Altair
Skip to content

Trading Analysis Report: NKE

Analysis date: 2026-10-21 · Model: claude-opus-5-5 · AI-generated, not financial advice.

  • Analysis date: 2026-10-21
  • Rating: Underweight
  • Generated: 2026-10-21 13:13:48
  • 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 at 37.42: reduce exposure now, but do not exit. Cut to a residual equal to your tolerable loss (as a share of a standard position) divided by 0.20. That defaults to one-third of standard if you cannot name a number, and should never exceed half. Sell the excess immediately; at most a fifth of standard may rest on a single 38.05 limit for one week (to about Oct 28), after which it sells at market, with mid-November as the hard backstop ahead of the likely dividend declaration. If flat, do not initiate (no resting bid at 36.56); if already at or under the cap, do nothing. Hold the residual through the late-December Q2 print whatever price does: halve it on a daily close below 34.47, exit the rest on a close below 31.97 or on any dividend cut, and rebuild only after the print on pre-written gates.

Current Price: 37.42

Price Target: 33.6

Confidence: Medium

Time Horizon: 2-3 months

IV. Risk Management Team Decision

Aggressive Analyst

Aggressive Analyst: I'll open since nobody else has, and I'll take on the objections I expect from both of you.

The Sell is the aggressive call here, not the timid one. The stock is up about 10.5% off the 10-02 close of 33.87, retail is roughly 93% bullish among labeled StockTwits posts, and people are posting "40 this week" and "shorts can't make this red." The trade is to sell into that, at 37.40 to 38.80, not to wait for a stop to do it later. Selling into crowded enthusiasm is where the edge is.

The setup supports it. NKE is in a weekly and monthly downtrend, with stops at 42.88 and 60.64. The bounce has run into a falling 50 SMA at 37.18 and the upper Bollinger band at 38.05. The daily TD-9 is at -9, and OBV is lower than it was on 09-21 even though price is higher, so volume isn't confirming the rally. The payoff map also looks poor. The nearest ceilings are 38.05 and 38.77, only about 1.7% and 3.6% away. The downside levels are 36.56, 35.68, 34.47 and the 31.97 low, which are 2% to 15% away. To get paid for holding a full position, price has to clear 38.77 and OBV has to turn up. Neither has happened.

To the neutral analyst, who I expect to say the signals conflict: the monthly TD-9 of +9 is real, but the October bar is still forming, so it's provisional. The weekly tier carries more weight, and it says down. MACD crossing zero and RSI at 58 are lagging confirmations of a bounce I'm already selling into. Conflicting signals are why we trim and don't go flat or short. If I'm wrong and it squeezes toward 42.88, the residual half still participates.

On the improving margins, gross margin was up 60bps, but gross profit dollars fell $145M and revenue was down 4.3%. That is margin improvement on a shrinking base, not demand recovery. Forward EPS of $1.60 is below TTM EPS of $2.08, so 23.4x forward is the multiple to use. The 18x TTM figure is distorted because the TTM numbers don't reconcile with the quarterly data. Paying 23x for falling gross profit isn't cheap.

To the conservative analyst, who may point to the 4.3% yield and the insider buying as reasons to hold or buy: the dividend is covered about 0.98x by forward EPS and about 136% of TTM FCF, and Q1 FY27 free cash flow was negative. A high yield on a stock down 47% from its high is a warning as much as an attraction. The insider buying was real, about $8 to $9M, but the buyers are underwater, at 35-39% down on the December buys and about 12% on the April buys. Nobody has bought in six months, including on the slide from 42 to 37 and the dip to 32, and officers have been selling modestly since June. Insiders also haven't shown conviction near these levels.

If the conservative analyst argues for a full exit, the balance sheet is why I disagree. There's $6.9B of cash and a 2.08 current ratio, so the dividend isn't in imminent danger. Going to zero would also give up the turnaround optionality. Moves like -10.5%, -15.5% and +4.9% have hit this stock on heavy volume around roughly quarterly events. That cuts both ways. I'd rather keep a half position sized for a 15% overnight gap than sit in a full position or sell everything and miss a violent upside gap.

There are gaps. I don't have the 10-Q, the Q1 call, consensus, the Q2 date or tariff exposure. A conservative analyst may say that means don't act. I'd say unknowns on a stock with double-digit earnings gaps are a reason to carry less risk into the print, not more. The stop at 34.47 is the daily supertrend. A close below it puts all three timeframes down, and the residual should go then, or earlier on a dividend cut.

I can see a way I'm wrong. A daily close above 38.77 with OBV turning up would mean the bounce has real buyers behind it, and I'd revisit the sizing. Until then, with no position context, my guidance is: if you're at or above standard weight, trim toward half of standard into strength before the Q2 print. If you're flat, don't initiate, and don't leave a resting bid at 36.56. The plan is to sell the strength decisively and keep a small, defined-risk stake in the turnaround. Aggressive Analyst: You both landed real hits, so I'll give those up first.

Calling the leftover half "defined-risk" was wrong. 34.47 is a daily close-below trigger about 8% under the price, on a stock that has gapped 10.5% and 15.5% on over 100 million shares. It tells you when to leave, not what you lose. The dividend-cut exit has the same flaw, because a cut arrives as a gap. Only size limits the damage, and I'll stop calling it anything else. You're also right that I anchored on a fill at 38.05 when 37.42 is already inside my own 37.40 to 38.80 zone. A plan that needs one price can miss, so start selling now. I'm dropping the crowd-fade too. 14 labeled posts out of 15, from one afternoon with Reddit off, isn't an edge. The Sell stands on the weekly and monthly trends, the OBV divergence, the falling revenue and the 23.4x forward multiple. It doesn't need the daily 9 either, and I wouldn't hang timing on it.

Now where I'm holding. The half-versus-quarter fight is smaller than it sounds. Those 7.5% and 5% losses are percentages of a standard allocation, not of the portfolio. The 31.97 washout low is 14.6% below the current price, and the 42.88 weekly stop is 14.6% above it. So the gap between a half and a quarter is about 3.7 points of allocation in either direction. That's a judgment about which tail is likelier, not a reason to flinch.

The conservative's skew argument also proves too much. If three big gaps (-10.5%, -15.5%, +4.9%) show the stock only gaps down, the logical endpoint is zero, not a quarter. Nobody wants zero, because four events don't fix a skew, and October 2 showed buyers absorbing a flush. It opened at 32.55, hit 31.97 and closed 33.87 on 142.7M shares. If that's enough doubt to keep something, keep something that matters.

The residual is the turnaround option, and one print can move it. Net income fell only $15M on $145M less gross profit, after three quarters of declines of 30% or worse. As rough math, assume gross margin dips about 1.6 points from Q1 to Q2, as it did last year. Then revenue down another 4% means gross profit about 3% lower. If revenue just goes flat, gross profit is about 1.5% higher. That swing is what an upside gap would feed on, and the institutional headlines are about as bearish as they get, so I wouldn't bet that there's no upside surprise.

So I'll take the neutral's formula, loss tolerance divided by 0.20. It's the honest version of sizing for the gap. I'd plug in 8% to 10%, which gives 40% to 50% of standard. The conservative plugs in 5% and gets a quarter. That's the real disagreement, a risk budget and not a thesis, and it's the caller's number.

I'd keep the neutral's three tranches with two changes. First, front-load them. Make the first piece the biggest, not a third. The 50 SMA is still falling, and OBV has recovered only 119M of the roughly 406M shares it lost between 09-21 and 10-02, under 30%, while price has more than recovered. Resistance is 1.7% and 3.6% away, and the supports are 2.3%, 4.6% and 7.9% away. If I sell at 37.4 and I'm wrong, the cost is a squeeze to 38.77. If I'm right, it pays much more.

Second, the re-entry rule. Setting it now is right, but a weekly close above 38.77 with OBV back above -1.31B needs another 287M shares of net buying. That's about two and a half times what this whole bounce produced. It will probably trigger well above 39, which is the same lag you held against the conservative's fundamentals rule. I can live with that, because it's what confirmation costs. But I'd use it as the signal to cancel any unsold tranches, not to buy back. A buyback waits for the weekly supertrend flip above 42.88, or a print where the revenue decline narrows and gross margin holds near 42.8%. That avoids selling at 38 and buying at 39, because the residual participates in the meantime.

On 34.47, cutting the residual by half on a close below is fine, and the rest goes if the 33.30 lower band fails. That's a review level, and I won't pretend it protects anyone.

On the missing data, the 10-Q, consensus, Q2 date and tariff read are all fetchable. Don't use them as an excuse in either direction. Don't hold the first sale for them, and don't make the last sale without them.

I'm wrong if NKE closes a week above 38.77 with OBV climbing, and then I stop selling. Until then, sell the strength now, size the residual to a number you can survive a 20% gap with, and keep it big enough to matter if the turnaround is real. Aggressive Analyst: I'll start with what I'm giving up, because some of it was my error and not just emphasis.

First, insiders. I said nobody bought on the dip to 32. That low was October 2, and the insider feed stops at September 15, so I can't see that window at all. The slide from 42 to 37 mostly happened before the print, when insiders are usually blacked out. That silence told me nothing, so I'm withdrawing it. The conservative's reverse reading has the same flaw, because a feed that can't see October can't show an absence of October buying. Pull the Form 4s since October 1. If directors and the CEO bought into 33 to 36 the way they did after December 19 and April 1, that's the strongest bullish fact in the file and I'll say so. If there's nothing three weeks into an open window, the conservative's reading wins.

I'm also dropping the bearish headlines as contrarian fuel. They're titles with no bodies, the same grade of evidence as the 14 labeled posts I already gave up. OBV goes too. If 255M of the 406M it lost came in two days, and it books a 142M-share day as selling even though the stock closed well off the low, then the divergence is mostly two outlier days, not distribution I can read. I'm also retracting "stop selling above 38.77." A cap is a risk limit, not a market view, so a breakout changes where I finish selling, not how much. If it squeezes to 42, that's what the cap costs. And the conservative is right that 31.97 isn't a floor, since a 15.5% gap from here prints around 31.6.

Now something I think changes the debate. The fundamentals report already carries Q1 FY27, the quarter ended August 31, so that print has happened. Nike normally reports in late September, and the only event volume in that stretch is October 1 and 2, at 112M and 143M shares. So we've likely been arguing about numbers the market already priced, and the next scheduled gap is late December, about nine weeks out. That's unconfirmed, and it's the first fetch. If it holds, it cuts against the conservative's reading that the flush was the market rejecting the improvement. NKE is at 37.42 against 36.10 on September 21. The market sold the print for two days and bought it back over thirteen sessions. That's the hardest fact for my own side. It's why I'm selling a retrace into resistance and not shorting, and why there's a residual at all.

On size, the neutral's arithmetic settles most of it. A third at the worst gap we've seen costs about 5.2% of a standard allocation, and a quarter at a 20% stress costs 5%. So quarter versus third is just whether you stress beyond history, worth about a point of allocation. I'm moving off half. The default is a third, a quarter if the caller says 5%, and half only if they really name 10%. The conservative calls the residual insurance sized by what you can afford. Fine, that's exactly what dividing by 0.20 does, and the input belongs to the caller. What I won't do is drop to a tenth by reflex. That's an exit with extra steps, on a tape that just rose 10.5% on a quarter whose numbers were already out. The sentiment feed also mentions an investor event, unverified, which would be an upside gap before December that doesn't wait for a print. Add it to the fetch list.

On rebuilding, the conservative's best hit is that "the decline narrows with margin holding" is the story I rejected. The neutral's version is mine plus a free cash flow gate, which is an improvement. Operating cash flow was already positive at $135M last quarter, so the conservative's gate wouldn't gate anything. I'd tie the first step to what the Sell is actually about, which is gross profit dollars. The Q1-to-Q2 margin drop was 1.8 points two years ago and 1.6 last year. A Q2 margin near 41% means flat gross profit needs revenue down no worse than roughly 1% to 1.5%. That's my estimate, not a forecast. So the first step back, to half of standard, needs gross profit dollars up year over year and free cash flow positive. Full weight needs real revenue growth or a weekly flip above 42.88. Since the residual stays on, I'm not buying back from zero at 42.

So here's where we stand on NKE. Sell. Take about half the trim now around 37.4, work the rest into 38.05 to 38.77 if it's offered, and finish by mid-November at any price. Cap the residual at loss tolerance divided by 0.20, with a third as the default, and hold that cap through the December print whatever price does. Cut it in half on a close below 34.47, and the rest goes if the 33.30 band fails. Fetch in this order: the print date and reaction, the Form 4s since October 1, the 10-Q for capex and cash, consensus, the tariff read, and any investor event before December.

What I won't give up is the shape: sell early and decisively, and don't confuse a risk limit with an exit. What's left is a number only the caller has. Aggressive Analyst: You two have moved me in several places, so I'll say where first.

On pace, the conservative is now more aggressive than I was, and he's right. I argued for front-loading, then left a 38.05 limit sitting in the middle of the plan. I've dropped the daily 9, OBV and the crowd. What's left is weekly and monthly trends down and gross profit falling, and none of that improves at 38.05. The money is also small. If someone goes from full weight to a third, the slice waiting at 38.05 is about a fifth of a standard position, and 3% on that is under a point of allocation either way. That's a reason to stop deliberating. Far over the cap, sell about two-thirds of the trim now. Near the cap, half. Leave one limit at 38.05 for a week, then sell whatever's left. Use dates, not prices.

I also concede two things the neutral landed. The upside is a tail, not a coin flip. Flat revenue needs a 10.8% sequential step from Q1, against 6.0% and 6.6% in the last two years, and the normal step puts Q2 revenue about 4.4% below last year again. So a third is the default, and I'll stop describing the residual as something that has to be big enough to matter. It's a hedge against lag. On insiders, a fresh October cluster would show what directors think, not when they're right, since both earlier clusters were followed by lower lows. I'd read it and not resize on it.

The conservative's round-trip point also helps the Sell more than it hurts. The stock closed 38.10 on September 8, officers sold at 37.59 the next day, and it's 37.42 now. So I'm selling a full retrace into the zone where it traded before the print, with the 38.77 September closing high just overhead and both higher-timeframe trends unchanged. That tells us the market was neutral on the print, and I'll stop claiming it read the print either way.

Now where I still push back. Conservative, your cap is a ratchet. If it can only fall on new information and rise only after the print, you're resizing on things that don't change the gap, which is what you objected to in the crowd and OBV arguments. Only the caller's tolerance, the date, and the gap size you assume should move it, in either direction. Your step to half on a weekly flip above 42.88 also breaks your own rule that nothing raises the cap before the print. The neutral's version, with the rebuild only after the print, is the consistent one.

You're right that free cash flow positive doesn't discriminate. Q2 and Q3 of last year made about $670M combined, so a flat quarter clears it. I'll drop it and use operating cash flow up year over year, with the comparator from the 10-Q. For gross profit, I'd write the number down now so the gate can't be rewritten when results land. Q2 FY26 gross profit was $5.045B. If Q2 gains the same 60bps as Q1, margin is about 41.2%, and clearing $5.045B takes revenue down no worse than about 1.5%. If revenue repeats minus 4.3%, gross profit lands near $4.9B, about 3% lower. That's a hard gate, but it tests a real claim. Then use the neutral's two keys after the print. One key, the fundamentals or a weekly flip above 42.88, takes you to half of standard. Both take you to full. If the print gaps up and you rebuild above it, that's the cost of the rule, and it's what the residual is there to cover.

The 34.47 and 31.97 closes are the tripwires, at fixed prices. They're after-the-fact exits. A 15.5% gap from here prints near 31.6, through both in one session, which is why the cap exists.

The neutral raised the dividend declaration, and it matters. The file shows $0.41 payments dated December 1, March 2, June 1 and September 1. If the declaration comes a few weeks ahead of that, as is usual (I haven't verified it), it lands in mid-November, before the print. That makes mid-November a hard backstop, not a target. A cut is unlikely with $6.9B of cash, but a cut is a gap, and the calendar may hold more than one date.

My call is still Sell. Size the residual at the caller's loss tolerance divided by 0.20, a third if they can't name a number, never above half. If they're already under that size, do nothing, and if they're flat, stay flat. Get there by early November and hold it through the print whatever price does. Fetch the December date and the size of the October gap, the Form 4s since October 1, the 10-Q, consensus, the tariff read, any investor event and the dividend declaration date.

What I won't give up is the shape: sell early and decisively, and don't treat a risk limit as a view or as an exit. The number is the caller's. Aggressive Analyst: There's very little left, so I'll take what the two of you landed, correct one thing I said, and push on a few spots.

On the stress input, I'll take the neutral's framing: 20% is 1.3 times the worst gap in the sample, a maximum and not a one-way valve. Here's a wrinkle. The October 2 open at 32.55 was 9.8% below the September 21 close. I can't split that between drift and gap without the September 30 and October 1 closes, but it suggests October was another large down move. If so, the count is three big down moves and one modest up move, which favors the conservative on skew. That's why I'm comfortable at a third and not a half, and it doesn't change the 20%.

I'm retracting the round-trip claim. My baseline was September 8, three weeks before the event days. Against the last pre-event close we have, 36.10, the stock is up 3.7%. Without the September 30 close, the October gap size and the reason for October 15, "neutral on the print" and "rejected the print" are both unsupported. The Sell stands on the weekly and monthly trends and on gross profit dollars.

I'll take the neutral's ladder. The numbers say Q2 isn't an easy comp. Q1 gross profit of $4.80B plus the normal 2.1% to 2.5% sequential step lands near $4.90B to $4.92B. The $5.045B gate asks for $125M to $150M more than a normal quarter. If it passes, my thesis is wrong, and rebuilding higher is what that costs. The conservative's base-effect worry belongs on the second print, so full weight needs year-over-year revenue growth. I also take standalone Q2 operating cash flow as a veto, sourced from the 10-Q, since the file has no six-month figure.

On the dividend, I recall the same pattern the neutral does, with a declaration in mid-November and ex-dates near the start of each quarter. I haven't verified it. Early November is the target under either reading, and the declaration date goes at the top of the fetch list.

Now my pushback. First, the conservative's October 15 point is good, but it shouldn't hold anything up. If it was a squeeze, we're selling into it. If it was news, it may be the same thing as the unverified investor event on StockTwits, so fetch those together. An earlier confirmed event moves the finish date up by a week and doesn't change the cap.

Second, the neutral is right that nine fetches and five rungs is too much machinery. Only two things can move the plan before the print: the calendar (December print, dividend declaration, any investor event) and the size of the October gap. The Form 4s, 10-Q, consensus and tariff read inform the rebuild and a thesis check. A fetch that shows I misread the quarter would cancel the remaining slice. A fetch that only adds color wouldn't.

Third, a practical point nobody made. If the caller is only slightly over the cap, with an excess under about a tenth of standard, sell it all now and skip the limit order. A staged exit on a tiny excess is deliberation the evidence doesn't justify.

Fourth, I want to be honest about the cost of being wrong. With a third on, a 10% upside gap misses about 6.7 points of standard allocation, and a 15.5% downside gap avoids about 10. That's cheap against the gaps we've seen, and I won't fight the conservative's point that a 20% loss needs a 25% gain to recover. It's also why the residual is a hedge against lag and not a bet.

What I won't give up is the shape. Sell now, don't wait for a price, don't mistake a tripwire for protection, and keep enough of a stake that an upside surprise isn't a total miss. The cost of dithering is real too. The excess stays exposed for every day of debate, and the first slice shouldn't wait on any fetch.

My call is still Sell. The cap is the loss you can absorb divided by 0.20, a third if you can't name a number, never above half. If you're more than double the cap, sell about two-thirds of the excess now. If you're closer, sell half. Leave one limit at 38.05 for a week, then sell the rest at market. Be at the cap by early November, with mid-November as the backstop. Hold it through the December print whatever price does. Halve it on a close below 34.47 and take the rest on a close below 31.97. Rebuild only after the print. Either gross profit at or above $5.045B with operating cash flow not down, or a weekly flip above 42.88, takes you to half. Both take you to two-thirds, and full weight needs revenue growth. If you're already at or under the cap, do nothing, and if you're flat, stay flat. The number is the caller's.