Altair
Skip to content

Trading Analysis Report: LLY

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

  • Analysis date: 2026-10-21
  • Rating: Hold
  • Generated: 2026-10-21 12:18:58
  • 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: Hold

Executive Summary: Hold LLY at 1273.01 and do not buy at market ahead of the unconfirmed Q3 print (last year's was 10/30). Existing holders keep up to standard weight. Read the Q2 10-Q by tomorrow's close, then trim any overage from the +50% run back to standard (trim anyway at that close if the read isn't done), and never go below standard before the print. Any flat-book starter is off by default. It is allowed only if every gate item passes and is logged and the print is at least 7 sessions out on a confirmed date. If so, use a day order at or below 1230.66 (zone 1217-1230.66), placed after the open and sized so a gap of twice the implied move costs no more than about 3% of a full position. Every holder, standard weight included, cuts toward half on a daily close below the live daily SuperTrend (1173.96 now). The 6-12 month valuation target is $1,327.50, and an upgrade to Overweight comes only after a clean Q3.

Current Price: 1273.01

Price Target: 1327.5

Confidence: Medium

Time Horizon: 6-12 months

IV. Risk Management Team Decision

Aggressive Analyst

Aggressive Analyst: Since neither of you has spoken yet, I'll get ahead of what I expect to hear. I'm backing Hold on LLY, but as a loaded position: patient, with a plan to press when the setup pays. Hold doesn't mean the bulls are wrong. It means the price of the next entry is about to be set by an event, and we should make the event work for us.

I expect the conservative analyst to lead with valuation: 42.7x trailing, 14x sales, a 1% FCF yield. But the trailing multiple is a rear-view number for a company whose H1 revenue is up 51% and H1 net income is up 72%. At 26.9x forward with a PEG of 1.22 and a beta of 0.45, the market isn't paying a bubble price for a franchise with 85.8% gross margins at a period high. Debt is the same story. Roughly $55B against $41.7B of EBITDA is about 1.3x, which is not a stressed balance sheet. The $23B liability jump looks like a funded acquisition or investment, and Q2 operating cash flow of $10.7B was 151% of net income, which can fund that. Yes, some of that may be timing, but a company with this growth rate isn't going to struggle to refinance.

The insider-selling point is mostly noise. Sixteen thousand shares and $18.8M, spread over two months, against a $1.1T market cap. Nothing since August 17. The four director grants on 10/19 are fixed-dollar compensation. Lilly Endowment's overhang has gone quiet. If officers selling into $1,150 to $1,190 were a signal, the stock would not be 7% above their sale prices.

The macro argument is the weakest. Polymarket pricing a 78% chance of a December hike is a striking claim that our own report says it couldn't verify, because every FRED call failed. And the stock has just gone up about 10% in five sessions with that backdrop in plain view. A low-beta, defensive-growth name leading a healthcare rally with recession odds at 8% is not behaving like a casualty of higher rates.

CagriSema is a headline about a subgroup, women only. We haven't seen the full dataset, and retatrutide isn't needed for the 2026 thesis. The more interesting part of the news flow is the volume-versus-price debate. Revenue is up 48% while prices fell, which tells me volume is overwhelming price erosion. Q3 commentary on that is what could extend this run.

Now the neutral analyst. I expect you'll say wait for confirmation, because the breakout is marginal: 3.32 points above the August close, with the latest print back below it. I'd call that a coiled spring. The band compressed from 09-24 to 10-09, then expanded sharply. ADX went from 19 to 39.81 in eight sessions, all three SuperTrend tiers are up, and OBV hit a 60-day high. A weekly z-score of +2.03 isn't a fade signal in this setup, and the technical report itself says so. In a trend like this, stretch resolves through time or a shallow pullback, not a collapse. The plan's pullback zone of 1217 to 1232 is 3% to 4.5% below the current price, and the 1173.96 daily stop is about 2.7 ATR away. That's a normal digestion range.

The risk/reward figure the plan cites, +3% against -7.8%, measures upside only to the last high. That is the wrong yardstick for a breakout. From 1231.86 the stop is 4.7% away, the 1311.75 high is 6.5% up, and a rough measured move off the 1114-to-1284 range points toward the low 1450s. That's my own arithmetic, so treat it as a sketch, but it's very different from a 3% cap. And the earnings-looking gap days I can read direction on were up 9.8% on 4/30 and 4.9% on 8/5, with the caveat that we haven't confirmed those were earnings. Gaps cut both ways, which is exactly why the starter is capped at one-third of standard, so a 10% gap costs about 3.3% of a full position.

The most aggressive thing in this plan is the discipline around the gate. The Q2 10-Q is unread, and that's an hour of work, not a research project. If it shows the net margin drop from 37.4% to 30.9% came from one-off acquired IPR&D or deal costs, we have a clear green light for the starter before the print. I'd push the team to clear it fast. If a bold position is waiting on homework, the homework is the slowest part of the process.

Two other notes. First, the 10/21 bar looks partial, with about 1.03M shares against a typical 2M+, so 1273.01 may not be a final close and I wouldn't read much into the red tape. Second, retail sentiment isn't a crowded-long warning. The 89% bullish ratio is padded by one author with five of the bullish tags, and five negative @1Kof posts are untagged. The sample is small, and it argues against reading either extreme.

So Hold now, a one-third starter once the gate clears, build to half on a pullback into 1217 to 1232, full weight after a clean Q3, and cut to half on a daily close below 1173.96. I wouldn't trim a +50% leader below standard allocation just because the chart looks tall, since weekly and monthly SuperTrend are still up and higher timeframes outweigh the daily noise. Cutting winners ahead of a catalyst they may beat is how you lose the next leg.

FINAL TRANSACTION PROPOSAL: HOLD Aggressive Analyst: Let me start with where you each landed a clean hit.

Conservative, you're right on the funding math. Operating cash covered about 70% of the $15.4B investing outflow, and the rest was borrowing. I said it could fund the jump, and it didn't. You're also right that "the stock is 7% above their sale prices" is hindsight, and that nobody was proposing to trim below standard. I argued against a position no one took. Neutral, you're right that "coiled spring" describes something that already sprang, since the upper band went from about 1190 to 1281 in two weeks. You're also right that the plan's entry is a limit near 1232, not a market buy at 1273. At 1273 I don't want it either, so we agree there.

Now the pushback.

On the Q2 margin, I think the story is less ominous than the conservative makes it. Net margin was 21.7% in Q1'25, then 36.4% the next quarter. It was 31.7% in Q3'25, then about 34%, then 37.4%. This company has had dip quarters, and each snapped back. A 30.9% print is not new territory. Neutral is right that some of the $3.8B below the gross line scales with revenue and tax. What's left is a hypothesis, not a finding: one discrete item could explain the margin drop, the $15.4B investing outflow, the $25B balance sheet jump and the $8.4B financing inflow together. The 10-Q confirms or kills that.

On cash flow, H1 operating cash was $16.0B against $14.5B of net income, about 110%. Q1 was 72% and Q2 was 151%, so the timing largely washes out. The $11B rise in current liabilities may also include short-term borrowing, which would sit in financing, not operating cash. So "lifted by $11B" probably overstates it. And even if the entire $8.4B was new debt, that's about 0.2x EBITDA. Leverage stays near 1.3x. That's not a stress case.

On volume versus price, I'll own that Q3 commentary is what decides it. But gross margin hit a period-high 85.8% in the same quarter the headline says prices fell. If price erosion were biting, that line is where it would show first.

On macro, an unverified number argues for ignoring it, not for sizing down. Otherwise any number we can't check shrinks any position. The FOMC in that market resolves December 9, well after the print.

On gaps, you said both gaps I cited started from range lows, which is fair. But every large day the technical report names this year was an up day: +9.8%, +7.1%, +4.9%, +3.6%, +3.5%. The one real drawdown, 12.9%, was a grind over nearly four weeks, not a gap. The report lists 10/30/2025 as a high-volume gap day but gives no direction, and the stock went from about 795 to 1,100 over the following month. I wouldn't call it a counterexample without seeing it. I do accept that near highs the downside gap risk is higher than from lows.

On the 8/19 to 9/11 drop, that pullback stayed far above the weekly stop at 1039.50 and made a new closing high within five weeks. That's why the plan cuts to half at 1173.96 instead of exiting. The daily stop is a reduce line, and the weekly stop is where the thesis breaks.

Neutral, I want to push on your forward-multiple point because it's the sharpest framing and I think it's slightly off. You said the multiple is 26.9x if margins snap back and closer to 39x if Q2's margin is the new normal. But the 39x assumes Q2 EPS never grows at all. The gap from about $32 annualized to $47.24 is roughly 48%. Moving net margin from 30.9% to 34 or 35% is worth only about 10 to 13% of that. The other third or more has to come from revenue continuing to grow. So the 10-Q is a smaller question than either of you has made it. Q3 revenue, pricing and the guide are the real swing factor, and that's the part of the story that has been working best.

On sentiment, you used the StockTwits thread to say retail is positioned for beat-and-raise, then dismissed it when it showed stops clustered at 1260 to 1270. It's one sample of about 18 authors, with one call promoter. I'd give it little weight in both directions.

On risk/reward, I'll accept your 1.4 to 1 from 1232. But 1311.75 is the highest price in the data we have, so measuring upside only to that high caps a trade that is at the edge of its range. I'm not claiming a measured move. The target isn't confirmed, and I said so.

Here's where I land versus the conservative. If we drop the starter and re-engage after the print, we buy after the gap, with a thinner reward and the same exposure to the next leg. Neutral already made that point and I agree. A resting limit in the zone costs about 3.3% of a full position on a 10% gap, and it only fills if the stock gives back roughly 3.5% first. That's a better price to hold into an event than 1273.

My version is mostly neutral's, with two tweaks. Cancel the resting order on a daily close below 1216.71, the bottom of the unfilled gap, so we don't catch a falling knife. And size the starter to the options-implied move, not to the stop, since the stop can be jumped.

Everyone should read the 10-Q and confirm the Q3 date, since last year's was 10/30. Trim overweights to standard as the plan says. After the print, build toward half only if there's a beat and a raise and day one holds the gap. Don't average down on a miss. Take full weight only after a clean Q3 and a real answer on volume versus price.

We're not trying to call the print. We're setting up so that a good print helps us and a bad one doesn't hurt much.

FINAL TRANSACTION PROPOSAL: HOLD Aggressive Analyst: Let me take my lumps first, because there are several.

Conservative, you're right on the drawdown. I called 12.9% the one real drawdown, and the report shows 1104.60 on November 25 falling to 848.53 on April 29, about 23%. My point that the big down moves have been grinds rather than gaps survives, but the number was wrong. You're also right that my 795-to-1,100 figure starts nine days before the October gap, and that all three high-volume gap days came off range lows. We have no observation of a gap from within 3% of the high after a 10% run, and I'll size for that. You and Neutral are also right that the $55B already includes the Q2 borrowing, that an 85.8% gross margin after a 390 basis point jump proves nothing about pricing, and that my cancel-on-a-close rule can't work when the fill happens intraday. Neutral's day-order protocol is better, so I'm taking it.

Now where I push back. On cash, Conservative, you set the $15.4B investing outflow against $11B of free cash flow and got 1.4x. But free cash flow is already net of capex, so that comparison partly overlaps with itself. The arithmetic that interests me more is that Q3'25 operating cash flow was about $8.8B (nine months less H1) and H1'26 was $16.0B. That's roughly $24.9B across three of the last four quarters, before Q4'25. Against TTM free cash flow of $11B, that implies at least about $14B of capex over the year, if the vendor figure is right. The capex line in the feed stops in 2022, so I can't verify it. But it suggests the 41% is a reinvestment rate, not a cash conversion problem, since H1 operating conversion was about 110%. Whether that spending earns a return is the real question, and I grant we can't see it.

On the 10-Q, I concede it isn't an edge, since it's been public since August. It's a screen. But I don't accept your one-sided veto framing. Your own $1.9B to $2.5B of unexplained cost is the size of a plausible one-time charge. If it were a non-deductible acquired R&D item, Q2 net income would have been roughly $9B to $9.6B, a 39% to 42% margin, which would be a record in our table. That's a hypothesis, not a finding, but it shows the filing can come back either way. And the "14x sales" worry misses that LLY didn't pay 14x sales for whatever it bought. The 14x is what you pay for LLY. As Neutral said, the question is how good, not how big. Neutral's ordering point also holds: if the filing is a veto, you can't trim an overweight before reading it. It's an hour of work, so read first, then trim.

On the multiple, you've both beaten me. $47.24 can't be calendar 2026 with H1 at $16.20, so it's a later or rolling number nobody has confirmed. On this year's EPS, probably the low-to-mid $30s, the stock is in the high 30s. I'm dropping "26.9x, not a bubble price." What I'd keep is narrower: the price demands about 40% EPS growth into next year from a company that just printed 48% revenue growth. That's a demanding bar, not an absurd one.

Now the real disagreement, the starter. I agree there's no edge at 1.4 to 1 and a third size, and Neutral's "cheap option on staying involved" is fair. But look at the structure. It's a resting order in the 1217 to 1232 zone, placed after the open and re-armed each morning after reading the news. It's dropped if the stock opens below 1216.71, sized from the implied move, and only used if the print is more than a couple of sessions away. If the stock runs straight to the print, it never fills and we're where you are. If it fills, we bought 3% to 4% lower with the stop 4.7% away instead of 7.8%. Your objection that it fills in the Novo-or-rotation selloff is real, which is why the news check is in the rule.

On paying up after the print, you were honest that you're paying for the guide and the volume-versus-price answer. But that information can cost 8% to 10%. At a third size, an 8% gap you missed is about 2.7% of a full position, the same order as the loss on a bad gap. So the starter's value is small in both directions, and my case for it isn't a forecast of the print. My lean is modest, because a 10% run in five sessions has pulled some good news forward. That's why I only want to buy a dip, never the high. The daily TD count is at 6, and if the next three closes hold above 1257, 1240 and 1283.66, it completes a 9 around the 26th. Either the stock digests and fills our order cheaper, or it's strong enough that we weren't chasing anyway.

So I land here. Hold, and nobody buys at market near 1273. Today, read the 10-Q, confirm the Q3 date and the period behind $47.24, pull the implied move, and check the actual fed funds rate, which settles the 78% in five minutes. Holders above standard trim to standard after the read. A flat or underweight holder with a clean read and a print more than a couple of sessions out can place the day order in the zone, sized to the implied move, with no averaging down. After the print, build toward half on a beat and raise that holds the gap, and go to full weight after a clean Q3 and a real answer on volume against price. Cut toward half on a gap through the stop. We don't need to call the print. We need a good one to help, a bad one to be survivable, and not to sit out a +50% leader over an hour of homework.

FINAL TRANSACTION PROPOSAL: HOLD Aggressive Analyst: Let me take what's right from each of you first, because it moves the plan.

Conservative, you're right that I skipped the branch where the order fills and keeps falling. A resting buy fills when the stock is weak, and Novo, Meta, Tesla or a rotation back to tech can all land while it's live. That's adverse selection, and the stop and the size are the answer to it, not pretending it away. You're also right that an option has capped downside and a third-size stock position doesn't, so I'm dropping "cheap option" even though I borrowed it. On TD-9, a daily 9 on the 26th with the weekly count at 3 is a reason to watch digestion, not a signal in either direction. Neutral, you're right that I understated the margin question. Ten to thirteen percent of EPS at a constant multiple is a bigger swing than the whole trade, so the filing earns its hour and I'm retiring the homework line.

Now the capex arithmetic. You said it cuts your way, Conservative, but I don't think the numbers pick a side. Your logic was that trailing capex of at least $14B is about $3.5B a quarter, so most of Q2's $15.4B isn't routine. That assumes the spending is spread evenly, and the visible quarters say it isn't. Total investing outflow, capex included, was only about $3B in Q3'25 and $3.9B in Q1'26. If trailing capex is at least $14B, the other two quarters have to carry at least $7B between them. With any normal Q4 operating cash flow in the mix, it's well over $10B. Q2 had $15.4B. So the vendor's free cash flow figure fits a quarter that was mostly a plant build just as well as it fits a deal. I'm not saying it was a plant. The arithmetic doesn't pick, and the capex feed stops in 2022. That's why we run the filing today and don't argue about it.

On the organic question, Neutral's bound is the right one. Even at five times sales, $15B buys under $1B a quarter against a $3.2B sequential step. And a 16% sequential jump isn't foreign to this company. It did 22% in Q2'25 and 13% in Q3'25. The slowdown from 56% to 48% year over year is real and I'm not hiding from it, but the revenue line doesn't need an acquisition to explain it.

Here is the part I'd like you both to answer. The plan, in your version and Neutral's, has every holder carrying a standard-size position through the print. For that holder, a 10% gap is 10% of the position. The same plan says someone flat or light can't carry a third of that. If the event is acceptable at standard size, it can't be disqualifying at one-third. What differs is the entry price and the sunk position, and neither changes what the print does to the stock. Neutral is right that this depends on a mandate we don't know, so call it a risk budget. And the budget is symmetric in size. A third-size starter costs about 1.6% of a full position if the stop works and about 3% on a gap twice the implied move. Missing an 8% gap costs about 2.7%. Conservative, you call one realized capital and the other opportunity, and for an absolute-return book that's fair. For a book measured against a target weight, it isn't. Staying light a +50% leader is an active position too.

I'll concede the calendar. Your final-week exclusion is reasonable, because the closer the print, the less the stop protects you and the more the starter is just an earnings bet. If the date is 10/30, that leaves almost no window, and I can live with that. The starter was always the smallest part of this plan. The thing I'd defend to the end is that nobody goes below standard. Cutting a leader to underweight ahead of a catalyst it may beat is how you lose the next leg.

Where I think the plan is actually weakest is after the print. "Holds the gap" needs a definition we write down now, something like a day-one close above the prior close and in the upper half of the day's range. And "clean Q3" for full weight can't quietly mean waiting for Q4. The volume-versus-price answer comes from the call and the filing within days. If we do the Q2 read today, the Q3 version is a checklist. The step from half to full should be decided inside that week. If a beat and raise gaps the stock 8% and our plan is to wait, we've built exactly the cost Conservative says we can absorb, and we've done it without being paid for it.

On macro, check fed funds. It takes five minutes, and I'll let the answer move sizing either way. A real hike path matters for a stock on high-30s earnings, and a false one drops out.

So where I land: Hold, and nobody buys at market near 1273. Today, everyone reads the 10-Q for opex, IPR&D, tax, interest and what the $15.4B bought. They also confirm the Q3 date and the period behind the $47.24, pull the implied move and check fed funds. Overweight holders trim to standard after the read, and never below. A flat or underweight holder can use Neutral's starter only if the filing is clean, the date is confirmed and far enough out under the final-week rule, and the implied move can be sized around. That's a day order in the 1217 to 1231 zone, placed after the open, sized so a gap twice the implied move costs about 3% of a full position, with no averaging down. After the print, with the gap definition written in advance, we press on a beat and raise that holds, move to full weight within days once the volume-versus-price answer is in hand, and cut toward half on a miss or a gap through the stop.

FINAL TRANSACTION PROPOSAL: HOLD Aggressive Analyst: We're down to a short list, so I'll take what's right in your last rounds and then argue the three things that aren't settled.

Conservative, you're right that a dip order doesn't capture the gap. If the stock runs into the print, the starter never fills and we miss it exactly as the no-starter plan does. The missed-gap cost was never an argument for the starter, so I'm dropping it. The only case left is a slightly better price with a tighter stop, and we've all put that edge at a few tenths of a percent. I also accept that the cut to half on a daily close below the live SuperTrend stop applies to every holder, standard weight included. That rule protects the plan from me as much as from you, because it stops anyone arguing "the higher timeframes are still up" while the daily line is breaking.

Neutral, you're right that a close above the prior close is nearly automatic after an up-gap. I'll take a close above the prior day's high, in the upper half of the range, judged at the close. I'll also take day three for full weight instead of day one. That's still inside the week.

Now the three open items.

First, the 10-Q gate. Neutral is right that the strict version belongs on the starter and not on holders. But Conservative, "shown to be non-recurring" is a bar no filing can clear. A 10-Q quantifies what happened and doesn't promise the next quarter. A gate nobody can pass is a veto in disguise, so let's make it testable. By my arithmetic, if Q2's costs below the gross line had stayed at Q1's share of revenue, they'd have been about $2.4B lower. Matching the trailing 33.5% net margin takes only about $0.6B of identified items in net income terms, and getting back to Q1's 37.4% takes about $1.5B. So the gate passes if identified items lift adjusted Q2 margin to at least the trailing level, and it's a clear pass if they get near Q1. If the drop turns out to be recurring operating spend with no clean explanation, the starter is off and I won't argue.

Second, Neutral's new number changes the picture, partly against me. Assets rose $25.7B in Q2. About $15.9B of that is non-current, $3.7B is cash, and roughly $6.2B is other current assets, which is what a 16% sequential revenue step looks like in receivables and inventory. So most of the money became long-lived assets, and my single-item story was too tidy. If the margin hit were an expensed acquired-R&D charge, it wouldn't sit on the asset side. That's accounting background, not something in our data. The margin drop and the asset build may be two separate questions, and the filing has to answer both.

The liability side answers part of Conservative's funding worry, though. Derived liabilities rose about $23B, and current liabilities took $11.1B of that. That leaves roughly $11.9B on the non-current side, so this isn't simply short-dated paper funding a long-dated outlay. It doesn't prove the maturity ladder is comfortable, and the debt footnote stays on the gate list. But it tilts away from the short-funding reading.

Third, the plan has holes that matter more than the starter. Neutral is right that the starter deserves less airtime. If the print is 10/30, the window is mostly shut, and the plan is hold, trim, read, wait. Here is what I'd fix:

  1. The read needs a clock. "Trim when the date is confirmed if the read isn't done" means trimming in about five minutes, which makes the read pointless for holders. I'd say the read is finished by tomorrow's close, and if it isn't, the overage is trimmed at that close.
  2. There's a cut rule and no restore rule. If everyone goes to half on a daily close below the live stop while the weekly stays up, the plan needs a way back. The mechanical one is to restore to standard when the daily SuperTrend flips back up with the weekly intact. Without it, a whipsaw leaves a holder at half weight in a leader because nobody wrote the exit.
  3. The post-print tree needs every branch written in advance. That includes the uncomfortable one, a beat and raise where the stock sells off. Under our own gap test that is a fade into the lower half, so no add and no chasing, and nobody improvises at the open. Neutral is right that the prior-day-high level has to double as the exit for any post-print add. That caps the risk of the add at a number we choose in advance, not at a stop 10% away.

Two small items. On the fed funds check, Conservative, I agree that if it shows a real hike path, the starter is dropped. An edge of a few tenths of a percent shouldn't carry a macro flag. Holders are a different case, because they already have the cut rule. Neutral, "already priced because the stock ran 60%" isn't established either. We don't know when that path got priced, so the check settles it, not the argument.

Where I land: the point of this whole machine is to be correctly sized for the leg after the print, not to win the pre-print argument. Weekly and monthly SuperTrend are up, OBV is at a 60-day high, and revenue is up 48% year over year. If the filing is explainable and the print confirms it, we want to be at full weight within three days, not waiting for Q4.

So Hold. Nobody buys at market near 1273. The 10-Q is read by tomorrow's close against the testable gate above, with the funding structure and what the $15.4B bought. We confirm the Q3 and Novo dates and the period behind the $47.24, pull the implied move, and check fed funds. Overweight holders trim to standard, never below. Everyone cuts toward half on a daily close below the live stop and restores on the daily flip back up. The starter is optional, with a fixed 1217 to 1231 zone, a day order after the open, and off on Novo's report day or a confirmed hike path. It's sized to the implied move, with no averaging down. After the print, we build on a beat and raise that passes the gap test, move to full weight around day three if the call answers volume versus price, and cut toward half on a miss or a gap through the stop.

FINAL TRANSACTION PROPOSAL: HOLD