CALX— AI Stock Forecast & Price Targets

Revenue trajectory is genuinely accelerating: $265M → $272M → $280M → $293M across the last four quarters, with Sales Y/Y TTM +27.8% and Q/Q +21.3%. Operating income has expanded sequentially from $9.3M → $12.7M → $21.8M, and operating margin has tripled from 3.4% to 7.4% over three quarters — a real operating-leverage signal tied to the platform mix shift. However, gross margin has compressed from 57.7% (Q4'25) to 54.6% (Q2'26), confirming memory-cost pressure is biting into the software-mix tailwind. The balance sheet is a genuine differentiator: $68.9M cash, $13.98M total debt, current ratio 2.83, $710M equity, and $61.5M TTM FCF ($11.9M Q2 FCF). EPS Y/Y TTM +294% flatters off a low base but is directionally correct. Capital allocation just completed a $431M buyback and filed a $51.6M ESOP shelf — the buyback at these levels is shareholder-friendly. What's broken: gross margin trend and the fact that trailing P/E of 48-49x still bakes in flawless execution; the forward multiple of 16x is only defensible if the operating leverage story continues to compound.

The tape is unambiguously ugly. 1wk chart shows a sustained downtrend from ~$56 in February to $37, -47.9% off the 52-week high of $71.22 and only 8.3% above the 52-week low of $34.26. Daily chart shows a base attempt near $36-37 with a modest bounce from July lows, but SMA200 sits -25% away and SMA50 -3%, RSI 45 — no momentum reversal signal yet, just a stabilization. Support: $34.26 (52w low) and the $36 psychological/base level. Resistance: $40 (recent July high and prior breakdown level), then $45 (April congestion), then $50 (May breakdown). The model's forecast bands are aggressively bullish across all timeframes ($43.58 1h, $53.89 4h, $50.02 1d, $45.71 1wk), but realized directional accuracy on this name is 36% vs 64% naive — the model is BEATEN by naive on this name in this regime, so I discount those forecast levels heavily. The realistic technical read is a range trade $36-$42 until the Q3 print, with an insider sell ($2.7M by a director on 7/27) reinforcing the near-term overhang.

• AI-native Calix One platform + Agent Workforce Cloud adoption at three new rural BSPs (Conexon, CBS, Tombigbee) announced July 24 — validates land-and-expand model
• Record Q2 software & services revenue disclosed in the July 20 8-K — the mix shift is happening in the numbers, not just the narrative
• Sequential operating margin expansion 3.4% → 4.5% → 7.4% over three quarters — operating leverage inflection is measurable, not aspirational
• $431M buyback completed at depressed levels — capital return at ~$37 is accretive if fundamentals hold
• BEAD rural broadband subsidy disbursement pace remains a multi-year tailwind for BSP capex directed at Calix platforms

• Gross margin compression from DRAM/NAND memory costs: 57.7% → 54.6% in two quarters; further deterioration on Q3 would break the operating leverage narrative
• Insider sell of $2.7M by a director on 7/27 — poorly timed and adds to near-term overhang
• Next-year EPS growth estimate cut from 40% to 33.8% (-6.2pp) suggests sell-side is trimming despite constructive posture
• Model's directional accuracy on this name is 36% vs 64% naive — momentum signals from the AI forecast are unreliable in this regime
• Historical pattern: post-earnings margin-related sell-offs create persistent overhangs, not V-bounces; the $34 low may retest
• Trailing P/E of 48-49x leaves no cushion if operating leverage story stalls; short interest 10% signals crowded skepticism
• Customer concentration in rural BSPs dependent on BEAD subsidies — funding delays or policy changes are single-point-of-failure risks