DKS Cracks Athletic Retail; Intuit Drops 10% After the Close
US Stocks · Options · News · Views
The Indexes Closed Green; The Internals Did Not
Tuesday, August 25, Treasury yields and crude fell together, and the four major indexes finished higher. The Dow gained 0.30% to 53,577. SPX rose 0.32% to 7,677. Nasdaq rose 0.66% to 26,151. Semiconductors joined: SOX gained 1.44% to 11,588, with 24 of 30 names up. NVDA bounced from 208.48 to 213.05, up 2.19%. AMD rose 4.91% to 479.
The close was tidy. The vote was not. Inside SPX, 207 stocks rose and 294 fell. Only six of eleven sectors finished higher. On days when SOX holds up, the rest of the tape often looks thinner. Volume was light. The chip bounce was light too. The 10-year yield eased to about 4.64%. Bonds helped lift the averages. Single names did not confirm.
The two macro prints were not bright. August consumer confidence fell to 89.4 from 90.2, below the 90.2 forecast. The present-situation index rose from 114.4 to 121.2. The expectations index dropped 5.8 points to 68.2. Jobs look a bit better now: 27% said jobs are plentiful, versus 24.4% in July; those who said jobs are hard to get fell from 21.7% to 19.5%. The look-ahead on business, labor, and income turned more pessimistic. July new-home sales plunged 10.5% to a 607,000 annual rate, below the 620,000 estimate. Confidence cracked in the outlook. Housing did not find a bid.
Consumer stocks opened the gap. DKS cut the outlook hard and finished down about 30%, dragging Nike and LULU with it. In software, INTU was hit to about 321 after hours, down roughly 10%. Indexes up, consumer cracked, chips bouncing on light volume. Wednesday brings NVDA after the close and PCE the same day. Tuesday’s green candle is not a direction.
Athletic Retail Split First; Nike Still Holds 38.40
DKS printed Q2 sales of $5.59 billion, a touch below about $5.64 billion expected, and adjusted EPS of $3.53 versus $3.76. Full-year EPS was cut to $10.94–$11.94 from $13.27–$14.27. Sales were cut to $21.9–$22.2 billion from $22.1–$22.4 billion. Core stores still work. Newly acquired Foot Locker comps fell about 3.6% in the quarter, and the full-year comps guide moved from growth to flat to down 2%. Promotions got sharper. Legacy and retro footwear missed. Consumers are more cautious. The channel marked prices down first. Brands get priced with it.
NKE fell 3.12% to 39.48 and is still holding 38.40. Into May 2027, management’s signal is clear: stabilize first, then talk about lift. After the five-year compound growth rate rolled over, the old high forward multiple no longer fits. It looks cheap until you cut the multiple. Then the price looks more reasonable. Lose 38.40 and the next map is 2014 support. A decade-old line is not a high-confidence floor. With inflation still high, discretionary spend keeps getting cut. Money that went into data centers is money that did not buy sneakers. Consumption used to be 70–80% of US growth. A large slice is now AI infrastructure investment. If consumption slows further and that investment stops, GDP gets ugly.
LULU sat near 118, down about 4%. The current fiscal year is still a recession print. Whether growth returns in a year depends on the numbers coming back. 109 to 125 is a newly formed shelf. Below that, there is almost no usable step. Everything above 118 is supply. The weekly chart stepped down with few real bounces. Overhead supply sits in layers. If you hold it, budget one to two years. The wait is for earnings, not one rebound bar. This is no longer one retailer’s accident.
Intuit Cut Price to Defend Share — and Changed the Yardstick
INTU’s fourth quarter was not the problem. Adjusted EPS was $4.03 and revenue $4.35 billion, above $3.58 and $4.27 billion. The dump was the new year. Fiscal 2027 revenue was guided to $23.3–$23.5 billion, up 9–10%, below about $23.7 billion expected and below the 14% growth just printed in fiscal 2026. Non-GAAP EPS of $22.88–$23.12 sits far under the old-basis Street number near $27.34.
The gap is not only a sudden operating break. From August 1, non-GAAP no longer excludes share-based compensation. The company treats it as a recurring cost. Fiscal 2027 non-GAAP EPS includes about $5.81 of SBC. Old-basis $27.34 and new-basis $23 do not compare. The accounting moved first. Q1 is soft too: revenue $4.29–$4.31 billion versus about $4.36 billion expected; non-GAAP EPS $2.44–$2.48, against an old-basis figure the Street had near $4. Even aside from the convention, growth slowing from 14% to 9–10% is a downshift. AI software is getting stronger. Share has to be defended with competitive pricing. That means cutting price. Price cuts without a higher revenue-growth guide send margins lower without volume filling the hole. The fundamental case shifts from still-accelerating to a new year that is likely flat, low-single-digit, or slightly worse.
The cash session closed 357.46, down 3.37%. After hours it was hit to about 321. The 371 stop was already triggered in February; the book moved into MSFT. 321 is stacked support: it cannot clear overhead, and the downside pocket is not wide. It looks slow. The July 2027 forward band is 459 to 689, median 574. After the reset, fiscal 2027 versus fiscal 2026 is unlikely to deliver another major raise. Software still needs MSFT to lead. INTU already changed the growth story.
Buybacks May Squeeze CTAs; Do Not Sleep Through September
The 10-year and 30-year yields both pulled back. Bonds rallied; yields fell. The Treasury General Account is about $950 billion, and long-end buybacks have already been sized up. If buybacks become routine, Treasury credibility takes a hit. The short-term read is positioning. CTA global bond shorts are about $155 million DV01, near a multi-year extreme. Longs are already thin. If bonds keep falling, there is little left to sell. A roughly two-standard-deviation rally could force about $150 million DV01 of covering plus re-buying inside a month. Bessent ran a hedge fund. He knows how shorts get squeezed by their own book. Suppressing the long end and lifting bond prices may be an attempt to force that cover. That is a motive, not a published order.
If the two-sigma move prints, the 10-year could be pushed from about 4.64 toward about 4.3, the prior low. Midterms are a little more than two months out. When yields drop, mortgage and credit rates feel easier. That is political credit before the vote. It is also friendly for equities in the short run. After the election, perpetual buybacks are not a policy. Reality returns. Treat September through November separately from the post-election tape.
August SPX volume is thin, with about a week left. September and October are historically the higher-vol months in the sample since 1974. Midterm years make that sharper. In the all-year sample, October is often the peak. In election years, September already exceeds August, and October jumps again. Vol is two-way. It is not an automatic crash and not an automatic melt-up. Down then up, or up then down, both count.
NVDA put in a bounce pattern near 213. Before the print, it does not mean much. Some will bet 0DTE on a large open. The direction is still Wednesday after the close: does the guarantee continue, and does spend keep expanding. July PCE hits during the day. Today’s light-volume chip bounce only handed the question to Wednesday night. Consumer already cracked. The index is still grinding near the highs. August can sleep a few more days. From September, do not treat vol as a surprise.