhere other than a man shouting on the internet. Its question is what’s worth a look.

And underneath all of it, the spreadsheet: the connective tissue. The cost-basis maths the tracker got wrong, the watchlist the screener can’t hold, the allocation targets nothing else tracks. It quietly becomes the real system of record, precisely because it’s the one place you can bend to fit your process.

None of these four is the wrong tool. Each earns its keep. The problem is that you have four of them and they refuse to talk.

Where the seams leak

Each tool is sound on its own little island. The trouble is the water between them, and the fact that you are the one ferrying every fact across it. There is no single source of truth. The tracker says the position is worth one number, the spreadsheet says another, and you can’t tell which is lying, or whether the gap is a stale price or a trade you logged in one place and forgot in the other. A book you can’t reconcile is a book you can’t trust, and you have somehow manufactured four of them. The moment two numbers that should match don’t, you quietly stop believing both, and now you’re managing a rumour you keep repeating to yourself.

Then the reasoning, which lives nowhere near the thing it’s reasoning about. Your thesis is a note in a doc, a cell in a spreadsheet, or, let’s be honest, a vibe you had eleven months ago. By the time a holding moves and you need to remember why you own it, the why is in a file you haven’t opened since you last reorganised your sock drawer, so you reconstruct it from how the trade turned out, which is the precise moment you stop being an investor and start being a person inventing a flattering story.

Then the analysis, which is a black box. The screener hands you a score in a cell and no opinion on where it came from, what went into the composite, or who chose the weights. So you choose between trusting a number you can’t inspect and ignoring a number you paid for. Neither is a process.

And every seam is manual labour. One real-world trade becomes four chores: copy it to the tracker, reflect it in the spreadsheet, note it in the journal, update the watchlist. Four acts of transcription, each a fresh chance to drift, and the spreadsheet that started life as helpful glue slowly mutates into the liability, with no live valuation and formulas that rot as the book grows. The cost was never the subscriptions. It’s that you make real decisions on a book no single tool can vouch for.

The case for one reconciled workbench

So here is the way out. The stack exists because three honest jobs got handed to three tools that don’t share data: the book, the reasoning and the analysis. Put all three on one reconciled ledger and the seams don’t get easier to manage. They stop existing, and you can’t leak a fact across a seam that isn’t there.

The book becomes one immutable ledger you enter yourself. Yes, by hand. Manual entry is the thing that makes the book reconcilable, because holdings, lots, cash and P&L are all derived by replaying that ledger. Nothing to keep in sync. Nothing left to drift. The reasoning moves onto the position: a thesis pinned to the ticker, with the invalidation rules that flip it to AT-RISK the instant a signal breaches, plus a journal recording plan versus actual against that same book. That’s the discipline loop, and it only works when the reasoning and the position share an address. And the analysis turns transparent on principle, because every factor-grid cell opens up into the exact signals, weights and normalisations behind it. No more scores you can’t pull apart.

Which leaves you with one source of truth, and that’s the whole difference between a book you act on and four files you spend your weekends reconciling.

What to keep, what to consolidate

People hear consolidate and panic, so to be clear: you keep what each assistant was for. A view of the whole book. A record of your decisions, and somewhere to rank a universe and do the analysis. What you consolidate is the seams: the manual copying, the thesis stranded in a doc, the screener score you can’t crack open. Nobody ever wanted those. They’re just the tax you pay for running four systems that were never built to share a book.

A workbench only earns the name when the number in the header, the thesis on the position and the score in the grid all trace back to the same entries you made. If you’ve been holding the whole contraption together by hand, by being the tired fourth assistant, that’s the seam worth closing. You can start your own book with nobody left to copy notes between rooms.