Blog

SanDisk doubled in two months. Our QScore never left HOLD — here's why.

SNDK ran from ~$930 to a $2,274 close (about +140% at the peak) while its QScore sat at HOLD, MEDIUM confidence the entire time. Momentum scored 92, but risk scored 18 and value/growth lagged — so a doubling netted to a dead-neutral 50. A factor-by-factor look at why the model didn't chase the run.

Over roughly two months, Sandisk Corporation (SNDK, the flash-memory business Western Digital spun off into a standalone public company) went from about $930 in late April to a $2,274 close on June 22 — up around 140% at the peak, and roughly a double on almost any window you pick inside the run. It is one of the largest moves in our coverage universe this quarter.

The obvious question — “should I buy it?” — isn't one a quant model answers. The more useful question for anyone who scores stocks is sharper: what did the model do while the price doubled? The answer is the whole post. As SNDK ran, its QScore barely twitched — it sat at HOLD, MEDIUM confidence, parked between 49 and 51 the entire time. The model watched a stock double and refused to chase it. Here is why that's a feature, not a bug.

The score didn't move with the price

Every QScoring snapshot is frozen to git the day it's computed — an append-only, no-look-ahead record of what the model said in real time, not a backfit. So we can put the daily QScore next to the daily price and watch them diverge. These are our own committed snapshots, not a hindsight reconstruction:

Snapshot dateSNDK closeQScoreSignal
May 26$1,589.5549HOLD
Jun 5$1,559.3249HOLD
Jun 12$1,980.1051HOLD
Jun 18$2,184.7550HOLD
Jun 22$2,273.7350HOLD

Price rose about 43% across that window alone; the composite moved a single point. A score that tracked price would have screamed higher. This one didn't, because price is one input among many — and the others were pulling the opposite way.

Why HOLD: the factors disagree

The QScore blends five factor categories. Splitting SNDK into its five scores shows exactly why a doubling nets to neutral:

FactorSNDK scoreWhat it's saying
Momentum92Screaming. Trailing returns, RSI, and the 50- vs 200-day moving average all line up behind the trend. The model fully sees the run.
Risk18Bottom-tier. A ~$930-to-$2,274 run with double-digit single-session swings produces enormous realized volatility, which the risk factor penalizes hard.
Value41Below average. After a double, the multiples are stretched — price ran faster than the fundamentals that anchor a valuation.
Growth41Below average. Real, but not keeping pace with the share price — the gap between the two is exactly what a stretched value score reflects.
Profitability59Slightly above average — the one genuinely supportive pillar, and not enough on its own to offset the value and risk drag.

Momentum at 92 and risk at 18 are almost mirror images: the same violent move that makes the trend irresistible to a momentum signal makes it terrifying to a risk one. Net them against below-average value and growth, add a modestly supportive profitability, and you land at 50 — dead neutral. The split also shows up in the horizon scores: SNDK's short-term score (57) sits above its long-term score (43), the model's way of saying “the trend is real right now, but the longer you hold, the more the valuation and volatility matter.”

A model that printed BUY here would just be a momentum tracker wearing a composite. The point of blending five factors is that they're allowed to disagree — and when a stock doubles on momentum while value, growth, and risk lean the other way, the honest output is HOLD, not a victory lap.

What the disciplined read actually is

Unlike a fresh IPO, where the model is starved of data and the right answer is “not enough history yet,” SNDK has years of prints. Its confidenceis MEDIUM, not LOW — the inputs are all there. This isn't the model saying “I can't see it.” It's the model saying “I see it clearly, and the factors don't agree.” That distinction matters:

  1. A high momentum score is a description, not a recommendation.It tells you the trend exists; it doesn't tell you it's durable. Paired with a risk score of 18, it's a flashing “this is a fast, dangerous tape” more than a green light.
  2. Watch whether the laggards catch up.The bull case isn't “momentum stays at 92” — it's value and growth climbing as earnings grow into the price. If the fundamentals close the gap, the composite rises for a good reason. If they don't, the move was multiple expansion that risk was right to flag.
  3. Let the volatility be the warning it is. A risk score in the teens after a double is the model telling you the position size that feels fine on the way up is the one that hurts on the way down.

None of this is a call on whether SanDisk is a good investment. It's the opposite: it's the model refusing to let a big number override the four factors that weren't along for the ride. The entire reason to score a stockinstead of reacting to its chart is so that a doubling and a HOLD can sit on the same page without contradiction — you can see the run, and see exactly why the model still isn't chasing it.

The data

Prices and QScores in this article are drawn from QScoring's own committed daily snapshots through the June 22, 2026 market close, cross-checked against settled end-of-day prices. SNDK refers to Sandisk Corporation, the flash-memory company that began trading independently after its separation from Western Digital. The score is recomputed every session — the live SNDK page always shows the current read.

Related reads

This article is for informational purposes only and is not investment advice. QScoring does not hold a position in any security mentioned.

Discussion

Comments are powered by GitHub Discussions. Sign in with GitHub to join the conversation.

← All posts