I’m back! This feels long overdue, but after moving across the pond and switching roles (I’m now at a family office), I finally have a bit of creative time again to keep writing here. To everyone who stayed subscribed: Thank you for hanging on!A lot has happened in the second half of 2025 (yes, I have many notes), but before I jump into a full-blown deep dive, I want to share a shorter observation that I think is genuinely useful.I’m still working through another piece, specifically, how Amkor Technology might finance “Project Arizona” given the size of its current balance sheet and the fact that Apple appears to be the anchor customer so far. But that’s for another note ;)Today’s note is about something much simpler: how a stock can look absurdly cheap… until you read the footnotes. Photo by Rob Wingate on UnsplashThe “Wait… is this real?” momentPicture this: you’re at your desk, scrolling for stock ideas. You’re eager to ride the Santa rally people keep talking about, and maybe you’re even thinking about buying something nice for your partner with those gains.Then you stumble on .~10x P/E.You blink. Twice.“Am I dreaming? Isn’t robotaxi competition coming? Doesn’t Uber have huge market share? Is this a once-in-a-decade mispricing?”And that’s where people do the most dangerous thing in investing:They stop thinking and start clicking.If you love reading my post, don’t forget to subscribe!Hold your horsesA low P/E can mean “cheap”… but it can also mean the “E” (earnings) is inflated by a one-off accounting item.If you look at Uber’s 2024 numbers, there’s a major clue hiding in plain sight.In 2024, Uber reported a very large income tax benefit, not because its core business suddenly became a tax-rebate machine, but because of an accounting change.Uber released a valuation allowance against U.S. federal and state deferred tax assets after concluding it was “more likely than not” that these tax assets would be usable, based on sustained profitability. Uber disclosed an income tax benefit of $5.8 billion for 2024, including a $6.4 billion benefit from releasing that valuation allowance (with exceptions such as certain California research-and-development credits). They further broke it down as $5.2 billion released on U.S. federal deferred tax assets and $1.2 billion on U.S. state deferred tax assets (again, with some exceptions). What matters for investors: this kind of tax benefit can materially inflate net income, making the P/E ratio look much lower than what the underlying operations would suggest.So what happens if we “normalize” earnings?If we adjust earnings by removing that unusually large tax benefit, then recalculate earnings per share (EPS), the P/E snaps back to something far less “too good to be true.”In your math, that adjusted P/E lands around ~29.4x.That’s not necessarily “expensive” or “cheap” by itself but it’s definitely not a “10x P/E bargain” you can rely on from a quick screen.So… is Uber cheap?I’ll stop here, because whether Uber is actually cheap is a bigger question (and honestly, a fun one). But the point of this short note is simpler:Stock screeners don’t read footnotes.Headline multiples don’t tell the whole story.And the fastest way to lose money is to mistake a one-time accounting boost for sustainable earnings power.Until next time!Thanks for reading! Subscribe for free to receive new posts and support my work.