NFLX Technical Appendix: The record, the model and its sensitivities

The NFLX research family (page counts are those of the approved PDF editions, which this site does not host):

• Deep Dive: The margin story is real, and the price already pays for most of it (59 pages). Public preview; the Deep Dive page says how to read the full report.
• Research Note: The margin story is real, and the price already pays for most of it (6 pages).
• Technical Appendix: The record, the model and its sensitivities (32 pages).
• Learning Companion: Reading the Netflix research (8 pages).

---

Technical Appendix · Blockhaven Rating 3 of 5, BALANCED, at $62.34 on 11-02-2026 · An excerpt of the Deep Dive: the dated comparison since the valuation date, the eight-year record, the forecast mechanics and the three paths year by year, valuation, terminal economics and sensitivities, capital allocation and the balance sheet, unknowns and sources

Since the valuation date

The valuation in this report is dated 11-02-2026, when Netflix closed at $62.34, and nothing in it has been recalculated since. Two things in the market have moved. The stock closed at $63.95 on 11-09-2026, the settled close of that session, which is $1.61 or 2.6% above the valuation close and $0.54 or 0.9% above the $63.41 close of 11-06-2026, and the Treasury 10-year par yield fell from 4.91% to 4.86% by 11-06-2026, five basis points, having ranged between 4.84% and 4.95% on the days in between. We captured no yield after 11-06-2026. The 11-09-2026 close is the only new observation in this report: the filing index we retrieved on 11-07-2026 lists only insider filings, Forms 4 and 144, for the dates from 11-03-2026 to 11-05-2026, with no 10-K, 10-Q or 8-K among them, and we have not retrieved it again, so we say nothing about filings after that retrieval. The date of the fourth-quarter report is not confirmed in any document we captured, so we do not give one. Everywhere else in this report, "the price" means the $62.34 valuation close.

The intrinsic values are those of 11-02-2026. They have not been rolled forward to 11-09-2026, and the model's rate is still the 11-02-2026 build, so the only thing that differs in the table is the price each value is set against. Against the 11-09-2026 close the reference path is 6.8% above the price where it was 9.5% above at the valuation close and 7.7% above on 11-06-2026, the skeptic's path is 1.3% above where it was 3.9% and 2.2%, the downside is 24.5% below where it was 22.6% and 23.9%, and the upside is 32.9% above where it was 36.3% and 34.0%. Both earnings cross-checks, $47.33 and $42.80, are still below the price. None of the four values in the table, and neither cross-check, has crossed the close in either direction: the close sits 43% of the way from the downside value to the upside value, a position in an unweighted range and not a probability, with the reference $4.34 above it and the skeptic's path $0.84 above it.

The same reference path can also be read against each close by asking what the model needs in order to match it. The table solves the model's own reverse valuation twice, once at the valuation close and once at the latest settled close, changing one input at a time and holding the rest of the reference fixed, including the market weights and the share count of the valuation date. It is a statement about arithmetic and says nothing about what investors believe or which combination is likelier.

The higher close moves each of these break-evens toward the assumptions we hold. At the valuation close the model matched the price at a cost of equity of 10.49%; at $63.95 it matches at 10.31%, 0.45 points above the built 9.86% where the gap was 0.63 points, which is a beta of 1.20 where it was 1.24 at the same premium. Margins would have to be 2.0 points lower, not 2.7, or growth scaled by 0.83, not 0.76, so the latest close is matched by a smaller miss on either axis, and the reference's lead, though still positive, is that much thinner against the rate we judged.

The later sections read their stored values against the $62.34 valuation close, as they say, and four of those values lie between that close and the latest one. The upside at a beta of 1.30 and a premium of 5.5%, $62.56, the reference with the low growth path, $62.77, the reference with growth scaled by 0.8, $63.21, and the reference with margins one point lower and growth scaled by 0.9, $63.62, are each above $62.34 and below $63.95. The reference with margins two points lower, $63.96, is above the latest close by less than a cent. None of these values has been revised; they sit on the other side of a higher price.

The comparison does not change our reading. The stored sensitivities give the scale of the yield's move that we can see: the reference is $81.10 with the discount rate one point lower and $58.88 with it one point higher, so five basis points on the rate is worth about $0.56 a share. That is a scale and not a revaluation, a lower yield would add to the values rather than take from them, and it applies only to the move to 11-06-2026, since we captured no yield after that date. The reference path and the skeptic's path were above the price on the valuation date and are above the latest close, and the BALANCED reading rests on how small the reference's lead is against the two judged inputs to the discount rate, which a lead of 6.8% still is: at the 4.5% premium a beta of 1.20 would now remove it, where 1.24 would have at the valuation close. The rating stays as proposed at the 11-02-2026 valuation close, and the comparison with the 11-09-2026 close adds no argument for a different level.

Historical financials: eight years, FY2018 to FY2025

A forecast is only as good as the record it extends, so this section sets out eight fiscal years from the filings, with the accounting adjustments that change what a year means and a reading of what each stretch teaches. Every figure is read from Netflix's 10-Ks for FY2019, FY2020, FY2022, FY2023, FY2024 and FY2025 and their XBRL facts; per-share figures before the ten-for-one split of 11-14-2025 are shown on the post-split basis, which for FY2018 to FY2022 means the figure as filed divided by ten (shares multiplied by ten), so a cent is the limit of their precision. Where a statement shows a dash we say that no amount was reported and do not read it as zero.

The income statement

Revenue compounded at 16.2% a year over the seven years, but the path was not smooth: growth of 27.6%, 24.0% and 18.8% through FY2021, a stall to 6.5% and 6.7% in FY2022 and FY2023, and a return to 15.6% and 15.9% in FY2024 and FY2025. Operating income compounded far faster, at 35.3% a year, and the reason is visible in the cost lines. Cost of revenues, which is mostly content amortization, fell from 63.1% of revenue to 51.5%, 11.6 points; marketing fell from 15.0% to 7.3%, 7.7 points; technology and development fell by 0.2 points and general and administrative rose by 0.2, so the 19.3-point gain in operating margin, from 10.2% to 29.5%, is almost entirely content and marketing. Marketing did most of its work early, falling from 15.0% to 8.9% by FY2020 and by only 1.6 points more since; content amortization has done its work later, and it is the line that the forecast leans on. Operating margin fell in one of the seven year-to-year steps, in FY2022, from 20.9% to 17.8%, and the change in operating income per added revenue dollar that year was -29.3%, which is what a content bill that has been committed ahead of a revenue slowdown looks like.

Net income rose from $1.21 billion to $11.0 billion, a little more than nine-fold. Earnings per share rose slightly faster, from $0.27 to $2.53 on the post-split basis, because the diluted share count fell 3.7%, from 4,512 million to 4,344 million. The book tax rate was 1.2% in FY2018, an unusually low charge that we have not attributed, 9.5% in FY2019, and between 12.4% and 14.7% from FY2020 to FY2025. Interest expense rose from $420 million in FY2018 and $626 million in FY2019 and has been flat at $0.7 billion to $0.8 billion since FY2020, and operating income covered it 3.8 times in FY2018 and 17.2 times in FY2025.

Where the revenue comes from

The regional record is the clearest account of where the growth came from and where it will have to come from. The United States and Canada were 53.7% of streaming revenue in FY2018 and 44.2% in FY2025; Europe, the Middle East and Africa rose from 25.7% to 32.1% and Asia-Pacific from 6.1% to 11.8%, while Latin America slipped from 14.5% to 11.9%. The slowdown of FY2022 and FY2023 was broad: no region grew faster than 9.3% in either year except Latin America's 13.8% in FY2022, and EMEA grew 0.5% in FY2022, a year in which the 10-K's constant-currency figure for revenue per membership there (up 6%) against the reported figure (down 5.5%) says that currency took roughly 11 to 12 points off it. The recovery reached every region by FY2025, though Latin America was still slower in FY2024, at 8.8% against 9.3% in FY2023, and in FY2025 Asia-Pacific grew fastest, at 21.3%, followed by EMEA at 17.2%, the United States and Canada at 15.0% and Latin America at 10.7%. The regional mix matters for the forecast because revenue per membership differs so much by region, as section 6 shows.

Cash flow and the content cycle

The cash flow statements show how far cash and earnings have diverged over the eight years and why. A dash in the repurchase row marks a year in which the filing shows no repurchase figure, which we read as unreported and not as zero.

Cash tells a different story from earnings in two ways, and both come from content. In FY2018 and FY2019 Netflix paid out 1.60 and 1.59 times its content amortization in cash and generated negative operating cash flow, $2.68 billion and $2.89 billion, financing the difference with new debt of $3.96 billion and $4.47 billion. Since FY2020 the ratio has fallen to between 0.93 and 1.43 times, free cash flow has been positive in every year but FY2021 (a deficit of $132 million, the year content cash ran at 1.43 times amortization), and in FY2025 it reached $9.46 billion, 20.9% of revenue. The FY2023 ratio of 0.93 times was a strike-year low, which is why free cash flow in that year was 1.28 times net income against 0.79 and 0.86 times in FY2024 and FY2025. The second way is tax: cash taxes were a larger share of pre-tax income than the book rate in FY2022 and each of the three years after it (15.4% against 14.7%, then 18.6%, 16.5% and 17.5% against 12.9%, 12.6% and 13.7%), so profit is converting to cash more slowly than the income statement suggests. Stock-based compensation, which the model treats as a cost, fell from 2.0% of revenue in FY2018 to 0.8% in FY2025.

The balance sheet, the share count and the return on capital

Net debt, which is total debt less cash and equivalents, was $9.74 billion at its highest year-end in this record, FY2019, and $5.43 billion at its lowest, FY2025. The content asset on the balance sheet grew from $20.1 billion to $32.8 billion while the amortization that runs through the income statement more than doubled, so the library now equals 2.00 years of the year's amortization against 2.67 in FY2018, which says the asset turns over faster than it did. Return on opening content assets and property, measured on the model's 16% tax rate, rose from 10.7% in FY2019 to 32.9% in FY2025; it is a return on the capital a content business must carry on its balance sheet and not a return on all capital, and we use it in section 9 as an independent check on the terminal value. The share count at year-end peaked at 4,453.5 million in FY2022 and was 4,222.2 million in FY2025, 5.2% lower.

FY2025 and the first nine months of 2026, in detail

Margin. The margin record needs two readings, because FY2025 carries a $619 million expense for Brazilian non-income-tax assessments booked in Q3 2025. Reported, margin rose 6.1 points in FY2024, 2.8 points in FY2025 and a guided 2.0 points in 2026. Before the Brazil expense, which is our adjustment and not a company figure, FY2025 margin was 30.9% and the steps are 4.2 points in FY2025 and only 0.6 points in the 2026 guide, or 1.0 point if the roughly 30 basis points of headcount-reduction cost in the 2026 guide are also set aside. The third quarter shows the same pace: reported margin of 34.6% against 28.2% a year earlier, but 33.6% a year earlier before the Brazil expense, a gain of 1.0 point. Over nine months the like-for-like gain is 0.3 points. The honest reading is that the large gains came early and that the recent like-for-like gains run from 0.3 points (the first nine months of 2026) through 0.6 points (the 2026 guide, or 1.0 if the headcount cost is also set aside) to 1.0 point (the third quarter), after 4.2 points in FY2025. Our reference path adds 1.0 point in each of 2027, 2028 and 2029 and half a point in each of 2030 and 2031, so for its first three years it sits at the top of that recent range and not at its center.

Earnings. In the first nine months of 2026 Netflix earned $12.3 billion, or $2.90 a diluted share, on $12.6 billion of operating income. Interest and other income includes the $2.8 billion termination fee that Paramount Skydance paid to Netflix on behalf of Warner Bros. Discovery when that company ended its merger agreement with Netflix and entered one with Paramount Skydance. The effective tax rate for the nine months was 17.3% against 13.7% in FY2025. Whether the fee raised it depends on how the fee was taxed, which the filings do not say: a fee taxed at a rate above the rest of the year's income would lift the nine-month rate, and one taxed below it would lower it. Taking the fee out at an assumed 21% rate, which is our assumption, FY2026 net income of $15.1 billion becomes about $12.9 billion, or $3.04 a share on the 4,251.3 million diluted shares weighted over the first nine months, and the $62.34 valuation close is 20.5 times that.

Cash. Free cash flow, which Netflix defines as operating cash flow less purchases of property and equipment, was $9.5 billion in FY2025 and $9.5 billion in the first nine months of 2026, split $5.1 billion, $1.5 billion and $2.9 billion across the three quarters. The first quarter carries the fee. Without the gross fee, and with $729 million of non-routine Brazil payments added back, nine-month free cash flow is $7.4 billion, and that is before any tax on the fee. Management now expects about $12.0 billion for the full year, down from $12.5 billion because of a real estate transaction expected to close in the fourth quarter; its size is not disclosed.

Content cash. The cash Netflix spends on content, which we construct from the cash flow statement as additions to content assets less the change in content liabilities, was $17.7 billion in FY2025 against $16.4 billion of amortization, a ratio of 1.08x. In the first nine months of 2026 it was $15.0 billion against $12.8 billion, a ratio of 1.17x, while management says it expects about 1.1x for the year. The ratio matters because cash spent ahead of amortization is real cash that the margin does not show. Content obligations stood at $25.2 billion on 09-30-2026, of which $19.6 billion is not on the balance sheet.

Adjustments that change what a year means

Seven items stand between the reported statements and the run rate, and each has a cash consequence that is not the same as its income-statement consequence.

Content is the first. The income statement carries amortization, and the cash flow statement carries additions to content assets less the change in content liabilities. We construct "cash paid for content" from those two lines because the company does not publish a cash content spend figure in the filings we hold. The gap between the two is the reinvestment that the model charges against free cash flow, and it is the largest single judgment in the forecast.

The stock split is the second. The ten-for-one split of 11-14-2025 was restated in the FY2025 10-K for FY2023 to FY2025; earlier years are divided here by the stated ratio of ten, and a check that earnings per share times shares equals net income holds within the rounding of a cent for each year.

Foreign exchange on euro debt is the third. Netflix's euro-denominated notes are remeasured each quarter, with the loss or gain in interest and other income and added back in the cash flow statement as a non-cash item. It was a loss of $533 million in FY2020 and a gain of $431 million in FY2021 and $353 million in FY2022, so it moved pre-tax income by as much as half a billion dollars without touching operating income. The model starts from operating income and so is unaffected, but net income and earnings per share in those years are not clean, and the interest line should not be read as the cost of the debt.

Brazil is the fourth. A $619 million expense for non-income-tax assessments sits in cost of revenues in the third quarter of 2025, which lowers FY2025 operating margin by 1.4 points to 29.5% from 30.9%; the cash side is $729 million of non-routine payments in the first nine months of 2026, inside operating cash flow, and the January letter says a deposit of about $700 million in the same dispute shifted from 2025 into 2026, which is part of why FY2025 free cash flow beat the company's own forecast.

The termination fee is the fifth: $2.8 billion in interest and other income in the first quarter of 2026, before any tax, with the tax unknown. Taxes are the sixth: the book rate has run at 12.4% to 14.7% since FY2020 while the cash rate ran between 8.7% and 19.4% over the eight years and has stayed between 15.4% and 18.6% since FY2022, which is why the model's 16% sits where it does. The seventh is the end of the DVD business, which Netflix says it discontinued in 2023: it was $366 million of revenue in FY2018, $83 million in FY2023, and nothing reported in FY2024, which makes revenue comparable on a streaming basis from FY2024 on and means the early regional percentages are shares of streaming revenue and not of total revenue.

What each stretch teaches

FY2018 and FY2019 were an investment phase financed by debt. Content cash ran at 1.60 and 1.59 times amortization, operating cash flow was negative, total debt rose from $10.4 billion to $14.8 billion in FY2019, and operating margin was 10.2% and 12.9%. The lesson for the forecast is that the content ratio can swing widely, from 1.60 times in FY2018 to 0.93 times in FY2023 in this record, and a reference path that holds it at 1.10 times is a statement that the company has stopped growing its slate faster than its revenue.

FY2020 and FY2021 were the pandemic years. Revenue grew 24.0% and 18.8% and margin reached 18.3% and 20.9%, but content additions jumped from $11.8 billion in FY2020 to $17.7 billion in FY2021, content cash reached 1.43 times amortization and free cash flow fell back to a deficit, and FY2021 carries the first repurchase ($600 million) and the first acquisitions ($788 million) that the statements in this record report; the FY2018 to FY2020 statements carry no such lines, which we read as unreported and not as zero. The lesson is that a one-year surge in spending preceded the stall that followed.

FY2022 was the test of the fixed-cost model. Growth fell to 6.5%, net membership additions were 8.9 million, the lowest of the seven years for which they are reported, and the United States and Canada lost 0.9 million memberships; amortization rose 14.7% in a year when revenue grew 6.5%, margin fell 3.1 points and the cash flow statement reports no repurchases. The downside scenario is this year extended: a business whose content cost keeps rising while revenue growth fades and margin never recovers.

FY2023 showed how much of the cost base can be flexed. Revenue grew 6.7% but amortization grew 1.2%, and margin rose 2.8 points to 20.6%, and content cash dropped to 0.93 times amortization in a strike year, so free cash flow of $6.93 billion was flattered by a one-time slowdown in spending. Repurchases resumed at $6.05 billion.

FY2024 and FY2025 were the monetization years: revenue grew 15.6% and 15.9%, margin rose 6.1 points and then 2.8 points (4.2 on a like-for-like basis), added operating income per added revenue dollar was 65.6% and 47.1%, and repurchases were $6.3 billion and $9.1 billion, 90% and 96% of free cash flow. Content amortization grew 7.8% and 7.3% in those years against revenue growth of about 16%, and that gap of roughly eight points is the thing the reference path has to keep alive: if amortization grows nearly as fast as revenue again, margin stops rising.

The first nine months of 2026 add one more observation. Revenue grew 13.7% and operating income 21.6%, but content amortization grew 9.5% and fell 1.3 points as a share of revenue (to 33.9% from 35.2%) while cost of revenues, which also carried the $619 million Brazil expense in the comparison period, fell 2.8 points to 47.8% (0.9 points before Brazil), and technology and development rose 0.4 points and general and administrative 0.2 points. So like-for-like margin rose only 0.3 points in nine months: the content leverage was real, and most of it was spent on other lines. Content cash ran at 1.17 times amortization in the nine months against 1.08 times a year earlier, which is why we treat the 1.1 times that management guides for the year as a test.

Forecast mechanics: the drivers, the links and the three paths

The model has five operating drivers (revenue growth, operating margin, the ratio of content cash to amortization, capital intensity and working capital), one tax rate, one discount rate and a set of terminal assumptions. This section states each driver for all three scenarios, sets the evidence for its level and path beside it, shows how the drivers link into cash, and prints the year-by-year schedules of the three paths, so that a reader can rebuild any scenario value from this page.

The drivers and the evidence for each

Revenue growth. The evidence for the starting point is the quarterly path the company reports.

Reported growth fell from 17.6% in the fourth quarter of 2025 to 11.8% in the third quarter of 2026, and management forecasts 11.5% for the fourth. The reference starts 2027 at 11.0%, which is the company's own constant-currency rate for the fourth quarter (11%), and steps down by one point a year to 7.0% in 2031. The regional detail shows why a step-down is the central case and not a collapse.

In the third quarter of 2026 the United States and Canada grew 11%, EMEA 11% (9% at constant currency), Latin America 17% (12% at constant currency) and Asia-Pacific 12% (15% at constant currency); every region has decelerated from its fourth-quarter 2025 rate at constant currency, by 7 points in the United States and Canada (18% to 11%), 6 in EMEA, 8 in Latin America and 4 in Asia-Pacific. The same growth rates read differently in dollars of revenue added per year.

Netflix added $5.3 billion of revenue in FY2024, $6.2 billion in FY2025 and, on the company's guide, $5.9 billion in FY2026E. The reference adds between $5.1 billion and $5.7 billion in each of the five years, which is to say that its percentage deceleration is mostly arithmetic: it asks Netflix to go on adding about what it has been adding, on a larger base, with the dollar add shrinking from 2029 and most in the last two years. The downside adds $4.1 billion in 2027 and $2.2 billion in 2031, close to the $1.9 billion and $2.1 billion of FY2022 and FY2023 by the end, which is why we describe it as an enduring impairment. The upside adds $6.4 billion to $6.7 billion in every year, more in each year than the largest annual add from FY2019 to FY2025, FY2025's $6.2 billion.

The guidance itself has been stable where it matters for this driver and revised where it does not.

The 2026 revenue guide has stayed within 0.2% of its January midpoint, $51.1 billion against $51.2 billion, and the margin guide has been 31.5% in all four letters. The revisions were to free cash flow (raised in April because of the termination fee and trimmed in October, for a real estate transaction, to about $12.0 billion, which is 9.1% above January's figure) and to advertising, which fell 10% from the April figure. A year of guidance is not a record of forecasting accuracy, and nothing here should be read as one, but it is evidence that management's revenue and margin forecasts for the year in progress are not being revised each quarter.

Operating margin. Management's stated intent, in the January letter, is to grow operating margin each year, with the size of the gain varying as it balances reinvestment against profit. That is an intention, not a result, and the evidence we weigh against it is in section 5: like-for-like margin gains of 6.1, 4.2 and 0.6 to 1.0 points in FY2024, FY2025 and the 2026 guide, and 0.3 points in the first nine months of 2026. The reference adds a point in 2027 and a point in each of the next two years, then half a point, to 35.5%; the downside holds the 2026 guide of 31.5% flat; the upside rises to 38.5%. The incremental-margin test in section 6 puts the reference at 41.2% to 45.6% of each added revenue dollar, below the last two years and above the latest nine months like for like.

Content cash against amortization. Management guides about 1.1 times for 2026, FY2025 was 1.08, FY2024 1.11, and the first nine months of 2026 ran at 1.17. The reference holds 1.10 in every explicit year; the downside holds 1.17 for two years and then 1.12 and 1.10; the upside converges on 1.05. Amortization is held at 33.5% of revenue in all three. The model's share enters only through the content cash that exceeds amortization, so the choice is not a margin assumption: at the 35.3% that management's guided 10% growth in amortization implies for FY2026E, the reference would be worth $0.27 less. The terminal ratio is not a driver in this sense: it is set by growth and a lag, as section 9 explains.

Capital intensity. Capital spending of 1.6% of revenue and depreciation of 0.9% give net capital spending of 0.7% of revenue, against 1.5% and 0.7% in FY2025 and 1.8% and 0.8% in the first nine months of 2026. The assumption is close to the record and small in the cash flow, at $0.4 billion to $0.6 billion a year in the reference.

Working capital. The model charges 2% of each year's revenue increase. The five working-capital lines of the cash flow statement absorbed 7.4% of the revenue increase in FY2025, 0.6% in FY2024 and 5.5% in FY2023, 4.5% pooled over the three years, and 39.5% in FY2022 on a revenue increase of $1.9 billion. The 2% is therefore below three of the last four years, though above FY2024's 0.6%. It matters little: at the pooled 4.5% the reference is worth $0.27 less, because working capital is about $0.1 billion a year against free cash flow to the firm of $13 billion to $20 billion. We keep the accepted 2% and report the sensitivity.

How the drivers link into cash

Growth enters the cash flow in three ways, and the schedules show how little of it comes through the most obvious route. Working capital is charged on the increase in revenue, which is small. Capital spending and depreciation scale with revenue, not with growth. The large link is content: the model sets content cash above amortization at (ratio minus one) times 33.5% of revenue, so each point of revenue growth adds amortization and, through the ratio, content cash.

In 2031 the reference reinvests 14.0% of after-tax operating profit, of which content cash above amortization is $2.63 billion of $3.29 billion, about four-fifths. Free cash flow to the firm is therefore 86.0% of after-tax operating profit and 25.6% of revenue in 2031, against 84.4% and 22.3% in the downside and 92.2% and 29.8% in the upside, where the content ratio has converged to 1.05 and reinvestment is a little over half the reference's. This is the reason the content ratio, and not growth, is the dial that moves the share of profit that becomes cash, and why the ratio's behavior in the terminal year (section 9) carries so much of the value. In value per share the explicit-year ratio is the smaller dial: moved to the downside's and upside's paths after their growth and margin (section 8), the content ratio is worth $0.66 and $0.94 a share, against $9.99 and $4.44 for the growth path. As a check against the record, FY2025's reported free cash flow, put on the firm basis the model uses, was 22.4% of revenue, and the reference's 23.1% in 2027 is a step up from it that is consistent with the margin path.

The three paths, year by year

The schedules below show every line from growth to present value. Revenue is the prior year's revenue times one plus growth; operating income is revenue times margin; after-tax operating profit applies the 16% rate; content cash above amortization is (ratio minus one) times amortization; net capital spending is revenue times 0.7%; working capital is 2% of the revenue increase; and the present value applies the 9.57% rate with the mid-year convention from 11-02-2026. All amounts are $ million. Years run across each schedule, and the last column is the first terminal year, which shows the figures the terminal value is built from (growth, margin, the content ratio, after-tax operating profit and free cash flow) and a dash in the other lines; the terminal value itself is in section 9.

Downside: an enduring impairment, not a trough. Growth fades to 3.5%, margin never rises above this year's guide, and content cash stays at 1.17 times amortization for two years before returning to the 1.10 the guide implies. Nothing recovers later. The path assumes nothing the record contradicts, since FY2022 and FY2023 show that growth of about 6.5% has happened to Netflix before, though on a much smaller base, and it assumes that the growth of the last two years does not recur.

Reference: built from the evidence, not a most-likely forecast. Growth steps down by a point a year from the guided fourth-quarter run rate, margin gains a point a year to 2029, at the top of the recent like-for-like range, and then half a point, and content cash follows the guided 1.10 times. What it assumes beyond the record is that the 2026 content ratio of 1.17 times normalizes, that margin gains of a point a year survive the loss of the 2025 content leverage, and that the dollar add to revenue holds between $5.1 billion and $5.7 billion a year.

Upside: advertising scales and margin reaches 38.5%. Growth fades from 12.5% to 8.5%, margin rises to 38.5% and the content ratio converges to 1.05. What it assumes beyond the record is a dollar add to revenue in every year above that of any year from FY2019 to FY2025, a margin 9.0 points above FY2025's reported 29.5%, and a content ratio of 1.05 by 2031, lower than every year since FY2020 except the strike-year 0.93 times of FY2023.

Valuation and scenarios

Method. We discount free cash flow to the firm: operating income after a 16% tax rate, less the excess of content cash over amortization, less net capital spending (capex of 1.6% of revenue against 0.9% depreciation), less 2% of each year's revenue increase for working capital. Stock-based compensation stays in operating income as a cost, so we do not also dilute the share count for it. Content amortization is held at 33.5% of revenue and the margin path carries any further leverage. Valuation date is 11-02-2026, the fourth quarter of 2026 is discounted from year-end, the five explicit years are discounted mid-year, and the terminal value is discounted on the same mid-year convention.

Discount rate. One rate of 9.57% is used in all three scenarios, because scenario risk belongs in the cash flows and not stacked on the rate. It blends a 9.86% cost of equity (94.4% of capital at market) with a 4.61% after-tax cost of debt (5.6%). The cost of equity is built: the Treasury's 10-year par yield for 11-02-2026, 4.91%, plus a beta of 1.10 times an equity risk premium of 4.5%. Beta and the premium are our assumptions, and we print their band, beta from 0.90 to 1.30 and the premium from 4.0% to 5.5%. We did not measure beta. We have not collected a market-benchmark series, and the other companies we cover are not a market proxy; measuring beta would need a split-adjusted daily benchmark series, taken from the same source of settled market data as the price, and a stated window and method. The cost of debt is the Treasury yield plus the spread at which Netflix issued: its 5.25% ten-year notes of 07-22-2026 priced when the 10-year Treasury yielded 4.67%, a spread of 0.58 points, which we assume still holds, so the pre-tax cost of debt is 5.49%. The coupon is the cost of Netflix's own debt and is never used as a risk-free rate.

The previous edition built the cost of equity as the 5.25% coupon plus 4.25 points, 9.50%, with a discount rate of 9.22%, because no dated Treasury yield had been captured. The table shows both, and what each combination of beta and premium does to the same cash flows.

Moving to the built central rate lowers the downside by $2.63, the reference by $4.03, the upside by $5.41 and the skeptic path by $3.80, with the paths unchanged. The break-even cost of equity at the price is 10.49%, 0.63 points above the built 9.86%, and the reference equals the price at a firm rate of 10.2%.

The reference path is also printed here in the compact form that the chart in this report is drawn from; section 7 gives every line of all three paths.

How each scenario is valued

The three scenarios are one calculation run on three sets of cash flows. Each adds the present value of the fourth-quarter 2026 flow, the five explicit years and the terminal value, subtracts net debt, and divides by the share count.

The fourth-quarter 2026 flow, $2.6 billion in each, is the same in all three because it comes from management's full-year free cash flow guide and not from the path. The explicit years add $50.7 billion, $64.8 billion and $76.0 billion, and the terminal value $153.8 billion, $222.7 billion and $280.7 billion, which is 74%, 77% and 78% of enterprise value. Net debt of $6.9 billion is the carrying value of the senior notes, $15.1 billion, less cash, restricted cash and short-term investments of $8.23 billion at 09-30-2026. Operating leases ($2.3 billion) and content obligations ($25.2 billion, $19.6 billion of it off the balance sheet) are not deducted, because the cash cost of both is already inside operating income and the content cash line, and deducting them would count it twice. The share count is the 4,106.3 million outstanding at 09-30-2026 plus the 40.7 million that the 120.2 million options at a $41.24 average strike add at the $62.34 price by the treasury method. Every cash flow after 09-30-2026, the fourth quarter's included, is attributed to the claims and shares as they stood on that date, which is why repurchases after it are not counted. The reference bridge in billions of dollars reads as follows.

Why the reference is the best-supported path and not a midpoint

The reference is not the average of the other two paths. Their arithmetic midpoint is $66.62, the reference is $1.67 above it, and we attach no weights to any of the three. It is the path in which each driver sits where the evidence puts it at the valuation date: growth starts at the company's own fourth-quarter run rate and steps down at a pace that adds about the same dollars of revenue each year, margin rises a point a year to 2029, at the top of the recent like-for-like range of 0.3 to 1.0 points rather than its center, and then slows, and content cash follows the guided 1.1 times. The paths are not symmetric around it, and the next table says where the asymmetry comes from. It moves the reference to each path one driver at a time in a fixed sequence, growth first, then margin, the content ratio and terminal growth, so each row is measured after the rows above it and the rows add up exactly to the gap; only the growth row is the same as moving that driver alone.

The downside is $20.03 below the reference and the upside $16.68 above it. Taken in that sequence, in the downside the growth path accounts for $9.99 of the gap and the margin path for $6.77; in the upside the margin path accounts for $6.36, terminal growth for $4.94 and the growth path for $4.44. The later rows depend on the order: terminal growth, taken last, is worth $2.61 and $4.94 here, against $3.60 and $4.19 when it is moved by itself (section 10). The downside sits farther from the reference because its growth path falls further below the reference's (to 3.5% in 2031 against 7.0%, a gap of 3.5 points) than the upside's rises above it (to 8.5%, a gap of 1.5 points), and because its margin never rises, ending 4.0 points below the reference's 35.5% where the upside ends 3.0 points above it. We built it that way from the record: the downside uses what has already happened to Netflix (growth of about 6.5% in FY2022 and FY2023, and a year in FY2022 in which margin fell), while the upside needs a dollar add to revenue that no year from FY2019 to FY2025 reached, five years running. That is a judgment about which departures from the reference are better evidenced, and it is the reason the paths are not weighted: a reader who disagrees can move the weights as they like, and the paths and the rate table in this section show what each choice is worth.

What changed from our first build, and from the previous edition. An independent recomputation of the first build reproduced its arithmetic and found two method inconsistencies, both corrected in the previous edition, which moved the reference from the $65.98 first calculated to $72.32. The first carried the 2031 content ratio of 1.10x into a terminal year with far lower growth, which overstated reinvestment and is worth $3.86 on the reference. The second discounted the terminal value as if its cash arrived at year-end while the explicit years used mid-year, which is worth $2.48. The same two corrections moved the downside from $45.83 to $50.89 and the upside from $87.18 to $90.38. They are corrections of method, not new views, and they do not change the paths.

This edition makes one further change, of a different kind. The dated Treasury yields, captured for this edition, are new evidence, and a built rate replaces the coupon-plus-premium proxy, which is a new method. Together they lower the reference by $4.03, the downside by $2.63 and the upside by $5.41, to $68.29, $48.26 and $84.97. It is not a correction of our scenario judgments, and no path, assumption about the business or share count changed.

Terminal economics

Most of the value is in the years after 2031, so the terminal year is where the model should be read most carefully. In the reference the terminal value is $341.0 billion, valued as if the first terminal year's cash arrived mid-year like every explicit year, so that it is discounted from the middle of 2031; it is worth $222.7 billion on 11-02-2026, which is 77% of enterprise value, and the undiscounted amount is 1.53 times its present value. The terminal year is built with the same formulas as 2031 at the terminal growth rate, so the question is what changes between the last explicit year and the first terminal year, and whether each change is earned.

Terminal economics, stated openly. In the first terminal year content cash runs ahead of amortization by the lead time between spending and amortizing, so the ratio is (1 + terminal growth) raised to that lead. The captured history implies a lead of about 1.4 years in FY2024 and 1.1 years in FY2025; we use the longer, which puts the reference's terminal ratio at 1.04x, below the 1.10x it carries in the explicit years because growth has fallen from 7% to 3%. The two are consistent, since 1.10x at 7% growth is itself a 1.4-year lead. The reference terminal year reinvests 7.3% of its after-tax operating profit, which at 3.0% growth implies a return on new capital of 41%. That is a high figure, and a diagnostic and not a plug: it says the model's value depends on Netflix needing little new capital to keep growing, and we test it below.

What changes at the step, and what it is worth

Three things change at the step into the terminal year. Growth falls, from 7.0% to 3.0% in the reference (3.5% to 2.5% in the downside and 8.5% to 3.5% in the upside), and that is the definition of a terminal year. Margin is held, so the step introduces no new margin gain. And reinvestment falls from 14.0% of after-tax operating profit to 7.3%, because growth is less than half as fast and the content ratio drops from 1.10 times to 1.042 times, which raises free cash flow relative to after-tax operating profit from 86.0% to 92.7%. The terminal year's free cash flow is therefore 11.1% above 2031's on 3.0% growth in profit, and that fall in reinvestment is worth $3.76 a share: it is the difference between the reference and the value with the terminal content ratio held at the 1.10 times of the explicit years, $64.53. The fall is a consequence of the content cycle and not an extra assumption, since the captured history implies a lead of cash over amortization of 1.1 to 1.4 years, and at 3.0% growth that lead gives a ratio of 1.04 times, but it is also the single step in the model that turns a cash flow that is 86.0% of profit into one that is 92.7%, and we have tested it at lead times of 1.0 and 2.0 years in the sensitivity table (the reference is $69.09 at one year and $67.08 at two).

Three checks that do not use the model's own terminal arithmetic

The first check is the return on capital. The terminal year's reinvestment implies that each new dollar of capital earns 41% (terminal growth divided by the reinvestment rate), against 50% implied by 2031, 34% in the downside and 47% in the upside. The measure of return we can build from the filings we hold is the after-tax operating profit at the model's 16% rate over the content assets and property that Netflix carries at the start of the year, which was 32.9% in FY2025 after rising from 10.7% in FY2019. The reference's 41% is 8.3 points above that. The two are not the same measure (the history excludes working capital and other assets, and FY2025 has the highest margin of the eight years), but the direction is clear: the terminal year assumes new capital earns more than the existing capital did in the best of the years we measured, FY2019 to FY2025. Setting the terminal return on new capital at the FY2025 level would lower the reference by $1.06, and at the 30% used in the earnings cross-checks by $1.57, so the reference remains above the price on either; we regard that as the size of this assumption and not a reason to change it.

The second check is the multiple. The terminal value is 12.2 times 2031 operating income and 14.1 times the first terminal year's after-tax operating profit, against 16.3 times FY2026E operating income at the price on 11-02-2026 (enterprise value of $262.9 billion, with equity valued on the 4,106.3 million shares outstanding at 09-30-2026; on the 4,147.0 million diluted shares that the per-share values use, the market multiple would be slightly higher). The model does not rely on today's multiple holding: it assumes the multiple compresses by about a quarter on the outstanding-share basis, and a little more on the diluted one, as growth decays from 13% to 3%, which is the pattern of a business reaching maturity and is not what a high terminal multiple would signal. The third check is the earnings cross-checks of section 10, which capitalize a single year of normalized earnings at 3.0% growth and a 30% return on new capital and give $47.33 and $42.80. They are lower than the reference because they omit its explicit margin expansion after the year each capitalizes, so the gap between them and the reference is a measure of how much of the value rests on the margin path, which is the point of section 10.

What the terminal share assumes

A terminal value that is 77% of enterprise value means that the first five explicit years and the fourth quarter together are worth 23%, $67.4 billion of $290.1 billion, and that the reference value is chiefly a view on the economics of Netflix from 2032 onward: growth of 3.0%, a margin of 35.5% that has been held, and a content cycle that has settled at 1.04 times. We trust those more than any single forecast year, because they ask only that the business stops improving and keeps its economics, but that is a reading of the economics and not evidence, and the sensitivities say how much each is worth: terminal growth of 2.5% or 3.5% moves the reference to $64.69 or $72.48, a terminal content ratio at the first build's 1.10 times lowers it to $64.53, and discounting the terminal value from the end of 2031 instead of the middle lowers it to $65.89. The terminal values of the downside and upside are 74% and 78% of their enterprise values, so the concentration is a property of the method and not of the choice of path.

Reverse valuation and sensitivities

The price of $62.34 is consistent with many sets of assumptions, and the useful question is which sets, so this section turns the model around: it says what combinations of the driving assumptions match the price, how far each assumption can move before the reference reaches it, what an independent challenge made of each assumption, and what a cross-check built a different way says. Everything here is conditional. A break-even is a statement about arithmetic, not about what the price requires or what investors believe. The price in this section is the $62.34 valuation close; section 2 gives the same break-evens at the 11-09-2026 settled close and names the stored values that sit on the other side of it.

What the price fits

Reverse valuation. Holding the rest of the reference fixed, the model matches $62.34 when every margin from 2027 to 2031 is 2.7 points lower (an exit margin of 32.8%), when growth is scaled by 0.76 (8.4% in 2027 falling to 5.4% in 2031), or at a cost of equity of 10.49%, which a beta of 1.24 gives at a 4.5% premium and a premium of 5.08% gives at a beta of 1.10. At the previous rate the same break-evens were 4.4 points, 0.62 and 10.5%. Read the other way, a buyer at the price earns about 10.2% a year on the whole enterprise, against the 9.57% discount rate, and about 10.49% on the equity, against the 9.86% built cost of equity, provided the reference path happens and the terminal value is realized. These are conditional break-evens, not statements of what the price requires or what investors believe.

The single break-evens are one dimension each. The price is also matched by combinations, and the table sets out the margin and growth combinations that reach it.

With the reference's margins, growth has to be 0.76 of the reference's to match the price, 8.4% in 2027 falling to 5.4% in 2031. With margins one point lower in every year the price needs growth at 0.85 of the reference; with margins two points lower, 0.93; with margins three points lower, 1.02, which is to say the reference's growth. Each point of margin across 2027 to 2031 is worth $2.17 a share, so margins 2.7 points lower and the reference's growth are the single-line break-even, and a path that takes margin one point lower and growth a fifth lower is $61.20, below the price. In the other direction, a path with growth a fifth below the reference's but the reference's margins is $63.21, still above the price. The market-implied view is therefore not one number but a frontier: a margin shortfall can be offset by faster growth and the reverse, and the reference sits above that frontier, 9.5% above the price.

Cost of equity is the other axis, and it interacts with margin.

At the built cost of equity the reference with margins two points lower is $63.96, above the price; at a cost of equity 0.75 points higher it is $57.45, below it. Margin and the rate are the two inputs that together decide the sign of the reference's lead, and the rate is a judgment (beta and premium) that we have not measured.

Sensitivities from the reference

Sensitivities. The reference moves most with the rate. The table prints each sensitivity at the built central rate beside the previous edition's, so the effect of the rate and the effect of the assumption can be read separately. Each row changes only what it names: most change one assumption of the reference, the row for both slower paths changes two together, and the upside and downside rows start from those paths and not from the reference.

If content cash stayed at 1.17x in every year, including the terminal year, the reference falls to $58.48, below the price, which is the clearest single statement of how much the price depends on content spending normalizing toward the guided 1.1x. The terminal value share is 77% in the reference, 74% in the downside and 78% in the upside (75%, 78% and 79% at the previous rate), and the implied return on new capital in the terminal year, a diagnostic and not a plug, is 41%, 34% and 47%.

What the independent challenge found

An independent challenge of the first build tested eight of its assumptions against a low and a high value each and said, for each, whether it agreed. The table repeats the challenge's own base, low and high values and the change in the reference it measured, which was at the earlier rate of 9.50% and 9.22% before the discount rate was rebuilt from the Treasury yield, and then re-runs the same low and high settings through the model at the built central rate.

The challenge's conclusions on each, and what we did, are these.

Cost of equity. Base 9.50%, low 8.75% and high 10.25%; the challenge agreed with the assumption but found that the reference equalled the price at a cost of equity only 0.37 points above it, so that the sign of the reference's lead was not robust to the rate. At the built central rate the same plus or minus 0.75 points gives $76.91 and $61.36, changes of $8.61 and $6.93, and the break-even is now 0.63 points above the central. The revision that built the rate from a dated Treasury yield answered the challenge's point, and the rate remains the judgment that moves the reference most.

Terminal content cash against amortization. Base 1.10 times, low 1.03 and high 1.06; the challenge disagreed materially, because carrying the 2031 ratio into a 3.0% perpetuity implied a lead of cash over amortization of 3.2 years that no captured year shows. We changed the model: the terminal ratio is now derived from terminal growth and a lead of 1.4 years, giving 1.042 times in the reference, and the 1.10 times is kept as a sensitivity ($64.53). At the built rate the reference is $69.09 at 1.03 times and $67.14 at 1.06 times, up $0.80 and down $1.16 from where it is now.

Operating margin path. Base exit margin of 35.5%, low 34.0% and high 36.5%, shifting the whole path; the challenge disagreed, but not materially, and found the material problem in how the margin record had been described, which we corrected in section 5. At the built rate the low and high paths give $65.04 and $70.46, changes of $3.25 and $2.17.

Revenue growth path. Base 11% in 2027, low path 10%, 8%, 6.5%, 5.5% and 5%, high path 12%, 11%, 10%, 9% and 8%; the challenge disagreed immaterially. At the built rate the low path gives $62.77, which is $0.43 above the valuation close and the closest any single change in the table comes to it, and the high path $71.23. We kept the reference and show a gentler slow path (10.5%, 9%, 8%, 7%, 6%) as the skeptic's.

Terminal growth. Base 3.0%, low 2.5% and high 3.5%; the challenge agreed. At the built rate the values are $64.69 and $72.48, changes of $3.60 and $4.19.

Operating tax rate. Base 16%, low 14% and high 18%; agreed. At the built rate the values are $70.04 and $66.54, changes of $1.75 each way; the tax check below tests the rate against the measured record.

Content cash against amortization in the explicit years. Base 1.10 times, low 1.08 and high 1.17; agreed. At the built rate the values are $68.72 and $66.80, changes of $0.43 and $1.49; the 1.17 times is also the first nine months of 2026.

Terminal value discount timing. Base the end-of-year convention (zero), low half a year earlier; the challenge disagreed, immaterially, because the explicit years are discounted mid-year and the terminal value was not. We adopted the mid-year convention, so the reference is the low setting, and the end-of-year convention is the sensitivity ($65.89, $2.40 lower).

Two assumptions that no challenge named are also worth stating. Working capital at 2% of each year's revenue increase is below the 4.5% that the cash flow statement shows pooled over FY2023 to FY2025; at 4.5% the reference is $0.27 lower. And content amortization at 33.5% of revenue is below the 35.3% that management's guided growth in amortization implies for FY2026E; at 35.3% the reference is also $0.27 lower. Neither changes the reading.

The 16% tax assumption, checked. Book tax rates were 12.9% in FY2023, 12.6% in FY2024 and 13.7% in FY2025, and cash taxes paid over pre-tax income were 18.6%, 16.5% and 17.5%. The first nine months of 2026 ran at 17.3% and the third quarter alone at 15.1%. The tax actually due on the $2.8 billion termination fee is unknown, so the nine-month rate cannot be divided into the fee's share and the rest, and the fee may have raised it or lowered it. We show two illustrative cases, both our assumptions: taking the fee out with 21%, the US statutory rate, charged on it gives 16.5%, and with no tax charged on it gives 21.3%. Neither is a proven endpoint, and the fee could have been taxed at another rate. For FY2025, setting aside the Brazil expense gives 13.1% if it gave no tax relief and 14.0% if it was deductible at 21%, which are two illustrative cases of the same kind. The 16% sits above the three full-year book rates and just below the three full-year cash rates and the two illustrative nine-month cases of 16.5% and 21.3%. That is a spread in the evidence and not a case for a different figure, because the nine-month rate of 17.3% would lower the reference by about a dollar, so we keep 16%. The reference is worth $71.27 at the FY2024 book rate, $70.31 at FY2025's, $69.05 at the third quarter's, $67.02 at FY2025 cash taxes, $67.15 at the nine-month rate and $63.64 at the 21.3% case, against $68.29 at 16%, $70.04 at 14% and $66.54 at 18%.

A cross-check built a different way

Steady-growth earnings cross-checks, and a stress. Capitalizing a single year of normalized earnings as a steady-growth perpetuity gives $47.33 from the reference's 2027 year and $42.80 from FY2026, which depends on no 2027 assumption. Each is that year's earnings, times one minus the ratio of growth to the return on new capital, divided by the cost of equity less the growth rate, per diluted share; the FY2026 earnings, taken before the termination fee and the fourth-quarter headcount cost, are first grown one year at the terminal rate. Both use the 9.86% cost of equity, 3.0% growth and an assumed 30% return on new capital, so each includes growth and the reinvestment it needs, and neither is averaged with the DCF. What they omit is the reference's explicit multi-year path, including its margin expansion from 2028 to 2031, and the gap between them and the reference reflects that path and not an absence of growth. The upside depends most on its 38.5% exit margin: with the upside's growth but the reference's margin path it is worth $78.15 and not $84.97. As a stress that is not the selected downside, putting the downside's margin at the FY2025 level of 29.5% throughout gives $44.75.

The cross-check and the return-on-capital check of section 9 test different things. The cross-check capitalizes what Netflix earns now, without the explicit margin path, and gives values below the price, which says that the reference's lead depends on margin and growth continuing beyond what one year's earnings show. The return-on-capital check lowers the reference by $1.06 to $1.57 and leaves it above the price, which says that the terminal economics are not what carries the lead.

Where the evidence points the other way. Two path assumptions are the least supported by history, and we show them rather than adopting them. Growth decays one point a year after 2027 on the reference. The record is one annual fall, not a rate of decay: annual growth was 15.6% in FY2024 and 15.9% in FY2025 and is guided at 13.2% for 2026, a fall of 2.7 points in one year, while quarterly growth fell 5.8 points over three quarters, from 17.6% in the fourth quarter of 2025 to 11.8% in the third quarter of 2026. A path of 10.5%, 9%, 8%, 7% and 6% gives $65.76. The margin path adds one point a year to 2029, which is the top of the recent like-for-like range of 0.3 to 1.0 points and not a conservative step; a path that fades from 32.4% to 35.0% gives $67.28. Both together give $64.79. We keep the reference as the evidence-built path and read $64.79 as the figure a skeptic would defend. The two together cost about $3.50 a share, which leaves the skeptic's figure still above the price, by 3.9%, but at a beta of 1.20 with a premium of 5.0% it is $55.96, below the price, and that is why the reference alone does not carry the rating.

The hardest stack of the one-at-a-time sensitivities, content cash held at 1.17 times in every year including the terminal year together with both slower paths, is $55.35, 11.2% below the price. We do not assign it a probability, and it is not a floor: it is a measure of how far three unfavorable and evidenced changes to the reference's operating assumptions together move the value. Other constructions in this report go lower: the reference is $51.20 at a beta of 1.30 and a premium of 5.5%, and the downside is $44.75 with its margin at the FY2025 level and $36.85 at that rate build.

Capital allocation and the balance sheet

Netflix says its approach is unchanged: it puts reinvestment in the business first, organically and through selective acquisitions, keeps a healthy balance sheet and ample liquidity, and returns the excess to shareholders through repurchases. The record shows what that has meant in practice, what the balance sheet looks like after it, and where the pressure is.

How cash has been generated and used

Over FY2023 to FY2025 Netflix generated $23.3 billion of free cash flow and repurchased $21.4 billion of stock, 92% of it. The ratio rose each year, from 87% to 90% to 96%, while debt was repaid in two of the three years and issued in one; the year-end share count fell from 4,453.5 million in FY2022 to 4,222.2 million in FY2025. Before FY2023 the record is thin: the first repurchase the statements report, $600 million in FY2021 (the FY2018 to FY2020 statements carry no such line, which we read as unreported and not as zero), none reported in FY2022, and acquisitions of $788 million and $757 million in those two years. The first nine months of 2026 changed the proportions. Repurchases were $10.8 billion, $1.3 billion in the first quarter, $4.7 billion in the second and $4.8 billion in the third, the largest of the three quarters of 2026, which is 113% of reported free cash flow and 145% of free cash flow without the gross termination fee and with the non-routine Brazil payments added back. They retired 3.1% of the year-end 2025 share count gross, and the count at 09-30-2026 was 2.7% lower net of exercises and awards. The board added $25 billion of authorization in April, and $22.3 billion remains.

Senior notes carry at $15.1 billion (principal $15.3 billion, fair value $15.2 billion) against $8.2 billion of cash, restricted cash and short-term investments, so net debt is $6.9 billion. Netflix issued $1 billion of 5.25% ten-year notes on 07-22-2026. At the $62.34 valuation close, market capitalization is $256.0 billion and enterprise value $262.9 billion, both on the 4,106.3 million shares outstanding at 09-30-2026; the per-share values divide by 4,147.0 million diluted shares, on which both would be slightly higher.

Capital return is the second leg of the thesis and the one that has run furthest. Netflix repurchased $9.1 billion of stock in FY2025 and $10.8 billion in the first nine months of 2026, including $4.8 billion for 63.9 million shares in the third quarter, the largest of the three quarters of 2026, with $22.3 billion of authorization remaining. Repurchases in the nine months exceeded underlying free cash flow by $3.35 billion, and that gap was funded by the termination fee itself, the new notes and cash on hand. Shares outstanding fell from 4,222.2 million at year-end to 4,106.3 million on 09-30-2026, and 120.2 million options at a $41.24 average strike add 40.7 million shares by the treasury method at the $62.34 price, giving the 4,147.0 million shares we use per share. Q4 repurchases are not known. We do not count them, and we do not need to, because every cash flow after 09-30-2026 is attributed to the claims and shares as they stood on that date.

Debt, liquidity and obligations

The senior notes are fixed-rate, unsecured and spread across 2026 to 2054, with 67.1% of the principal maturing by 2029.

The $1.0 billion of 4.375% notes due in November 2026 is the only maturity in 2026, and the company has said the 5.25% notes it sold in July were issued to fund it, so the repayment will not change net debt, which at 09-30-2026 already reflects both the new notes and the cash they raised. The next three years are heavier, $1.5 billion in 2027, $3.5 billion in 2028 and $4.3 billion in 2029, and Netflix has refinanced before: it issued $1.8 billion in FY2024 and $1.0 billion in 2026 while repaying $1.8 billion in FY2025. A portion of the notes, €4.7 billion, is in euros and is remeasured every quarter, which moved pre-tax income by as much as $533 million in a year without touching operating income. Operating income covered interest 17.2 times in FY2025, interest was 7.7% of operating cash flow, and net debt was 0.4 times operating income against 3.7 times in FY2019, so the capacity to carry the debt is not in doubt on any of the measures we hold.

The commitments that matter more than the debt are the content commitments, because they are the operating leverage in another form.

Netflix is committed to $25.2 billion of content, of which $11.8 billion falls due within twelve months and 77.7% is off the balance sheet, and its debt service in the next twelve months, principal and interest, is $3.2 billion, 39.0% of its cash, cash equivalents and restricted cash of $8.2 billion at 09-30-2026. Together with leases the near-term commitments are $15.5 billion against $12.0 billion of guided free cash flow for 2026, which is not an alarming comparison because content cash is already inside free cash flow, and which is the reason we do not deduct the content obligations from enterprise value. The model's treatment of the balance sheet is deliberately plain: net debt of $6.9 billion at carrying value, the 09-30-2026 share count with options by the treasury method, no repurchases after 09-30-2026 and all flows after it attributed to those claims, no dividend (the filings say none has been declared or paid), acquisitions outside the forecast and treated as neither creating nor destroying value, and all liquidity treated as available.

Management's own account is that the strategy is to invest in exclusive entertainment, technology and monetization, to reinvest "organically and through selective M&A" and then return excess cash through repurchases, and that capital allocation is unchanged. The one large departure was the agreement to merge with Warner Bros. Discovery, signed on 12-05-2025, restated on 01-20-2026 and terminated on 02-27-2026, with a $2.8 billion fee paid to Netflix. We treat acquisitions as outside the forecast: Netflix spent $585.7 million on them in the first nine months of 2026, we assume that spending neither adds nor destroys value, and none of the growth we model comes from it. Management has also announced headcount reductions that cost about 120 basis points of fourth-quarter margin, roughly $161 million by our estimate (the company gave no dollar amount), and says the advertising tier will launch in 15 new countries in 2027, starting in March.

Where the pressure is

The balance sheet is not the risk. The risk is the use of cash in 2026, when repurchases ran at 1.45 times free cash flow without the fee, and the fee, the new notes and a fall of $0.8 billion in cash made up the difference. A company that is returning more than it generates can do so for a time and its authorization of $22.3 billion would take at the third quarter's pace roughly four and a half more quarters of buying, but the more it relies on the balance sheet the less room it has for an acquisition like the one it attempted this year, and the less margin of safety it has if growth disappoints. Whether the pace continues will be visible in the fourth-quarter and first-quarter repurchase figures, which we do not have and have not estimated. Where the filings are silent we have left the quantity unknown: repurchases after 09-30-2026 are not known, and the tax on the termination fee is not known, and each bears on how much of the 2026 return of capital was funded by earnings and how much by one-time cash.

What we do not know

Where something is missing we have left it unknown and have not set it to zero or estimated it. We do not know the tax paid on the termination fee, the size of the fourth-quarter real estate transaction, or repurchases after 09-30-2026. We do not know Netflix's membership count or its revenue per membership after FY2024: the company has not reported either since the FY2024 10-K, and the only later disclosure is that paid memberships crossed 325 million during the fourth quarter of 2025. We do not know churn, the share of members on the advertising tier, what an advertising member earns, the cost of an hour of viewing, or the company's own definition of cash content spend. We hold two half-year totals of view hours and no longer series, and no audience-measurement data, so we make no claim about share of viewing. We do not know the date of the fourth-quarter report or of the FY2026 Form 10-K, because none is stated in a document we captured. We hold no consensus estimates, since no approved provider is in use, and no peer's content cash, only the expense and cost lines they report. We read constant-currency changes in revenue per membership for four of the seven years in which the company reported memberships, and none for the years since.

Several inputs are our judgments and not measurements. We did not measure beta, which we judge at 1.10 with a band of 0.90 to 1.30, because we have not collected a market-benchmark series; the equity risk premium of 4.5%, with a band of 4.0% to 5.5%, is a judgment too, as is the assumption that the spread at which Netflix issued its notes in July still holds. The operating tax rate of 16%, the content lead of 1.4 years behind the terminal ratio, the shares of revenue for amortization, capital spending and depreciation, and the 2% working-capital charge are assumptions that the record brackets and does not fix. The earnings cross-checks, the 21% tax on the fee and the 30% return on new capital are assumptions as well. Peer filings give context and nothing more: they are segment disclosures on each company's own definitions.

Sources

Reported figures for the valuation come from Netflix's Form 10-K for FY2025 (filed 01-23-2026), Form 10-Q for the quarter ended 09-30-2026 (filed 10-21-2026), the Q3 2026 shareholder letter (10-20-2026) and the 8-Ks of 12-05-2025, 01-20-2026, 02-27-2026 and 07-22-2026. Management claims are labeled as such: guidance in the Q3 letter, including the fourth-quarter forecast, the 2026 margin and free cash flow guides, the 1.1x content ratio and the advertising forecast, and the statements in the 2026 letters about engagement, churn, live programming and the advertising launches. The valuation price is the NASDAQ regular-session close of 11-02-2026, taken from settled market data. The risk-free rate and the spread at which Netflix issued come from the U.S. Treasury's daily par yield curve for 2026 (the 11-02-2026 and 07-22-2026 rows, 10-year column); the daily real yield curve was captured as context and sets no input.

The eight-year record, the memberships, revenue per membership and regional figures, and the cash flow, balance sheet and share data of sections 3, 5, 6 and 13 are read from Netflix's Forms 10-K for FY2019 (filed 01-29-2020), FY2020 (01-28-2021), FY2022 (01-26-2023), FY2023 (01-26-2024), FY2024 (01-27-2025) and FY2025, with their XBRL facts; the guidance record and the engagement and advertising statements of sections 3, 6 and 7 are read from the shareholder letters of 01-20-2026, 04-16-2026, 07-16-2026 and 10-20-2026. The peer figures come from Disney's Form 10-K for fiscal 2025 (filed 11-13-2025), Warner Bros. Discovery's Form 10-K for 2025 (02-27-2026), Paramount Skydance's Form 10-Q for the quarter ended 06-30-2026, and the Forms 10-Q of Comcast (filed 10-22-2026), Alphabet (10-29-2026) and Amazon (10-30-2026) for the quarter ended 09-30-2026. The additional Netflix and peer filings and the SEC filing indexes that located them were retrieved on 11-07-2026 between 23:16:44 UTC and 23:18:27 UTC, and the indexes show no later quarterly report for Warner Bros. Discovery, Paramount Skydance or Disney than those we hold. Section 2 uses later captures: the settled close of 11-06-2026, the Treasury 10-year par yields for 11-02-2026 to 11-06-2026 and the SEC's filing index for Netflix, all retrieved on 11-07-2026, and the November 9, 2026 regular-session closing price, retrieved at 22:22:49 UTC on 11-09-2026. They set no input of the valuation. The Netflix filing index is the SEC's EDGAR listing of the company's filings, at sec.gov under Netflix's central index key 1065280.

Dates: the evidence cutoff is 11-09-2026 at 22:22:49 UTC, the valuation date and rating price date are 11-02-2026, the latest settled close is that of 11-09-2026, and no capture was retrieved after the cutoff (the November 9, 2026 regular-session closing price at 22:22:49 UTC on 11-09-2026; the earlier market-data snapshot at 01:57:52 UTC, the Treasury table and the Netflix filing index at 01:40:56 UTC, and the filings and peer documents between 23:16:44 UTC and 23:18:27 UTC, all on 11-07-2026), so nothing observed after the cutoff was used. Only the 11-09-2026 closing price is newer than 11-07-2026: the filings, the model, the valuation and the disclosures are as of the earlier dates. Reported figures are re-read from the filings each time the model is built, and the model's declared reconciliations hold: they prove the arithmetic is consistent, not that the economics are right. Calculations, assumptions and judgments are separated in the text: where we write "our adjustment" or "we assume", it is ours. Figures read from the filings are exact as filed; figures we construct, such as cash paid for content, the return on content assets and property, and Netflix's margin on the peers' nearest basis, are our constructions and are labeled.

Disclosure

Blockhaven research is not investment advice, and a Blockhaven Rating is not a recommendation to buy, hold or sell. Blockhaven™ is a mark in use; this is not a claim of registration. Holdings and conflicts for this publication are stated under Positions and conflicts.

Important disclosures

Blockhaven research is provided for general informational and educational purposes. It is not individualized investment advice and does not take into account any reader's financial circumstances, investment objectives, risk tolerance, tax situation or portfolio.

Nothing in this report is an offer or a solicitation to buy or sell a security or other financial instrument. Reading or accessing Blockhaven research does not by itself create an investment-advisory, fiduciary, brokerage or client relationship.

Investing involves risk, including the possible loss of principal. Scenario values, terminal values, valuation ranges, sensitivities and forward-looking assumptions are analytical tools, not guarantees or promises of future performance, and actual results may differ materially.

Blockhaven uses information from company filings, company publications, market-data providers and other sources it believes to be reliable, but that information may be incomplete, delayed, revised or incorrect.

Opinions, ratings and interpretations reflect Blockhaven's research judgment as of the report date and may change as new evidence becomes available.

Positions and conflicts

As of November 6, 2026, Blockhaven's owner held no shares, options, or other direct derivatives in Netflix, Inc.; had no material positions in Netflix, Inc. through entities the owner controls; and had no paid financial relationships with Netflix, Inc.

© 2026 Blockhaven LLC. All rights reserved.