The research question
Good business momentum can meet a demanding starting price
Micron’s June 2024 report is a useful test of the phrase “it sold off despite good earnings.” The business question is how much of the recovery came from memory pricing, product mix and accounting effects. The market question is what happened after the announcement relative to the price already reached beforehand. Those questions need separate evidence.
This historical dossier studies June 26, 2024 in depth and connects it to 40 quarterly earnings events across the ten-year window. September and December 2024 illustrate contrasting paths within that larger archive. Each quarter has an original release link and a reproducible price study. The complete financial and call narrative below focuses on June 2024; it is not presented as 40 completed conference-call postmortems.
Inside the report
Reconcile reported and adjusted earnings
Fiscal Q3 revenue was $6.811 billion. GAAP operating income was $719 million and net income $332 million; adjusted net income was $702 million. The $370 million net-income difference reconciles to $217 million of stock compensation, $3 million of other adjustments and $150 million of tax adjustments. GAAP EPS was $0.30 versus adjusted $0.62. The two measures also use different diluted share counts.
Operating cash flow was about $2.48 billion. Micron reported $2.06 billion of net capital investment and $425 million of adjusted free cash flow. Its next-quarter revenue guidance was $7.60 billion ± $200 million and adjusted EPS $1.08 ± $0.08.
HPT interpretation: capital investment absorbed most of the operating cash generated. The adjusted earnings bridge makes dilution and tax treatment visible instead of hiding them inside a single headline. A growth thesis must eventually reconcile the cost of supplying that growth with the cash it produces.
| Q3 FY2024 bridge | USD millions |
|---|---|
| GAAP net income | 332 |
| Stock compensation adjustment | 217 |
| Other adjustment | 3 |
| Tax adjustments | 150 |
| Adjusted net income | 702 |
Inside the report
Open the notes: pricing, bits and inventory accounting
The subsequent Q3 10-Q reports DRAM revenue up 13% sequentially, with pricing approximately 20% higher and bit shipments down mid-single digits. NAND revenue rose 32%, with pricing approximately 20% higher and bit shipments up high single digits. Inventory previously written down contributed a $382 million cost benefit in Q2 and none in Q3. These details appear in MD&A, pages 30–31.
HPT calculation: $382 million divided by Q2 revenue of $5.824 billion is approximately 6.56 percentage points. That illustrates the size of the prior-quarter inventory benefit relative to sales; it is not a complete pro-forma margin restatement. Q3 margin improvement therefore cannot be explained simply by continuing that same write-down benefit.
HPT interpretation: separate price × shipment volume from product mix before extrapolating revenue. A business can generate more sales while shipping fewer DRAM bits if pricing improves enough. Sustainable margin analysis asks what survives if price increases slow, cost reductions normalize, or lower-margin end markets return to the mix. The filing is later post-report evidence, not a pre-announcement input.
Evidence: Micron · Q3 FY2024 Form 10-Q ↗
Inside the report
What management committed to—and what remained uncertain
In prepared remarks, CEO Sanjay Mehrotra described HBM as sold out for calendar 2024 and 2025, with most 2025 pricing contracted. He anticipated several hundred million dollars of FY2024 HBM revenue and multiple billions in FY2025. CFO Mark Murphy described pricing, mix and cost execution. The introductory remarks specify that financial discussion is generally non-GAAP unless stated otherwise.
Management also expected FY2025 capex around the mid-30s percentage of revenue. PC and smartphone customers had built inventory in anticipation of pricing and supply conditions. These are management statements and expectations, not a guarantee that future demand or cash conversion would match them.
HPT interpretation: contracted capacity helps define visibility, but it does not remove qualification, yield, delivery or broader memory-cycle risk. Customer inventory accumulation may support shipments today and reduce replenishment later. A proper follow-up asks how contracted units become recognized revenue and then cash, rather than treating “sold out” as synonymous with unlimited incremental upside.
“Our HBM is sold out for calendar 2024 and 2025”
Sanjay Mehrotra, CEO · Q3 FY2024 prepared remarks · Data Center, pages 3–4
Inside the report
Read the questions as well as the prepared story
In the third-party Q&A transcript, Vivek Arya asked about HBM’s wafer trade ratio and whether record sales could mean record gross margins. Mehrotra retained an approximately three-times HBM3E trade ratio versus D5 at mature yields; Murphy described improving price and mix while withholding full FY2025 guidance. Thomas O’Malley asked about capital intensity; management said spending was dominated by DRAM and HBM, with NAND a smaller share. These are paraphrases of a third-party transcription, distinct from Micron’s official prepared remarks.
HPT interpretation: an answer about a favorable industry setup is not a numeric commitment. The working draft should record the question, the actual answer, and what is still unresolved in separate fields. That makes it possible to compare the next call against the original commitment without upgrading an optimistic phrase into a forecast.
Evidence: ROIC · transcript of Micron Q3 FY2024 call and Q&A ↗
Inside the report
June, September and December did not produce the same path
Before the June report, the close was above the 12-, 22- and 55-day exponential averages, with those averages ordered upward. That is measured trend alignment, not proof of a bull flag. The first post-report close was down 7.12%; five sessions later the loss had narrowed to 3.89%, but after 20 sessions it had widened to 24.52%. SMH lost 9.33% over that same 20-session window. The semiconductor backdrop explains context, not all of Micron’s underperformance.
September’s first close rose 14.73%, yet its five-session gain was only 4.26%. December’s first close fell 16.18%, but its 20-session return was positive 5.27%, and a daily high first touched the pre-report close on January 7, 2025. The interactive event selector below shows each path without collapsing them into one average.
Across these selected cases, a negative first day did not always become a persistent decline, and a positive gap did not ensure immediate follow-through. Later prices incorporate additional news and market conditions. This design measures outcomes; it does not claim that the initial report caused the entire 20-session path.
Evidence: Yahoo Finance · MU historical daily prices ↗Yahoo Finance · SMH historical daily prices ↗Yahoo Finance · SPY historical daily prices ↗Micron · September 25, 2024 earnings release ↗Micron · December 18, 2024 earnings release ↗
Inside the report
Make the next draft answer specific questions
For the next Micron research draft, start with memory pricing versus bit shipments, HBM revenue actually recognized, gross-margin drivers, customer inventory, capex and free cash flow. Record the prior guidance and the date of each consensus snapshot if one is available. Distinguish beating company guidance from beating analyst consensus.
Before the release, freeze the options snapshot and technical references. After the release, replace no assumptions silently: add the financial results and call review as a separate dated publication. After the first regular session, record the gap, range, closing position and participation. Follow up after 5, 10 and 20 sessions when those observations exist.
A reclaim of a broken reference, a failed retest, or weakness against SMH can each be tested with completed bars. None identifies a market maker’s inventory by itself. “Active tape” here describes unusual participation; proving dealer positioning would require appropriately sourced options and inventory assumptions beyond these daily bars.