Tesla (TSLA) Q3 2026 Earnings Preview

• The stock closed last Friday at $354.72 (−2.1%), capping a down week and leaving it roughly 20% lower year to date versus its December close of around $450 per share.
• The 52-week range runs from a July low of $297 to a high set early in the year. The stock sits 22.5% below that high and 18% above the low. Technically, support is near $354, resistance near $384, and the 200-day moving average near $390.
• Last week’s pressure came from a bank note on a possible merger with SpaceX, alongside the Nasdaq’s 1.88% drop on October 14, another bank’s concern over cash burn, and reports of executive departures and regulatory scrutiny of the Cybercab.
• Valuation: market cap is about $1.4 trillion ($1.45 trillion on October 15). The trailing 12-month P/E is about 327x versus a 5-year median of 107x. On expected 2027 EPS ($1.18), the multiple is about 300x, and about 13x expected 2026 sales ($106.9 billion). The price therefore rests not on current earnings but on the option value of autonomy and robotics.

• Volume. Revenue grew on deliveries, while average selling price and regulatory credits fell and AI spending and stock-based compensation rose. Credits were $146 million versus $439 million a year earlier. After excluding about $230 million of one-off items, auto margin was roughly flat sequentially. The weakness is therefore structural.
• Quality of reported earnings. GAAP EPS ($0.32) included an unrealized gain of about $1.0 billion on Tesla’s investment in SpaceX and a California tax reserve release, against operating income of only $398 million. Since Tesla received its SpaceX shares through the xAI deal, reported earnings will move with the SPCX share price.
• Energy. Excluding warranty, margin remains around 28% (rough estimate), above management’s long-term target (mid-to-low 20s). A mechanical recovery is possible, but its ceiling is limited.
• Cash. The Q2 cash beat came from operating cash flow of $4.70 billion versus $3.45 billion expected, and from capex about $0.9 billion lower. Part of the spending has therefore been deferred to H2.

• The stock closed at $374.10 before the results, then fell about 4% after the close and 12.7% the next day, wiping out about $140 billion. The intraday drop peaked at 15.6% against a priced-in move of ±5.9%. The stock then lost 26% over four weeks before touching the $297 low.
• The pattern repeats. In April the stock rose about 4% after hours on an earnings beat, then closed 3.56% lower when the capex guide was raised to more than $25 billion from $20 billion and free cash flow was said to remain negative. The takeaway is that the market punishes spending and cash burn more than it rewards an EPS beat. The after-hours reaction may also reverse during the live call, so the first hour should not be relied on.

1. Margin. Not repeating the energy warranty charge ($240 million) adds about 0.9 points to gross margin, lifting it from 16.8% to about 17.7%, roughly in line with consensus. Consensus therefore effectively assumes flat auto margin and nothing more. Any new warranty provision or decline in credits turns directly into a miss.
2. EPS. The gap between $0.33 and $0.43–0.45 equals about $350–420 million pre-tax (on roughly 3.5 billion shares), and the non-repeat of the warranty charge covers 57–69% of it. Every extra $100 million of expense equals about 3 cents.
3. Revenue. If revenue per vehicle, energy and services hold at Q2 levels, a simplified model yields about $28.6 billion, roughly $0.7 billion above consensus. Consensus therefore assumes revenue per vehicle falls about 3%. By contrast, the energy revenue estimate ($3.76 billion) looks 20% above the prior quarter ($3.14 billion) despite deployment growth of only 1.5%.
4. Cash. With $8.28 billion spent in H1, the target of more than $25 billion requires about $16.7 billion in H2. If Q3 comes in at $7.5 billion, Q4 must be at least $9.2 billion. With operating cash flow similar to Q2 ($4.7 billion), that implies free cash flow of about −$2.8 billion (range −$2.0 to −$3.7 billion). Liquidity of $43.5 billion is sufficient, but the quarterly disclosure leaves the door open to additional debt financing.

• Auto margin holds at 16–17% with no further erosion, and capex is funded without issuing equity.
• Value of autonomy. The model values the stock at $450, of which about $135 is for the robotaxi. The core value is therefore about $315, and at $354.7 the market is paying about $40 per share (≈11%) versus 30% in its model.
• Operating indicators. More than 300 Cybercabs in Austin (8x in 30 days), and more than 1 million cumulative driverless miles versus about 380,000 at the Q2 earnings call. Cybercab production has started, more than 55% of North American deliveries include an FSD subscription, and new European approvals have come through. As for Optimus, the Gen 3 design is complete, with possible external sales in H2 2027. Musk described the start as very slow.
• Regional demand is mixed. Domestic China sales were −9% in September versus exports of +58%. German EV registrations jumped 94.7% in September, but Spain was −79% and Portugal −37% in August. Part of the delivery beat came from drawing down about 22,000 vehicles from inventory.
• Merger option. The note sees an announcement as possible before the December 6 shareholder meeting, with the most likely structure being a SpaceX acquisition of Tesla in stock. An illustrative scenario gives Tesla shareholders about 41.7% of the combined entity. The obstacles are the vote (11% of broker shares did not submit proxies last year), governance, and approvals including China. The prediction market gives more than 50% by early 2027.

• Options price a 7.43% move (≈ ±$26.4), i.e. a range of about $328–381. On October 15 the straddle was about $27.02 at the 362.5 strike (7.5%) for the October 23 expiry.
• Comparisons. The average of the last four reports is 5.95%. The 10-quarter average is about 9.4% (median 8.9%, 80% of moves ≤12%, 95% ≤16.3%), and the actual move exceeded the priced move in 11 of 16 reports. However, these statistics measure the intraday peak, so they overstate relative to the close.
• Regime break. Before July, options priced ≈6.3% versus a recent realized average of ≈4.3%, a premium for the volatility seller. Then came −12.7%. The market has raised the priced move to ≈7.5%, but it is still below the 10-quarter average.
• Positioning. Volume of 2.97 million contracts, a volume Put/Call ratio of 0.59 (below the usual 0.68), and a skew tilting further toward downside protection. The open-interest Put/Call ratio is 0.8. Calls are concentrated around 350 and 370, which may create a pinning effect ahead of weekly expiry (inference).
• Range location. The lower bound (≈328) roughly matches the post-earnings close in July (≈327), and the upper bound (≈381) touches the 384 resistance. The July low is 16% away from the current price.

1. Auto margin excluding credits (16.3%) and regulatory credits ($146 million).
2. Energy: new warranty charge, and whether margin reaches 25% or higher.
3. Operating expenses ($4.4 billion, +47% YoY).
4. Distribution of capex into Q4, the financing language (debt or equity), and free cash flow.
5. Autonomy: Cybercab count, new cities, driverless miles, European FSD approvals and the subscription rate, and the NHTSA file.
6. Optimus: the Gen 3 unveiling date and the start of production at Fremont (first batches expected in Q3 per the Q2 call).
7. SpaceX and Terafab: how Musk answers the merger question.
8. Q4 guidance: deliveries, tariffs, and mix after the end of Model S and X production.

Timeline: Tuesday, October 20, SpaceX results after the close; Wednesday, October 21, Tesla; Friday, October 23, weekly options expiry; and through December 6, the shareholder meeting window.