IBM is not worth buying unless it reinvents itself. It has some dominant businesses, but since the financial crisis it has relied on financial engineering rather than real innovation. Even buying at the Covid low or the 2022 low produced a poor return, cushioned only by a roughly 5% dividend. x.com  x.com  x.com 

IBM is a valuation trap, with a high-conviction sell target of $200 by December 2026 as its multiple compresses toward its historical average. It lags in AI and quantum, its premium P/E lacks the structural growth to justify it, and its CY26 EPS estimate of $12.35 is under threat. A macro slowdown could push enterprises to cut nice-to-have transformation projects, and missing hyped quantum milestones would speed a rerating as Google and Microsoft pull ahead. x.com  x.com 

IBM looks like a long-term buy with record long-term accumulation behind it. Its last candle showed $245B of six-month high-volume accumulation, and the stock is retesting both EMAs and $232.90 support. A $14M write of long-dated $300 calls reads as a hedge by large holders who accumulated near $300, not a bearish bet; a retest of $300 is the next test, with $200 the level that would invalidate it. x.com 

IBM is extremely expensive, and its latest announcement adds caution: enterprises are prioritizing hardware and memory purchases, which pressures software spending. Quantum computing is a distant option, software is too small a part of the business to drive upside, and the stock carries a significant Trump premium. The same shift supports even stronger conviction in semiconductors and memory. x.com 

IBM trades at an AI premium with no real AI growth behind it. The company forecast only 1% growth yet more than doubled from $150 to over $320 on AI hype, while its AI business is effectively non-existent. Its Q2 earnings warning came from consulting and infrastructure, not software, and the stock has fallen back to its May level around $220. x.com  x.com  x.com 

IBM looks attractive because investors are treating a timing slip like a demand collapse. x.com 

IBM is one of the strangest and most expensive companies today because its re-rating rests on a software transition its numbers don't show. Bulls say IBM is becoming a software company, yet deferred revenue was lower as a share of total assets in FY25 than in FY16. x.com 

IBM remains bearish on trend heading into earnings, which will be a damage report rather than a preview. It already pre-warned on July 15 with preliminary results well below estimates, BofA cut its target to $280, and the stock is down about 34%. Because much of the bad news is already out, a bounce is possible if it resets the bar and stops the bleeding. x.com 

IBM could reach a $2 trillion market cap as a sovereign AI play, per Hugh Hendry. Global banks refuse to trust unproven AI apps and want IBM's legacy structural safety, and it also offers a pure US foundry play on quantum computing; Hendry rates it a buy-and-hold long toward its 2012 highs near $210. x.com 

Hardware is the new software as clients redirect capex toward supply-constrained servers, storage and memory, benefiting Dell and SanDisk. IBM fell 25% after saying clients shifted late-June capex toward those purchases ahead of expected price increases, a reprioritization bigger than it anticipated. x.com