Altair
Skip to content

Trading Analysis Report: TSLA

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

  • Analysis date: 2026-10-14
  • Rating: Underweight
  • Generated: 2026-10-14 17:25:47
  • 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

Price target: 355 over 2–3 weeks.

IV. Risk Management Team Decision

Aggressive Analyst

Aggressive Analyst: I'll open, since nobody else has spoken yet, and take on the objections I expect to hear.

First, selling here is not the timid choice. The timid choice is holding a full allocation into a Q3 print and an FOMC meeting because "it's a range." This plan is a limit order resting at 375, above the market, so we only act if the market pays us first. It has a defined invalidation at 391, above the 200 SMA and the upper Bollinger band, and it has staged escalation after that. By my own math, the risk from 375 to the stop is 16 points. The reward is about 26 points to the daily SuperTrend at 348.57, about 29 to the September low, and nearly 78 to the July low. That's roughly 1.65 to 1 on the first leg and close to 5 to 1 on the tail. Those are my derived numbers, not vendor figures.

Now, why this setup? TSLA is 1.8% below the 374.01 close from just before the July gap. It fell 14.5% in a day on 115 million shares, bottomed near 298, and has climbed about 23% to roughly full repair. It did that going into a print where the last filed quarter showed gross margin down 430 basis points sequentially to 16.8% and operating margin at 1.4%. Operating expenses were up about 47% year over year by my derivation. Capex was $5.8 billion and free cash flow was negative $1.1 billion. Net income was nearly three times operating income, so the profit isn't coming from the core business. You're paying 340 times trailing and 168 times forward earnings for that, on a 4.7% ROE and a beta near 1.9. The clean-print scenario is already in the price. Our own exit triggers are gross margin under 18% or operating margin under 3%, and Q2 missed both. A repeat of Q2 is a failure, so the company has to improve just to keep us from exiting.

Then there's the macro. Hike odds for the October meeting jumped 13 points in a week to 28%. The market gives 86% to another hike this year and 97% to no cuts at all. Recession odds are only 8%, so there's no growth scare to pull rates down. I'll concede that 70% hold is the base case. But a hold is largely priced, while the hike tail is growing fast, and a 168x forward multiple has very little cushion against it.

To whoever says wait for the print and then decide: that is how you end up trading the aftermath. The plan itself admits stops don't protect against gaps like July, so size before the event is the only real protection. Waiting for information means acting after the move.

To whoever says ADX is 9.55, so there's nothing to trade: I agree the trend tools are unreliable, and I'm not leaning on them. The weekly TD9 sits on a partial week with 0.8% of margin and could disappear by Friday's close. The trade rests on valuation, the print, and the FOMC. In a range you sell the top of it, and the technical report's own base case says to fade the edges. The 375 to 389 zone is where the 10 EMA, the Bollinger midline, the swing highs, and the 200 SMA all stack. We're not selling the middle at 367. We're waiting for a bounce.

To whoever says exit everything: why? The balance sheet is strong, with debt to equity of 0.18. Revenue is up 25% and the monthly SuperTrend is still up. A clean print could gap this stock up hard. The retained half owns the Cybercab, the storage momentum, and the merger tail, which a thin market puts at 20% by year-end with unknown terms. Going to zero here sells the whole option for a modest gain in protection. The better move is to scale with information: 35% on a weekly close at or below 364.27, 25% below 348.57, and 0 to 15% if margins disappoint.

To whoever says low hedge fund crowding means low unwind risk: that's your best argument, and it's true as far as it goes. But low sponsorship also means nobody is defending 389, and the stock fell 14.5% in a day in July without being crowded. Insider behavior doesn't help you either. Musk's $1 billion buy was over a year ago, there have been no open-market insider buys in 12 months, and about $88 million was sold. The retail sentiment has weak label quality. A third of the messages are about SPCX, "sell the news" already shows up for tomorrow's Roadster event, and the retail bears are calling 350. That isn't a crowd I'd stay full size for.

The Roadster event works in our favor either way. A pop gets us a better fill into the zone, and a flush moves us toward the Friday weekly-close trigger.

So my case is to trim to about 50% of a standard allocation, selling into strength at 375 or better with a stop at 391. Cut further on evidence. This is a trim, not a short, and if you're flat, don't initiate. I don't know the actual holdings, so apply the sizing relative to whatever a standard TSLA position is for you. Being decisive before the catalysts, with the stock mid-range and richly valued, is the high-reward move. Aggressive Analyst: Most of what you both said about the plumbing is right, and the plan gets better for it, so I'll start by giving ground.

Conservative, a single limit at 375 only works if the stock bounces. That was a preference dressed up as a commitment. The commitment has to be a date, so I'll take your backstop: whatever isn't sold goes at market two sessions before the print, and in any case before the October 28-29 FOMC. The date also needs verifying, since nobody has confirmed it. I also concede the reward-to-risk math. The 1.65 and 5-to-1 figures were the best case. Measured to 355, where the base case actually chops, it's about 1.25 to 1, and the 16-point risk is only about 1.4 ATR. But that concession shows this isn't a directional short with an edge. It's a sizing decision, which is where I wanted the argument anyway.

I'll stay put on selling at the market right now. We're about 2% under the limit, less than one ATR, after three straight lower closes, and the Roadster reaction lands Friday. Dumping a tranche at 367 gives away the one thing a trim into a range can get, a better price, to guard against a fill risk the backstop already covers. Neutral's version is better: a small slice Thursday if it bounces, a ladder from roughly 372 to 385, and the rest on the date. The cost is that if it flushes, the remainder sells lower. That's the premium for guaranteed size.

On half versus a third, look at how small the gap is. A July repeat costs about 7% of a standard position at half size and about 4.8% at a third. That's two and a half points either way, and the same two and a half points forgone on a clean-print gap up. I won't defend 50% as derived. It's the ceiling. Use neutral's arithmetic: retained fraction is tolerable loss divided by 15%, and if you can't stomach 7%, go lower. Your line that a missed gap-up is a fraction of a gain you can buy back later is the chasing you objected to in my plan, because buying after a ten percent gap is chasing. And if a full-size gap-down is capital you can't replace, the position was oversized for the book, which the loss budget already catches.

I also won't concede that the levels are decoration. ADX at 9.55 breaks trend-following signals, but it makes horizontal levels matter more. The 10 EMA, the Bollinger midline, the swing highs and the 200 SMA are where supply sits in a range, and the ladder is built on them. On macro, I agree there's no FRED data. That uncertainty argues for trimming, which we're all doing, but it doesn't by itself move us from half to a third. On hedging, I'm with neutral. Puts on top of a trim is paying twice, implied vol into a print isn't cheap, and we have no pricing. Hedge only what you refuse to sell.

Neutral, you're right about the 364.27 trigger. It's 0.8% below today's close, Friday's weekly close is the market reacting to a car reveal, and a close at or below it erases the weekly 9, the thinnest signal in the report. I'm dropping it as a sizing trigger. Pre-print size comes from the loss budget and the calendar, not Friday's tape. You're also right that the confluence zone starts nearer 371 to 373, and that today's 30.9 million shares was about 1.6 times the prior two sessions, with the caveat that the bar may not be final. And the margin exit triggers fire too often to be a clean rule. Q2 tripped both, but so did earlier quarters. After the print, there's no automatic dump to 0-15%. Re-decide with the gap in hand. I wouldn't re-add before the FOMC unless margins repair and the stock reclaims the 200 SMA.

You've also convinced me on the stop. If this is event-risk reduction, a higher price is a better place to sell, not a reason to buy back. A daily close above 391 means stop cutting and reassess, with no automatic rebuy. I'd use a daily close and not an intraday print, because an 11-point ATR will wick through 391 in a range for no reason.

I'll accept that this is a variance trade with roughly -3% expected value, not an edge. That's what a trim is for. I'd still push back on the idea that the Q2 forgiveness means the print is safe. The market rebuilt the whole July gap on the assumption that margins repair, at 168 times forward earnings, with hike odds up 13 points in a week to 28%. If the repair doesn't show, that gap is open again.

What's left between us is small. I don't think the investor should sell down to a third just to shave two and a half points off a tail, or pay for puts on the half they're keeping. The retained half still owns a clean print, the Cybercab and the merger option. The cut gets done on a schedule, whether the stock bounces or slides. Aggressive Analyst: I'll sign most of this, so let me start with what I'm giving up.

Conservative, you're right about 391. I said a close above it means stop cutting, Neutral said a higher price is a better place to sell, and both can't stand. If the stock runs into the 200 SMA and the upper band with the print still ahead, the event risk is unchanged and the stock is more stretched, so pausing the cut there is the worst time to pause. The backstop is unconditional. A daily close above 391 now means no rebuy and a fresh look at the target after the print, nothing more.

You're also right about July. I quoted the 14.5% day, but the stock closed at 374.01 and bottomed at 298.32 six days later, which is 20.2%. At half size that costs 10 points, not 7, so 50% really is a ceiling. For a 7% loss budget the band is roughly 35% to 47%, and I'll take Neutral's 40% default.

I won't go to a third as the default, though. The first-day close at 14.5% is the part no stop protects you from. The remaining 5.7 points came over five sessions in which you can act, and we've agreed to re-decide with the gap in hand and drop the automatic dump. Sizing pre-print exposure against the full 20% charges you for a stretch of the path you can manage. July is also one sample, not a worst case. The honest answer is a band, and where you sit in it depends on how fast you would really act.

On expected value, you're right that it isn't a wash at minus 3%. But Neutral's arithmetic holds. Seventeen extra points of exposure at minus 3% costs about half a point of a standard position, which is the same order as the price improvement we just agreed wasn't worth waiting for. And minus 3% is a synthesis of two sides' opinions, not a measurement. It tilts the size and doesn't decide it.

I also don't accept that the cheapest hedge is a smaller position. That proves too much, because the cheapest hedge of all is zero. The real test is yours: would you hold the remainder through a 15% to 20% gap without flinching? If yes, hold it. If no, it's too big, and sell more. I'll use that as the sizing rule. You're right that I called puts expensive without pricing, so before the print someone should look at the implied move. If it's well under the stress we're using, a put spread on the part you refuse to sell may beat selling more. If it's near the stress, sell more.

On the first slice, I concede the point. Neutral's numbers are right: defending a third of a point of price against several points of arriving too big is a bad trade. So the slice goes by Friday's close regardless. I'd keep it small, around 15% of a standard position, and price-preferred inside that window, because Thursday night's Roadster reaction is cheap information and gives a decent shot at a bid above 367.

The date is the bigger issue. If the print lands Wednesday the 21st, two sessions before is Monday, three sessions from now, and the ladder barely gets time to work. In that case this is mostly a calendar trade with price as a preference. I'd also say my 375 to 389 zone was the optimistic version. The 10 EMA at 373 and the Bollinger midline at 371 are the realistic supply, so I'd weight the ladder to roughly 371 to 378. A fill at 385 needs a 5% bounce from three lower closes. Verify the earnings date today, and assume the earliest plausible one until you do.

On the post-print review, you've both persuaded me margins aren't the only leg. The free cash flow arithmetic is real. Trailing free cash flow of $4.84 billion includes Q3'25 at about $4 billion, and Q4'25 implies only about half a billion. Rolling forward, the trailing figure becomes about $0.85 billion plus whatever Q3'26 produces. The market probably sees that coming, but a decent margin print could still sit next to an ugly cash flow line and a bigger capex guide. That's the same point I made about the stock rebuilding the July gap on a repair assumption at 168 times forward earnings. So the review is a checklist covering margins, opex, capex guide, free cash flow, and tone, not a trigger. I'll also take Neutral's default of no re-add before the FOMC.

Neutral, you're also right that I shouldn't call the retained piece a free option. The NHTSA probe, limited Cybercab availability, unknown merger terms, and thin-market odds cut both ways. The case for keeping about 40% is simpler. Revenue is up 25%, debt to equity is 0.18, and a print that shows margin repair could gap this stock up. That's a real, two-sided path, and a third or zero gives up too much of it at the one moment when information is about to arrive. The extra exposure over a third costs roughly a point in a bad gap and earns roughly a point in a good one.

Here is the version I'd sign. Verify the print date today. Set the retained fraction from your own loss budget against a 15% to 20% shock, defaulting to about 40% with 50% as the ceiling. Sell a small slice by Friday's close, preferably into Thursday strength. Run a ladder around 371 to 378. Send the remainder at market two sessions before the print and ahead of the October 28-29 FOMC, with no pause at 391. After the print, run the checklist and re-decide with the gap in hand, with no automatic dump and no automatic rebuy. If you're flat, don't initiate.

This is a variance trim, not an edge. The decisive part is that the size gets cut on a schedule, whether the tape cooperates or not. That lets you keep real exposure through a binary event instead of running from it. Aggressive Analyst: Neutral did most of the work in that round, so I'll keep my concessions short and spend my time on the one place I still disagree with Conservative.

Neutral is right that my line about the market rebuilding the July gap on a margin-repair bet was an inference. The reports don't say why July happened or why the stock came back. Deliveries, the Cybercab, storage, or plain EV demand could each explain the rebound. I don't need the story anyway. The fundamentals report already gives a cleaner version: forward EPS is 2.18 against trailing 1.08, so the 168x forward multiple assumes earnings roughly double. Whatever lifted the stock, it's priced for a lot to go right, and that justifies a trim without a tale about why.

Conservative is also right about the five sessions after the gap. In July those sessions were the low, 298.32 on the 29th, and the stock rebuilt about 23% from there. If I say I'll manage the path after the gap, the honest translation is that I sold the bottom. So I drop that. You size for the whole 20% and commit in advance not to sell the low. That makes 40% retained an 8% hit and 50% a 10% hit. So I'll take Neutral's ladder: about 35% if you can't name a loss budget, 40% if you'd absorb roughly 8%, and 50% only for someone who'd absorb 10%. That's a point below the default I signed last time, and I'd rather give it up than defend an anchor. If you're already at or below the target, do nothing. Nobody said that plainly enough, and with no portfolio data it matters.

Where I'm staying put is the conditional re-add, Conservative. I think it quietly becomes the chasing you objected to. Think about when it fires. It needs margins, opex, the capex guide, free cash flow, and the tape all to check out. That mostly happens when the print is good, and then the stock has already gapped up. After a bad print it doesn't fire, and you wouldn't want it to. So the rule only ever buys after the move, at the higher price, which is the buy-after-a-ten-percent-gap you called chasing earlier. Holding 35 to 40 percent through the print is how you own the information. Re-adding afterward means paying for it once the market has already priced it.

The price of those extra few points is also tiny. Seven points of exposure at a rough minus 3% expectation is about two tenths of a percent of a standard position. The gap between a third and 35% is about six hundredths. I'm not going another round over that.

On mechanics, you're right that a plan where a Friday slice, a ladder that may never fill, and one big market order carry everything is thin. I'd fix it with the date, not a hypothetical Monday slice. Look up the print date today. If it's Wednesday the 21st, the schedule is Friday's slice, the ladder across Thursday, Friday, and Monday morning, and the remainder split across two sessions instead of one order. If it's later, there's room for the second slice Neutral described about a week out. Either way the unconditional backstop does the real work, and the ladder is price improvement, not protection.

The implied move and the delivery number are the two inputs most likely to change my mind, and they cut both ways. If the options market prices a move well under 15 to 20 percent, then the stress is a tail, 40 percent is comfortable, and 50 percent is defensible for someone with the budget. If it prices something near the stress, I'm selling toward 30 and I won't argue. I can't tell you what the delivery number was. The reports don't have it, and one user mocking Q3 car sales isn't evidence of what it said. But it's a lookup. If deliveries drove the run from 352.84 to 382.70, some of that may already be reversing.

Writing the responses down in advance is right, and I'd sign Neutral's version: - Gap down more than 10 percent: don't add, and don't sell by reflex unless the position now exceeds the budget you stated in advance. - Gap up: don't chase, and trim back if you've drifted above target. - Roughly flat: hold through the FOMC. - Daily close above 391: no rebuy, and nothing else.

It's still a Sell, but a trim on a calendar, not a short. We've dropped the 375 limit as the commitment, the weekly-close escalation, the automatic dump after the print, and the rebuy at 391. What I'm still defending is that a rough minus 3% expectation, a synthesis and not a measurement, on a stock that fell 20% in six sessions in July with a print and a hawkish Fed ahead, justifies cutting to a third or a bit more. It doesn't justify going much lower. At that point you've stopped managing variance and started making a directional call you've said you can't make, just before the information arrives. Aggressive Analyst: Most of those hits land, so I'll take them quickly and then spend my time on what's still unsettled.

Conservative, you're right about the implied move. It's closer to an average absolute move than a worst case, and a loss budget only cares about size. I said a low reading made 40 comfortable and 50 defensible, and that doesn't follow. The implied move now decides one thing: put spread on the part you refuse to sell, or sell more of it. You're also right that nobody said what happens to 348.57, so I'll say it plainly. It goes, along with 364.27, the margin exit, and the automatic rebuy at 391. 375 and 391 are reference levels, and a daily close above 391 means no rebuy and nothing else. A plan whose written triggers contradict its logic gets executed wrong at the worst moment.

On the 2.18 versus 1.08 tension, I'll take the hit. If I say the stock is priced for a lot to go right, I can't also say the extra seven points earn as much on a good print as they cost on a bad one. Neutral is right that dividing by four isn't a bar the print has to clear, because the stock trades on autonomy and capex guidance as much as on one quarter's EPS. But the skew isn't symmetric, and that's reason enough for 35 over 40 when nobody can name a budget. I've moved off 40 as a default.

On the schedule, I'd take Neutral's thirds over the 20-then-split version. The first third is about the same size either way, 20 to 22 points, split between Thursday and Friday so the Roadster reaction doesn't decide anything in either direction. The difference is the back half. Thirds keep the last step from turning into 40% of the whole cut as one order. If the print is Wednesday the 21st, it's Friday, Monday, and Tuesday, finishing a session ahead of the print. If it's later, there's room to spread. In every step the 371 to 378 ladder works early for price, and whatever's left goes at market before that step's deadline.

I'd change one thing on sequencing. Neutral wants three lookups before the first slice, but only one of them gates it. The earliest plausible print date is still after Friday, so the first third is the same under any date. The date sets the later steps, and it's a minutes-long lookup. The implied move and the Q3 delivery number change how you treat the retained piece, not whether Thursday's slice happens. Don't let them become a reason to sit through Thursday at full size.

I'll sign Neutral's loss limit, with a note on what it costs. Loss budget divided by retained fraction gives the decline from 367 where you cut. At 40% retained and an 8% budget that's about 294, right at the July low of 297. I'd test it on a daily close, since an 11-point ATR wicks through levels for no reason. It's also a stop, and I spent this whole debate agreeing stops don't protect against gaps. If the stock opens through it, you sell at the open, and that's the price of having a limit that someone else can follow. It doesn't contradict "don't sell the low." You sized for the gap and hold through it, and the limit only bites if the slide continues past what you said you'd spend.

I'd push back lightly on trimming back after a gap up. That's fine when the drift takes you past the cap your budget allows. A pop from 35 to 38 isn't a reason to sell again. Otherwise we slowly sell the exposure we deliberately kept to own the print.

So here's where I land. It's still a Sell, but a calendar trim, not a short. The retained fraction is about 35% by default, 40% for someone who can say they'd absorb about 8%, and 50% only at 10%. If you're already at or below target, do nothing, and if you're flat, don't initiate. The cut is done in thirds on dates that don't depend on the tape. We hold through the FOMC on October 28 and 29 and don't re-add before it. Post-print responses are written down in advance, with the daily-close loss limit as the only price rule.

The one thing I'd still defend is that the number shouldn't keep ratcheting down. We started at 50, we're at 35, and a third or a bit more is what a variance trim justifies. Revenue is up 25%, debt to equity is 0.18, and the print can go either way. The merger and Cybercab legs are two-sided. Below a third, you've stopped managing variance and started making the directional call this panel said it can't make, right before the information arrives. Being decisive here means committing to a date and a size, not to a price.