The Magnificent 7 Derating: What the Multiples Tell Us

The Magnificent 7 — Apple, Microsoft, Alphabet, Amazon, NVIDIA, Meta, and Tesla — collectively lost over $3 trillion in market capitalization between late 2021 and late 2022. While prices have since recovered for most names, the valuation multiples have not. This is the derating story, and it matters more than the price action.

The Multiple Compression

Let's look at the numbers. The group traded at a median EV/EBITDA of ~35x at the 2021 peak. Today, the median sits closer to 25x — a 30% compression despite strong fundamental performance.

What's Driving the Derating?

Three forces are compressing multiples simultaneously:

1. The rate regime shift. Higher discount rates mechanically reduce the present value of future cash flows. With the 10-year Treasury sustained above 4%, the "TINA" (There Is No Alternative) premium that inflated tech multiples in the zero-rate era has largely evaporated.

2. Revenue deceleration. The pandemic-era pull-forward is fully digested. Cloud growth has slowed from 40%+ to 20-25% across major players. Digital advertising has matured. Even AI revenue, while growing rapidly, isn't large enough yet to offset the base rate deceleration.

3. Capex intensity. AI-related capital expenditure is consuming an increasing share of free cash flow. Microsoft's capex has roughly doubled since 2022, and Amazon's AI infrastructure spending continues to accelerate. Markets are rightly discounting the returns on this investment until they materialize.

Historical Context

Derating cycles in technology are not unprecedented. The last comparable episode was the 2000-2003 period, when mega-cap tech multiples compressed by 60-70%. The current compression of ~30% is more moderate, which makes sense — today's big tech companies have far more durable business models, stronger cash generation, and actual monopoly-like market positions.

Relative Value Assessment

Within the group, dispersion has increased. Meta trades at a meaningful discount on EV/EBITDA despite best-in-class margins, likely reflecting ongoing skepticism about metaverse capital allocation. NVIDIA commands the highest premium, though it has compressed significantly from its 2023 peak multiple.

What to Watch

The derating thesis resolves one of two ways. Either multiples stabilize at current levels and returns become a pure function of earnings growth (the "growth at a reasonable price" scenario), or multiples re-expand as rate cuts materialize and AI investments begin generating returns (the "re-rating" scenario).

The key variable is AI monetization. If the ~$200B in annual AI capex across the group generates returns that justify the investment within 2-3 years, the current multiples will look cheap. If it takes longer, or if the returns disappoint, multiples could compress further.

For investors, the implication is clear: the easy money in big tech is likely behind us. The next leg of returns will require more selectivity and a deeper understanding of which companies are best positioned to monetize the AI transition. The era of buying the index and riding multiple expansion is over.