Clearly demand is tight until 2028, so why are memory stocks falling every day?

This is a question from the backend memory guy. It's also the most frequently asked question in the backend over the past couple of days.

"Memory shortage until 2028, so why have Micron and DRAM been dropping like this recently? I'm holding and feeling anxious—should I cut my losses and run?"

I completely understand that contradiction. The fundamentals are ridiculously strong, but the charts are awful, and when two signals don't line up at all, people get shaky.

Everyone knows the memory fundamentals. Every company's earnings and every institution are raising their forecasts, and DRAM demand growth needs to reach 12% to keep up, but actual supply growth is only 7.5%.

Apple itself couldn't absorb the memory price hikes and ended up turning to CXMT for chips. If even Apple is scrambling to find supply worldwide, you tell me whether there's a shortage or not.

On the fundamentals side, the outlook points to three years of certainty.

So why are the charts still falling?

Because the charts reflect the current chip structure.

That 2x leveraged SK Hynix ETF listed in Hong Kong has ballooned to $15B, becoming the world's largest single-stock leveraged product.

You need to know that leveraged ETFs tracking Micron, NVIDIA, or Tesla have never exceeded $10B. Only this Hynix one has hit $15B.

A massive number of retail investors are using 2x leverage to bet on a short-term memory rally. As soon as the price pulls back slightly, leveraged positions start getting liquidated, the liquidations hammer the price even lower, triggering more liquidations.

So the short-term crash you're seeing has nothing to do with whether memory is in shortage. Fundamentals drive three years, but leverage might drive three days? Three weeks? You're holding a three-year thesis but losing sleep over three-day volatility. It's not necessary.

Once you understand this, what to do becomes clear.

If, like me, you're holding DRAM or the underlying stock without leverage, then this sell-off caused by leveraged cascading is not a bad thing—it's others helping you wash out weak hands.

If you're itching to jump in with 2x leverage at this level to catch the bottom, I'd advise you to cool down. You might be right on direction, but you could die on timing. Leverage won't give you three years to wait—it only takes one pullback to take you out, even if you're ultimately right.

This is the cruelest part of the current memory cycle. The direction is the most certain in the entire market, but that certain direction is packed with leveraged traders, turning it into a casino with terrible odds. Being right doesn't guarantee winning—only being able to hold does.