Central Puerto S.A. (CEPU)

By both methods, the value today (DCF $31 · Multiples $22) exceeds the market price ($14).

Today — fairly valued: at $14 trades ~37.1% below its value discounted to today (~$22); the discount is positive but does not reach the margin of safety we require (≥38%).

At 5 years — Infravalorado: the target price ($28) plus dividends yield above the required average return (10%) — the business compounds.

The bridge: the return at 5 years exceeds the risk-free rate (4.5%) — but the discount does not reach the required margin of safety (≥38%). To require a 15% annual return, it would need to be bought at ~$14.

Educational / informational. Does not constitute investment advice.

The inverse approach: instead of projecting growth to obtain the value, the market price ($14) is taken as given and it solves for what annual owner-earnings growth would need to hold for 5 years for the present value —at the method's rate (4.5%, no-growth terminal)— to equal that price. It is the disconfirmation test: the expectations the price already pays for, contrasted against the method's projection.

The market discounts less growth (-3.5%/year) than we project (14.0%/year) → if the base case holds, there is margin: perception is more pessimistic than the estimated reality.

That growth implies ~ARS 135.2 bn of owner earnings in year 5 (vs ~ARS 311.2 bn of our base case). It recalculates if the DCF assumptions are edited.

Year-by-year projection of the selected scenario. From each year, two versions of the flow are derived: growth FCF (operating flow − total capex, the cash surplus) and maintenance FCF (the owner earnings: what the business yields if it only sustains its capacity). The flow is returned almost in full (dividend + buyback) or redeployed into the operation, so that EV stays roughly flat and multiples compress because the metric grows, not because of cash accumulation. The valuation is done on Utilidad neta atribuible a la controlante (P/E). In edit mode, revenue, margins, capex, and exit multiples can be adjusted.

Model in billions of pesos, at the equity level: cash and debt are zero in the common block because funding is already charged inside profit through the interest line, and the multiple is applied to profit attributable to the parent. It is the same mold used to value the rest of the merchant generation archetype. Year 0 is the fiscal year closed December 31, 2025, the annual-reporter convention: Central Puerto files Form 20-F and its 6-Ks seal the period, so the pipeline could not identify the earnings release on its own. It was searched for and incorporated by hand: the consolidated interim statements at June 30, 2026 arrived via the August 19, 2026 6-K and are cited in the sources. There is no quantitative guidance published for the current fiscal year — the filing states that its forward-looking statements are subject to regulatory restrictions — so the path is anchored in the trajectory and in the already-filed half-year, and that gap is disclosed here rather than assumed closed. Normalization of year 0, which is this record's central work. The reported operating income of 370.373 carries inside it four items that are not extrapolable: an impairment reversal on property, plant and equipment and intangibles of 60.950, which changed sign three years running (plus 126.029 in 2023, minus 134.287 in 2024, plus 60.950 in 2025) and is non-cash; a net foreign-exchange difference of 73.136 classified within other operating income, which existed because the functional currency was the peso while receivables with the market administrator are pegged to the dollar; an insurance recovery of 18.851; and penalties collected from suppliers of 9.968. Subtracting all four, the fiscal year's recurring operating income comes to 207.468, a margin of 18.9% against the 33.8% reported. Below operating income, the result from investments measured at fair value is also removed, which contributed 134.632 against 3.306 the prior year and corresponds to the revaluation of the minority stake in a mining project. The monetary and financial block, however, is left intact and not broken into parts: its net for the fiscal year, after the 59.699 equity stake in associates, is negative 38.540, comparable to the negative 32.284 of the annualized 2026 half-year — a reasonably even level between the two periods, and stripping the exchange gain on the asset while leaving the loss on the liability would have skewed the base. With those adjustments, normalized pretax income falls from 453.028 to 228.627, and profit attributable to the parent, at the filing's own 22.1% effective rate and net of the non-controlling interest, comes to 174.8 against the 346.4 reported: half. The model's year-0 net margin is that 15.93%. Comparability between the two filings is limited by the company itself, and that must be said. Effective January 1, 2026, the functional currency became the dollar, and all assets, liabilities, revenue, expenses and equity items were converted at a rate of 1,482 pesos per dollar; Note 2.4 of the interim statements warns that comparative information is presented as it was included in prior statements and that those circumstances limit comparability. That is why the half-year columns are not chained to fiscal year 2025 through any restatement index. What does make them comparable in level is that the closing exchange rate at December 31, 2025 was 1,455 pesos per dollar, per the same 6-K's table of financial assets and liabilities in foreign currency: the December-2025 pesos and the pesos converted at 1,482 are within 2% of each other, and the two series are read on that basis. The year-1 path is not a projection but a filed half-year annualized to twelve months. First-half-2026 revenue was 1,040.177, nearly as much as the full 2025 fiscal year, and operating income was 307.279, a 29.5% margin. The jump has two causes and only one is economic: Resolution 400/2025, in force for transactions since November 1, 2025, shifted fuel management and procurement onto generators — which the market administrator previously delivered at no charge — which enlarges revenue and cost at once without proportional percentage margin, and at the same time introduced a marginalist scheme with dollar prices tied to variable production cost, which is a genuine repricing. Added to that is the Piedra del Águila concession, renewed for thirty years and taken over on January 9, 2026, with tariffs now denominated in dollars. The regulatory repricing is built into the base; what is not extrapolated is the half-year's seasonal spot price — the Argentine first half includes the summer demand peak —, the period's hydrology, and the 1.5% effective tax rate the half-year reports, which is an artifact of the functional-currency change, and that is why the model uses the structural effective rate of 22.1%. The half-year is annualized by multiplying by two, with no seasonal factor, because the comparatives are not homogeneous and there is nothing to calibrate against: this is disclosed rather than inventing the adjustment. From there the path smoothly decelerates toward 5% by year 5, which is what a mature generator sustains with system demand, the entry of the two battery-storage projects — 150 MW at Central Puerto and 55 MW at Central Costanera, with an estimated capital of US$130 to 140 million; the 20-F projects completion by mid-2027 and the Management Report in the August 12, 2026 6-K moves it up to the fourth quarter of 2026, a tension that is disclosed rather than resolved in favor of a single source — and real price. Operating margin compresses from 28.9% to 26.5%: the Resolution 400/2025 framework involves progressive normalization, and Decree 450/2025 opens a two-year transition toward free contracting, so the capital now entering the sector for high margins ends up competing them away. Net margin falls from 20.2% to 18.0% also on financial cost: financial debt rose from US$337.8 million at December 31, 2025 to US$671.9 million at June 30, 2026, to fund the hydroelectric concession and storage. Maintenance capital is set at 58% of total invested capital, matching it to depreciation and amortization. It is not a mute shortcut: under Greenwald's criterion, growth capital is approximated by the ratio of fixed assets to sales multiplied by the change in sales, and here sales are growing strongly and the capital in excess of depreciation is buying identifiable, dated expansion — the US$245 million Piedra del Águila concession and the two storage projects — not replacement. Year-0 depreciation and amortization is taken from the difference between the half-year's adjusted earnings before interest, taxes, depreciation and amortization (393.762) and its operating income (307.279), annualized: the issuer's taxonomy does not publish the concept, and this is a subtraction of two facts from the same filing, not a derived figure. No buyback is modeled: the share count fell from 1,514.022 million to 1,513.770 million ordinary shares during the fiscal year, with 252,034 shares canceled, i.e. 0.017% annually, well below the materiality threshold. The record is written in depositary receipts: each represents ten ordinary shares per the 20-F's cover page, so the 1,513.770 million ordinary shares are 151.377 million receipts, which is the denominator of every per-share value in this record. The common block's market value is expressed in pesos using the 1,482 exchange rate the company itself discloses in the 6-K, as the most recent one citable from a filing; the anchor price is in dollars, which is the quotation currency.

Today's elevated multiple is the price of growth: if the business grows, the entry point cheapens on its own going forward (the metric grows while EV stays roughly flat). The exit multiple at 3 years is higher than the terminal at 5 years —at 3 years there is more growth still ahead—, so the value curve shows whether value creation is concentrated in the early or the later years. The required return is applied to the base scenario.

They are valued separately, with their own rationale, and are not incorporated into the base or the verdict (they are excess return). When assigning them value — in Editmode —, the total with optionalities updates live, without moving the base.

The verdict, the base CAGR, and the margin of safety are always calculated on the base; optionalities do not alter them (with optionalities at $0 they do not move).

Each card fixes a required annual return and answers: if the business is worth $28 in 5 years, what is the maximum that can be paid today to obtain that return? Since it now trades at $14, the margin of safety is how much cheaper the market is than that maximum. The three thresholds: 4% covers inflation (the floor), 10% is the long-term average return, and 15% is the level of a great investment.

With a target price of $28 in 5 years and a required return of 4.5% annually, the maximum to pay today is $22. Against the current market price ($14), the margin of safety is 37.1% (trades below the maximum → there is margin) and the total return at that price would be 14.6% annually.

• Entry multiple. It enters at 18× on the fiscal year's normalized profit and compresses to 7× on year 5's.
• Normalized base. Recurring attributable profit for 2025 is 174.8, not the 346.4 reported: half of the result is not extrapolable.
• Exit multiple within band. The blend of the three pieces comes to 12.0 times, within the archetype's band of 10 to 14, without lowering it out of caution.
• Convergence of the two lenses. The multiples valuation and the present-value one converge within a reasonable margin despite jurisdiction risk.

Each margin over sales, year by year: historical (solid line) → projection (dotted).

• ✕ ROIC exceeds the cost of capital (~10%)
• ✓ CFROIC backs up the ROIC (100%, cash vs. accruals)
• ✓ Healthy balance sheet (low corporate debt)
• ✓ Durable competitive moat (multiple advantages)

• Predictable cash generation. Operating cash flow stayed between 340 and 411 billion pesos while the reported result swung from 259 to 969.
• Return on capital. 5.83% on the balance sheet's invested capital in 2025, well below the 10% bar, although the level in the first half of 2026 clears it comfortably.
• Reinvestment runway. 205 MW of awarded storage and a thirty-year hydroelectric concession just taken over give retained capital an identifiable destination.
• Quality of the reported result. Four straight fiscal years with non-cash impairments and reversals that move operating income within a range of nearly four to one.

Annual levels from the official filing (10-K); the % over each bar is the year-over-year (YoY) growth vs the prior year.

The economic driver of a generator is volume times price: megawatt-hours generated and sold, plus megawatts of availability, against the price of each sales modality. Central Puerto generated 18,603 GWh in 2025 with 6,933 MW installed, 13.03% of the system's supply, but its taxonomy does not publish the annual generation series, so the level is cited and the series does not exist: the gap is disclosed rather than reconstructed. The three drivers that do have a full series show what decides this record: a notably stable operating cash flow, a capital investment that multiplied ten and a half times between 2023 and 2025 funding the hydroelectric concession and storage, and reported operating income that swings from 259 to 969 billion pesos on the effect of impairments and reversals that are not cash.

The % are the annual (year-over-year) growth: each year —historical and projected— vs the prior one; the TTM (trailing 12m) vs the TTM of a year ago, to avoid overlapping windows. The historicals are exact figures from the official filings; the projected years come from the year-by-year model (the intermediate years without their own series are interpolated between the anchors). The projected columns (+1y…+5y) are 12-month windows counted from the TTM close (31-dic-2025): the projection starts from the most recently reported data, not the fiscal year. The projected base is realistic and unbiased — the risk discount is applied at the end, via the required return. The rationale for each metric is in the (i).

• Regulatory repricing. The marginalist scheme with dollar prices has been in force since November 2025 and the following first half already shows the new level filed.
• Volume growth. Generation depends on dispatch and hydrology: Piedra del Águila fell from 5,173 to 2,683 GWh between 2023 and 2025.
• Capacity under construction. 205 MW of storage: the 20-F projects mid-2027 and the August 12, 2026 6-K moves it up to the fourth quarter of 2026; the tension between the two sources is disclosed.
• Durability of the jump. Part of the revenue increase is an accounting-perimeter effect, from the shift of fuel procurement onto the generator, and it does not contribute proportional margin.

• Undifferentiated product. Energy is dispatched by cost and price is set by a resolution: there is no product differentiation or pricing power.
• Leading scale in the private segment. 15.5% of private installed capacity against 11.3% for the second-largest operator, and 13.03% of system generation.
• Recurring, growing demand. The system posted a historic demand peak of 30,257 MW in February 2025, surpassing the prior year's record.
• Single counterparty. The market administrator is the counterparty for approximately 82% of 2025 revenue (50.0% spot, 25.6% forward under contract, and 6.2% from the RenovAr program, per the customer table in Item 4 of the 20-F), with a history of collections that exceeded 90 days in the past and that since March 2024 have averaged 2 to 5 days beyond the regulatory term.
• High operating leverage. Dominant fixed costs on long-lived assets: a price improvement passes through almost entirely to the result, and so does a compression.

Not all debt is equal: only the structural needs refinancing; the rest is operational (self-liquidating).

Structural debt is what is exposed to the contraction phase of the cycle; operational debt (leases, matched funding) self-liquidates with the business.

• ✓Leverage (net debt / EBITDA)Net debt / EBITDA 0.4x
• !Interest coverage (EBIT / interest)4.1x
• –Liquidity (current ratio)no data
• ✓Cash quality (CFROIC vs ROIC)CFROIC backs 100% of ROIC
• ✕Value creation (ROIC − 10% bar)-4pp
• ✓Malinvestment test (capex vs incremental ROIC)Capex/D&A 1.8x
• –Float / working capitalNeutral WC
• ✓Dilution (SBC % of revenue + shares)SBC 0.0% of revenue

A traffic-light interpreted by the method (not generic): float (negative WC) adds up, capex is judged by incremental ROIC (malinvestment test), and a lender is not subjected to corporate solvency. The (i) shows the derivation of each number.

• Leverage. Debt of 493.025 billion pesos against equity of 2,616.118 at the close of 2025, with the company describing it as relatively low.
• Liquidity. Cash and other current financial assets of 337.861 billion pesos, equivalent to US$231.50 million.
• Recent increase in debt. Financial debt rose from US$337.8 to US$671.9 million in a half-year to fund the concession and storage.
• Currency composition. 98.17% of financial liabilities are in foreign currency, today matched with the new dollar functional currency.

How cash comes in and how it is deployed. In green, the business's own cash (the owner-FCF it generates and reinvests); in gray, the float and credit — customer and funding money, which is not the shareholder's.

Capital allocation shifted decisively toward reinvestment: capital investment rose from 28.2 billion pesos in 2023 to 295.4 in 2025 while the dividend was cut from 62.8 to 1.0, on top of which come the US$245 million hydroelectric concession and the US$130 to 140 million storage projects. Buybacks are marginal, at 0.017% of the share count annually. On August 31, 2026 the board additionally authorized a buyback program of up to US$30,000,000 (up to 10% of capital, 180 days), an order of magnitude above the historical pace of cancellations, which is not yet reflected in the capital-allocation cascade because it was not executed as of the close of the incorporated filing.

• Capital allocation. Cut the dividend from 62.8 to 1.0 billion pesos and multiplied capital investment by ten and a half to fund a concession won in a competitive tender.
• Price discipline in the acquisition. The US$245 million for thirty years of a 1,440 MW plant were won in a national and international tender, not a private negotiation.
• Transparency of disclosure. Discloses the change in functional currency and explicitly warns that comparative information limits comparability between periods.
• Alignment and incentives. No proxy statement filed with the Securities and Exchange Commission: detailed share ownership and compensation remain a disclosed gap.

• Missing buyers by jurisdiction: it is an Argentine company that trades through depositary receipts in New York, with exchange controls in force that the filing itself flags as a limitation on remitting dividends to holders.
• The reported result is illegible without work: four straight fiscal years with impairments and reversals that swing operating income from 259 to 969 billion pesos, plus a change in functional currency that the company itself warns limits comparability between periods.
• The regulatory framework that reprices energy in dollars has been in force since November 1, 2025 and has no proven track record: the market discounts its reversibility before having seen a full fiscal year under it.
• On August 31, 2026 the board, with a report from the Audit Committee and the Supervisory Committee, approved a buyback program of up to US$30,000,000 (up to 10% of share capital, 180 days, funded with discretionary reserves), on the explicit grounds that «the market quotation of the shares does not adequately reflect the fair value of the Company's underlying assets nor the economic potential derived from its operations». The buyback ceiling, US$16 per depositary receipt, sits above the market price of US$14.04.

The source of the discount is the combination of a hard-to-read reported result with a new regulatory framework. Both are real, but neither is a deterioration of the business: operating cash flow stayed between 340 and 411 billion pesos across the four fiscal years in which the reported result swung nearly four to one. What is being discounted is uncertainty over the durability of the repricing, and the disconfirming test for that reading is the second half of 2026.

The annual return (CAGR at 5 years) in each scenario, with the total period return below — the margin of safety made visual: upside range wide, downside range narrow.

Even in the bear scenario, the return holds at ~0%/year (-2% total): the margin of safety protects the downside. The bull (+25%/year, +207% total) exceeds it comfortably — a favorable asymmetry, with a narrow downside range and a wide upside range.

• Regulatory reversibility. The marginalist scheme is a revisable administrative act, with no full fiscal year of history under the new regime.
• Fuel risk. The filing itself warns that if it fails to pass through or hedge the higher fuel cost, the result could be materially and adversely affected.
• Dilution in the fund companies. Two stakes have already fallen from 30.9% to around 10%; the third, consolidated today, faces the same mechanism with at least 70% state ownership.
• Currency and remittance risk. The peso depreciated 41.35% in 2025 and exchange controls could limit dividend remittance to depositary-receipt holders.
• Hydrology. Generation at Piedra del Águila fell 48% between 2023 and 2025 due to variability in the Limay river's flow.

Each thinker's analytical framework applied to our data.

• Buffett / Graham Quality + margin of safety

  Fails the quality gate: ROIC 6% does not clear the 10% bar.

• Peter Lynch Growth at a reasonable price (GARP)

  A fast grower growing 14% at a PEG of 1.3 → reasonable for its growth.

• Joel Greenblatt Cheap and high-return (Magic Formula)

  Earnings yield 7% (EBIT/EV) + ROIC 6% → falls outside the Magic Formula.

• Howard Marks Perception vs reality + cycle

  The price discounts -4% vs our 14%: perception is more pessimistic than reality, with an identified source of the discount.

• Seth Klarman Capital protection (bear scenario)

  Bear-scenario floor -0%/yr, bull-scenario ceiling +25%/yr over 5y → capital protected, asymmetry in your favor.

• Pat Dorsey Moat strength (Five Rules)

  A narrow moat, stable; sources: efficient scale, cost advantage, intangibles → partially passes the Five Rules.

• Aswath Damodaran Expectations implied by the price

  Justifying the price requires discounting -4%, within what we project (14%) — the story squares with the numbers.