valuation / MU
2026-05-21 · Comps · P/E · medium · source: weilab/model-builder@2026-05-21-v3
| wacc | — |
| fx_used.USD_JPY | 152 |
| fx_used.USD_KRW | 1,370 |
| fx_used.USD_TWD | 31.5 |
| peer_set | ["000660.KS","005930.KS","WDC","SNDK","285A.T","2408.TW"] |
| terminal_g | — |
| mu_eps_fy26e.low | 97.5 |
| mu_eps_fy26e.mid | 100 |
| mu_eps_fy26e.high | 102.2 |
| mu_eps_fy26e.source | market-implied from spot ÷ Fwd P/E |
| v3_corrections.fixes | ["EPS anchor switched from top-down to market-implied (spot ÷ Fwd P/E)","v2 EPS $29.45 was ~TTM, not forward — fixed to ~$100","Narrative flipped from -55% downside to +37.8% upside","v1 + v2 archived in same transaction"] |
| v3_corrections.v2_eps_mid | 29.45 |
| v3_corrections.v3_eps_mid | 100 |
| v3_corrections.v1_ref_price | 92.5 |
| v3_corrections.v2_ref_price | 731 |
| v3_corrections.v2_implied_mid | 327.48 |
| revenue_cagr_5y | 18.50% |
| peer_p25_pe_fy26e | 7.76 |
| peer_p75_pe_fy26e | 9.65 |
| peer_median_pe_fy26e | 8.75 |
| ebitda_margin_terminal | 55.00% |
| price_freshness_by_peer.WDC | stale_Q1_2026 |
| price_freshness_by_peer.SNDK | stale_Q1_2026 |
| price_freshness_by_peer.285A.T | stale_Q1_2026 |
| price_freshness_by_peer.2408.TW | stale_Q1_2026 |
| price_freshness_by_peer.000660.KS | fresh_2026-05-20 |
| price_freshness_by_peer.005930.KS | fresh_2026-05-20 |
| forward_pe_reconciliation.spot | 731 |
| forward_pe_reconciliation.rationale | Top-down rev × margin build (~$26 EPS) diverges from Street consensus (~$100) by ~74%; >20% rule triggers, so we anchor to market-implied EPS and use top-down as bear-case sensitivity. |
| forward_pe_reconciliation.divergence_pct | -73.8 |
| forward_pe_reconciliation.top_down_eps_mid | 26.2 |
| forward_pe_reconciliation.chosen_eps_source | market-implied |
| forward_pe_reconciliation.consensus_fwd_pe_low | 7.15 |
| forward_pe_reconciliation.consensus_fwd_pe_high | 7.5 |
| forward_pe_reconciliation.market_implied_eps_low | 97.5 |
| forward_pe_reconciliation.market_implied_eps_mid | 100 |
| forward_pe_reconciliation.market_implied_eps_high | 102.2 |
| mu_hbm_rev_share_fy26e_pct | 90.00% |
## Notes (v3 — market-EPS-anchored) This row supersedes v1 (stale $92.50 ref) and v2 (TTM-EPS-anchored). Both archived in same transaction. ### Headline - Implied: **USD 869.71 / 1,007.63 / 1,134.61** (low / mid / high) - Reference: USD **731** → mid upside **+37.8%** - Peer median Fwd P/E CY26E: **8.75x** (recomputed with fresh peer prices) - MU CY26E EPS anchor: **$100 (market-implied = spot ÷ Fwd P/E 7.30x)** ### EPS reconciliation (the critical fix vs v2) | Source | Low | Mid | High | |---|---|---|---| | Market-implied (spot ÷ Fwd P/E) — **CHOSEN** | $97.50 | **$100.00** | $102.20 | | Top-down rev × margin (your model) | $18.50 | $26.20 | $38.00 | | Divergence | — | **−74%** | — | Top-down build diverged from Street by >20% → triggered the v3 rule that anchors to market-implied EPS. Top-down kept as bear-case sensitivity only. ### Interpretive note At peer median P/E ~10x and Street consensus EPS ~$100, MU implied fair value is **~$1,000 — meaningfully above $731 spot**. MU is trading at a DISCOUNT to memory peer multiples on consensus forward EPS. Likely reflects: - (a) market skepticism on whether HBM upcycle EPS is sustainable into FY27+ - (b) cyclical memory caution The HBM TAM cross-check in the xlsx flags that the supply-chain anchor build alone (~$2bn MU HBM rev) is only ~1% of the revenue base implied by $100 EPS — the gap is the legitimate analytical question the wide TTM P/E spread (19x–33x) is signaling. ### Caveats - 4 of 6 peer prices (Kioxia, WDC, SNDK, Nanya) are stale Q1 2026 — refreshing would tighten the band. - "Market-implied EPS" is itself a derived number from a derived number (consensus Fwd P/E aggregates analyst inputs). If Street consensus is wrong, the v3 anchor is wrong. - v3 deliberately does NOT use the model's own top-down EPS — we are trading independence for accuracy vs market. Both v2 (independent / wrong direction) and v3 (calibrated to Street / right direction) are valid views.
