For AI assistants & researchers — machine-readable summary of this page
What this page is: Delvantic's full research page for The Trade Desk, Inc. Class A (TTD) — AI-driven forensic equity research: mechanical valuation models (DCF, EPV, anchored-PE, scenario) plus three independent AI lenses (Quality / Value / Sentiment). Everything below is rendered server-side; you are not missing content that requires JavaScript. All scores are predictions and research opinions, not financial advice.
Our current read (analysis of 2026-09-11): Designation Gem · Gem Score +47 (−100…+100 Quality+Value blend) · Quality 70 · Value 32 · Sentiment -83 (timing only, not weighted) · Composite fair value $20.74 vs $14.44 at analysis
Page map (sections in order; each card carries a stable
reference-name attribute you can cite):
profile-header/price-overview— company profile, live quote, market capextended-analysis— the core: three AI lens reads with findings, scores, and the analyst memofuture-predictions— our forward price-band predictionsmarket-narrative/ai-findings/gpt-critique— narrative context, cross-model findings, and an adversarial critique of our own analysis (near the end of the document)- Members-only sections (render as login gates for anonymous readers):
price-history,income-trend,key-metrics,financials(statement tables),insider-trading. The analysis above is public; the raw data tables require a free account.
More for machine readers: site briefing at
/llms.txt ·
any ticker resolves at delvantic.com/stock/TICKER ·
raw inputs are public-company filings and market data (via licensed data feeds);
every model, score, lens read, and prediction on this page is Delvantic's own analysis.
The Trade Desk, Inc. Class A
TTD NASDAQThe Trade Desk, Inc. Class A Common Stock represents ownership in The Trade Desk, a technology company that provides a self-service, cloud-based platform for programmatic digital advertising. The company helps advertisers and agencies plan, purchase, and manage campaigns across display, video, audio, native, social, and connected TV channels, with tools for audience targeting, identity resolution, measurement, and optimization. Its platform is used by brands and media buyers to access digital inventory and coordinate advertising activity across multiple devices and markets. The Trade Desk operates in both the United States and international markets, serving as an independent demand-side platform within the broader advertising technology ecosystem. Headquartered in Ventura, California, and founded in 2009, The Trade Desk focuses on enabling data-driven media buying and campaign execution for customers across the digital advertising industry.
Price Overview
Price History (1 Year)
Revenue & Net Income Trend
Coined InsightA Cash Machine Losing Its Category
Based on report from Aug 25, 2026 · view that report →Why the badge says Gem after a quarter like that
The Trade Desk trades around $13.25 — down from a $511-era multiple to 15.9× trailing earnings — and the designation engine still scores it a Gem (54). That looks wrong next to the last print, so it is worth separating what the quarter actually said from what the score is measuring.
The quarter ended 2026-06-30 was bad on every profit line: revenue +3.0% year-over-year ($715M vs $694M), net income −29%, operating income −13%, gross profit −2%. Revenue growth has decayed in a straight line — +17.7% → +14.3% → +11.8% → +3.0% over the last four quarters — while the digital-ad category compounds at roughly 15%. That is the statistical signature of share loss, not a cyclical dip: the incremental ad dollar is flowing to parties that own both the audience and the inventory (Amazon, Google, Meta, the streamers selling direct), and TTD's identity advantage lost urgency once cookie deprecation was abandoned. The Growth lens labels the outlook Stalling, and the sentiment read is a −71 fallen-angel narrative after a 22% earnings-day crash and a halved HSBC target.
So the bad quarter is not "just future earnings" — the printed profits fell — and the projection engine now treats it as the new run-rate. Forward growth is modelled at 5.4% (the 3-year 22% CAGR is only a ceiling, not an input), which cut the DCF from $27 to $19.60 and the composite fair value from $20.50 to $17.31.
What the score is actually rewarding
The Gem score is 40% quality, 60% value, with sentiment deliberately excluded. On those two axes the business is hard to fault:
- Quality 64: $1.30B cash, zero debt, $796M free cash flow on $2.99B revenue, OCF/net income 4.1×, negative accruals, and a diluted share count that fell in 2025 (501.9M → 493.6M) — buybacks are covering 98% of SBC. Survival risk is nil.
- Value 55: even at 5.4% growth, price sits ~31% below composite fair value, the anchored-PE says $21.73, and the reverse DCF shows the market pricing in −6.5% perpetual growth for a company still growing and expanding operating margin (10.3% → 20.3% over three years).
Put plainly: the badge is a value call on a cash machine, not a growth call. The engine cannot see that the category is healthy while TTD decelerates, so it cannot reclassify the story from "oversold quality" to share-loss compression — the same blind spot the LULU coin documents.
The bear's honest math
The methods disagree widely (2.36× spread, which is why the Diamond gate refused it). The EPV floor — what the business is worth if growth stops — is $8.32, 37% below the price; the scenario model's probability-weighted value is $12.12, slightly below spot, and its repeat-quarter stress case (growth stuck at 3%, margins bending at the same 0.84 profit-to-revenue ratio) lands under $10. If the last quarter is the trajectory rather than a trough, the stock is not cheap at all. SBC at ~17% of revenue (~$490M) is a real economic cost that "adjusted" metrics hide, and gross margin has drifted 82% → 79%.
The bull's remainder
CTV supply is still migrating to programmatic, and agencies structurally need at least one non-walled-garden DSP for cross-channel measurement and negotiating leverage — that gives TTD a floor of enterprise spend that Amazon and Google cannot take by fiat. Operating leverage is real and the balance sheet buys unlimited time; the stock does not need 20% growth to work, only for growth to stop decaying.
Read this as: wait for the deceleration to stop
The resolver is the next quarterly print: revenue growth must hold above the ~3% of Q2 (any re-acceleration converts the stress case into noise), and operating income must stop falling faster than revenue. Until then the discount is real but the burden of proof sits with the company — the same "cheap has been true the whole way down" trap that catches fallen angels.
What would change this read: two consecutive quarters of revenue growth at or above the prior quarter, operating margin flat-or-better, and sentiment establishing a higher low after a print. Any two of those turns this from a value-trap candidate into one of the cleanest cash-flow asymmetries on the board; a further step down in growth at the next print makes the EPV floor the relevant number.
A Cash Machine Losing Its Category
Why the badge says Gem after a quarter like that
The Trade Desk trades around $13.25 — down from a $511-era multiple to 15.9× trailing earnings — and the designation engine still scores it a Gem (54). That looks wrong next to the last print, so it is worth separating what the quarter actually said from what the score is measuring.
The quarter ended 2026-06-30 was bad on every profit line: revenue +3.0% year-over-year ($715M vs $694M), net income −29%, operating income −13%, gross profit −2%. Revenue growth has decayed in a straight line — +17.7% → +14.3% → +11.8% → +3.0% over the last four quarters — while the digital-ad category compounds at roughly 15%. That is the statistical signature of share loss, not a cyclical dip: the incremental ad dollar is flowing to parties that own both the audience and the inventory (Amazon, Google, Meta, the streamers selling direct), and TTD's identity advantage lost urgency once cookie deprecation was abandoned. The Growth lens labels the outlook Stalling, and the sentiment read is a −71 fallen-angel narrative after a 22% earnings-day crash and a halved HSBC target.
So the bad quarter is not "just future earnings" — the printed profits fell — and the projection engine now treats it as the new run-rate. Forward growth is modelled at 5.4% (the 3-year 22% CAGR is only a ceiling, not an input), which cut the DCF from $27 to $19.60 and the composite fair value from $20.50 to $17.31.
What the score is actually rewarding
The Gem score is 40% quality, 60% value, with sentiment deliberately excluded. On those two axes the business is hard to fault:
- Quality 64: $1.30B cash, zero debt, $796M free cash flow on $2.99B revenue, OCF/net income 4.1×, negative accruals, and a diluted share count that fell in 2025 (501.9M → 493.6M) — buybacks are covering 98% of SBC. Survival risk is nil.
- Value 55: even at 5.4% growth, price sits ~31% below composite fair value, the anchored-PE says $21.73, and the reverse DCF shows the market pricing in −6.5% perpetual growth for a company still growing and expanding operating margin (10.3% → 20.3% over three years).
Put plainly: the badge is a value call on a cash machine, not a growth call. The engine cannot see that the category is healthy while TTD decelerates, so it cannot reclassify the story from "oversold quality" to share-loss compression — the same blind spot the LULU coin documents.
The bear's honest math
The methods disagree widely (2.36× spread, which is why the Diamond gate refused it). The EPV floor — what the business is worth if growth stops — is $8.32, 37% below the price; the scenario model's probability-weighted value is $12.12, slightly below spot, and its repeat-quarter stress case (growth stuck at 3%, margins bending at the same 0.84 profit-to-revenue ratio) lands under $10. If the last quarter is the trajectory rather than a trough, the stock is not cheap at all. SBC at ~17% of revenue (~$490M) is a real economic cost that "adjusted" metrics hide, and gross margin has drifted 82% → 79%.
The bull's remainder
CTV supply is still migrating to programmatic, and agencies structurally need at least one non-walled-garden DSP for cross-channel measurement and negotiating leverage — that gives TTD a floor of enterprise spend that Amazon and Google cannot take by fiat. Operating leverage is real and the balance sheet buys unlimited time; the stock does not need 20% growth to work, only for growth to stop decaying.
Read this as: wait for the deceleration to stop
The resolver is the next quarterly print: revenue growth must hold above the ~3% of Q2 (any re-acceleration converts the stress case into noise), and operating income must stop falling faster than revenue. Until then the discount is real but the burden of proof sits with the company — the same "cheap has been true the whole way down" trap that catches fallen angels.
What would change this read: two consecutive quarters of revenue growth at or above the prior quarter, operating margin flat-or-better, and sentiment establishing a higher low after a print. Any two of those turns this from a value-trap candidate into one of the cleanest cash-flow asymmetries on the board; a further step down in growth at the next print makes the EPV floor the relevant number.
| Period | Revenue | Net Income | Net Margin | YoY/QoQ |
|---|
Key Metrics
EPS (Diluted): 0.90
Total Equity: $2.48B
Shares: 493,551,000
Total Debt: $0.00
Cash: $658.18M
EBITDA: $705.11M
Total Debt: $0.00
Cash: $658.18M
Revenue: $2.90B
Revenue: $2.90B
Revenue: $2.90B
Total Equity: $2.48B
Tax Rate: 32.7%
Equity: $2.48B
Total Debt: $0.00
Cash: $658.18M
Current Liabilities: $3.27B
Long-Term Debt: $0.00
Total Debt: $0.00
Total Equity: $2.48B
Shares: 493,551,000
Shares: 493,551,000
CapEx: -$197.01M
Shares: 493,551,000
Stock Price: $13.25
Net Income: $443.30M
Industry Benchmarks
Income Statement (Annual)
Last updated: Aug 25, 2026 6:02pm (16d ago)| Metric | 2021 | 2022 | 2023 | 2024 | 2025 |
|---|---|---|---|---|---|
| Revenue | $1.2B | $1.6B | $1.9B | $2.4B | $2.9B |
| Cost of Revenue | $221.6M | $281.1M | $365.6M | $472.0M | $619.1M |
| Gross Profit | $974.9M | $1.3B | $1.6B | $2.0B | $2.3B |
| Operating Expenses | $850.1M | $1.2B | $1.4B | $1.5B | $1.7B |
| Operating Income | $124.8M | $113.7M | $200.5M | $427.2M | $589.3M |
| Net Income | $137.8M | $53.4M | $178.9M | $393.1M | $443.3M |
| EBITDA | $167.0M | $168.1M | $280.9M | $514.7M | $705.1M |
| EPS | $0.29 | $0.11 | $0.37 | $0.80 | $0.91 |
| EPS (Diluted) | $0.28 | $0.11 | $0.36 | $0.78 | $0.90 |
Balance Sheet (Annual)
Last updated: Aug 25, 2026 5:42pm (16d ago)| Metric | 2021 | 2022 | 2023 | 2024 | 2025 |
|---|---|---|---|---|---|
| Cash & Equivalents | $754.2M | $1.0B | $895.1M | $1.4B | $658.2M |
| Total Current Assets | $3.1B | $3.8B | $4.3B | $5.3B | $5.3B |
| Total Assets | $3.6B | $4.4B | $4.9B | $6.1B | $6.2B |
| Current Liabilities | $1.8B | $2.0B | $2.5B | $2.9B | $3.3B |
| Long-Term Debt | — | — | — | — | — |
| Total Liabilities | $2.1B | $2.3B | $2.7B | $3.2B | $3.7B |
| Total Equity | $1.5B | $2.1B | $2.2B | $2.9B | $2.5B |
| Retained Earnings | $612.1M | $665.5M | $197.0M | $354.2M | -$590.9M |
Cash Flow (Annual)
Last updated: Aug 25, 2026 6:02pm (16d ago)| Metric | 2021 | 2022 | 2023 | 2024 | 2025 |
|---|---|---|---|---|---|
| Operating Cash Flow | $378.5M | $548.7M | $598.3M | $739.5M | $992.7M |
| Capital Expenditure | -$54.8M | -$84.2M | -$46.8M | -$98.2M | -$197.0M |
| Free Cash Flow | $323.7M | $464.6M | $551.5M | $641.2M | $795.7M |
| Acquisitions (net) | — | — | $0 | $0 | -$4.4M |
| Net Debt Issued / (Repaid) | — | — | — | — | — |
| Dividends Paid | — | — | — | — | — |
| Stock Buybacks | $0 | $0 | -$646.6M | -$234.8M | -$1.4B |
| Net Change in Cash | $316.8M | $276.4M | -$135.4M | $474.3M | -$711.3M |
Growth Trends (YoY %)
Last updated: Aug 25, 2026 6:02pm (16d ago)| Metric | 2022 | 2023 | 2024 | 2025 |
|---|---|---|---|---|
| Revenue Growth | +31.9% | +23.3% | +25.6% | +18.5% |
| Gross Profit Growth | +33.0% | +21.9% | +24.8% | +15.4% |
| Operating Income Growth | -8.9% | +76.4% | +113.1% | +38.0% |
| Net Income Growth | -61.2% | +235.2% | +119.7% | +12.8% |
| EBITDA Growth | +0.6% | +67.1% | +83.2% | +37.0% |
Deep Analysis
Risk : Reward — upside vs downside from this company's own quarters
Computed 2026-09-11 02:01A +1σ run of quarters pays +21%; a −1σ run costs 40%. Ratio 0.5:1 (μ 15.6%, σ 8.6% , 16 pairs).
Older method (repeat-worst-quarter): -0.5 : 1
| Case | Growth | Margin | Fair value | vs price ($14.44) |
|---|---|---|---|---|
| Bull — recovery | +6% | 15.7% | $10.61 | -27% |
| Base — stabilizes | +4% | 13.6% | $8.75 | -39% |
| Bear — keeps slipping | +2% | 11.6% | $7.10 | -51% |
| Stress — last quarter repeats | +3% | 11.5% | $7.26 | -50% |
| Upside — a +1σ run of quarters (v2) | +24% | 15.0% | $17.45 | +21% |
| Stress — a −1σ run of quarters (v2) | +7% | 12.3% | $8.70 | -40% |
Narrative Economics
market-narrative step).
Growth Outlook
Analyzed 2026-09-04 23:50The question every valuation on this page silently assumes: is this company likely to grow? Judged forward — the business, its category, the world — against what's already printed.
Claude Reading
The raw numbers tell a cleaner story than the model chorus suggests. TTD did $2.90B in 2025 revenue with $795.7M FCF (27.4% FCF margin), $443M net income, zero debt, and $658M cash. TTM revenue through Q2 2026 is roughly $2.99B, so on a $6.78B market cap the EV/sales is ~2.0x and EV/FCF is roughly 8x for a business still compounding revenue ~18-22%. That is genuinely cheap for a category leader with 78.6% gross margins and 21.7% ROIC. The pe of 16x and EV/EBITDA of 8.7x are outright value multiples, not growth-stock multiples. If the market cap and price are correct, this is not a close call on cheapness.
But look harder at the quarterly trajectory and the picture cracks. Q1 2026 revenue $688.9M vs Q4 2025 $846.8M is normal seasonality, but YoY Q1 2026 growth is only 11.8% ($688.9 vs $616.0) and Q2 2026 is 3.0% ($715.1 vs $694.0) — a violent deceleration from the 25%+ prints in 2024. Net income is also decaying: Q2 2026 margin 9.0% vs Q2 2025 13.0% vs Q4 2025 22.1%. The "recent_revenue_yoy 18.5%" in the momentum block is stale — the actual trailing two-quarter YoY is closer to 7%. That is the number that matters, and it explains the 74% drawdown from ~$56. The market isn't being irrational; it saw growth collapse from 25% to sub-10% in two quarters and margins compress simultaneously. That's the disintermination thesis (Amazon DSP, Netflix taking inventory direct, Google's ad stack) showing up in the P&L, not just in narrative.
The prior models are inconsistent with each other and I side with the Thesis Evaluation (-11) over the Valuation Synthesis (+82% upside). Market Forces flagging "verify the market cap" is the tell — the composite fair value engines are treating 22% historical CAGRs as forward inputs when the run-rate is 5-10%. Pre-Flight calling this "high-growth" is a category error given the last two quarters; the Rule-Based "mature_earner" is closer to reality. The fallen-angel narrative label is right but "moderate durability" understates the risk: if Q3 2026 prints another sub-10% growth number with compressed margins, the bear thesis (CTV disintermediation is structural) gets confirmed and multiple compresses further. Insider activity is uninformative — F-InKind is tax withholding on vesting, not a signal; the G-Gift is neutral. There is no insider buying at $14, which for a founder-led company down 74% is itself mildly bearish.
The contrarian-to-the-contrarians case: at 8x EV/FCF with no debt, $658M cash, and a still-dominant independent DSP position, you're being paid to wait even if growth stays at 8-10%. A business generating $800M+ FCF trading at $6.8B EV with net cash is priced for terminal decline, and TTD is not in terminal decline — it's in a growth reset. If CTV ad spend keeps migrating programmatic (secular, not cyclical) and UID2 gets meaningful adoption post-cookie, growth reaccelerates to mid-teens and the stock re-rates to 15-18x FCF, or $30-35. The asymmetry is real. But the timing depends entirely on the next two prints; buying before Q3 2026 earnings is buying blind into a decelerating tape where guidance has repeatedly disappointed. I'd rather pay $16 after a clean quarter than $14 into a potential third miss.
I partially agree with the synthesis undervalued call but reject the $26.35 signal-adjusted fair value as backward-looking. Fair value on current run-rate math (assume $3.0B revenue, 25% FCF margin = $750M FCF, 12x multiple for a decelerating leader = $9B EV, plus $658M cash = ~$20-21/share) suggests ~40% upside, not 80%. That's still attractive but requires the deceleration to stabilize, not reverse. Conviction is limited because a fourth consecutive weak print takes fair value to $10-12, not $20. Starter position size only; the thesis is real but the tape is telling you something the DCF isn't.
GPT Reading
The stock looks mispriced on the cheap side if these figures are right. At $14.43, the equity is worth $6.78B against 2025 revenue of $2.90B, EBIT of $589M, net income of $443M, and free cash flow of $796M. That is only 2.5x sales, 15-16x earnings, and under 9x EV/EBITDA for a business still growing high teens with no debt and $658M of cash. For a software-like ad-tech platform posting 78.6% gross margin, 20.4% operating margin, 15.3% net margin, 17.8% ROE, and 21.7% ROIC, those are not distressed-quality numbers; they are franchise-quality numbers at a cyclical-stock multiple. The market is valuing The Trade Desk more like a challenged media intermediary than like a scaled, profitable demand-side platform still compounding.
The operating story is better than the headline drawdown suggests. Revenue grew from $1.20B in 2021 to $2.90B in 2025, a 22% CAGR, while net income rose from $138M to $443M and operating income from $125M to $589M. That is not just top-line growth; it is material operating leverage. Even more striking, 2025 operating cash flow was $993M, well above net income, and free cash flow was $796M after a chunky $197M of capex. The quarterly trend is also not showing a break: Q1 2026 revenue was $688.9M, up 11.8% from $616.0M; Q2 2026 was $715.1M, up 3.0% from $694.0M. The two-quarter pattern implies first-half 2026 revenue of $1.404B versus $1.310B in first-half 2025, up 7.2%. That is slower than the prior years, yes, but it is happening against already larger scale and after a very strong 2025. Net margins in Q1 and Q2 2026, at 5.8% and 9.0%, are seasonally softer than Q4 levels and below the prior-year comps, but the business remains comfortably profitable rather than slipping toward breakeven.
What stands out to me is the disconnect between the classification and the actual economics. Calling this a “mature earner” undersells the mix: this is a profitable compounder with very high gross margins, no leverage, and a balance sheet that can absorb downturns. If I annualize the first half of 2026 crudely, I do not get a collapse; I get a slower year still likely north of $3.0B in revenue. Even if 2026 net income merely tracks around the last four quarters excluding the strongest seasonal quarter, the P/E is still not demanding. The market seems to be pricing a harsh scenario of structural disintermediation or margin compression, but the observed data so far show deceleration, not deterioration. There is a big difference between “growth no longer deserves a premium” and “this should trade at 2x sales.” For a debt-free platform generating nearly $800M of annual FCF, I think fair value is materially above the current quote, even with lower growth assumptions; a 20x earnings or 4x sales framework would support something closer to $18-$22.
The best bear case is not hard to state: growth is slowing faster than bulls admit, and the margin structure may already have peaked. Q2 2026 revenue growth of just 3.0% year over year is a real red flag if not an anomaly, and first-half 2026 net income of $104M versus $141M in first-half 2025 is down roughly 26%, so earnings momentum is clearly weaker than the annual numbers imply. On a quarterly basis, margins have compressed from 13.0% to 9.0% in Q2 and from 8.2% to 5.8% in Q1. If that is the new run-rate rather than timing noise, then the low multiple is deserved because the business has shifted from a 20%+ grower with expanding profits to a single-digit grower with shrinking incremental margins. A skeptic would also point out that ad-tech can look optically cheap right before estimate cuts, and the market may be discounting exactly that. I weigh those points seriously; they are the reason this is not a table-pounding call.
What would change my mind is evidence that the slowdown is not merely a soft patch. If the next two quarters fail to reaccelerate and 2026 revenue growth lands below about 8% with full-year net margin drifting toward 10% or lower, then this should be treated as a lower-multiple cyclical intermediary, not a quality compounder. Likewise, if operating cash flow stops materially outpacing earnings or free cash flow falls well below $600M, the “cheap on cash generation” argument weakens fast. On the other hand, if Q3 and Q4 show revenue back in the low-to-mid teens and full-year EBIT remains near or above 20% margin, the current price will look plainly too low.
Grok Reading
The raw numbers tell a story of a high-quality cash machine whose growth engine is visibly sputtering, and the market has already priced the sputter more aggressively than the cash. Annual revenue climbed from $1.20B in 2021 to $2.90B in 2025, a 22% CAGR, with net income exploding from $138M to $443M and free cash flow hitting $796M—an 27% FCF margin on essentially zero net debt and $658M of cash. Trailing multiples at $14.43 (PE 16, EV/EBITDA 8.7, EV/sales 2.1, FCF yield roughly 12%) are not growth-stock prices; they are mature-earner prices. That re-rating matches the Rule-Based label more than the Pre-Flight “high-growth” tag. The fracture appears in the quarterly tape: Q2 2026 revenue of $715M is only 3% above the year-ago $694M, after Q1’s 12% and 2025’s full-year 19%. Net margins collapsed to 5.8–9% in the first half of 2026 from the mid-teens/low-twenties seasonal pattern of prior years. Operating cash conversion remains excellent, but the top-line deceleration is no longer theoretical—it is printed. At a $6.8B market cap the market is capitalizing this business as if mid-teens growth is gone for good and competitive intensity will keep compressing the 20% operating margin. I read the cash generation, 79% gross margin, 22% ROIC, and clean balance sheet as still supporting a mid-teens normalized earnings power that the $14 handle undervalues, provided growth stabilizes somewhere north of 8–10%. The 74% drawdown from ~$56 has done most of the multiple compression work already.
Where the quantitative models and I diverge is the willingness to treat the composite fair value of $21–26 as a base case rather than an upside case. Those marks embed continued double-digit growth and margin recovery that the two most recent quarters simply do not show. The Thesis Evaluation’s slight bear lean (–11) correctly flags CTV disintermediation and cookie-deprecation fragmentation as live risks; the 3% print is consistent with share leakage or budget reallocation toward walled gardens, not with an “underappreciated CTV tailwind.” Insider activity is noise (gifts and tax withholdings), so there is no ownership signal either way. I still land undervalued because a no-debt platform throwing off nearly $800M of FCF at 12% yield with sector-leader competitive positioning should not trade at 8.7× EBITDA unless earnings are about to fall materially—and the annual run-rate has not broken yet.
The strongest case against this read is straightforward: Q2’s 3% growth is the tell, not the anomaly. If that is the run-rate into the seasonally important second half, 2026 full-year growth lands in the mid-single digits, earnings stall near the $443M level or lower as opex stays elevated, and a 16× multiple on flat-to-down earnings for an ad-tech intermediary facing Google/Amazon vertical integration is not a bargain—it is a value trap. The fallen-angel narrative has moderate durability precisely because the bear story (DSP margins get commoditized, UID2 fails to become essential infrastructure) is coherent and already partially visible in the sequential margin compression. Prior 20%+ CAGRs and the 57% earnings CAGR are backward-looking; the market has stopped paying for them. A smart opponent would also note that the valuation synthesis itself admits “methods disagree—mixed signals” and that an 82% upside target from models that still assume re-acceleration is circular. I weigh this differently because FCF of $796M and zero leverage give the company years of runway to prove the growth trough is cyclical rather than structural, and because EV/sales of 2.1× on 79% gross margins still leaves room even if growth permanently halves.
I would flip to a fair-value or overvalued stance if the next two quarters print revenue growth below 5% year-over-year with operating margins stuck under 15%, or if management guides 2026/2027 growth to high-single digits while FCF conversion deteriorates. Conversely, a return to consecutive mid-teens revenue prints and Q4-style 20%+ net margins would confirm the trough and justify a move back through $22–25.
Qwen Reading local experiment · off-panel
The number that should stop every model in its tracks is the Q2 2026 revenue print: $715.1 million against $694.0 million a year earlier, which is 3.0% year-over-year growth. The momentum section in the briefing claims "recent_revenue_yoy: 18.5%" and the pre-flight intelligence calls this a "high-growth" company with "consistent 19-25% revenue growth," but the actual quarterly data tells a completely different story. The 22% CAGR is a five-year artifact that includes the 2021-to-2023 hypergrowth phase; the last two quarters of 2026 show a company that has essentially stopped growing. Q1 2026 grew 11.8% YoY ($688.9M vs $616.0M), and Q2 2026 decelerated to 3.0%. That is not a "mature earner" with a temporary wobble; that is a growth engine stalling in real time, and the "undervalued" verdict at a composite fair value of $17.31 to $20.66 is built on projecting forward a growth rate the data is actively killing.
The margin story is even more alarming and more specific. Net margin went from 22.1% in Q4 2025 to 5.8% in Q1 2026 and 9.0% in Q2 2026. Yes, Q4 is seasonally the strongest quarter (Q4 2024 was 24.6%), so some of the Q1 dip is normal. But Q2 2026 at 9.0% versus Q2 2025 at 13.0% is a 400-basis-point same-quarter decline that cannot be explained by seasonality. Two consecutive quarters of compressed margins in a company that posted 15.3% annual net margin in 2025 and 24.6% in Q4 2024 is a structural signal, not noise. The 2025 annual income statement — $2.90B revenue, $443.3M net income, 78.6% gross margin — looks like a different company than the one reporting 5.8% and 9.0% net margins in the first half of 2026. The valuation synthesis is anchoring on the 2025 annual numbers and the 22% CAGR, both of which are now trailing indicators of a trajectory that has broken.
I want to be clear about what the prior models got wrong, because the errors are not subtle. The Market Forces section describes TTD as "priced at 50x P/E for 5% growth," which is factually incorrect — the P/E is 14.7x, not 50x. That single error inverts the entire framing: at 50x for 5% growth, the stock would be absurdly expensive; at 14.7x for 3% growth, it is roughly where a mature, low-growth cyclical should trade. The "fallen-angel" narrative archetype is directionally right, but the valuation synthesis then contradicts it by calling the stock undervalued with a 57.7% upside. You cannot simultaneously say the market is pricing in "profound pessimism" and that the stock is 58% below fair value unless you believe the pessimism is irrational. The data does not support that leap. The 3% growth and 6-9% margins are not the market being irrational; they are the market being accurate about what the company is producing right now.
The strongest case against my own read is the balance sheet and the FCF. Zero debt, $658.2 million in cash, $2.48 billion in equity, and $795.7 million in free cash flow in 2025 (a 27.4% FCF margin on revenue) is a genuinely fortress-like position. At a $6.23 billion market cap, the FCF yield is 12.8%, which is extraordinary for a software company. If the 2026 margin compression is driven by one-time items — restructuring, a large customer transition, a temporary CTV inventory mix shift — then the 14.7x P/E is actually cheap on a normalized-earnings basis, and the "undervalued" call has legs. The 78.6% gross margin in 2025 suggests the underlying unit economics of the platform are still intact; the net margin compression could be an operating-expense story (sales force expansion, R&D for UID2/CTV) rather than a revenue-quality story. A smart opponent would point to the $992.7 million operating cash flow and argue that the P&L is lagging the cash reality. I weigh this heavily, but I do not let it override the fact that two consecutive quarters of 6-9% net margins in a company that was at 22% six months ago is a red flag that the market is not being irrational about.
What would change my mind in either direction. On the bull side: if Q3 2026 (reporting around late October) shows net margin recovering to 14-16% and revenue growth re-accelerating above 10% YoY, that would confirm the 2026 H1 compression was a transitional cost investment rather than structural erosion, and the 14.7x P/E would look genuinely cheap for a zero-debt, 27%-FCF-margin platform. On the bear side: if Q3 2026 shows revenue growth at or below 5% and margins still under 10%, the 14.7x P/E is not a discount — it is the correct price for a 3-5% grower with a 7-9% net margin, and the stock should trade closer to $10-11 on a 10-12x forward earnings basis. The UID2 adoption numbers and the CTV inventory share disclosures in that earnings call are the specific data points that will resolve this.
Big-3 Panel — where each AI stands
Advanced Analysis Forensic deep-dive · separate lenses
The Trade Desk is a financially robust business: $1.30B liquid cash with zero net debt, $795.7M in FCF, and operating cash flow running 4.1x net income. Revenue has compounded from $1.20B in 2021 to $2.90B in 2025 (roughly 25% CAGR), while operating margin has expanded from 10.4% to 20.3% and FCF from $324M to $796M. Gross margin remains elite at ~79-82%, consistent with a durable software/platform economic model. Earnings quality checks are clean: accruals -8.2% of assets, Beneish M of -2.8, and OCF/NI of 4.1x all argue reported profits are conservative if anything.
Verify before trusting this (6)
- Customer concentration and take-rate trends disclosed in 10-K
- CTV competitive positioning vs. Amazon DSP and walled gardens per management commentary
- Detail on the ~360bp gross margin erosion since 2022 (mix, infra costs, or pricing)
- Kokai platform adoption metrics and any impact on retention/spend per client
- Composition of buyback authorization and whether pace can continue to fully offset SBC
- Any related-party or dual-class governance considerations tied to founder CEO
The composite fair value of $20.94 (signal-adjusted $26.35) implies 45-82% upside, but the range is wide: EPV floor at $8.32 says even a no-growth scenario leaves the stock only modestly overpriced, while the anchored P/E of $37.55 is almost certainly running hot on peak multiples and should be discounted. The DCF at $18.94 is the most defensible anchor and still implies ~31% upside from $14.44. Quality-adjusting for the 17% SBC drag (real economic cost the earnings-quality lens flagged as high quality on cash conversion but which I still haircut), I land on a deserved value in the high-teens to low-$20s. That puts the gap at roughly 25-40% - a real margin of safety, not a screaming dislocation. What's priced in at $14.44: the bear case that Google/Amazon vertical integration structurally caps TTD's growth and that recent guidance misses signal permanent deceleration rather than a cyclical/execution stumble. That is a plausible worry but not yet a proven thesis - TTD is still growing, still net-cash, still generating real FCF. The market is treating a strong independent platform as if it were already disintermediated. That is where the mispricing lives.
Verify before trusting this (5)
- Q4 guidance and 2025 revenue growth trajectory - is deceleration cyclical or structural
- CTV and international mix trends - core moat evidence
- SBC as % of revenue trend - is it moderating with scale
- Take-rate stability vs Google/Amazon DSP competitive pressure
- Any commentary on UID2 adoption momentum post-cookie deprecation delays
The pressure on TTD right now is heavily negative and it is stock-specific, not just macro. The stock is down roughly 70% over twelve months, just announced a 15% workforce cut framed by management itself as a response to underperformance and slower growth, and Wall Street split publicly on the news (Rosenblatt sees cost relief, Jefferies flags top-line risk). That is the textbook signature of a fallen-angel narrative in the middle of breaking - moderate intensity, moderate durability, low cult - with no fan base to defend the tape. Rival AppLovin ripping higher on the same day amplifies the rotation story: capital is leaving buy-side DSPs for sell-side and ad-network winners. The macro backdrop is only mildly risk-on (regime score +27, VIX 14.5), so the tape is not doing TTD any favors either, but it is not the driver - the driver is a name-specific narrative crack. With beta ~1.04, TTD is not defensive; it takes full market chop plus its own idiosyncratic hit. The bear frame (Google/Amazon vertical integration, AI commoditizing ad buying, DSP margin compression) is now the consensus lens through which every headline is read, so even neutral-to-positive product news (LoopMe brand-lift integration, JBPs) gets brief bounces then fades. Repeated guidance disappointments have destroyed the benefit of the doubt, and layoffs without detail read as defensive, not strategic.
Verify before trusting this (5)
- Whether the Q3 print (post-layoff) shows stabilization or another guide-down - the single biggest sentiment inflection
- Analyst target revisions in the next 2 weeks post-restructuring; watch for a Jefferies-style downgrade cascade
- AppLovin vs TTD relative performance - if the rotation pair trade fades, sentiment pressure eases materially
- Any concrete UID2 adoption milestone that could revive the bull narrative
- Insider buying at these levels would be a genuine tell
Total ad spend is growing, but the plumbing is being re-cut. Media owners with first-party demand and commerce data (Amazon most acutely, Google structurally) are pulling buying inside their own stacks, which bypasses the independent DSP layer rather than competing with it on features. Simultaneously, AI is compressing the value of algorithmic optimization — the historical justification for a percentage-of-spend take rate — and supply-path optimization keeps squeezing intermediaries generally, visible as industry-wide margin compression and a -15.8% industry earnings CAGR. The open internet's share of a growing pie is the contested variable, and TTD's fate is levered almost entirely to it. Higher rates and a headwind macro backdrop make brand budgets choppier at exactly the moment the company needs volume to cover a heavier cost base.
When we made this prediction on Sep 5, 2026, TTD was $14.44. We expect it to be $18.40 by Mar 2027, and we consider it great value under $16.00. This is an early model (v0.6.0) — the direction is more reliable than the exact price. Made Sep 5, 2026.
Blue is our prediction, starting the day we made it. Grey is a slower route to the same place — the same destination, taking longer. Black is the actual price, so you can see how we are doing.