{
  "@context": "https://frontierpicks.com/schemas/dossier.v1.json",
  "ticker": "BAND",
  "name": "Bandwidth Inc.",
  "url": "https://frontierpicks.com/dossiers/BAND/",
  "json_url": "https://frontierpicks.com/dossiers/BAND.json",
  "status": "DORMANT",
  "current_conviction": "LOW",
  "graded_conviction": "LOW",
  "archetype": {
    "code": "a6",
    "n": 6
  },
  "current_thesis": "Late-stage small-cap CPaaS squeeze: +360% YoY into the mid-$70s, ~5% below the $79.08 high, with Citizens ($86, 2026-07-20). Riley ($85) both lifting targets in July. A confirmed Q2 print on 2026-07-29 is a binary two trading days out — the tape is extended into the catalyst, not based under it.",
  "invalidation_trigger": "A weekly close below $60 breaks the June momentum leg and opens mean-reversion toward the ~$53 pre-convert base; a 2026-07-29 Q2 print that misses on revenue or EBITDA guidance and gaps the tape lower confirms the stand-aside.",
  "catalyst_date": "2026-09-15",
  "outcome": "INVALIDATED",
  "outcome_date": "2026-07-23",
  "invalidation_fired": true,
  "themes": [
    "gpu-cloud-neoclouds",
    "ai-datacenter-infrastructure",
    "small-cap-value-rotation"
  ],
  "tags": [],
  "sources": [],
  "notes": [
    "2026-06-16: priced $275M 0% convertible senior notes due 2032 (announced 6/15, stock lower pre-market). Proceeds for buybacks + debt repayment — 'debt repayment' likely refinances the 2026 convertible maturity wall. Watch follow-on 8-K (~T+5) for capped-call / note-repurchase + share-count impact. Convert-arb delta shorting adds near-term supply.",
    "Dual-class structure: the buyback authorization applies to Class A shares; Class B carries superior voting rights.",
    "At 2026-06-30 the balance sheet carried $169.9M of cash against $330.8M of convertible senior notes.",
    "The $275M 0% convertible priced 2026-06-16 carries a $72.64 conversion price, well above the 2026-08-14 close of $53.97.",
    "Sector classifiers often file BAND under networking/optical. It is CPaaS/cloud communications; the comparison set is TWLO, RNG, EGHT and Sinch.",
    "No AI-attached revenue line is disclosed. Voice AI and Maestro attach are management commentary, not a reported segment."
  ],
  "body_markdown": "## Current Thesis\nThe leg that ran from April to July is finished, and it ended on the catalyst rather than drifting out. Bandwidth reported Q2 2026 before the open on 2026-07-29: revenue $219.9M, +22% YoY, ahead of the ~$217M consensus figure circulated post-print; adjusted EBITDA $27.8M, +27% YoY at a record ~18% margin; free cash flow $23.7M; net income $2.4M against a $4.9M loss a year earlier. Full-year guidance went up to $900–910M revenue, $123–125M adjusted EBITDA and $1.71–1.79 non-GAAP EPS, with Q3 guided to $231–235M against a ~$224M Street number. The stock fell roughly a quarter anyway — StockStory headlined a 26% drop, Seeking Alpha described a 28% crash, ChartMill 22% — because the beat leaned on low-margin pass-through charges (gross margin 36% versus 40% a year earlier) and management flagged decelerating second-half topline. The prior published break level, a weekly close below $60, fired that week. What remains at the 2026-08-14 close of $53.97, 31.2% under the $78.44 52-week high with RSI(14) at 46.2, is a company guiding to ~$900M of revenue and a Voice-AI-attach story that has no disclosed revenue line, and no binary event to resolve it before the ~November Q3 print.\n\n## Bull Case\n- **2026-07-29: FY26 guidance raised on all three lines** — revenue to $900–910M, adjusted EBITDA to $123–125M, non-GAAP EPS to $1.71–1.79. The raise happened into the gap down, so the fundamental trajectory and the price trajectory separated on a dated, checkable event.\n- **Q3 guide of $231–235M sits above the ~$224M consensus** cited in post-print summaries. Management set a bar above the Street while simultaneously warning on deceleration.\n- **Operating leverage is showing in cash, not just adjusted metrics:** $23.7M free cash flow and a record ~18% adjusted EBITDA margin in Q2 2026, with GAAP net income positive at $2.4M for the first time against the year-ago $4.9M loss.\n- **Capital return was widened rather than paused:** $20M of Class A stock repurchased in H1 2026, and the credit agreement amended to lift the restricted-payments basket from $20M to $40M or 20% of trailing-twelve-month EBITDA.\n- **The washout was fully retraced.** StockStory reported a $38.40 pre-market print on 2026-07-29 against the 2026-07-28 close of $52.25; the 2026-08-14 close of $53.97 is above the pre-print reference. Eleven sessions absorbed a 26% gap.\n- **Street targets did not follow the price down:** B. Riley reiterated Buy with an $85 target after the print, and the five-analyst median sits at $73 with a $38–$86 range and a 4 Buy / 1 Hold split.\n\n## Bear Case\n- **Gross margin fell 400bp YoY to 36%.** Revenue grew 22% while the mix shifted toward pass-through and surcharge dollars that carry no margin — the reason a beat-and-raise was sold.\n- **Management guided the second half slower.** The deceleration commentary on the 2026-07-29 call is the specific item the tape reacted to, and nothing between now and the Q3 print revises it.\n- **Voice AI is commentary, not a segment.** Management tied every $1M+ customer win in the quarter to the Maestro orchestration platform, but no AI-attached revenue line is disclosed in the release or the 8-K. The re-rating case rests on an undisclosed number.\n- **Leverage against the cash pile:** $330.8M of convertible senior notes versus $169.9M of cash at 2026-06-30. The $275M 0% convert priced 2026-06-16 carries a $72.64 conversion price — far above the current tape, so the dilution cap that framed June has become a distant strike and the notes are now straight debt in economic terms.\n- **The 52-week distribution is wide.** A $78.44 high, a 26% single-session gap, and a full retrace inside three weeks describe a name whose price is set by positioning in a small float, not by a discounted stream of $900M revenue.\n- **Target dispersion is the widest it has been in this cycle:** $38 to $86 across five analysts. That spread is a statement that the covering desks do not agree on what the business is worth.\n\n## Setup & Price Structure\nReference close 2026-08-14: $53.97. 52-week high $78.44 (-31.2%); three-month return 0.0%; RSI(14) 46.2 — mid-range, no momentum signature in either direction. The structure that matters is a repeatedly-visited shelf at $52–54: the 2026-06-15 close of $52.83 that set the convert's $72.64 strike at a 37.5% premium, the 2026-07-28 pre-print close of $52.25, and the current $53.97. Price has now tested that band from above and below and held it, which is a different situation from July, when the tape was vertical into a binary with nothing beneath it.\n\nAbove, the first real resistance is the $60 area the momentum leg lost on the print week, then the $72.64 conversion price. Below, the 2026-07-29 pre-market print of $38.40 reported by StockStory marks the washout zone; it is a pre-market reference, not a settled daily low, and should be treated as approximate.\n\nThe narrative is **dead**, dated 2026-07-29. The narrative an investor was buying in June and July — a reflexive small-float advance into a print, with sell-side targets chasing from $27 in April to $85–86 in July — failed on the event itself, and the published break level went with it. What is forming underneath is a different, second-order proposition (voice infrastructure consumed by AI agents), and it has not earned a label: there is no disclosed AI revenue line, no higher-low sequence beyond three weeks, and no catalyst that could produce either before roughly November.\n\nCrowding and positioning, as observables rather than verdicts: retail-sentiment coverage clustered in mid-May (Benzinga's 2026-05-09 \"Stock Whisper\" Index, the 2026-05-14 small-cap list built around a +381% mover) and the sell-side chorus peaked in July with Citizens at $86 on 2026-07-20 and B. Riley at $85 on 2026-07-09 — both dated before the print that took the stock to $38 pre-market. The convert priced on 2026-06-16 issued $275M of paper into strength at a $52.83 reference close. No Form 4 activity appears in the current filing set. Price sits at neither an extended premium to nor a capitulation discount from its recent range, and no earnings date falls inside 30 days.\n\n## Catalyst Calendar (next 30 days)\n- **2026-09-15** — management presents at the Piper Sandler Growth Frontiers Conference in Nashville (webcast plus one-on-one and small-group investor meetings). The only scheduled public forum between the gap and the Q3 print.\n- **2026-09-30** — Q3 2026 quarter end. Nothing is disclosed on the date itself, but it closes the period the $231–235M guide covers.\n- **~2026-11-04 (est.)** — Q3 2026 results. Date not yet confirmed by the company; prior 2026 prints landed 2026-04-30 (Q1) and 2026-07-29 (Q2).\n\n## What Would Change Our Mind\nThe $52–54 shelf is the whole structural argument right now — it is the level that held on 2026-06-15, on 2026-07-28 and again on 2026-08-14, and the reason the post-print gap reads as absorbed rather than ongoing. Lose it and the read has nothing left: a weekly close below $49 forfeits that band and puts the 2026-07-29 washout zone back in play, with no scheduled company event to interrupt the move before roughly November.\n\nTwo secondary conditions would break the constructive half of the case independently of price. First, a Q3 print at or below the $231M low end of guidance, or an implied Q4 that drags full-year revenue under the $900M floor, would confirm that the deceleration management flagged on 2026-07-29 is steeper than the guide. Second, a Q3 gross margin at or below the 36% recorded in Q2 would establish pass-through mix as a structural drag rather than a quarter's noise.\n\nWhat would flip the read the other way is narrow and specific: a disclosed AI-attached or Maestro revenue figure — at the 2026-09-15 conference or in the Q3 materials — that is large enough to be modelled, alongside a reclaim and hold of the $60 shelf lost on the print week. Absent that number, the second-order AI framing stays an assertion.\n\n## Correlation Notes\n- Correct comparison set is CPaaS and cloud communications: TWLO, RNG, EGHT and Sinch. Sector classifiers that file BAND under networking or optical (LITE, COHR, AAOI) produce misleading correlations — the revenue is voice, messaging and 911 APIs.\n- Twilio's de-rating from 30x+ EV/S in 2021 to low-single-digit EV/S is the group's multiple anchor, and A2P price competition from Sinch and Infobip plus carrier-fee pass-through is the mechanism that keeps it there. BAND's own margin print of 36% in Q2 2026 is the same pass-through economics showing up in the P&L.\n- Small-float momentum unwinds cluster: the mid-May Russell 2000 momentum-basket coverage that first put BAND in front of retail also carried the names that unwound alongside it. Correlation to IWM and momentum factor exposure is likely higher than correlation to the company's own fundamentals over multi-week windows.\n- Convertible arbitrage links the equity to the $275M 0% notes due 2032. With the stock far below the $72.64 strike, hedge delta is low, so the mechanical short-into-strength supply that capped June and July is largely absent at these levels — an inference from the structure, not a disclosed flow.",
  "first_seen": "2026-04-19",
  "last_analyzed": "2026-08-16T14:24:41+00:00",
  "last_synthesized": "2026-08-16",
  "last_update_source": "watchlist_research",
  "license": "Content © FrontierPicks. Cite the canonical URL."
}