Why did Alphabet's P/E drop to 17.8 the same quarter net income hit $112B?
Alphabet's trailing P/E just fell from **21.7** to **17.8**. In the same quarter, net income more than doubled to **$112.193B** and diluted EPS came in at **$9.11**. If your first reaction is that the stock got cheap, hold that thought.
No position in GOOGL.
Quick context for anyone who doesn't follow this one closely: this is Search and YouTube ads plus Google Cloud, with a growing portfolio of equity stakes in other companies that's becoming relevant to the headline numbers in a way it wasn't a couple years ago.
Trailing twelve month revenue grew **20.1%**. Trailing twelve month net income grew **111.3%**. Those two numbers should not diverge that hard on the same income statement, and I went looking for why. It's one line: a **$98.999B** non-cash, mark to market gain, tagged Debt And Equity Securities Gain Loss on the filed 10-Q (period ending June 30, 2026, filed July 23, 2026). It's tied to that stakes portfolio, which the balance sheet shows growing from **$36.2B** in September 2024 to **$131.5B** now. That book has more than tripled in under two years, and swings in its marked value are now moving reported earnings more than the actual ad and cloud business does.
Here's the reconciliation, off the same filing, quarter over quarter:
* Operating income: **$39.696B** → **$40.77B**
* Debt and equity securities gain: **$36.804B** → **$98.999B**
* Net income: **$62.578B** → **$112.193B**
* Diluted EPS: **$5.11** → **$9.11**
Take this quarter's **$138.753B** pretax income and subtract the **$98.999B** gain directly. You land at **$39.754B**, within rounding distance of the **$40.77B** of operating income alone. Nearly the entire gap between operating income and pretax income this quarter is one line item.
General lesson, not just a GOOGL thing: whenever a company's net margin jumps double digits in a single quarter, it's worth checking the raw income statement tag for a one-off gain or loss before trusting whatever margin number a screener shows you. Net margin especially is the kind of number that gets scraped and repeated without anyone opening the actual filing to see what's driving it.
Cash backs this up in a way net income can't fake. Operating cash flow this quarter was **$39.069B**. Capex was **$44.924B**. That's negative **$5.855B** of free cash flow, the first negative quarter I can find anywhere in this filing history, against a positive **$10.116B** the quarter before. The trailing twelve month free cash flow figure (**$53.273B**, down from **$64.429B**) is still positive and will stay that way for a while. A single bad quarter takes time to drag down a trailing sum, which is worth remembering if this name shows up healthy on a screen built off TTM numbers alone.
**The number that actually surprised me: ROIC fell**
I calculate return on invested capital as after tax operating profit divided by average invested capital (total equity plus total debt, minus excess cash). Running that here, ROIC (TTM) fell to **22.6%** from **26.3%** the prior quarter, in the same quarter net income more than doubled. NOPAT, built off operating income rather than net income, grew a modest **5.9%**, from **$113.806B** to **$120.465B**. But invested capital jumped **31.3%**, from **$526.634B** to **$691.651B**, because total equity grew **$161.734B** in a single quarter, largely the same paper gain landing in retained earnings. The gain doesn't just inflate the earnings side. It inflates the capital base you're supposed to be measuring returns against, so the "return" on that capital actually compresses. This is also why the falling P/E at the top of this post is a trap: the denominator (trailing net income) got inflated by the same gain, so the multiple looks cheaper without the stock, or the business, actually getting cheaper.
Worth being direct about the limits here. I'm working off one quarter's filing, and a mark to market gain on private stakes can reverse next quarter just as easily as it can compound further. Nothing in this filing tells you which way it goes from here. Net debt also crept up, to **$42.254B** from **$39.438B**, which I'd call a minor factor sitting next to the equity swing, not a story on its own.
Push back if you think stripping the gain out is the wrong read. I'm treating operating income as the cleaner proxy for the actual business, but if you think the mark to market gain deserves to count as real economic value creation rather than noise, I'd like to hear the case. Alphabet's stake portfolio is big enough now that the argument isn't obviously wrong either way.