$SMCI — the framework I use for Super Micro Computer. Reference price: $45.20 | FY26 rev: $21.5B | NI: $1.2B | Shares: 584M 5-year bull / base / bear model: Bear Base Bull 2031 price $7–10 $46–55 $128–170 Annual return ~-25% ~1–3% ~23–30% Equity value $4–6B $27–32B $75–100B FCF multiple 12× 18× 25× Probability 30% 50% 20% Bear case for SMCI (30% probability): Margin compression from competition (Dell, HPE, OEM partners). Revenue stalls at $18-20B, FCF drops below $500M. Stock trades at 12× earnings. Auditor issues linger. The risk here isn't small. Super Micro Computer has real structural problems that won't resolve quickly. Base case (50%): AI server demand stays elevated but growth normalizes. Revenue $28-30B, FCF $1.5-1.8B. Stock re-rates as profitability proves durable. Auditor clean. This is the "nothing special, nothing terrible" scenario. SMCI survives and muddles through. Bull case (20%): AI capex cycle extends through 2027-28. SMCI maintains design wins with hyperscalers. Revenue grows to $45B+, FCF expands to $3-4B. Stock rerates to growth tech multiple. This is where the real upside lives. But it requires everything to go right. The critical assumption The entire thesis rests on: GPU supply stays tight AND SMCI holds hyperscaler design wins. Lose either one and the thesis breaks. Numbers to watch over the next 4-6 quarters: • Quarterly GPU allocation from NVDA • Hyperscaler order book changes • Gross margin trend (target: 14-15%+) • Auditor situation resolution My rough probability weighting: Bear 30% / Base 50% / Bull 20% Rough probability-weighted value: ~$48.00 The interesting exercise: what does Super Micro Computer look like at 30-40% below current prices? Because at that level, the risk/reward changes dramatically — even if the bear case plays out. SMCI — high conviction name if you believe the thesis. Value trap if you don't.