Delek's 117% Monster Just Bit Its Own Tail

# Delek's 117% Monster Just Bit Its Own Tail

Let's talk about Delek US, the refining-name that ran up roughly 135% year-to-date and then decided to introduce itself to reality.

The stock closed around $68–69 at the end of September. It hit a 52-week high near $82. So in the space of about six weeks, shares went from "what was this thing?" to "what is this thing?" and back again. If you bought the story in August, you're feeling it now. If you didn't, you dodged the worst of it and are watching from the outside at a discount.

Here's the setup, because it's actually instructive.

Delek reported a monster second quarter: $5.48 in earnings per share against consensus somewhere in the $2.20-to-$2.70 range, on revenue of $4.09 billion, up nearly 48% year-over-year. The driver was a roughly 197% surge in refining crack spreads and the system running at over 100% of nameplate capacity. That's not a good quarter. That's a quarter that makes you believe the cycle has permanently shifted in Delek's favor.

Then, last month, the company priced $400 million in zero-percent convertible notes due 2031. Markets read that as opportunistic de-levering. The tape read it as dilution. The stock dropped 6% on the news and hasn't recovered.

So where do we actually stand? Not at the top. Not at the bottom. In the messy middle of a stock that's forcing a decision.

The bull case is real and I won't pretend it isn't. Delek runs four inland refineries with about 302,000 barrels per day of nameplate throughput, holds a controlling stake in Delek Logistics Partners, and entered the second half of 2026 with its full refining system online and no planned turnarounds. DKL posted a record ~$143.5 million in adjusted EBITDA. Several firms raised price targets into the $81-to-$90 zone. On sales, it trades near 0.4x versus a rough industry average closer to 2x. Nothing about the operating business is broken.

Now the part that keeps me up at night, and it's the same part every ref-margin trader wakes up sweating about: none of it is durable.

Q1 2026 showed a negative gross margin and a $201 million net loss. The Q2 miracle was powered by crack spreads that cannot stay elevated forever—that's not a forecast, that's arithmetic on a cycle. And Delek is loaded. Total debt around $3.26 billion against stockholders' equity of roughly $186 million. That's a debt-to-equity ratio in the high teens, a current ratio under 1, negative working capital. When margins compress, that balance sheet doesn't absorb the hit—it amplifies it.

Add a few more things to the ledger and the picture gets less seductive. Insiders sold roughly 168,000 shares, about $11.3 million, in the trailing quarter. EVP Mark Hobbs and Director Gary Sullivan Jr. took out roughly $3.2 million of their own in August alone, all under pre-arranged 10b5-1 plans—so yes, technically routine. It still stings when the people who know the business most intimately are the ones cashing into a parabolic run. And the analysts who are supposed to be the adult room aren't agreeing on much: targets range from the low-$80s down toward $55, and even the EPS estimates for Q3 are all over the map, from around $2.05 to $5.26 depending on who's holding the pen. That's not a consensus. That's a debate with a scoreboard.

Technically, the stock already told you something. It rejected hard off the $80 area, broke below its 20-day moving average, and sits about 16% off its high. Relative strength is neutral-to-weak, not oversold. The model that's been tracking this name has been about as accurate as a coin flip in this regime—59% directional accuracy, which is exactly its naive baseline, meaning it has no real edge here. So don't treat any downtrend band as a price target; treat it as a scenario. The levels that actually matter are the ones humans will watch: support at the $60 round number, then $50; resistance at the $72-to-$75 zone where prior highs and analyst targets cluster.

My read? This is a momentum stock that ran on cycle euphoria and then tripped over its own leverage. The business is fine. The balance sheet is fragile. The tape has already corrected most of the froth, but "most" isn't "all," and the next catalyst—Q3 earnings on November 6—lands exactly where the disagreement will crystallize. If spreads hold, Delek prints another impressive quarter and the analysts who called $5.26 look like geniuses. If they roll over, that high-teens debt-to-equity stops being a footnote and becomes the story.

I'd rather trim into strength or wait for the $50-to-$52 zone to retest than chase a name that just proved how fast it forgets a good quarter. The people who own it clean will be the ones who understood, early, that a record refining margin is a weather event, not a climate.

Delek isn't broken. It's just telling you, in the most honest way a balance sheet can, exactly how much it needs the good weather to keep standing.