Altair
Skip to content

Trading Analysis Report: GEV

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 11:17:31
  • 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 GEV. Trim existing holdings to about 40% of a standard allocation (33-50% band, up to 60% only for a diversified holder facing real tax friction) before the Q3 print, open no new position, and do nothing if you already hold less. Confirm the earnings date first: if it is 10/22, finish the trim at market today; if it is 10/28, sell down to about 60% now and finish by the 10/27 close, treating any bounce into 937-965 as a bonus rather than a condition. Treat a daily close above 1,022.43 as the level where this stance is wrong (not a stop) and a weekly close below 837.76 as a prompt to re-read the numbers; the target is $884 over 3-6 months, and rebuilding happens only in thirds after reading the release, capped near 75% of standard until a second quarter of margin progress or a lower price earns full weight.

Current Price: 933.07

Price Target: 884.45

Confidence: Medium

Time Horizon: 3-6 months

IV. Risk Management Team Decision

Aggressive Analyst

Aggressive Analyst: Since neither of you has spoken yet, I'll set the frame. People hear "Sell" and assume timid. I think the timid move is carrying full size into a print on a $248.5B company at roughly 138x trailing operating income while the daily trend is broken. The bold move is trimming to half, selling into a stack of resistance, and being ready to pay up later for proof.

Start with what the price demands. Trailing operating income is about $1.8B on $41.4B of revenue, a 4.3% margin. The research plan's math says margin has to roughly double, to about 9.1%, just to break even on a 35x exit, and even the bull couldn't support better than 45% odds on the 12% margin case. I respect the operating story. Gross margin has risen four straight years, from 11.7% to about 20%, and Q2 revenue grew 21.9%. But the market is already paying for the version where margin doubles.

The headline numbers also flatter the story. Trailing EPS of $34.88 is distorted by non-operating gains, and Q2 GAAP EPS of $2.47 annualizes to about $9.90, or roughly 94x. First-half operating cash flow of $10.7B against about $0.8B of operating income looks like customer advances. Current liabilities went from $35.3B to $55.8B and the current ratio fell to 0.85. That's my inference, since the tools don't break it out. That cash is a loan from customers, repaid in delivered work, so it doesn't support the multiple the way a 5% FCF yield suggests. The $4.3B Q1 investing outflow is also unexplained, so some of that 21.9% growth may be acquired. That's why the plan makes the 10-Q's acquired-revenue disclosure a trigger.

The tape agrees. GEV has closed under the 200 SMA five sessions running. Today's +4.8% bounce came on about 60% of the sell-off's average volume. MACD is still sliding after the bounce, at -10.18, and RSI at 45.5 hasn't reclaimed 50. There was no capitulation either: RSI never went oversold, and TD-9 shows no exhaustion on any timeframe. The sell zone is where supply sits: the 200 SMA at 936.91, the 10 EMA at 940.40, the 50 SMA at 958.61 and the Bollinger midline at 964.70. We're selling a bounce into overhead supply, not panicking at the lows.

Then there's the event. The gap history is +13.7% in April and -8.7% in July, and no stop covers that, with ATR already at 4.5%. If the July 22 gap was the Q2 print, which the cadence suggests but I can't confirm, then a quarter with 21.9% growth and a 5.9% operating margin still got sold 8.7%. Good news was already priced in. From the 6/30 close to the 9/14 close the stock fell 25.5%.

To the conservative analyst: you'll want a full exit, and I'll tell you why that's too cute. The weekly and monthly SuperTrends are both up, with stops at 837.76 and 635.30, so the primary trend is intact. Revenue is accelerating, and the market gives about 70% odds, on thin volume, that orders top $22B. September showed GEV can rip 17.7% in three weeks, from 874.76 to 1,029.21. Zero weight is a bet that the gap goes down, and you'd own the regret if April repeats. Half size is a barbell. The plan already has the escalation path: cut to 25% if acquired revenue exceeds about 5 points of growth, and to 0-25% on a weekly close under 837.76 or gross margin under 21%.

To the neutral analyst: holding full weight isn't neutral. It's maximum event exposure with no edge. You'll say the 20% pullback reset expectations. It's still about 6x sales and nearly 37x forward earnings, and the stock has repeatedly sold off near 1,000 to 1,030. You'll also point to the 5-to-2 bullish StockTwits tags, but that's seven labeled posts, and the text reads "can't tell where this guy is going" and "needs a catalyst." Macro adds little help. There's a 97% chance of no Fed cuts, AI-bubble odds are up 3.5 points on the week to 10%, and the AI-infrastructure read-throughs, the Barron's 40% selloff and the AI IPO slide, are shaky. GEV trades on AI-power capex, so that matters.

Trimming isn't free insurance. With half the position, a 13.7% gap up costs me about 6.9 points of foregone gain, and an 8.7% gap down saves about 4.4. I accept that because the odds and the valuation favor the downside, and because the proceeds are dry powder. I can rebuy the 855-890 cluster if it breaks, or chase above 1,022.43 on rising volume if the print confirms.

I'm not being stubborn. Confirm the earnings date first. Polymarket's 10/28 resolution suggests on or before that, but nothing I have verifies it, and the 10/21 bar may be provisional. If price closes above 1,022.43 on rising volume with fundamentals to match, I'm wrong and I'll say so. The rebuild to full size needs gross margin around 21.5% or better, opex growing slower than revenue, organic growth near 20%, and orders of $22B or more. Then I'd happily pay a higher price.

FINAL TRANSACTION PROPOSAL: SELL Aggressive Analyst: I'll start with where you've moved me, because it's more than a little, and then say where I still won't budge.

You're right about the gross margin line. The last four quarters ran 19.0, 21.2, 19.1 and 21.3, and Q3 last year was the 19.0. If Q2 was about a point better than a year ago, a normal Q3 lands near 20, so a hard 21% trigger would likely fire on an in-line print, and a 21.5% rebuild bar asks for 250 basis points of year-over-year expansion. I'll switch to year-over-year. At or below last year's 19% is a miss and cuts me toward a quarter. Around 20 is in line and I hold the half. About 20.5 or better, with orders at $22B or more and opex lagging revenue, is the rebuild signal.

You're also right that 1,022.43 isn't a stop. It's the level where I'm wrong. On July, I undersold the sequence. 985 to 900 is another 8.6%, so half size avoids about 8 points, not 4.4. And I'll drop "good news was already priced in" as a pillar, since it leans on July's gap being the print, which I couldn't confirm. The case stands on price, tape and gap structure without it.

On timing, neutral, the Wednesday catch is the best point in this debate. All three spikes were Wednesdays, tomorrow is a Thursday, and 10/28 is a Wednesday. One wrinkle: today is also a Wednesday, and the stock opened 3.5% higher. I don't think that was the print, because the orders contracts were still open and trading when the report was pulled, and volume was thin. But the bar may be provisional, so the date check is step one. I'll take your two-step. Sell about half the trim now, since 933 is 0.4% under the 200 SMA and effectively the bottom of the zone. Sell the rest into 937-965, or at the close before the print once the date is confirmed. Conservative is right that the first tranche shouldn't wait on a bounce.

Now the disagreement over quarter versus half. Run the arithmetic at the margin. Going from half to a quarter sells another quarter of the position. If April repeats, you give up another 3.4 points of the position's value on the gap. If July repeats, you save 2.2 on the gap and about 4.1 on the full week. The break-even is the same 39% to 55% band neutral computed for the full position. Nobody here can place the true odds inside that band. We have two prints, and the orders contracts trade $100 to $616. When every extra unit you sell is a coin flip with an unknowable bias, you stop at half. Going to a quarter feels safer, but it's its own directional bet, that the move is down and you can rebuy cheaper.

Conservative, I also don't accept that the way back in is "known and bounded." You used July's continuation to size the downside, so apply the same logic to April. It gapped, faded about 6% by mid-May if that insider print is representative, then ran to 1,175 by June 30. Rebuilding three-quarters of a position into that path costs real money. Half size at least lets me rebuild a smaller piece.

Zero isn't the absence of a bet either. It's a bet against a weekly and monthly uptrend, on a company growing revenue 22%, from a price that's already 20% off the high.

On the loss budget, neutral is right that 0.5% was a number you picked, and the answer depends on a standard weight we don't know. The honest rule is that post-trim weight times about 17% is what you're accepting on a July repeat. If half of your standard weight breaches that, go lower, but that's a fact about your book. For GEV, I'd say half. If you're already at or below half, there's nothing to sell, and if you hold none, don't open.

I agree with neutral on the put spread. Implied volatility into a print charges you for the same gap you're sizing down for, and I haven't seen option prices. Selling shares is a hedge with no vol premium, and the proceeds are the optionality.

Neutral, on valuation, you're right that GEV has been expensive all the way from 600 to 1,175, so multiples can't time anything. That's why I'm not at zero. But a thin margin of error is exactly what a print tests. On the reset argument, the stock is about 13.5% below where it sat going into July, and that's real cushion. It's also why I stop at half and not a quarter. But the multiple is still about 6x sales and 37x forward, and about 94x on Q2 annualized GAAP EPS. A price move isn't a numbers move.

I'm not buying the higher low. 871.16 versus 877.44 is a 0.7% difference with ATR at 4.5%, which is noise. The 10/15 reversal to 926 was followed by three lower closes. Today's bounce came on 60% of sell-off volume, MACD is still sliding, RSI bottomed at 35, and TD-9 shows nothing near exhaustion on any timeframe. It isn't a free-fall, but nothing in the data looks like a bottom yet.

On customer advances, I'll grant they're partly a demand signal, and the orders number isn't independent evidence for selling. My point was narrower. $10.7B of first-half operating cash against about $0.8B of operating income means the 5% FCF yield isn't earnings power. Both things can be true. The buyback inference is plausible, since $2.74B of financing outflow minus roughly $140M of quarterly dividends lines up with the $2.6B equity drop. That narrows one unknown. The $4.3B Q1 investing outflow remains, and it's why the acquired-revenue disclosure matters.

On the 837.76 trigger, I'll make it print-aware. The reason I'm keeping the half is that the weekly and monthly trends are up. If the weekly flips, that reason is gone, but I won't sell the low twice. A weekly close under 837.76 with weak orders or guidance takes me to a quarter or zero. With orders at $22B or more and guidance intact, I hold the half one more week.

On rebuilding, you've persuaded me on two things. I won't buy the open on a gap-up day. I'll read the release and filing detail first and rebuild in thirds, starting only once volume confirms. And I'll buy the 855-890 cluster only after it holds, meaning a close back above it after a test. 837.76 is only about 2% under 855, and a break flips the weekly trend.

Here is the bold part. I'm not running from this event. I'm sizing so that I act on the print instead of reacting to it. If it gaps down with good orders, I have cash to buy. If it gaps up, I still own half and can rebuild in thirds. If orders and margin both miss, I go to a quarter with a plan. Cutting to a quarter or zero gives up the scenarios where GEV's uptrend reasserts, and full size gives up the cash that makes the other scenarios playable.

FINAL TRANSACTION PROPOSAL: SELL Aggressive Analyst: Most of the last round was fair, so I'll take what's right and then argue about the one thing that's still open.

Conservative, you're right that the Polymarket line shouldn't be a trigger. It trades a few hundred dollars. Orders against sell-side consensus plus company guidance is the right yardstick, and I'm swapping it in. Neutral, you're right that the tape can't carry the trim. MACD went from -15.6 on 9/21 to positive by 10/1, and the stock then ran 17.7% to the 10/6 close. MACD, RSI and TD-9 are reasons not to add and a guide to where I work the second tranche. They aren't reasons to sell. The trim stands on three things: a print with a 9-14% gap history, a multiple that needs operating margin to roughly double, and position sizing. That's enough.

Conservative, on the escalation path, I ran your math too. 933 times the July ratio lands near 779. But look at what triggers the sale. It's not the price. It's the release. If orders miss consensus, guidance softens and gross margin is at or below last year's 19%, I cut because the thesis got worse, which is a response to information and not a panic sale. If orders and guidance are fine and the stock is at 780 anyway, I hold the half and start planning the rebuy. A weekly close under 837.76 means go read the numbers. It isn't a sell button. Neutral had that right.

That leaves half versus a quarter. I'll concede that half isn't derived from an edge, because nothing here gives anyone an edge on which way the gap goes. My case for stopping there is about how the risk behaves. Event variance on a position scales with the square of its size. Going from full to half removes about three quarters of it. Going from half to a quarter removes only about another fifth of the original. The cost doesn't shrink the same way. Each slice you sell gives up the same share of the drift: a weekly and monthly uptrend, revenue growing 22%, and a rebuy that on Neutral's numbers lands 7 to 14% above where you sold if April repeats. Tax is a cost too, and Neutral is right to raise it. It scales with how much you sell and it doesn't come back when you rebuy higher. I don't know your account, but that's another reason not to default below half without a reason.

Conservative, you asked to stress the downside at 20% instead of 17%. Then stress the upside too. April went from the 990.79 pre-gap close to 1,174.86 by June 30, about 18.6%. On 933 that's roughly 1,105, and at a quarter you'd be rebuilding three quarters of a position into it. Stress both tails or neither.

On cash as optionality, I made that argument, and you're right that more cash is more optionality in the abstract. But it only has value if you deploy it, and my rebuild conditions are strict. I buy in thirds, after reading the filing, and only into a zone that holds or strength that confirms. A lot of that cash will sit waiting for a signal that may not come, and that's a cost.

Neutral, I'll take your calm-at-780 test as the decision rule. If you could read the release calmly with the post-trim position at 780, you're done. If you couldn't, go lower. If your standard weight already comes from a risk process, half of it is already a real cut. If you're at or below half, there's nothing to sell.

The plan, then, in one breath. Confirm the earnings date first, since the 10/21 bar may be provisional and the date is only inferred. Sell about half the trim now, with 933 at the edge of the 937-960 zone. Sell the rest into 937-965, with a hard finish at the last close before the print. After the print, judge year-over-year gross margin (around 19% is a miss, around 20% is in line, 20.5% or better is constructive) alongside orders against consensus and guidance. Rebuild in thirds after reading the filing detail, and buy 855-890 only once it holds. 1,022.43 is the level where I'm wrong, not a stop. A close above it on rising volume, with matching fundamentals, and I say so and pay up.

This is the bold version because I'm not hiding from the print or betting that it goes my way. I'm sizing so that either outcome leaves me free to act on what the numbers say.

FINAL TRANSACTION PROPOSAL: SELL Aggressive Analyst: Neutral, you're right about the tape, and it changes how I defend my number. I threw out the chart as a reason to sell and then leaned on the weekly and monthly uptrends as a reason to keep half, and I can't do both. So the uptrend is a review prompt, same as 837.76. What I keep half for is that nobody here has an edge on which way the gap goes, and that revenue is growing 22% on four years of gross margin expansion. I'd rather own a real piece of that than a sliver.

Your ATR math helps me. 34.32 over 41.94 is about 82%, so the volatility change alone takes standard weight to four-fifths. Half of standard is then roughly a 40% cut for the event on top of that. A third would be a 60% cut for an event whose direction nobody can call, and that's a lot to pay for a coin flip.

Conservative, take your "start at the number whose failure is cheaper" rule. Applied consistently, it doesn't stop at a quarter. A forced sale at the low costs even less when you own less, and the cheapest failure of all is zero. You reject zero because of the uptrend and the growth, which is the same tradeoff I'm making at half. So the rule doesn't pick a third over a half. It only says to hold less if the holder is the type who sells lows, and that's a fact about the holder.

You also caught me on the gap premise, and that's fair. I'm treating 4/22 and 7/22 as likely prints because they sit on the same cadence as last year's 10/22 volume spike, but that's an inference. The 11.3% drop in five sessions with no print shows the tail doesn't need one. So the case for the trim doesn't depend on those gaps being earnings. It depends on the print being close, which is why the date check comes first.

On the $4.3B Q1 investing outflow, I agree it's the biggest open question. But the plan already attaches an action to it: cut to 25% if acquired revenue is more than about 5 points of growth. If I sell down now against an unknown I've already assigned a response to, I'm paying for the same risk twice. Sizing covers the gap, and the trigger covers the aftermath.

Neutral, I'll take your point on the date too. The trim is an event hedge, so "unconditional" should mean unconditional on price, not on timing. If the print is Wednesday the 28th, there are four sessions left, and the last close on the 27th is the hard deadline. The first tranche now is fine, though a limit order in the zone would do just as well. If the print is two or three weeks out, nothing we agreed on supports selling today.

Your aggregate-exposure point is the best new idea of the round, and it's what should set the number. Someone holding GEV along with BE, VRT, ETN, CEG and CAT already has a much bigger position on this one print than the GEV line shows. For them, a third or less is right, and the proceeds shouldn't go into another AI-power name. Someone whose only exposure to the theme is GEV should hold half. That's a better dial than "unknown book defaults lower."

I'd also tighten the rebuild plan. Conservative was right that if July repeats and the stock lands near 780, the 855-890 zone is irrelevant. After a gap down with orders and guidance fine, I'd skip the price zones and rebuy in thirds on evidence. The first third goes in once the stock holds the post-gap low for a couple of sessions and the call and filing confirm the release. After a gap up on good numbers, the same thirds apply, and I don't chase the open.

On tax, I agree it's a checklist. One wrinkle is that the highest-basis lots are probably the ones at a loss, and those are exactly the ones a wash-sale rule disallows if you rebuy within 30 days. A taxable holder has to choose between harvesting the loss and rebuying inside the window, so plan the rebuy before the sale.

What's left between us is a sixth of the position, about 2.3 to 2.75 points either way. A single ATR day on the full position is bigger than that. So here is the rule. Confirm the date. Trim to half of standard if GEV is your only exposure to the theme, and to a third or less if you're stacked in AI power. Open nothing new, and if you're already below that, there's nothing to sell. Judge the print on orders against consensus, guidance, and gross margin against last year's roughly 19%. Treat 1,022.43 as the level where I'm wrong, and 837.76 as a prompt to read the numbers, not a sell button.

The bold part is that I'm not guessing the gap. I'm keeping enough exposure to profit if I'm wrong and enough cash to act if I'm right, with the response to the print written down in advance.

FINAL TRANSACTION PROPOSAL: SELL Aggressive Analyst: Neutral, your correction on the Wednesday pattern is the most useful thing said this round, because it adds a branch none of us had. If the rule is the fourth Wednesday, the print is today. If it's the 22nd, it's tomorrow. If it's the last Wednesday, it's the 28th. I still lean toward next week. The orders contracts are trading at 70% rather than pinned near 0 or 100, no headline in the window reports results, and the retail chatter reads like people waiting on a catalyst. But that market is thin, the 10/21 bar may be provisional, and "until Thursday" is too ambiguous to lean on. So the gate has to ask whether the print has already happened, not just whether it's soon. If the numbers are out, the pre-positioning case is gone, and selling now means selling a stock that's up 4.8% on news nobody has read. The gate then has four outputs:

  • If the numbers are already out, don't sell on event grounds. Read the release and use the content rules.
  • If the print is tomorrow, finish the whole trim today at market.
  • If it's the 28th, sell tranche one now and tranche two into 937-965, with the 27th close as the hard deadline.
  • If it's more than a week out, nothing we agreed supports selling today.

You're also right that my limit-order comment only works with several sessions to spare. A limit that doesn't fill leaves you at full size for the event.

On the $4.3B Q1 investing outflow, you've changed my mind. I treated it as something a post-print trigger could handle. Both 10-Qs are already filed, so it's homework to do before the sale. Here is why it matters. Revenue growth ran 11.8%, 3.8%, 16.3% and 21.9% over the last four quarters. If that Q1 outflow was an acquisition, Q2 would carry a full quarter of the bought revenue, which could explain part of the jump from 16.3% to 21.9%. That's a hypothesis, not a finding. But it means the filings, not a default, should choose between a third and half. If acquired revenue is more than about 5 points of Q2 growth, go to a third. If it's mostly organic and the cash story is advances and buybacks, half is defensible. That settles most of what Conservative and I have been arguing about.

Conservative, I'll keep one piece of my case. You said an extra slice has near-zero expected cost, so the only price is friction. Fine, but your "repair" asymmetry doesn't hold up against the two precedents. The gap leg can't be repaired in either direction. The second leg can, and it's larger in July. April went from 990.79 to 1,126.98 on the gap, +13.7%, and then only reached 1,174.86 by 6/30, about 4.2% more, with a fade toward 1,059 in mid-May if that insider sale is representative. July closed the gap day at 985.03 and was at 900.28 a week later, another 8.6% with several sessions to act on content. So sizing covers the gap leg and a content rule covers the second leg. If a miss on orders and guidance is read the next session, that's a sale near 850, not a forced sale at 780. The scenario where holding half makes me sell the low is the one where I ignore my own rule.

I'll also concede that your point cuts against me. After an April repeat, rebuilding in thirds recovers almost none of the move, because the money was made on the gap day. That's the real price of every slice you sell. It's why I won't defend half as a floor, and also why I won't go below a third by default.

On your point that advances and orders move together, I agree: a strong orders number and strong operating cash flow are probably the same fact counted twice. That's why Neutral's ranking is right. Orders and guidance against consensus set direction. Year-over-year gross margin and operating margin, against Q3'25's 19.0% and 3.7%, say whether the hold stays comfortable. Acquired revenue and the advances share change size, not direction. I'd make the margin test concrete. The 35x exit math needs operating margin to roughly double from 4.3% TTM to about 9.1%, and Q2 printed 5.9%. A Q3 at 6% or better keeps that path alive. Under 5% breaks it.

The point I most want to hold onto is how the trim gets used. Selling isn't the bold part. Having the cash and a written plan when a 9-14% dislocation hits a company growing revenue 22% is. So write the buy side now, like the sell side. After a gap down with orders and guidance intact, the first third goes in once the release and filing detail confirm and the stock holds the post-gap low for a session or two, not at a price zone I picked today. After a gap up on good numbers, the same thirds apply, and I don't buy the open. If the stock closes above 1,022.43 on rising volume with matching fundamentals, I'm wrong, and I pay up.

I'm taking Neutral's dial. Start near 40% of standard and let the 10-Qs and your theme exposure move it: toward half if acquisitions are small and GEV is your only exposure to the power trade, toward a third if the acquired share is large or you're stacked in BE, VRT, ETN, CEG and CAT. The proceeds go to cash or bills, and if you're already below the target, there's nothing to sell. The gap between a third and half is smaller than one day of ATR, so I'm done arguing it. The date, the filings and a pre-written playbook matter more.

FINAL TRANSACTION PROPOSAL: SELL