OSI Systems: The 52-Week Low Is Pricing the Delay, Not the Backlog

OSI Systems: The 52-Week Low Is Pricing the Delay, Not the Backlog

The market spent the last week turning a three-month shipping delay into a verdict on the whole company. OSI Systems reported revenue that came up short and initial guidance that underwhelmed, and the stock got sold to its lowest level in a year — falling somewhere in the high single digits to low double digits on the report, after roughly a 28% drubbing in the prior three months. The story the tape tells is that the security-screening business has stalled. The cash-flow path tells a different one: about $50 million of planned deliveries slipped a few months, and the market priced it as if the orders were gone.

The miss was a calendar question

Start with the miss, because that is what has everyone's attention. Fiscal fourth-quarter revenue came in at $484.1 million, down 4% year over year and short of the roughly $529 million Wall Street had modeled; the full-year total of $1.786 billion was still a record, just light of the company's own earlier guidance. The shortfall traces directly to conflict-related delays in the Middle East, where roughly $50 million of Security-division shipments scheduled before the June 30 year-end got pushed out, and the initial fiscal 2027 revenue forecast of $1.875 billion to $1.93 billion came in a couple of percent below consensus.

None of that is a demand problem. The delayed deliveries remain in the order book — the backlog ended the year at a record $1.9 billion, up from $1.8 billion a year earlier — with shifted delivery schedules, and a substantial portion of the deferred shipments now lands in the second half of fiscal 2027. Management said plainly that these do not reflect order cancellations. The revenue "miss" is a calendar question.

The quarter underneath the revenue line is where the story flips. Adjusted (non-GAAP) earnings — which strip out one-time items — came in at a record $3.78 a share, up 17% year over year and at least matching consensus for a fifth straight quarter. Full-year adjusted EPS of $10.35 was a record, up 11%. Margins expanded, and that is what mattered: gross margin reached 34.7% and adjusted operating margin 17.7%, up 200 basis points in the quarter. The healthcare division that had dragged on results swung from a 1% to a 10% adjusted operating margin, and service revenue grew 13% to $441 million, pushing the mix toward recurring, higher-margin revenue instead of one-off hardware.

The cash-flow path

Now the number I care about most: free cash flow. For the trailing twelve months it roughly quadrupled to about $227 million, and the company ended the fiscal year with $359.8 million of cash, up from $106.4 million a year earlier. One honest caveat: operating cash flow of $275.9 million, up from $97.6 million, is flattered by a $159 million collection in the fourth quarter from the company's largest Mexico customer, so I would not annualize the single quarter. But the direction is real, and management expects fiscal 2027 free cash flow to exceed 100% of net income — after already returning cash: the company repurchased 564,880 shares for $123.6 million during the fourth quarter, then added another million shares to the buyback authorization on results day.

The biggest reason the next twelve months look better than the last one sits in that Mexico compare. The legacy contract subtracted roughly $150 million from fiscal 2026 comparisons; the drag fades to less than $25 million in fiscal 2027. That is a swing of well over $100 million in the year-over-year math — invisible if you are staring at the deferred Q4 deliveries, but the single biggest number in the fiscal 2027 setup. Service revenue compounding in the double digits and a shrinking share count ride on top of it.

What the market is still pricing

Which brings me to the market's misread. After the selloff the stock sits just under $200. On trailing earnings it is trading near 21 times, but on the company's own fiscal 2027 adjusted-EPS guidance — midpoint $11.31, range $11.13 to $11.49 — it is roughly 17 to 18 times. That is the multiple of the lumpy, hardware-dependent security vendor OSI used to be, not of a company engineering record backlogs, more than 17% return on equity, and a free-cash-flow yield near 7% (about $227 million of trailing free cash flow against a roughly $3.3 billion market capitalization). At its 52-week high the stock traded near 28 times forward earnings; it now trades at roughly 17 times for the same order book.

The sell-side is not even arguing with the numbers. Analysts trimmed estimates after the report — Oppenheimer cut its target from $320 to $300 and B. Riley to $265, and both kept bullish ratings — while the Street had already cut the quarter's EPS consensus from $3.90 to $3.76 over the prior month, and no downgrades followed the print. AInvest's aggregate signal still labels the stock a Buy. When the analysts who own the base rates trim a few dollars off their targets and the stock falls another 12% anyway, the expectations reset has done its work while the business has not broken.

Target, timeframe, and the tripwire

So I will put a target and a tripwire on it, because the financial bridge is explicit. Price: the low- to mid-$240s over the next twelve months, roughly a quarter above the panic close. That only assumes a re-rate to about 21 times the midpoint of the company's own guidance — still below the roughly 28 times it carried at the high, and below where the sell-side targets, a third to a half above the current price, already sit. Timeframe: fiscal 2027, which management built deliberately back-half-weighted — conservative on the Middle East, and including only part of the U.S. Customs IDIQ ceiling (a multi-year framework for ordering scanners as needed), with much of the larger volume landing in fiscal 2028 and beyond. The four things that carry the bridge: the $1.9 billion backlog converting, the Mexico compare getting more than $100 million easier, service revenue compounding in the double digits, and a shrinking share count financed by rising cash generation.

The break condition is equally specific. If the backlog stops feeding revenue — a quarter or two of book-to-bill below 1.0 (book-to-bill is new orders booked relative to revenue shipped, so a reading above 1 means orders are outpacing shipments), the deferred Middle East orders converted into cancellations, or guidance pulled — then the market was right and I would cut the position without ego. Middle East execution is a genuine, recurring risk, and the fourth-quarter cash number was partly a one-time collection. I can be wrong again. But right now the market priced a few months of delay like it was a broken demand story, while the record backlog and the fading Mexico drag say the next twelve months improve. For anyone building or holding a position, the ledger is simple: enter or hold through the noise while the backlog converts, and treat a bookings rollover as the exit signal. I would rather own that mismatch than explain it later.