IRON MOUNTAIN INC (IRM) stock intrinsic value fair value estimate & owner earnings analysis

Iron Mountain is a real estate investment trust that stores and manages physical records, documents, media and other valuable assets for corporate, legal, medical and government customers worldwide, and over the past several years has built a second, fast-growing business leasing enterprise and hyperscale data center capacity. The original records-storage business is mature and very stable: customers pay ongoing rent to keep files in Iron Mountain's facilities for years or decades, often because retention is a legal requirement, and once records are placed they rarely move, which shows up in very high customer retention and steady revenue growth in the mid-single digits. The newer data center business requires large, up-front construction spending well ahead of revenue, and the company reports roughly 97% of its existing built capacity already leased, with substantial additional capacity under construction. We estimate what the shares are worth from owner earnings - the cash the business generates after the true cost of keeping its existing portfolio in working order, but before the extra money spent building brand-new data center capacity, which we treat as investment rather than a recurring cost. Iron Mountain discloses this split itself, breaking capital spending each year into 'growth investment' (new data center construction, real estate expansion, new initiatives) and a much smaller 'recurring' category meant to sustain existing buildings and equipment. That recurring figure has stayed close to $140-150 million a year for a decade even as total capital spending grew roughly sevenfold to over $2.2 billion, which is strong evidence that nearly all of the recent increase in spending is new investment rather than the ongoing cost of the business as it exists today. We used that disclosed split - rather than a generic formula based on revenue growth - to separate growth spending from maintenance spending, because a generic formula would have misclassified most of the new data center construction as a routine cost and understated the cash the existing business actually throws off; this is the single most consequential judgment call in the analysis. On this basis, owner earnings per share has grown fairly steadily from about $2.00 in 2016 to about $3.74 in 2025. We anchor our valuation on the median of the last three completed fiscal years (2023-2025, roughly $3.40 per share), a period we chose because it captures the business at its current, data-center-scaled level without leaning on the single strongest year. Owner earnings grew about 10% in 2025 versus 2024. Our base case assumes growth starts at 7.5% in year one - below that trailing pace - and gradually slows to about 3% by year ten, as the records business continues its steady mid-single-digit growth and the data center business keeps leasing up its built and under-construction capacity, with the pace of growth naturally moderating as the base gets larger and as financing costs on the company's substantial debt continue to rise. This is important context: total debt has grown to roughly $16.5 billion, up from about $13.8 billion a year earlier, funding the data center build-out, and the company's stockholders' equity is negative (a deficit of roughly $1.0 billion) even though the business itself is solidly cash-generative - a structure common among REITs that have paid out earnings as dividends for decades, but one that leaves less room for error if growth investments disappoint or financing costs rise. About a fifth of the debt carries a variable interest rate. Our bear case assumes the data center build-out does not pay off as quickly as hoped - hyperscale customers slow new leasing commitments for a period - while interest costs on the large, partly variable-rate debt load continue to climb, squeezing cash flow before the business stabilizes into low-single-digit growth. In this case, owner earnings actually contract slightly in the first two years before recovering to roughly 2% annual growth. Our bull case assumes data center demand stays robust and lease-up continues at a brisk pace, the records business sustains its recently above-trend growth through pricing and digital cross-sell, and earnings scale faster than debt so that financing costs stop being a drag - producing high-single-digit to low-double-digit growth in the earlier years, moderating over time. Because our base and bull cases lean on continued strength in a business line that is only a few years into its current scale, and our bear case leans on financing-cost pressure against a highly leveraged, negative-equity balance sheet, the gap between the cases is meaningfully driven by how well the data center investment converts into durable, growing cash flow versus how much of it gets absorbed by debt service. Beyond the ten-year forecast, we assume the business simply sustains its year-ten level rather than continuing to grow, treating both the records-storage real estate and the by-then-mature data center portfolio as renewing themselves through the recurring maintenance spending already included in owner earnings, with no further net expansion assumed. That flat continuation makes up a majority of the estimated value in all three cases (a bit over 60% in the base case), which is standard for a real-estate-heavy business with long-lived assets, but is worth naming plainly: our estimate rests heavily on the durability of both cash streams decades out, not just on the next ten years. Weighing all of this, our base case estimate is around $87 per share, with a bear case near $60 and a bull case near $105. Given the resilience we see even in the bear case, balanced against the real financial-leverage risk from a negative-equity, growing-debt balance sheet, the margin-of-safety price is around $70 or lower. We were not able to fully reconcile every individual line of the cash-flow statement to the penny in the very earliest years we used (2016-2018), because the company's own growth-versus-recurring capital spending categories were defined somewhat differently before 2020; we relied on the best available figures and note this as a minor, non-material data limitation rather than a reason for concern.