Capstone Energy Plus, Inc. (CEPL) stock intrinsic value fair value estimate & owner earnings analysis

Capstone Energy+ (formerly Capstone Green Energy, and before that Capstone Turbine Corporation) makes oil-free microturbine generators used for on-site "behind-the-meter" power at industrial and commercial sites, and is now positioning the same technology for data centers that need reliable power faster than the utility grid can deliver it. The company has shipped over 10,800 units worldwide over nearly four decades. In August 2025 it acquired its West Coast distributor (Cal Microturbine), giving it a direct sales presence across several western states. This is a business with a difficult recent history. Its stock was suspended and delisted from Nasdaq in October 2023 following a financial-statement restatement, and it has traded on the OTCQX market since. It has lost money in most years of the last decade, and its cash flow from operations has swung widely from year to year, driven largely by working-capital movements rather than a stable, repeatable level of profitability. We measured owner earnings -- operating cash flow after stripping out working-capital swings, stock-based compensation, and the portion of capital spending needed just to replace worn-out equipment -- at roughly $4.7 million in the fiscal year ended March 2025 and $12.1 million in the fiscal year ended March 2026 (helped by revenue growth from the Cal Microturbine acquisition). An earlier year (fiscal 2024) comes out around $7.5 million on a less complete data set. We used the median of this three-year window, about $7.5 million, as the starting point for our forecast, rather than the most recent year alone, because the swings between years reflect working-capital timing more than a change in the underlying business. The single most important fact about these shares is what happened on the last day of fiscal 2026. The company completed a major recapitalization: it raised $80 million by issuing a new class of convertible preferred stock to funds managed by Monarch Alternative Capital, and raised a further $43.7 million by selling new common stock, using the combined proceeds to buy out an older, similarly structured preferred investor for $83.5 million and pay down $8.3 million of debt. The new preferred stock sits ahead of common stock in the capital structure. It carries a dividend that starts at 5% a year and is scheduled to increase by roughly one percentage point every year regardless of how the business performs, rising even faster -- up to a cap of 13% -- if the company fails minimum earnings, leverage, and liquidity tests. Because that dividend is paid in additional preferred shares rather than cash, it never shows up in operating cash flow; it simply grows the preferred holders' claim on the company every year, compounding. On its scheduled path, that claim could roughly double over the next decade even if the business does nothing wrong. The preferred can also convert into common stock at $5.00 a share, which would add about 16 million new shares to the current base of roughly 30.2 million. That structure dominates what these shares are worth. We valued the whole business first -- the stream of owner earnings, discounted over ten years, plus a conservative view of what it sustains after that -- and only then worked out what is left for common stockholders once the preferred stock's claim is accounted for. In our base case, revenue and owner earnings grow modestly as the data-center opportunity develops and the Cal Microturbine acquisition beds in -- high-single-digit growth near term, fading to roughly 2-3% by year ten. That produces a whole-business value in the area of $195 million. Against that, the preferred stock's compounding claim -- which, on this path, the preferred holder would rather keep than convert, since it grows faster than the equity does -- comes to roughly $110 million in today's dollars. What's left for the existing 30.2 million common shares works out to about $2.90 per share. Our bear case assumes the AI/data-center opportunity does not scale as hoped, competition from batteries and other on-site power sources erodes margins, and the company periodically misses the preferred stock's financial tests -- something that has already happened once before, in 2024, when a separate debt covenant required a waiver. Missing those tests pushes the preferred dividend rate up faster, and owner earnings drift down roughly 5% a year instead of growing. On this path, the preferred's claim compounds to roughly $126 million in today's dollars, more than the roughly $57 million we estimate the whole business is worth on that trajectory. In plain terms, in our bear case the company would owe the preferred investor more than the operating business is worth, and the common shares would be worth nothing -- a shortfall of a little over $2 a share on paper, though shareholders can't actually lose more than what they paid. Our bull case assumes the data-center and energy-resiliency demand for on-site power genuinely takes hold, with double-digit owner-earnings growth for several years before settling into slower, sustainable growth. On that path the whole business would be worth roughly $390 million, and the preferred holder does better by converting into common stock (at $5.00 a share, adding about 16 million shares) than by keeping its preferred claim. Splitting the larger pool across the enlarged share count works out to about $8.40 per common share. More than two-thirds of this bull-case value rests on what the business sustains many years beyond our ten-year forecast -- an assumption that demand for on-site, behind-the-meter power at data centers and critical facilities keeps growing slowly and indefinitely, which is far from assured this early in that market's development. Putting these together, the margin-of-safety price sits at a meaningful discount to our base-case estimate -- around $1.45 -- reflecting how much of the outcome depends on staying current on the preferred stock's financial tests and avoiding further dilution. That price also assumes the company demonstrates, over a couple of consecutive quarters, that it is comfortably meeting those tests and that its recent owner-earnings improvement isn't just a one-time working-capital swing. A few things we could not fully verify from the filings: the exact non-cash items making up operating cash flow in fiscal 2024 were not broken out at the same level of detail as the two most recent years, so that year's figure is directionally useful but less precise. We also note that a deemed non-cash charge of $69.6 million was recorded against fiscal 2026 earnings, tied to the old preferred investor's payout at redemption; this is a one-time accounting entry from the recapitalization, not an ongoing cost, and we have excluded it from operating owner earnings for that reason -- though its size illustrates how much value that earlier financing arrangement had accrued to its holder over time, which is exactly the kind of dynamic our bear case assumes could repeat with the new preferred investor.