August 14, 2026
Written before entry.
Position
Long SNPS at a reference price of $415.50, taken intraday on August 14, 2026, with the bid at 414.93 and the ask at 415.99 against a prior close of $411.75. The fifty-two week range is $366.00 to $626.24, and the low went in inside the last thirteen weeks. The stock is down 11.5% year to date; the fifty-two week opening price was $616.50. Annualized historical volatility is 40.3%. I intend to hold six to twelve months.
The argument
Over the twelve months to early August 2026, Synopsys fell roughly 36% while the iShares Semiconductor ETF returned approximately 106%. If hyperscaler capital expenditure were the variable setting the price at $415.50, the stock would have tracked the group, and it did not. The portfolio already carries that factor through SKHY, which is why I want the second position responding to something else.
The business under that divergence is design software. Every advanced chip, whether it is an Nvidia accelerator, an Alphabet Tensor Processing Unit, a Broadcom custom application-specific integrated circuit, or a Chinese design nobody has qualified yet, gets laid out in electronic design automation software, and Synopsys and Cadence Design Systems hold that market between them. Revenue arrives as multi-year licenses written against customer research budgets and design starts, so rack installations in any given quarter barely touch it. Synopsys exited fiscal 2025 with backlog above $11 billion, up from $10.1 billion the prior quarter. Fiscal 2026 guidance is revenue of $9.625 billion to $9.705 billion, an operating margin target of 40.5%, roughly 320 basis points above fiscal 2025, and non-GAAP earnings per share of $14.72 to $14.80. At $415.50 against the midpoint, the forward multiple is approximately 28 times.
A capital expenditure pause reaches different companies at different speeds. If hyperscaler spending decelerates in 2027, inspection tools and optical transceivers see it in orders within two quarters. Design programs already underway do not stop, because the sunk cost of an abandoned tapeout exceeds the license fee. That is why I expect the correlation to be lower. It is reasoning from how the revenue is booked, and I have not measured it.
Catalyst
Third quarter fiscal 2026 results land on August 26, 2026, twelve days from writing. Design IP is the line to watch. It declined year over year in the second quarter while the company raised full year guidance on revenue, operating margin, earnings per share and free cash flow, and management has called fiscal 2026 a transition year for the segment. If Design IP falls again while guidance elsewhere goes up again, I will stop treating the problem as transitional.
Synopsys has also said it will detail artificial intelligence monetization at an upcoming Investor Day, which is the other dated event on my list.
August 26 is a crowded day for the book. HP reports fiscal third quarter results the same day, and that is the demand destruction checkpoint written into the SKHY memo, so two of my three invalidation tests resolve at once.
Risks
Three things went wrong at the company, and none of them are resolved. Two further risks here are mine rather than theirs: when I bought, and what I have assumed.
The Ansys acquisition put more than $10 billion of long-term debt on a balance sheet that historically carried almost none, and the associated intangible amortization of roughly $504 million annually distorts reported earnings badly enough that the trailing multiple near 81 times is not usable. Management has cited synergy targets of $400 million in annual run rate revenue. I have no independent basis for judging whether they arrive.
China is approximately 10% of revenue and falls under United States export restrictions on chip design software. Washington imposed restrictions in late May 2025 and rescinded them on July 2 of the same year, which is a five week round trip on a policy I cannot forecast in either direction. China revenue excluding Ansys fell 22% in fiscal 2025.
Elliott Investment Management’s managing partner joined the board after the second quarter. Someone concluded the company needed outside pressure, and I read that mainly as a comment on the years before the appointment. Nvidia separately took a $2 billion equity position as part of a strategic partnership. Both facts tell me who is now watching Synopsys, and neither of them is a valuation.
Entry timing is the risk I am least comfortable with. Three days before writing, the stock traded near $384. It is $415.50 now, so I am paying roughly 8% above the recent print and buying into a bounce. The fifty-two week low of $366.00 was set inside the last thirteen weeks, which means the selling is recent and I have no evidence it has finished.
The decorrelation case rests on twelve months of relative performance and an argument about revenue recognition. I have not calculated a beta to SKHY. If a broad artificial intelligence repricing arrives, everything in the book falls together whatever the toolchain business is doing.
Limits
The investment-mandate caps this position at 30% of the account, and I am running a trailing 25% stop. The thesis is invalidated if Design IP revenue declines year over year for two more consecutive quarters, if fiscal 2027 guidance implies revenue growth below 8%, if United States export restrictions expand to cover electronic design automation software beyond the current perimeter, or if Cadence Design Systems takes disclosed share in a competitive displacement at a named customer.