2 Stocks So Cheap It's Like Christmas in July

The article argues Nvidia is “surprisingly cheap” despite AI demand, trading at ~16x next year’s earnings (vs 23x this fiscal year) as revenue growth is projected to decelerate from ~82% this year to ~41% next year. Upbound is highlighted for a 7.4% yield while earning ~$4.00–$4.35/share (adjusted) and trading at just ~5x this year’s profit, with a low ~37% payout ratio at the midpoint. It flags key risks for both—margin sustainability and growth slowing for Nvidia, and economic sensitivity, substantial leverage, and possible regulatory scrutiny for Upbound—suggesting valuation support is real but confidence is mixed.

Analysis

NVDA is not a classic bargain; it is a scarcity asset whose multiple only looks low because the market is already discounting a sharp deceleration after an unusually strong cycle. The real question is whether hyperscaler capex and inference demand can keep earnings revisions positive for another 2-3 quarters; if yes, the stock can absorb a premium multiple re-rate, but if growth simply normalizes faster than consensus, the 16x forward P/E can compress further even with record profits. The second-order winner is likely the supply chain around advanced packaging, memory, and networking rather than rivals trying to outship Nvidia on raw GPU share.

UPBD screens cheap for a reason: this is a levered consumer credit instrument disguised as an operating company, so equity value is highly sensitive to small changes in charge-offs, delinquency roll rates, and refinancing access. The dividend is supportable only if the credit book stays benign; in an early recession or labor-market wobble, the yield becomes a trap and the market will reprice the equity off normalized earnings power, not the current payout. The bond market is the cleaner tell here: widening spreads or tighter ABS funding would likely hit the stock before the income statement does.

The consensus mistake is treating both names as simple valuation stories. For NVDA, valuation is being judged on peak-earnings optics instead of duration of growth; for UPBD, investors may be underestimating how quickly a high-yield equity can de-rate when unsecured consumers weaken. The near-term catalyst path is earnings and macro data, but the falsifier for the bear case is different: NVDA needs another quarter of upside revisions, while UPBD needs stable employment and credit loss metrics over the next 1-3 months to avoid a value-trap label.

Key Decisions for Investors

• NVDA: prefer staged long exposure on 5-10% pullbacks ahead of the next earnings cycle; 6-12 month horizon. Fails if next two quarters show growth falling below consensus and gross margin rolls over materially.
• UPBD: avoid initiating fresh longs here; if consumer credit data deteriorates or unemployment ticks up, consider a small 3-6 month put spread. The setup works only if charge-offs and funding costs stay contained.
• Relative value: long NVDA vs short a consumer-credit-sensitive basket until the market starts pricing a softer labor market. This isolates AI demand durability from consumer stress and reduces beta to the broader market.
• Watch item: if NVDA datacenter revenue guides below expectations or major cloud capex commentary softens, the 'cheap' multiple thesis is invalidated quickly; use that as the signal to de-risk.