MU — Reverse-LBO worked example.

The right discipline test for any public-equity thesis is to read the same name with a sponsor's mental model. Below: Micron through two scorecards — through-cycle and AI-cycle — with reverse- LBO output for each. The scorecards disagree by ~$60 of implied take-private price ($85 vs $146). That spread is the recommendation.

The setup

Same target, same exit horizon (5-year hold), same target sponsor IRR (20%). Two scorecards with deliberately different views on what mid-cycle EBITDA looks like for memory in 2026:

Both scorecards run the same engine: forward LBO with year-by-year debt schedule (mandatory amort, cash sweep, capex, NWC, taxes), bisection on entry multiple to hit the 20% IRR target, then convert max entry EV to $/share by netting net debt and dividing by diluted shares. The browser calculator runs both pulls in real time; the workbook ships both as worked tabs.

Reverse-LBO at every IRR threshold

Both scorecards, four target IRRs each — the price below which a sponsor at each underwriting hurdle would step in.

Numbers are illustrative output of the bisection LBO at each hurdle. They are not target prices. See disclaimer.

The read

Through-cycle scorecard: at $86 the public is paying essentially exactly fair value relative to a 20% IRR sponsor underwriting the cycle reverting — implied take-private comes in at $85, basically flat. Below $79 (a sub-15% drawdown), the 25%-IRR sponsor threshold opens up; above $93, even a debt-floor 15% IRR sponsor wouldn't bid. The through-cycle scorecard says: the public sits exactly on top of where patient capital with sponsor leverage would land. That's not a coincidence; it's how cyclicals tend to clear when consensus already prices the cycle reverting.

AI-cycle scorecard: the public is paying $86; a sponsor at 20% IRR would underwrite up to $146. That's a 70%-ish gap. Either (a) this is the public market disagreeing with the AI- cycle thesis (likely true in some part), (b) public-market investors can't run the leverage a sponsor would and so naturally pay less, or (c) the AI-cycle scorecard is too aggressive on EBITDA and exit multiple. Some combination of all three. The point isn't that the stock is "going to $146." The point is that if AI-cycle is the right scorecard, there's enormous headroom; if through-cycle is right, the public is roughly on it.

What does this say about the public-equity thesis?

The MU long lives at mu-through-cycle.html. The reverse-LBO read here adds three things to that thesis:

1. A floor. Through-cycle reverse-LBO at 25% IRR (an aggressive sponsor hurdle) lands at $79. That's a meaningful structural floor — if the stock drifts toward $79 and the through-cycle thesis holds, even high-IRR opportunistic sponsor interest becomes rational. The 20% threshold ($85) sits within a dollar of current, meaning the public market is pricing the name almost exactly at sponsor-territory under through-cycle assumptions.

2. A discipline check on the bull case. If you're bullish via the AI-cycle scorecard targeting $146, ask whether a sponsor at 20% IRR would actually underwrite at $146. The answer is yes, but only if you accept the peak-half EBITDA and the 10× exit multiple. If your DCF case requires the same EBITDA and a 12× exit, the sponsor wouldn't pay it. That's information about how aggressive your bull case is.

3. A read on what would change the math. The most sensitive input in both scorecards is the exit multiple, not the EBITDA. Run the through-cycle scorecard with exit at 7.0× instead of 8.0× and the implied take-private drops to ~$73. Run AI-cycle at 9× exit and implied price drops to ~$130. Multiple compression is the kill shot — not EBITDA. Same conclusion as the through-cycle long, just from a different angle.

What this isn't

It isn't a take-private prediction. There's no current sponsor interest in MU that I'm aware of. Memory is heavily strategic; regulators on both sides of the Pacific would scrutinize anything large; there are exactly four bidders worldwide who could underwrite this size deal at 4.5x leverage in current credit markets. Reverse- LBO ignores all of that.

It also isn't a bull or bear call on the stock from me. The point of the exercise is to put capital-structure discipline on a public- equity thesis. Around $79-$85 there's a real floor under through- cycle reasoning; above $146 even the AI-cycle scorecard at sponsor leverage caps out. Between those, you're picking a structural view. The MU through-cycle long takes a position; this page is the discipline check.

How to reproduce this

The math here is bisection LBO — not a closed-form approximation. To reproduce all four IRR thresholds for both scorecards in about 30 seconds:

Open the Reverse-LBO Calculator. Click "MU — through-cycle" in the worked-examples section. The reverse-LBO table on the page renders the four-row table above. Click "MU — AI-cycle" for the second scorecard. Or download the free 8-tab workbook — both scorecards are pre-populated and recalculate when you overwrite Drivers cells.

I may hold positions in MU. See disclaimer — particularly the section on take-private analysis. Nothing on this page is investment advice or a transaction signal. Reverse-LBO outputs are illustrative.