Altair
Skip to content

Trading Analysis Report: NKE

Analysis date: 2026-10-14 · Model: claude-opus-5-5 · AI-generated, not financial advice.

  • Analysis date: 2026-10-14
  • Rating: Underweight
  • Generated: 2026-10-15 00:28:11
  • TradingAgents 0.6.0: anthropic, deep claude-opus-5-5, quick claude-sonnet-5-5
  • Analysts: market, sentiment, news, fundamentals; research debate rounds 5, risk debate rounds 5
  • Data vendors: core_stock_apis yfinance, technical_indicators yfinance, fundamental_data sec_edgar,yfinance, news_data yfinance, macro_data fred, prediction_markets polymarket

V. Portfolio Manager Decision

Rating: Underweight

Executive Summary: Underweight NKE. Cut exposure to about 0.5x a standard allocation, with a floor of 0.33x, selling in slices of one-sixth of a standard allocation and adding no new money at $35.30. If you hold none, don't buy. Today, once the open settles, sell any excess above 1.0x plus the first sixth, using a limit a few cents through the live bid. After reading the 10-K, 10-Q, Q1 release and call this week, rest the second sixth at 36.4-36.9 and the third at 37.2-37.6. A clock forces those out by mid-November and Dec 1, or earlier if the confirmed Q2 report date is earlier. Also sell one sixth on each of the first and second closes below 33.85; all of these count against a single three-sixth budget. A weekly close above 37.59 with RSI above 50 and OBV confirming pauses the selling but never cancels the fundamental tripwires. A failed early-November hold test or a dividend cut takes you to the 0.33x floor, and a close below 31.97 forces a full re-underwrite. Horizon is 3-6 months, through the late-December Q2 FY27 print and one more quarter, with proceeds held in T-bills.

PRICE TARGET: $33.60, a valuation objective. Forward EPS of $1.68 x 20x = $33.60. I use 20x, one turn below today's roughly 21x, which is the top of the 17-20x bad-branch range both research sides called arguable. It also reflects multiple compression on a shrinking-revenue turnaround with the 10-year above 5%. That is about 4.8% below 35.30. Adding about $0.82 of dividends over six months gives roughly -2.5% total return, consistent with expected value below the hurdle. Cross-checks: the lower Bollinger band at 33.82, the 10-02 closing low at 33.87, and the HSBC $34 target cited in the debate. The bear branch (clean EPS of $1.75-1.80 at 15-18x) gives $26-32. The bull half-recovery gives $42-46.

Current Price: 35.3

Price Target: 33.6

Confidence: Medium

Time Horizon: 3-6 months

IV. Risk Management Team Decision

Aggressive Analyst

Aggressive Analyst: I'll start with the obvious objection, because I expect it from both of you: "Sell a stock that's already down 50% from its 52-week high, with 4.6% yield, $6.9B in cash and washed-out sentiment? That's the textbook place to be patient." I think that's the comfortable answer, and it's wrong here.

Start with what you're being paid to hold this. The expected value in the research plan is roughly -1% to +4%, which is about the risk-free rate with the Fed funds rate at 3.75%. For that, you carry a stock with a 3% daily ATR and three gap-downs of -10.5%, -15.5% and a 7.4% open on 142.7M shares in the last year. And the 4.6% dividend isn't the cushion people think. The 10-year Treasury yields 5.21%, so you'd get paid more to own the government. The payout is also uncovered: roughly $2.4B a year against $2.18B of FY26 free cash flow, with Q1 FY27 free cash flow at about -$64M. That's about 98% of forward EPS, paid out of a cash pile that's already $3B below its FY24 level. I'm not forecasting a cut, because nothing in the evidence says one is coming. But I won't call it a safety net.

The valuation case for patience is also weaker than it looks. Trailing P/E of 17x depends on a Q4 FY26 quarter with a 49.2% gross margin, when every adjacent quarter ran 40-43%, and $1.07B of net income against about $675M for the other three. If that was a one-off, clean EPS is around $1.75-1.80, so you're paying about 20x for revenue that fell 4.3% last quarter and sits 10% below the FY24 peak. I'll concede the bridge isn't in the evidence, and that's exactly why this is a trim to about half and not an exit. If Q4 turns out to be durable, I still own half. But I'm not waiting for permission to act on the more probable branch. And "down 80% from the high" is an anchor on a 2021 price. It tells you nothing about what the business earns today.

For the contrarian bulls, I'd go through your evidence one piece at a time. The weekly and monthly TD-9 buy setups are real, but they have no confirmation signal and may be provisional until the bars close. Insiders did buy, about $9.2M, but they're 17% underwater on the April buys and about 40% underwater on December, and nobody has bought in six months. Insider buying has been a poor timing signal here. The "washed-out sentiment" point is also weaker than it looks, because estimate revisions are still falling. BofA downgraded and HSBC cut to $34, which is below spot. Stocks rarely bottom while the targets are still being marked down. And I'm not leaning on StockTwits, since 17 of 18 messages are one put-holder talking his book. I'd ask you not to lean on the one "hung tough on a red day" post either.

On the technicals, the stock is below the 50 and 200 SMA with the 50 under the 200, and SuperTrend is down on all three tiers. OBV hasn't moved while price rose 3.9%, and volume has faded to 19.4M, the lowest since August. That's a relief bounce, not accumulation. ADX falling from 36.7 to 22.4 is just a gap impulse being absorbed. It doesn't signal a reversal. And if the daily SuperTrend flips at 36.38, the weekly stop is still at 42.88, and higher tiers outrank lower ones.

The macro makes this worse. 97% odds of no cuts in 2026, 82% odds of another hike, a 10-year above 5%, oil near $103. Recession odds are only 8%, so there's no flight-to-quality bid coming. What hurts is valuation compression on long-duration turnaround stories, and NKE is exactly that.

Now the objection from the other direction, which I expect from the conservative side: "Sell it all now, at market." I'd say no, and the reason is price. Selling at 36.4 against a 37.59 invalidation line and a 31.97 downside marker gives about 4.4 points of downside against 1.2 of risk, nearly 3.7 to 1, and still above 2 to 1 if you only target the 33.85 area. Selling at 35.30 turns that into roughly 1.5 to 1. The ladder into 36.4-37.6 is what makes the trade good, and the Dec 1 deadline caps the cost if the bounce never comes. Keeping 0.4-0.6x also preserves a cheap option on the turnaround, paid for by selling the bounce. A full exit gives that up, and I'd only go to 0.25x or out if the bridge shows a one-time gain or we close below 31.97 on heavy volume.

To the neutral side, "wait for the Q4 bridge and earnings" sounds balanced, but the last three big drops arrived as gaps, and stops won't catch those. Waiting means taking another gap without compensation, with the next scheduled report likely late December and not yet in the evidence. Selling in tranches ahead of it is how you reduce exposure to that event without paying for it.

The bold part of this plan is being willing to cut a stock when the retail story says it's too cheap to sell, and doing it at a price that makes the math work. I'm sticking with Sell. Aggressive Analyst: Several of my lines didn't survive your arithmetic, so I'll fix those first.

Neutral is right that my 3.7 to 1 leans on a single intraday print at 31.97. The honest figure is about 2 to 1 on 33.85 from a 36.4 sale, and about 0.6 to 1 at market. I'm also dropping "cheap option," because a long with no floor isn't one. I'm dropping "stocks rarely bottom while targets are cut" too, since price often turns before revisions and my case doesn't need it. The 17x versus 20x argument goes as well. The market already pays 21x forward EPS of $1.68, so clean earnings aren't news the price is ignoring. The multiple says the stock is priced for a recovery that revenue down 4.3% hasn't delivered. That's a reason not to add, not proof that selling is right.

The earnings-cadence point is the best contribution in this debate, and it weakens my claim that waiting means taking another gap. If those gaps were report reactions, the next one is probably mid-to-late December, so I overstated the hazard for the next six weeks. I'd still hold the inference loosely. The October 1 bar carried 112M shares before the October 2 gap, which isn't what a clean after-the-close report looks like, and the August 17 break came on 58.7M shares with no report. Unscheduled risk is smaller, not zero. Still, nothing forces a market order at the midline with the daily z-score at -0.06.

Conservative, your front-loading cost is mispriced. The 1.5% is the cost of selling the whole half at market, and only in the roughly 57% of paths where the ladder would have filled. On Neutral's math, a one-third slice at market costs about 0.5% in those paths and saves about 0.7% in the rest. That's near zero in expectation, so what you're buying is variance reduction. Neutral's first rung at 35.8 to 36.0 buys the same thing for about a quarter percent and fills about seven times in ten before the lower band. I'm adopting that rung and the thirds, because it's a better plan than mine. Paying double to feel safer isn't bold. Neutral also caught a tension in my position, since a strong negative-drift view argues for faster execution. But my view is a tilt, not a conviction. An expected value of -1% to +4% is a no-edge view with a skew, and the first rung is the right compromise for that.

Your gap math also needs a correction. The 3.7% on the half for the smallest repeat uses the 7.4% open gap. On October 2 the stock opened at 32.55 but closed at 33.87 against a prior close of 35.15, which is 3.6% close to close. That scenario is closer to 1.8% on the half. The 10.5% and 15.5% gaps were close to close, and they're real. The half I keep takes that risk, and I'm choosing it knowingly. Your point about the December 1 deadline is fair: it caps time, not cost. But with thirds, only the last sixth of the original position is ever sold into weakness at the deadline.

The macro calendar, with the FOMC, the midterms and CPI, argues about the size of what I keep, not the speed of getting there. The retained half sits through every one of those dates either way. The rate headwind works on the multiple slowly, and beta of 1.10 doesn't capture that.

Why keep a half at all? Nobody here has an edge to be fully out. The weekly and monthly z-scores are -1.30 and -1.56, both TD-9s are complete, and the plan's own range tops out at +4%. The option I wanted is in the cash, not the shares. Sell the bounce into bills, and if December shows margins holding and revenue stabilizing, buy back after the gap with information, possibly below where I sold. If it doesn't, I sold the bounce. The price is forgone upside on the sold half. A collar on the retained piece would cap the upside that's the only reason to hold it, and a put spread leaves a 15% gap partly uncovered. Neither of us has seen option prices, and the cheapest protection is the smaller position. I'd revisit a hedge only if a large piece is still held heading into December.

On triggers, 31.97 is too far and I concede it. A close below 33.82 is only about 1.4 ATR away, and the stock closed at 33.87 eight sessions ago, so I'd rather not cut on one noisy print. I'd cut toward 0.25 to 0.3x on a second close below, or one on expanding volume.

On the bridge, I'll pull it this week, but the run rate is already visible. If 49.2% gross margin were the new normal, Q1 FY27 should show it. It printed 42.8% and $712M, in line with the adjacent quarters and the $1.68 forward EPS. One quarter isn't conclusive, so the filing still has to say why. But I'd want a structural explanation before going to 0.6x. This also makes 0.4 versus 0.5 moot, since the filing will be read well before the upper rungs could fill.

So I'm still at Sell, as a trim in thirds. The first rung goes at 35.8 to 36.0, the second at 36.4 to 36.9, and the third at 37.2 to 37.6. The working target is 0.5x of a standard allocation, with no new money at 35.30 and stragglers finished by December 1. Proceeds go to bills, and a weekly close above 37.59 with RSI over 50 and OBV confirming stops the selling. All sizing is relative to a standard allocation, since I don't know the actual position. If you hold nothing, this simply means don't buy here. Aggressive Analyst: Most of this is settled, so I'll say where you've moved me and then argue the few places I still disagree.

Conservative, you're right on the first sixth, and I was wrong to call it paying double to feel safer. Neutral's numbers settle it. A resting order at 35.9 fills about seven times in ten and gains about 0.3% of allocation, and the other three times the unsold sixth ends up about 0.7% worse. That's a wash, which is what any limit order on a driftless walk gives you. The rung only wins if the bounce keeps going, and I've spent two rounds arguing it won't. Your decomposition also lands. The three gaps compound to about -27% and the stock is down 51%, so ordinary days did roughly half the damage. I can't justify leaving a sixth sitting above the market while that drift goes on, so I sell it now. This changes the trader's 36.4 entry for that sixth only.

The upper rungs aren't a discount either, but they aren't a plan to wait for a rally. They're one door of two. If price rallies into 36.4 to 37.6, I sell there. If it breaks down, the triggers sell. Either way I end near 0.5x, and the Dec 1 deadline covers the path where it goes nowhere. The old 37.59 "stop" is really the line where I stop selling and reassess, since this is a trim and not a short.

On the triggers I'm with Neutral. After one sixth is sold I'm at about 0.83x, so a single close under 33.85 taking me to 0.4x means selling about 0.43 of the position at the next open. A bounce to 35.30 then costs about 2% of allocation. Your own argument was that whipsawing a sixth costs a few tenths of a percent, and that favors sixth-sized slices. The first close under 33.85 sells a sixth. A second close, or an expanding-volume close, sells another. A heavy-volume close under 33.82, or a bad filing, goes toward 0.3x. You're also right that a reactive order fills at the next open, so my close-to-close correction doesn't help a seller. I'll stop leaning on it.

On 0.4x versus 0.5x, three sixths leave you at 0.5x, so 0.4x isn't a default. It's an extra tenth the filings have to earn. Your cash-flow arithmetic is right, and I'd been using the full-year figure. Q4 FY26 free cash flow was about $1.5B against $685M for the nine months, so without it the run rate is under $1B annualized. Q1 operating cash flow was $135M against $222M a year ago. But FY25's Q4 was only about $363M, so a big Q4 isn't seasonal, and it's the same quarter with the 49.2% gross margin. That makes inventory and working capital the lines to read in the 10-K, and the CFO change gets an 8-K check. It kills the 0.6x branch for me. It doesn't take me below 0.5x by itself, with $6.9B of cash against a $2.4B payout. I'm still not forecasting a cut, but I'd add the dividend declaration ahead of the Dec 1 ex-date to the calendar.

On December I'll take Neutral's rule over Conservative's number. You're right that if I plan to rebuy after the gap, the half I hold through it is the half I hold without the information. But a 10% move on a half is a tail, and you're comparing it to an expected value. The tail has two sides, and the last three being down tells us to respect variance, not which way it goes. So I'll price the January put spread this week with the strikes you described and decide before the print. If the filings are clean and the quote is fair, I hold the half. If the filings are murky or the quote is rich, I sell another sixth, because implied vol into an earnings date charges you for exactly the gap you're hedging, and selling has a known price. I still wouldn't use a collar, since it caps the upside that's the only reason to hold the half.

Neutral asked for the rebuy rule in advance. These thresholds are my judgment, not derived from the evidence. If the next report shows revenue declining by less than Q1's 4.3%, gross margin around 42% or better with no one-time item, and operating cash flow back to positive, I rebuild toward a standard allocation in steps. I do that whether or not the price is above where I sold, because that is the cost of having been out for the gap. If revenue is worse or margin slips below about 41%, I stay at or under 0.5x.

The plan, then, sizing relative to a standard allocation since we don't know the book. I sell the first sixth now, read the 10-K, 10-Q, Q1 release and call this week, and then rest the other two sixths at 36.4 to 36.9 and 37.2 to 37.6. The downside slices fire as described. I aim for 0.5x and go to 0.4x only if the filings say so. I add no new money at 35.30, park proceeds in bills, finish by Dec 1, and stop selling on a weekly close over 37.59 with RSI above 50 and OBV confirming.

What survives from my first round is that the cash is the option, not the shares. Selling a bounce into bills costs me some upside on the sold half. That's a cheaper price than carrying full gap exposure at 21x forward earnings while revenue is still falling. It's still a Sell, as a trim. If you hold nothing, it simply means don't buy here. Aggressive Analyst: Most of the plan is settled, so I'll take what I'm conceding first, then argue the three places I still disagree.

Conservative, you're right about my rebuy rule. Q1 operating cash flow was positive at $135M. What went negative was free cash flow, about -$64M, because capex was $199M. I named the wrong line. But I'd take Neutral's fix over yours. Requiring 42.8% gross margin in Q2 asks for a 220 basis point jump from last year's 40.6% quarter, which is a seasonality error, not discipline. So I rebuy when gross margin is up year over year, the revenue decline is narrower than Q1's 4.3%, free cash flow is positive, and the outlook hasn't been cut. I rebuy in sixths, not all at once.

You're also right on two other things. The hedge decision belongs in the first half of November, because event premium is fattest at the print. And the dividend declaration is a dated event we hadn't calendared. Your gap math holds too. At 0.83x, a repeat of the 10.5% gap costs about 8.7% of an allocation, and the downside slices don't change that, because a close under 33.85 comes after the gap. Slices manage drift and only size manages gaps. That's the case for Neutral's clock, which I'm adopting. The second sixth goes at market by mid-November and the third by December 1. There is one budget, so each sixth is sold once, whichever door or clock gets there first.

Neutral, you're right on the bridge. The trader's trigger to cut to 0.25x if Q4 was a one-time gain is retired. The price is 21x a forward $1.68 that Q1's $0.48 already supports, so confirming the one-off isn't news. While reading the filings, check what the stock did the day Q4 was reported. I can't tie a date to it from the evidence, but the tape went from 46.18 on June 12 to a 40.32 close on June 26. If that was the reaction, the market saw a 49% gross margin and sold it, which says more about what it believes than a footnote will. I'm also keeping your earnings-cadence inference loosely. October 1 carried 112M shares before the gap, and August 17 broke on 58.7M with no report in our evidence.

Now the disagreements. First, Conservative, selling the next sixth at market because the filing calls Q4 a one-off is trading on news already in the price. If what you want is a third, argue it as a size, not as a reaction to a filing. And the target has ratcheted from 0.5x to 0.4x to a third while the tape sits at the Bollinger midline with a daily z-score of -0.06. Your cash-flow subtraction was real new information, and it moved me off 0.6x. It doesn't move me below 0.5x, with $6.9B of cash against a $2.4B payout.

Second, our sixths land on 0.83, 0.67, 0.5 and 0.33, so let's stop quoting 0.4x and 0.25x. The base is a half and the floor is a third. The floor moves only on a new fact, such as a dividend warning or a heavy-volume close under the 31.97 low, which would break the whole range.

Third, December. "No edge, so variance is unrewarded" argues for zero, not a third. The case for keeping something isn't edge, it's two-sidedness. The weekly and monthly z-scores are -1.30 and -1.56, and both TD-9 setups are complete. Q1 gross margin rose 60 basis points on falling revenue, and directors and the CEO put $9.2M in at much higher prices. None of that proves a bottom, and the insiders are 17% to 40% underwater. But a half carries about 7.8% of an allocation on a repeat of the 15.5% gap, and that is survivable. So I take Neutral's test, written down now so it isn't a feeling in December. Hold the half through the print only if three things hold. The Q1 outlook is consistent with $1.68, the dividend is declared unchanged, and the January put spread quote is inside a maximum premium I set before I look at it. If any one fails, sell the sixth. Nobody has seen option prices, so I won't pretend the quote is cheap. I still wouldn't collar the half, since that caps the upside that's the only reason to hold it.

Here's the plan, sized against a standard allocation since we don't know the book:

  1. Sell the first sixth now, around 35.30, and add nothing.
  2. This week, read the 10-K, 10-Q, Q1 release and call. The items to read are:
  3. outlook language
  4. inventory and working capital, behind operating cash flow of $135M versus $222M on flat earnings, and behind Q4's roughly $1.5B of free cash flow versus $363M a year earlier
  5. dividend wording
  6. the reason for the CFO change

A bad read sells the next sixth at market that day. 3. If the read is clean, rest the second sixth at 36.4 to 36.9 and the third at 37.2 to 37.6. 4. The clock sells the second sixth by mid-November and the third by December 1. 5. The downside slices sell a sixth on the first close under 33.85 and another on a second close or an expanding-volume close. 6. Price the put spread this week, and apply the hold test in the first half of November. Confirm the actual report date, since it isn't in our evidence. 7. A weekly close over 37.59 with RSI above 50 and OBV confirming cancels the clock and any extra sixth. 8. Proceeds go to bills.

If you hold nothing, it simply means don't buy here.

It's still a Sell, as a trim. The cash is the option, and I'd rather meet December holding bills and a half-sized position I can afford to be wrong on than hold full size on a hope. Aggressive Analyst: Since this is the last pass, I'll say what I'm taking from each of you, where I still see holes, and then put the plan in one place so it can be executed.

Conservative, you're right that the hold test as written couldn't fail. I'm taking Neutral's repair. Write down today's forward EPS of $1.68 now, and a drop of a few percent from it by the early-November decision date is a fail. The tolerance is my judgment, not a derived number. Vendor forward estimates may lag this week's BofA and HSBC moves, so a late drop counts as information. You're also right that a pass on the hedge means you buy the spread, not that you admire the quote. And a flat $0.41 is a pass, a cut or hedged wording is a fail, and a raise earns nothing extra on a payout near 98% of forward EPS.

On the reclassification, I'd note that it makes the bridge less valuable. A reclassification big enough to turn a 40-43% margin into 49% would also show up in Q1's 42.8%, and it doesn't. So Q4 is either a one-time item or an artifact of deriving it as annual minus three quarters. The price at 21x forward already treats it that way. A smaller reclassification could still contaminate the 60 basis points, though. So I'm dropping that number from my reasons to keep a half until the filing says the comparison is like for like. My case for keeping a half doesn't need it.

Neutral, you're right on the June check. I raised it as something to look up, not a pattern, and "the market has sold four prints" rests on prices nobody has tied to a date. It doesn't count until the report date is confirmed, and that goes first on the reading list because the whole December schedule hangs on it. You're also right that slices are sixths of a standard allocation, not of whatever you hold. The target is a level, a half. Someone at 1.5x has six slices to sell, and someone at or under a half sells nothing and keeps only the downside triggers.

The plan now names "a sixth" at five different triggers, so I want to pin down the budget. Three sixths take you from a standard allocation to a half, and the rungs, the clock, the closes under 33.85 and a bad filing read are only ways of deciding when each of those three goes. A bad read pulls the next sixth forward and doesn't add one. A fourth sixth, taking you to the floor of a third, comes only from a failed hold test or a genuinely new fact. Failing two legs of the hold test is still one sixth. Below a third means re-underwriting, which is also what a close under 31.97 means.

On December, I'll accept the burden of proof. I have to be able to write one sentence on why the retained piece earns its variance, and it can't use "down 80% from the high" or "sentiment is washed out," since we've retired both. Mine is that the outlook fits the $1.68 the price already reflects, the dividend is intact, and this is the piece that participates if December gaps up. If I can't write that on the decision date, the sixth goes. That's a stricter standard than I started with, and I'm fine with it. I still wouldn't collar the half, because that caps the only upside that justifies holding it. If the retained piece is too small for a clean 100-share spread, selling is the cleaner hedge.

On the rebuy, I'll take the cap of two sixths back after any single print, since one print is one observation by the same standard we applied to the gaps. The rest waits for a second confirming quarter. A weekly close over 37.59 ends the selling and rebuys nothing, because the weekly SuperTrend sits at 42.88 and higher tiers outrank a bounce. If December gaps up and I rebuy above where I sold, that's the cost of having trimmed. I wrote that down already.

Conservative, you're right that gap scaling cuts both ways. If the allocation is small, rushing costs little and so does waiting. If the book is concentrated, the front-loaded sixth and the floor are the whole point. None of us can see the book, which is why everything is in standard-allocation terms.

Here is what I'd stand behind. Sell the first sixth at today's open with a marketable limit near the 35.30 close, resting nothing above the market for it, and add no new money. This week, confirm the report date and read the 10-K, 10-Q, Q1 release and call. The reading list is: - the outlook language - inventory and working capital behind $135M of operating cash flow - dividend wording - the CFO change - whether the margin comparison is like for like

A bad read sells the second sixth at market that day. A clean read rests the second at 36.4 to 36.9 and the third at 37.2 to 37.6. The first and second closes under 33.85 each sell a sixth, the clock sells the second by mid-November and the third by December 1, and a weekly close over 37.59 with RSI above 50 and OBV confirming stops the selling.

In the first half of November, run the forward-EPS and outlook legs and buy the spread if the quote is inside a cap written down before looking. The dividend declaration is a tripwire, and a cut or hedged wording sells a sixth that day. Proceeds go to bills.

It's still a Sell, as a trim. I'd rather meet December holding bills and a half-sized position I can afford to be wrong on than carry full gap exposure at 21x forward earnings while revenue is still falling. If you hold nothing, it just means don't buy NKE here.