The fear running through Nvidia’s valuation right now has almost nothing to do with chip demand. The fear is that Nvidia has become the banker of the AI buildout and that, at some point, the credit losses show up in the income statement. Bank of America, in a note published Tuesday, argues that fear is miscalibrated. The resulting discount is, by its math, as large as 50%.

Where the Discount Lives

BofA’s framework is a sum-of-parts analysis built on free cash flow, not earnings multiples. The core finding: Nvidia trades at roughly 18 times estimated calendar 2027 FCF. Comparable AI semiconductor companies average approximately 38 times that same year. Even after applying a 50% haircut to the investment-related portion of cash flow to account for financing risk, BofA’s framework yields a valuation of 36 times 2027 FCF. The gap between 18 and 36 is where the 34% to 50% discount lives.

That cash flow base is not theoretical. In Q1 FY2027, Nvidia reported $48.6 billion in free cash flow, up from $26.1 billion a year earlier. Revenue hit a record $81.6 billion, up 85% year-over-year. The problem the market has is not with those numbers. It is what Nvidia has committed to do with the remaining cash.

The Commitment Stack

Nvidia has committed $105 billion to backstop OpenAI, a financing guarantee tied to an Ohio data center campus built through SB Energy. Add a $500 billion memorandum of understanding with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs, and KKR, where Nvidia acts as contingent co-investor at up to 25% per opportunity, and the total potential exposure becomes difficult to anchor to any single balance-sheet line.

BofA’s framing is direct: Nvidia is not a bank. A GPU that supports a failing project can be redeployed to a different customer. The CUDA ecosystem creates software lock-in that sustains residual hardware value across generations. Critics, including Michael Burry, have argued that a two-to-three-year GPU upgrade cycle makes five-to-seven-year useful life assumptions unrealistic. Neither view is unreasonable. What BofA is saying is that the market has already fully priced the bear case and is giving no credit to the bull case.

That financing question is becoming increasingly important because
the physical cost of the AI buildout is becoming impossible for investors to ignore. Nvidia is no longer simply selling accelerators into the buildout. It is increasingly helping make that buildout financeable.

Options Market and Trade Framework

NVDA’s near-dated expiries carry elevated implied volatility relative to longer-dated contracts, concentrating event risk into August 26. IV typically rises into NVDA earnings and falls immediately after. That volatility dynamic matters because
post-earnings volatility crush can punish long-premium traders even when they get the direction right.

Bull case. If you believe the BofA thesis compresses the discount, a call spread expiring after August 26, with the long strike at or near the money and the short strike near the $350 target, captures the re-rating thesis while capping premium at risk.

Bear case. For traders who believe the financing concern is underpriced, a put spread with the short strike near 80-day moving average support offers a defined-risk expression if guidance disappoints or commitment disclosures expand.

Neutral case. If you expect the realized move to fall short of the implied move, an iron condor outside the implied range can collect post-earnings volatility compression. Tail risk, both a significant beat and a guidance miss, carries meaningful probability given the scale of Nvidia’s recent commitment announcements.

In each scenario, elevated IV and NVDA’s tendency to see volatility fall after earnings both argue for spread-based approaches over naked long options.

Action Checklist

• BofA expects Q2 revenue of $94–95 billion against Nvidia’s own $91 billion guide. A beat in that range is not sufficient on its own. The Q3 guide matters more: BofA models $107–108 billion versus a Street consensus near $104 billion.
• Watch for any expansion of the financing commitment disclosure. Additional guarantees announced alongside earnings could widen the perceived discount even if chip demand numbers are strong.
• If Nvidia raises its FCF return target above 50% or accelerates buybacks, treat that as a positive re-rating signal. BofA identifies accelerated buybacks as the most direct mechanism for closing the FCF discount gap.
• Size defined-risk structures to account for outsized post-earnings gaps. The average event move is not the only risk input.