CAT Technical Appendix: How the valuation was built and tested

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

• Deep Dive: A great franchise priced beyond its modeled economics (44 pages). Public preview; the Deep Dive page says how to read the full report.
• Research Note: The business has moved up a level, the price has moved further (6 pages).
• Technical Appendix: How the valuation was built and tested (20 pages).
• Learning Companion: How to read the Caterpillar research (10 pages).

---

The method, the tests of the reference path, the limits of the evidence and the sources, as the Deep Dive states them. Priced at the 11-02-2026 close against intrinsic value at the 11-02-2026 valuation date, from an explicit forecast to the ninth forecast year with a terminal value; not a forecast of twelve-month price performance. Latest settled close 11-09-2026. Information through 11-09-2026.

Since the valuation date

The valuation in this report is dated 11-02-2026 and has not been redone. Since the earlier editions of this report we have added one observation, the settled close of 11-09-2026, and it sits beside the 11-06-2026 close and the valuation close below. The filings, the model, the valuation and the disclosures are as of their earlier dates, and the only thing new is that market snapshot. Caterpillar closed at $613.62 on 11-02-2026, at $558.63 on 11-06-2026 and at $598.88 on 11-09-2026, which is 2.4% below the valuation close and 7.2% above the close of the previous session. A settled close is the price at the end of a regular session and not a live quote. The second table sets the three values, which are present values at 11-02-2026 and are not rolled forward, against the valuation close and the latest settled close.

What moved is the price and, slightly, the yield. The price fell 9.0% by 11-06-2026 and recovered 7.2% by 11-09-2026, and we do not know why it did either, since no document we gathered explains the moves. The last Treasury row we gathered is dated 11-06-2026, when the 10-year yield was 5 basis points lower than on the valuation date, at 4.86% against 4.91%; we gathered no later row. It is an input to our discount rate, which we built from the 11-02-2026 row and have not rebuilt. The stored one-point sensitivity gives the scale: a discount rate one point lower gives the reference $399.19 and one point higher gives $307.24, against $346.89, so a move of five hundredths of a point is small beside it.

What did not move is the value. The three values are those of 11-02-2026. We have not rolled them forward, and we have not recalculated any scenario, break-even or sensitivity. At the 11-09-2026 settled close each of our three paths still sits below the price: the upside 22.2% below it, the reference 42.1% below and the downside 64.4% below. Put the other way, the price is 28.5% above the upside, 72.6% above the reference and 180.6% above the downside. The gap to the upside was 24.1% at the valuation close, 16.6% at the 11-06-2026 close and is 22.2% at the 11-09-2026 close, and each change is the effect of a different price and not of a better value.

One qualification belongs beside that. The most favorable of the joint sets we ran (the rate one point lower, Power & Energy margins higher and a lighter working-capital charge) gives the upside case $570.61, which was 7% below the 11-02-2026 close, was 2.14% above the 11-06-2026 close and is 4.72% below the 11-09-2026 settled close, while the same set leaves the reference 29.1% below it. That set is a sensitivity and not one of our three paths, and the refresh did not change it. It fell short of the price at the 11-02-2026 close, did not fall short at the 11-06-2026 close and falls short again at the 11-09-2026 close, so whether it reaches the price depends on the close it is compared with and on nothing we recalculated. The full stack of favorable ends, set out under how hard we tested the reference, puts the upside at $676.06, above each of the three closes and 12.89% above the 11-09-2026 settled close, and the reference at $517.55, 13.58% below it. We state it plainly because what we say about the gap covers the cases we ran and the close each was compared with, and not every setting a reader might try.

The comparison does not change our reading. The rating is a statement about the 11-02-2026 close and valuation date, a rating is tied to the price it is given at, and at the 11-09-2026 settled close the price is still above each of our three paths, which is the relation the rating rests on, and the gap to each is still wide. The break-even discount rates move with the price, and the table of what the price fits sets the 11-09-2026 figures beside the 11-02-2026 ones. The new company information we read, the third-quarter Form 10-Q and the Fabick transaction, is set out under subsequent events below and moves no forecast.

Forecast mechanics: the drivers, the schedules and the links between them

Each of three segments is forecast with its own sales growth and its own segment margin in each of nine twelve-month years, the first of which starts in the fourth quarter of 2026 so that the base is the reported last twelve months. Corporate items and the small All Other segment are held at their last-twelve-month ratios to sales. Machinery, Power & Energy profit is taxed at 23%, capital expenditure and depreciation follow a path set as a share of sales, and working capital absorbs 15% of each year's change in sales. Growth drives the rest. Faster sales need more capacity and more working capital, which is why the upside, with the most growth, spends $3.74 billion on capital expenditure and $1.09 billion on working capital in its first year against $3.20 billion and a working-capital release of $662.4 million in the downside. Free cash flow is what is left, and it is the only thing discounted. The paths below are the assumptions, and every one of them is a Blockhaven judgment.

Power & Energy carries the reference. Its sales grow 14% in year one, against the 17.7% of the third quarter and the $36.0 billion of the last twelve months, and the growth fades to 3% by year nine. The evidence for the level is the backlog, $78.1 billion at 09-30-2026 with $32.6 billion of it beyond twelve months, together with two competitors' filings that describe the same demand. For the fade there is no record to point to: our record has no segment margins before FY2023, so no captured document shows an earlier power cycle, and the fade is our judgment that scarcity pricing gives way as capacity is added, with nothing in a document saying when. The margin path is the harder call. We start at 22.5%, a little above the 21.03% of the last twelve months and below the 24.4% of the third quarter; both of those may include tariff recoveries, because the twelve months hold the second quarter's $392 million and the third quarter's $128 million and the filings do not state how they split by segment. We fade it to 20% from year seven. The review challenged exactly this, under the name Power & Energy segment margin path (reference fades to 20% from y7; P&E only), with a low of 18% and a high of 21.5%. Measured on the earlier model, the low value lowered the reference by $24.80 and the high value raised it by $18.60; on the current model the same settings move it by $23.21 down and $17.40 up. The review disagreed with the fade as immaterial, since holding the margin near the current level is at least as well supported, and an effect of that size changes no conclusion about the price.

The machinery segments carry the cycle. In the reference Construction Industries sales shrink in the first two years and then grow 3% a year, at a margin that eases from 19.5% to 19%; Resource Industries is flat to slightly up for two years, then grows 3% a year, with its margin rising from 14.5% to 16%. The evidence for the early shrinkage is management's statement at the second quarter that it expected dealers to reduce inventories in Construction Industries in the fourth quarter, the $500 million dealer destocking of 2025 and the cycle record of falls of 22.4% to 41.5% in consolidated revenue. The margins are set below the FY2023 and FY2024 levels of 25.44% and 24.22% in Construction Industries and 20.75% and 20.35% in Resource Industries, because FY2025 showed what happens to those margins when price turns negative: 18.66% and 15.94%. Our segment record is short, FY2023 to the last twelve months, and in Resource Industries it runs one way: the margin fell from 20.75% to 20.35% to 15.94% and then to 13.95% over the last twelve months, three falls in a row, with 13.48% in the third quarter, when price realization added only 2.40% of sales and the incremental margin was 12.66%. The reference's rise from 14.5% to 16% therefore assumes a reversal that the record does not yet show; it is our judgment of mid-cycle, and the path stays below the segment's FY2023 and FY2024 levels while starting above its latest ones. At the Machinery, Power & Energy level the upside ends year nine at 20.99%, near the 21.35% of FY2024, and the downside ends at 15.15%.

One of the review's two material challenges sits here, under the name MP&E through-cycle operating margin. Its base is the reference's own margin from year seven, 18.14%; its low is 15.53%, the FY2018 to FY2025 mean of Caterpillar's consolidated operating margin; and its high is 21.35%, the FY2024 margin of the machinery and power business. Applied as a uniform shift of every segment margin in every year, the low value lowered the reference by $60.78 and the high value raised it by $74.68, as the review measured them on the earlier model. On the current model the same two margins give the reference $294.56 and $411.19, $52.33 lower and $64.30 higher than $346.89. The review called this a material disagreement. Its argument was that a margin of 18.14% is close to what the long record supports and not conservative, since the consolidated margin reached 19% only in FY2023 and FY2024, and that the reference could not be labeled conservative-to-central. We accepted that. The reference is our central estimate of current economics, and the long-record reading is shown beside it as a stress and not substituted: with every segment margin lowered so that the ninth-year margin equals the long-record mean, the reference becomes $294.56. The case for leaving the reference where it is has two parts, that power and services now carry more of the mix and that the consolidated series carries Financial Products and every charge, and neither is something a captured document measures.

Working capital and cash taxes are the two reinvestment assumptions the review adjusted at the margin. The model charges 15% of each year's change in sales to working capital, which matches the measured ratio of the working-capital measure to sales, 15.42% at 09-30-2026 against 15.19% at the end of 2025. The review challenged it under the name Incremental MP&E working capital (fraction of the change in MP&E sales), with a low of 10% and a high of 18%. Measured on the earlier model, the low value raised the reference by $4.71 and the high value lowered it by $2.83; on the current model the low value gives $351.33, $4.44 higher, and the high value $2.67 lower. The reason to think the true charge is lower while the backlog grows is that the measure nets only current customer advances, and the third-quarter 10-Q puts $3.4 billion of contract liabilities in Other liabilities, outside it; the working-capital measure itself grew 17.30% in nine months. The review judged the disagreement immaterial, and so do we. Cash taxes are the other half. The model taxes book profit at 23% in every year, and the review's challenge, Cash taxes below book tax in years one to three (not modeled), tested a cash tax shortfall of $912 million a year, which raised the reference by $5.22 on the earlier model and gives $352.07, $5.18 higher, on the current one; the cause and duration of the shortfall are not stated in the documents we gathered, which is why we left it out of the base.

Capital intensity and Financial Products complete the mechanics. Capital expenditure starts near management's 2026 expectation of $3.5 billion and settles at 3.3% of sales, with depreciation climbing to 2.8%, and the return on operating capital in the reference falls from 49.99% in year one to 41.6% in year nine as the plant it builds enters the base. Financial Products is not forecast year by year. It is valued at its book equity times a multiple built from a sustainable return on equity of 15% in the reference against 16.39% over the last twelve months (measured on the average of two balance-sheet dates and including subsidies MP&E pays Financial Products), which the review challenged under the name Financial Products return on equity (reference). Measured on the earlier model, a low of 12% lowered the reference by $5.23 and a high of 16.5% raised it by $2.62; on the current model the same settings move it by $4.77 down and $2.38 up. The review agreed with the assumption.

The schedules show the result year by year, and all three scenarios are given in full because a reader who disagrees with a path should be able to see which year and which line is doing the work.

Three things follow from the schedules. Free cash flow in the reference rises every year, from $9.16 billion to $13.8 billion, and that is the discounted stream, so the near-term years matter less to value than the margin the path reaches and keeps. The downside's year-two trough in sales, $63.9 billion, is followed by recovery, and its terminal-year free cash flow of $8.79 billion is what the downside is actually valued on. And the upside's free cash flow reaches $19.2 billion in year nine on sales of $126.0 billion, at an operating margin of 20.99%, and it assumes, among other things, that Power & Energy holds a margin of 23% or more for the whole period.

How we value it

We value MP&E with a nine-year discounted cash flow, forecast in twelve-month years that start in the fourth quarter so the base is the reported last twelve months, and add Financial Products at the book value of its equity multiplied by what its return on equity earns above its cost. Three segments are forecast separately, each with its own growth and margin for every year, and the corporate and elimination lines are kept at their last-twelve-month ratios. The firm's cash flows are discounted at 10.27%, built from a cost of equity of 10.41% and a debt weight of 2.19%, with a terminal growth rate of 3% and a tax rate of 23%. Only one of the rate inputs is observed. The risk-free rate of 4.91% is the 10-year yield on the official daily Treasury par yield curve for 11-02-2026, the price date, and our pre-tax cost of debt of 5.41% is that yield plus a spread. The equity risk premium of 5%, the beta of 1.1 and the spread of 0.5% over the Treasury yield are our assumptions, and the rate is one of the two inputs, with the margins, that move the value most; which of the two moves it more depends on the step tested, as the section on how hard we tested the reference shows. Half a point less on the premium gives the reference $374.87 and half a point more gives $322.75; a tenth less on beta gives $372.14 and a tenth more gives $324.81; a point more on the spread gives $346.13, because debt is a small part of the capital. Capital expenditure starts near management's 2026 expectation of $3.5 billion for MP&E and settles to 3.3% of sales, with depreciation rising to 2.8% as new capacity enters service. Claims on MP&E are deducted once: net debt including lending to Financial Products of $2.35 billion and after-tax retiree obligations of $2.78 billion.

We state three conditional paths and attach no probability to any of them. The reference is the path we think current economics most support: backlog converts with some slippage, the power premium partly persists, Construction Industries fades to a mid-cycle margin below its FY2023 and FY2024 levels, and Resource Industries, whose margin has fallen in each period since FY2023, recovers part of the way to a mid-cycle that is also below those levels. The downside is a machinery trough in the first two years with recovery to a lower mid-cycle, and a permanent loss of the power premium as competitors' capacity arrives. The upside lets the power build-out run to about 2030 with its premium margin intact, and returns machinery margins to near their FY2023 and FY2024 levels.

The terminal year does not release reinvestment: capital expenditure of $3.66 billion sits above depreciation of $3.05 billion, and it is tied to the net property, plant and equipment the nine forecast years actually build, which reaches 19.22% of sales. Return on operating capital in the terminal year is 41.68% and on new capital 40.82%. Those are high by construction, because Financial Products finances the dealer receivables that MP&E would otherwise carry, so we treat them as a diagnostic that the cash flows are not implausible and not as a plug. The terminal value is 55.33% of MP&E value in the reference, and more than half in each of our three paths, which is why the terminal margin and the rate matter so much.

These are present values per share at 11-02-2026 on 461.4749 million shares, not future exit values and not price targets. The share basis is the 06-30-2026 count plus third-quarter dilution. The third-quarter Form 10-Q, filed after the valuation date, reports 457.594174 million shares outstanding at 09-30-2026, which is 0.84% fewer than our basis. We kept our basis, and the subsequent events below say what the reported count would change. Two share-count effects appear in this report and they differ because the counts differ: the sensitivity table's $347.95, $1.06 above the reference, takes our basis less an estimate of the shares bought back in the third quarter, while the dated addendum's $1.57 uses the lower count the 10-Q reports, with the same dilution.

Each valuation lens measures something different, and the differences are informative. The cash-flow model values the cash we forecast, with margins that fade, growth that slows and reinvestment that follows both. Capitalizing last-twelve-month operating profit with no growth values current earnings as if they were permanent and nothing more, and it gives $238.73 a share. Capitalizing last-twelve-month free cash flow grown once at the terminal rate gives $417.82, and it is the most flattered because that cash flow carries customer-advance timing. The reference of $346.89 sits between them, and the lenses are not averaged because they answer different questions. In every scenario the machinery and power business is almost the whole value: it accounts for 93.20% of the reference's equity value and Financial Products for 5.49%, so the argument about the price is an argument about MP&E margins and their duration, and the terminal value, which is 54.89% of MP&E value in the downside, 55.33% in the reference and 57.37% in the upside, is where most of that argument is settled.

The reference is not a midpoint, and a reader could be forgiven for thinking so. The midpoint of the downside and the upside is $339.74, which is $7.15 from the reference, and that closeness is arithmetic coincidence. The reference is built from its own drivers, each set where the record best supports it: Power & Energy growth near the rate it is running at, a power margin that stays above its FY2023 level for most of the period, machinery margins below their FY2023 and FY2024 levels, the highest in our short segment record (FY2023 to the last twelve months), and a cycle that turns once in the nine years without a deep recession. The downside and the upside are coherent alternatives built to test the reference from each side, and no probability attaches to any of the three. Where a reference driver falls between the other two paths, that is what the evidence suggested for that driver and not a rule of construction.

What the reference is, and how hard we tested it

We call the reference our central estimate of current economics, and we say plainly what that means. Against the reported run rate it looks conservative, and against the run rate without tariff recoveries it is not: its ninth-year MP&E margin of 18.14% is below the 18.85% of the last twelve months, but equal to the 18.14% those months earn without the $520 million of recoveries that we do not treat as a run rate, and the reference's first four years, at 19.26%, 19.3%, 19.16% and 18.99%, are above both. Against the long record it is generous, because the consolidated margin averaged 15.53% over FY2018 to FY2025. We keep it where it is because the franchise has plausibly changed, and we show the long-record reading beside it as a separate stress and do not substitute it. With every segment margin lowered by 2.4 points in every year so that the ninth-year margin equals that average, the same growth path is worth $294.56. Asking the same question of the upside gives a split answer. Its Construction Industries margin of 21% and its Resource Industries margins of 16% to 18% sit inside the record from FY2023 to the last twelve months, and for those two segments neither direction has been accepted on judgment while the other was asked for proof. Its Power & Energy path is another matter: at 24% falling to 23%, it is above each of the four points in the Power & Energy record, 17.63%, 19.88% and 19.93% for the full years and 21.03% for the last twelve months, as the bull case below says. The reference's Power & Energy path is above all four in its first four years, from 22.5% to 21.5%, and is 20% from year seven, below the twelve-month figure.

The downside is a bear case with recovery, and it is mild against history. Its worst year is 8% of MP&E sales below the year before and its low point is 13.5% below the last twelve months, at a trough margin of 13.59%. Consolidated revenue fell 22.4% from FY2019 to FY2020, 36.9% from FY2008 to FY2009 and 41.5% from FY2012 to FY2016. We therefore add two hypothetical stresses that sit below the downside and are not forecasts. In the deep-trough stress MP&E sales fall 30.1% in two years, to a trough margin of 9.55%; from year four it takes the downside's own margins and growth rates, so its margins converge on the downside's, but its sales do not catch up within the forecast: in year nine they are $67.9 billion against the downside's $80.8 billion, 15.90% lower. It is worth $175.00. In the no-recovery stress the same fall is permanent within the forecast: sales stay at $51.6 billion from year two through year nine, the margin settles at 10.99%, close to the long record's median of 10.95%, and the terminal value then grows at 2%; it is worth $105.52. Even there MP&E keeps earning more than its cost of capital on the capital it employs, so this is a stress on earning power and not on solvency.

The rate sensitivities are worth stating together. One point lower on the discount rate lifts the reference to $399.19 and one point higher lowers it to $307.24; margins two points lower or higher give $303.24 and $390.53; terminal growth half a point lower or higher gives $336.30 and $359.04. Two smaller items lean upward and are not in the valuation: a lower working-capital charge gives $351.33, and cash taxes running below book for three years, which is our own timing measure and not a company statement, gives $352.07. Holding everything else fixed, the model matches the price with the reference at a discount rate of 7.01%, or with every segment margin 12.22 points higher in every year, and no terminal growth rate up to 7% gets there on its own, since the reference is worth $548.07 at that rate. Those are conditional arithmetic, not a statement of what investors believe. Combining a point off the rate, Power & Energy margins higher by 1.5% of sales and a lighter working-capital charge, the most favorable of the joint sets we ran, gives $424.36 for the reference and $570.61 for the upside. That upside figure was below the 11-02-2026 close, above the 11-06-2026 close and is below the 11-09-2026 settled close, as the section "Since the valuation date" sets out.

That joint set carries only three of the favorable ends in the table below, so we also ran every one of them at once, on the reference and on the upside, as a labeled hypothetical stress with no weight. The firm rate is one point lower, at 9.27%, which is below the 9.3% that the table's low beta and low premium would build, and the cost of equity that prices Financial Products and affiliates is built from that low beta and premium, at 9.41%, so the rate is lowered once and not twice. Every segment margin is 2 points higher in every year, with Power & Energy a further 1.5% of sales higher; on the reference that still left the ninth-year MP&E margin below the review's high, so its shift is raised to 2.14 points, where the ninth-year margin equals the 21.35% of FY2024. Terminal growth is 3.5%, working capital takes 10% of the change in sales, cash taxes run $912 million a year below book for three years, the $1.20 billion import-overpayment receivable is added at face value, Financial Products earns 16.5% on its equity, and the share count is the 459.3942 million that the 10-Q's reported count and the same dilution give. Fabick's cash price is left out, as it is everywhere in the model, because this stack takes only favorable ends.

The stack lifts the upside to the price and beyond it: $676.06 is $62.44 above the 11-02-2026 close. It does not lift the reference there, which stays $96.07 short. At the 11-09-2026 settled close the upside stack is $77.18 above the close and the reference stack $81.33 below it. What the upside stack assumes beyond the record is the whole list at once: a ninth-year MP&E margin of 24.09%, above every MP&E margin in our record, a firm rate below the one we can build from the observed Treasury yield and our premium and beta, terminal growth above the long record's pace, three years of cash taxes below book whose cause no document states, and a receivable collected in full. We show it because the sentence "no combination we ran reaches the price" would otherwise be wider than the combinations it rests on; it is true of the reference and not of the upside.

Which input moves the value most depends on the step. On the table's steps, a point on the rate moves the reference by $52.30 up and $39.65 down, a span of $91.95, and two points on every margin by $43.64 up and $43.65 down, a span of $87.29, so the rate is larger on the low-rate side and smaller on the high-rate side. On the review's own settings rerun on the current model, the through-cycle margin moves the reference more in both directions, $52.33 down and $64.30 up, a span of $116.63, against $34.04 down and $54.42 up, a span of $88.47, for the review's rate builds.

A second method points the same way. Capitalizing last-twelve-month MP&E operating profit with no growth at our discount rate, then adding Financial Products and affiliates and deducting claims, gives $238.73 per share; at the FY2018 to FY2025 average margin it gives $198.96. Capitalizing last-twelve-month free cash flow grown once at our terminal rate gives $417.82, which is the highest of the three and also the most flattered, because that cash flow carries customer-advance timing. Looked at the other way round, the price capitalizes $36.9 billion of MP&E operating profit with no growth, which is 2.65 times the last twelve months. We do not average these with the cash-flow model, because they answer different questions. We gathered no analyst research and no consensus estimates, so the comparison with outside estimates is limited to these lenses and to the long record.

The one-at-a-time sensitivities are collected below, each applied to the reference with everything else held fixed, so that a reader can see which input moves the value and by how much. They are tested ranges and not forecasts of what the inputs will do.

The rate, the margins and the terminal growth rate dominate, and the rate is the only one of the three whose build we partly observe. The next table sets out what the price fits, and it is conditional arithmetic and not a statement of what investors believe: holding everything else fixed, each row gives the single change at which the model matches the close, or says that it does not. The first column is the 11-02-2026 close, the rating price, and the second is the 11-09-2026 settled close; the values compared are the present values at 11-02-2026. The second column is the model's own break-even routine run again with the new close as the target and nothing else changed, so the valuation date, the rates and the assumptions are those of 11-02-2026.

The stress matrix lays the labeled stresses beside the scenarios year by year, so that the depth of each trough and the level each path settles at can be read directly. The stresses are hypothetical and sit outside the three paths.

A cross-check built a different way asks what return on operating capital the value would need, instead of what cash flow. Machinery, Power & Energy ties up $23.1 billion of net property, plant and equipment and working capital at 09-30-2026 (this excludes goodwill and leased assets), and last-twelve-month operating profit after tax earns 46.37% on it. The reference values the business at 6.67 times that capital and the price at 11.95 times, against 4.18 and 9.00 for the downside and the upside. In a steady state at our terminal growth rate and discount rate, a multiple of capital is worth paying only if the business earns a return of about g plus the multiple times the gap between the rate and g, which is 51.51% for the reference and 89.85% for the price. Holding the rest fixed, the identity fits the reference at a return a little above what the business earned over the last twelve months, which the growth years supply, and fits the price at nearly twice that return, held for the whole life of the business. That is a statement about what the price fits and not a forecast, and it points the same way as the cash-flow model through a different identity.

Terminal economics

The last explicit year and the first terminal year are laid side by side below for each path. The terminal year does not release reinvestment, and the return it shows on new capital is high because Financial Products finances the dealer receivables that the machinery business would otherwise carry.

The terminal growth rate of 3% is the last assumption the review challenged, under the name Terminal growth of MP&E free cash flow, with a low of 2% and a high of 3.5%. Measured on the earlier model, the low value lowered the reference by $24.78 and the high value raised it by $15.43; on the current model they move it by $19.90 down and $12.15 up. Its evidence is the long record: consolidated revenue was $45.0 billion in FY2007 and $67.6 billion in FY2025, a little over two percent a year in nominal terms through three deep cycles, and the review put 3% at the top of what that record supports. Electrification and power demand argue for more than history and maturity argues for less. We kept it, and it is immaterial to the direction of the conclusion, since no terminal growth rate up to 7% lifts the reference to the price.

The steady-state identity gives a check on the terminal value that does not use the discounted cash flow. In a steady state, growth equals the share of profit reinvested times the return earned on it, and the reference reinvests 7.14% of its terminal profit at a return of 40.82%, which funds growth of 2.91%, a little below the 3% declared, and the same in all three paths. That agreement tests how we built terminal reinvestment, which is tied to the plant the forecast builds, and it is a consistency check and not an independent estimate of growth; the independent evidence on the terminal assumptions is the return-on-capital and earnings-power cross-checks in the section on how hard we tested the reference. The terminal value is 12.77 times terminal profit after tax, a multiple that we find high for a cyclical manufacturer and that is explained by a discount rate of 10.27% and growth of 3% together. The terminal value is 55.33% of MP&E value in the reference, which is why the margin and the rate matter more than any single forecast year.

The independent challenge, assumption by assumption

An independent review challenged every assumption that the model's value leans on, tested each against the model across a low and a high value, and recorded whether it agreed. The table gives the review's own names, bases and effects. The review ran on the model as it stood before three later changes: the discount rate rebuilt from the captured Treasury yield, the terminal value discounted mid-year, and terminal capital spending tied to the plant the forecast builds. Its effects are therefore measured on that earlier model, whose reference was $373.56, and they are not changes to the current $346.89. The discount-rate row carried a note in the review that the risk-free rate was still uncaptured; that has since been corrected.

The second table reruns the review's own low and high settings on the current model, so that each can be read against the current values. The discount-rate row keeps the review's beta and premium and uses the observed risk-free rate in place of the review's assumed ones; the trough-depth row is measured on the downside, every other row on the reference.

Of the eight, the review disagreed materially on two, the through-cycle margin and the depth of the downside's trough, and both are discussed above under their own names. The discount rate and the through-cycle margin are the two largest by dollar effect, and which is larger depends on the side and the step. Its base in the review was 9.62%, with a low of 8.64% built from a beta of 1, a premium of 4.5% and an assumed risk-free rate of 4.25%, and a high of 10.64% from a beta of 1.2, a premium of 5.25% and an assumed 4.5%; both risk-free rates are below the 4.91% since observed for 11-02-2026. On the earlier model the low value raised the reference by $65.79 and the high value lowered it by $50.89, against $60.78 down and $74.68 up for the margin: the rate's low setting moved the value more than the margin's low setting did, but the margin moved it more upward, more downward and more in total. Rerun on the current model with the observed risk-free rate, the review's beta and premium give the reference $401.31 and $312.85, and the same pattern holds. The review judged the rate reasonable and its disagreement immaterial to direction, while noting that it bears on the choice between the two lowest ratings. Since then the risk-free rate has been built from the Treasury yield for 11-02-2026, and our rate is 10.27%, from a cost of equity of 10.41%.

The review also found method issues, and each was corrected in the model before this report. The terminal value had been discounted at the end of the year under a mid-year convention, and the correction moved the reference from $373.33 to $383.23; that step also carries $0.23 from carrying intermediates at full precision, so the mid-year correction alone is $9.67. Terminal reinvestment had not been tied to the plant the nine years build, and tying it moved the reference to $381.62. Rebuilding the discount rate from the captured Treasury yield, the last step, brought the reference to $346.89. The steps are mechanical and attributed: the first two are corrections of method, and the third is new evidence together with a built risk-free rate and a spread tied to it.

Subsequent events

Caterpillar filed its third-quarter Form 10-Q on 11-04-2026, after our valuation date. The headline figures we compared with the third-quarter earnings release, sales and revenues, operating profit, profit and diluted profit per share, agree with it, and it reports no material change to the risk factors in the 2025 annual report. It also reports 457.594174 million shares outstanding at 09-30-2026, contract liabilities of $8.706 billion against $4.678 billion at the end of 2025, and unsatisfied performance obligations of $49.4 billion on contracts of more than one year. We have not changed any forecast or scenario for any of this, and none of it explains the October fall in the share price, which we still cannot document.

The filing also carries the terms of an acquisition our model does not include. Caterpillar announced on 09-29-2026 that it had agreed to acquire Fabick, a long-standing Cat dealer, with no price given, regulatory approval needed and closing expected within 30 days, and announced the completion on 11-02-2026, again with no price and no closing date. The 10-Q says the purchase completed on 10-31-2026, that $1.24 billion in cash was paid at closing and that an estimated $165 million will follow after certain holdback periods. The price is preliminary and subject to working-capital and other adjustments, and purchase accounting is not yet available. The deal closed before our valuation date but after the 09-30-2026 balance sheet our net debt uses, so neither the cash paid nor the dealer acquired is in the model.

Three approximate sensitivities, set out in a dated addendum of 11-04-2026, are illustrations we tested and not measurements of the full effect. They are hypothetical, they are not additive, they are not a combined bound and they are not a rerun of the model, and none of them is a bound on the economic effect of the acquisition or of the contract liabilities. First, if the Fabick purchase returned no value at all, the reference would be about $3.06 a share lower, 0.9% of $346.89. That is an extreme illustration and not a net adjustment: Caterpillar acquired an operating dealer with inventory, receivables and earnings that the model does not include, and at a price equal to value the net effect would be about zero. Second, holding the model's equity value fixed, the reported share count with the same dilution would raise the reference by about 0.45%, roughly $1.57 a share, the downside by about $0.97 and the upside by about $2.11.

Third, of the $8.706 billion of contract liabilities, about $3.4 billion sits in Other liabilities and $5.329 billion in current Customer advances. Our working capital is forecast as a flow of 15% of the change in sales, on a measure that nets only current advances; it neither deducts the stock of non-current advances nor assumes it unwinds. An unwinding of the non-current amount with nothing replacing it would be worth about $5.50 to $7.35 a share, 1.6% to 2.1% of the reference, and it bears on the downside and no-recovery paths more than on the reference. Our own working-capital sensitivity, 10% in place of 15%, gives the reference $351.33. None of the three, as tested, changes our reading of the rating. That is a statement about these illustrations and not about the full effect of the acquisition, which the model does not include, or of the contract liabilities, which it treats only in part.

The company's third-quarter analyst slides schedule its next results: fourth-quarter 2026 on 02-02-2027, first-quarter 2027 on 04-29-2027, second-quarter on 07-29-2027 and third-quarter on 10-28-2027.

What we measured, what we assumed and what we do not know

The figures for history come from Caterpillar's annual report, its 2026 quarterly filings and its third-quarter earnings release; where we compute a figure, such as a last-twelve-month total or a margin, it is arithmetic on those reported numbers. The price the rating is set at is the settled regular-session closing price of 11-02-2026, split-adjusted and not adjusted for dividends, and the 11-06-2026 close comes from a later series of regular-session closing prices of the same kind. The latest settled close, 11-09-2026, is the November 9, 2026 regular-session closing price, one market snapshot taken after that session settled; it is a closing price and not a live quote. The risk-free rate is the 10-year yield on the official daily Treasury par yield curve for 11-02-2026, as published, and the 11-06-2026 yield is from the same table. The valuation date and the rating price are 11-02-2026, the latest price observation is the settled close of 11-09-2026, and the evidence runs through 11-09-2026. What is new since 11-07-2026 is that market snapshot only: the filings and other documents we gathered, the model, the valuation and the disclosures are as of their earlier dates. The forecasts, the equity risk premium, the beta, the spread of our cost of debt over the Treasury yield, the margin paths, the stresses and the mechanical one-year roll-forward of value are Blockhaven assumptions and judgments and are labeled as such. We did not obtain the third-quarter earnings call, segment margins before FY2023, how the tariff recoveries are allocated between segments, or any outside estimate, and we found no document that reports machine or engine unit volumes, the amount of services revenue, market share, or the results of the non-US makers. The eight-year history and the peer comparison are read from the SEC company-facts series for Caterpillar, Deere, Cummins, GE Vernova and Terex, retrieved on 11-07-2026, consolidated and as first reported; a peer's fiscal year is named by the calendar year in which it ends, and a figure a company did not report for a year is shown as not available and never as zero. The price-realization bridges, the incremental margins and the allowance figures are read from Caterpillar's own tables and sentences, and the review's assumption table is read from the independent review's record. We read the third-quarter Form 10-Q and the company's analyst slides for the dated facts set out above, and we did not rerun the model on either. The filing says where the customer advances sit: current Customer advances and Other liabilities, as the subsequent events describe. We do not know why the shares fell in October.

Sources

Caterpillar Form 10-K for 2025 (02-13-2026); Form 10-Q for the quarters ended 03-31-2026 and 06-30-2026 (05-06-2026 and 08-05-2026); Form 10-Q for the quarter ended 09-30-2026 (11-04-2026); earnings releases for the fourth quarter of 2025 and the first, second and third quarters of 2026 (01-29-2026, 04-30-2026, 08-04-2026 and 10-29-2026); the 2026 proxy statement (04-30-2026); the 2025 Investor Day transcript and presentation (11-04-2025); Caterpillar's third-quarter 2026 glossary and analyst slides (slides retrieved 11-07-2026); Caterpillar's press releases on Fabick (09-29-2026 and 11-02-2026); SEC EDGAR filing index for Caterpillar; GE Vernova Form 10-Q for the quarter ended 09-30-2026 (10-28-2026); Cummins Form 10-Q for the quarter ended 06-30-2026 (08-04-2026); SEC XBRL company facts for Caterpillar, and for Deere, Cummins, GE Vernova and Terex (retrieved 11-07-2026); regular-session closing prices for CAT through 11-06-2026, and the November 9, 2026 regular-session closing price; U.S. Department of the Treasury daily par yield curve and real yield curve rates for 11-02-2026, and the par yield table through 11-06-2026.

Disclosure

This research is for information and education. It is not individualized advice, and the Blockhaven Rating is Blockhaven's own assessment on its own scale; it is not a recommendation to trade any security. Scenarios are conditional, unweighted and not forecasts.

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 Caterpillar Inc.; had no material positions in Caterpillar Inc. through entities the owner controls; and had no paid financial relationships with Caterpillar Inc.

© 2026 Blockhaven LLC. All rights reserved.