Bernstein just released what I think is one of the most important research notes on the memory industry this year. $纳指100ETF(QQQ)$ $闪迪(SNDK)$
Most headlines focused on one number:
A Bull Case valuation of $4,400 for SanDisk.
But in my opinion, that's not the real takeaway.
The report spends far more time explaining why this memory cycle may be fundamentally different from every cycle before it.
A year ago, when SanDisk was trading a fraction of today's valuation and most investors still viewed NAND as a deeply cyclical commodity business, Bernstein was one of the very first firms on Wall Street to publish a $1,000 price target.
Back then, many thought it was far too aggressive.
Looking back, they were simply early.
This time, the report focuses on one topic:
Long-Term Agreements (LTAs).
Many investors hear "LTA" and assume it's nothing new.
Bernstein argues the opposite.
Today's LTAs are completely different from the contracts that defined previous memory cycles.
1. Pricing has fundamentally changed.
Historically, memory pricing moved almost entirely with the spot market.
If prices collapsed, suppliers had little protection.
Today's contracts increasingly include price floors, price ceilings, or hybrid pricing mechanisms that limit both upside and downside volatility.
Micron has already disclosed that most of its Strategic Collaboration Agreements (SCAs) include explicit pricing bands.
SanDisk hasn't disclosed its floor price publicly, but Bernstein estimates the implied floor is roughly $0.29/GB, close to today's spot pricing.
That changes the economics dramatically.
Instead of ASPs collapsing every downturn, future pricing may become significantly more stable.
For the first time, memory pricing is beginning to look less like a commodity market.
2. Financial guarantees may be even more important than the contracts themselves.
This was probably the most interesting section of the report.
Traditional Take-or-Pay agreements always had one major weakness.
If a customer walked away, suppliers had to pursue legal action.
Expensive.
Time-consuming.
And often ineffective.
That's no longer the case.
Micron has already secured roughly $22 billion in cash commitments and letters of credit.
SanDisk has obtained more than $11 billion in financial guarantees.
But Bernstein makes an even more important observation.
The protection actually becomes stronger over time.
Imagine a $100 billion supply agreement backed by $22 billion of guarantees.
At the beginning, that only covers 22% of the contract.
Several years later, suppose only $30 billion remains to be delivered.
The guarantee is still $22 billion.
Now it effectively covers nearly 70% of the remaining obligation.
In other words:
The closer customers get to the end of the contract, the more expensive it becomes to walk away.
That's why Bernstein repeatedly emphasizes that the real protection isn't at the beginning of the agreement.
It's at the back end of the contract.
3. Contract duration has completely changed.
Historically, memory contracts were short.
A few months.
One year at most.
Revenue visibility barely existed.
Today, that's changing rapidly.
Micron's SCAs now extend up to five years, while SanDisk is increasingly signing three- to five-year agreements.
Capacity is reserved.
Pricing is protected.
Demand is committed.
That's something the memory industry has rarely experienced before.
This leads to Bernstein's biggest conclusion.
The real story isn't that SanDisk could be worth $4,400.
The real story is that the business model of the memory industry may be changing.
For decades, memory traded at low valuation multiples because earnings were unpredictable.
Margins swung wildly.
Cash flow disappeared during downturns.
If a growing portion of industry revenue shifts to long-term agreements,
revenue becomes more predictable.
Margins become more predictable.
Cash flow becomes more predictable.
And eventually, valuation multiples should change as well.
In Bernstein's view, memory is gradually evolving from a classic commodity business into a more infrastructure-like industry.
That's why they raised not only their EPS forecasts, but also the multiple they're willing to assign.
Their Base Case implies roughly $3,000 per share.
Their Bull Case points to approximately $4,400.
Whether those numbers prove achievable is still uncertain.
But one thing is becoming increasingly clear.
Wall Street is no longer debating only how much memory companies can earn.
It's beginning to debate whether the entire industry deserves to be valued differently.
Not financial advice.
Comments