Why I built my own portfolio app.

I used to look at my portfolio once or twice a year. That's not a lot. I wanted to see it more, and I wanted to see more than what my brokerage tells me.

So I built my own. It's called Vantage.

Why I didn't use one that already exists

There's a lot of information I want to see, and I want to put in my own algorithms to calculate things. Most apps don't let you do that. So I built one that does.

It's inspired by Simply Wall St, and it has a lot of different charts I added on top of that.

The charts I'm proudest of

The snowflake charts. I think they'll give me the best view of a company at a glance. They're also really hard to code, which is why I'm proud of them. This is the part my AI tools struggled with most. They didn't understand how to build charts and graphs at first, and it took a lot of back and forth.

My own algorithms

One of them is a fair value check. It uses a few different things to work out what a company is worth per share. Then I can easily compare that number to what I paid. I can also use it on companies I'm thinking about buying.

What it is

It's a Mac app, written in Swift, and about 13,000 lines of code. It's private, and everything stays on my computer.

I type my holdings in by hand. I didn't want it connected to my brokerage. It handles stocks, ETFs, crypto, cash and mutual funds. It reports in US dollars, but it can also hold Canadian and European stocks, and it converts them with the Bank of Canada's exchange rates.

One rule is built in: it never makes up a price. If it can't get one, it says so.

How it got built

I've built it with ChatGPT and Codex so far. I might switch to Claude once some of my bigger projects are finished, so a smarter model is working on it.

Here's what I've learned about building like this. The first five to ten audits of anything I build come back with a lot of bugs. You find things you didn't know were there, because you're not deep in the code. I'm not going to read every line, so I have AI tools check it for me.

The audits on Vantage found real problems. One was in the backup system. If the app had to restore from a backup, it would have thrown away the most recent changes. An audit caught that. Another was code to fetch old prices that existed but was never actually connected to anything. I had it connected.

Where it stands

It's fully built, but I haven't tested it enough. I haven't even put my own data in yet. I want to be sure it works before I trust it with my real numbers.

For me, working correctly means all the algorithms work, it pulls the right data, and it updates properly. Then I can just look at it and get insights.

I'll probably check it monthly, not daily. That's still a lot more than twice a year. It's private for now, but if it turns out well, I'll open source it.

This is not financial advice. It's just an app I built for myself.

Drafted and created with the help of AI.