It's not my intention to provide updates every Friday, and honestly I expected these updates to be way less frequent. But the markets are moving and so am I.

(Cash and Total positions assume cash held flat at $8,069.41 (79.29% of $10,177.38) with no yield applied. CRM, MSFT, and LEN dollar figures reflect actual shares held, with approximate intraday prices.)

MSFT (Microsoft) — My perspective on this company was a toss up prior to the July 29th earnings. You can read the full breakdown here. But in short, the unknown timeline from capex converting to cashflow meant the price could go anywhere. The payoff profile didn't lean one way or another, making the bet more of a coin toss than an investment.

Earnings resolved a good chunk of that ambiguity. Azure growth accelerated to 43% and crossed $100B in annual revenue. Copilot went from 20 million to 30 million paid seats in a few months. Commercial backlog hit $678B. None of that proves the AI spend earns an exceptional return, but it does confirm the demand is real rather than hypothetical, which shifts the question from "will this monetize" to "how profitably will it monetize." That's a materially better question to be underwriting.

The downside of this information was the stock got more expensive (up double digits in a single day). However, it was a tradeoff I was willing to accept.

There's an instinct worth naming here: the stock jumped roughly 15% overnight, and buying into that felt icky. Sort of like the boat had already left the dock.

But that instinct is about price, not about the bet. Once I ran the new information through the model, the reward profile looked better, not worse. I didn't get in as cheap as I'd have liked, but the asymmetry — how much room there is to run relative to how exposed I am if I'm wrong — is more appealing now than it was before earnings.

Worth being precise about what kind of position this is, though.

CRM and LEN are mispricing bets — the market is wrong, or underappreciating something specific. MSFT isn't either of those.

It's a durable compounder trading around fair value. The way I ended up sizing it was by asking what it was actually competing against in the portfolio — not CRM, not LEN, but cash.

My long term goal was to keep 25% of the portfolio in cash to take advantage of other opportunities while my other bets matured. But since it’s taking me so long to pull size my other bets, I’m sitting in cash for longer, which drags the portfolio down.

Instead, framed as "25% cash" versus "15% cash and 10% Microsoft," the second was the easier call. I don't need this one to be a bargain. I just need it to beat sitting in cash over five years, which feels like a reasonable bet for one of the best-run businesses in the world.

CRM (Salesforce) — I didn’t make any changes since last time, but the stock price did. Up ~17% from my last purchase, or up ~15% from my blended cost basis. Still, this is short term noise. The next catalyst (September 2 earnings) is still about a month away.

I had originally earmarked 25% of my portfolio to this bet, but I’m starting to doubt I’ll be able to make the whole bet before the price takes off. A combination of finding the opportunity too late and being too timid in my purchases. Still, I’d rather err on the side of caution and be right, than rush and be wrong.

Still watching the $145 load-up trigger on every pass through the watchlist.

LEN (Lennar) — No change. Down modestly since entry, tracking the broader housing/rate narrative more than anything company-specific. Still waiting to see what happens.

New line for the new tranche, per the rule — this ledger only grows, nothing gets edited after the fact.

If you’ve been following along over the past several weeks, you’ll know that I’m always interested in what Berkshire Hathaway invests in.

And now that I’ve invested in MSFT, you may be wondering why I haven’t invested in Google as well, since Berkshire admitted a gigantic stake in their last 13F.

A very good question, and one I intend to talk about next time.

Happy Investing.