META, 6-month History

Meta has been a volatile stock to trade over the past year. It spent much of that time swinging inside a wide channel.

Our Multi-model Signal caught the latest breakout. Meta has since pulled back from its highs, but our models have grown more confident: all four now agree on the long side.

In our backtest, the Multi-model Signal for Meta lagged buy-and-hold over the past three years. Over the past five years, it beat buy-and-hold by 125%. It protected capital in the 2022 bear market by staying out of the stock.

Meta sells advertising. Ads were $63.1 billion of Q3 revenue of $64.6 billion, or 97.6%. Meta sells them across Facebook, Instagram, Messenger, WhatsApp and Threads, which 3.64 billion people used on an average day in September.

How the ad engine grows. Ad revenue is the number of ad impressions times the price per ad. The CFO attributes the price gains to better ad performance and improved macro conditions.

The new bets. Muse, Meta's personal AI agent, passed 10 million sign-ups in the US and Canada in six weeks. About 25% of them use it on any given day. Meta earns nothing from Muse beyond subscriptions so far.

Meta guides Reality Labs (headsets and AI glasses) to lose $18 billion to $19 billion in 2026.

Revenue rose 26.1% YoY and 6.2% QoQ to $64.6 billion. Growth has eased from 33.1% YoY in Q1 and 28.0% in Q2.

The drivers:

Growth now leans on pricing. Daily users grew 3% YoY, while revenue per user rose 22.5% YoY to $17.72.

Meta raised the expense guide to include the $9.89 billion legal charge. It gave no 2027 CapEx or expense guide and expects to "continue investing aggressively".

GAAP operating margin fell to 20.1% from 40.1% a year earlier. A $9.89 billion legal charge related to an agreement with US state attorneys general over teen mental health explains most of the fall.

Excluding the charge, as Meta states it (not GAAP), the margin was 35.4%. That is still 466 basis points lower YoY, as cost of revenue (up 45.2% YoY) and R&D (up 32.4% YoY) grew faster than revenue. Meta cites depreciation, cloud capacity and AI token costs.

Free cash flow turned negative, at minus $1.2 billion on the chart (operating cash flow less CapEx). On Meta's own definition, which also deducts finance-lease payments, it was minus $2.1 billion (non-GAAP). A year earlier, it was $11.2 billion on the chart's basis and $10.6 billion on Meta's.

CapEx (purchases of property and equipment) rose to $33.9 billion, or 52.5% of revenue and 103.5% of operating cash flow. Including finance-lease principal, the basis of Meta's guide, it was $34.9 billion. The consensus estimate puts 2026 CapEx at $139.9 billion and 2027 at $211.3 billion.

Meta bought back no shares in the first nine months of 2026. Its last buyback on the chart was $3.3 billion in Q3 2025. Meta held the dividend at $0.525 a share. Shares outstanding rose by 8 million in Q3 to 2,556 million, from stock awards.

New shares and debt. Meta issued no debt in Q3, after $25.0 billion of notes in May.

After the quarter, it priced 40 million Class A shares at $625 under forward sale agreements, which settle by 19 November 2027. It also priced $12.5 billion of 6.50% mandatory convertible preferred stock. Together they add 2.2% to 2.3% to the share count.

Meta also plans an at-the-market programme of up to $60 billion of Class C shares, not before the first quarter of 2027. It has set no terms.

Here are the key financial metrics from the past four quarters. I designed the table this way so we can focus solely on the key top-line, profit, and cash flow metrics.

On the consensus estimate, revenue grows 22.5% over the next twelve months, while free cash flow falls to minus $62.4 billion as CapEx outruns operating cash flow.

EV/Sales divides enterprise value (market value plus debt less cash) by revenue. EV/EBIT and EV/FCF divide it by operating profit and free cash flow. P/E divides the share price by GAAP earnings per share. Each band is one standard deviation (sigma) of the multiple over the three years shown.

META trades 1.8 standard deviations below its three-year median on EV/Sales (6.58x against 8.88x). Its GAAP P/E is 2.1 standard deviations below (20.7x against 27.4x). On this basis, the stock looks attractive.

In my view, free cash flow is now the metric that matters most. As CapEx absorbed cash, trailing EV/FCF jumped to 55.5x.

This chart plots META's EV/Sales against the share price return over the following year, for start dates in the two years to November 2025. The relationship is close to linear, with an R-squared of 0.743 (a measure of fit, where 1 is a perfect line): the lower the EV/Sales, the higher the return a year later.

Over this period, the share price rose over the following year from every start date with EV/Sales below 8x (6.58x today).

On that history, today's valuation is attractive again. However, AI CapEx has changed Meta's risk profile. In Q3, CapEx exceeded operating cash flow for the first time on our chart. The consensus estimate shows negative free cash flow for the next twelve months, so Meta now funds part of the build with new shares and debt.

I value META using EV/EBIT. It has a long history and is neutral to how Meta funds itself. It also charges the data-centre build through depreciation.

For a 15% three-year EBIT growth estimate (CAGR), I see 21x as fair, a little below the three-year median of 22.8x. On next-12-month consensus EBIT, that gives a 12-month price target of $799, 32% above the latest close. The bear and bull cases, with EBIT 15% lower or higher, give $679 and $920. The consensus price target is $848.

Our inverse DCF asks what today's $606.73 price assumes. WACC is the blended return that shareholders and lenders require. At a WACC of 10.59% and 3% terminal growth (the yearly growth assumed after year ten), the model gives a DCF value of $623 a share, 3% above the price, with:

On these assumptions, free cash flow stays negative for the first four years. Terminal value, the value of all cash flows after year ten, is 85% of enterprise value. Each half point on the WACC moves the value by about 10%.

Since mid-2024, META has made higher lows. Since August 2025, it has made lower highs. In such a wide pattern, technical analysis has little to say. None of the key moving averages has given a reliable signal in either direction.

The stock now trades in its heaviest volume zone since May 2024. On this basis, the chart looks neutral.

Can Meta's ad engine pay for its AI build? Not on its own for now. Ad pricing is accelerating, but CapEx overtook operating cash flow in Q3. Meta has priced $37.5 billion of new Class A shares and preferred stock.

Even so, I lean bullish at this price. EV/Sales sits 1.8 and GAAP P/E 2.1 standard deviations below their three-year medians. In my view, the price already reflects much of the AI build and funding risk. After such a large fall in valuation, the risk-reward looks appealing to me.

Our inverse DCF shows today's price is consistent with the consensus estimate for 2026 revenue, then growth of 20% YoY or more for three years, with CapEx peaking at 67% of revenue in 2027. That is a reasonable bar. It leaves upside if growth reaccelerates or CapEx falls back sooner than the model assumes.

Our EV/EBIT target is $799, 32% above the latest close. The chart is neutral inside a wide range that dates back to mid-2024, while all four of our models are long. My view lines up with the models.

The next test is the 2027 CapEx guide on the Q4 call.