Agora screens as an asymmetric small-cap dislocation because the market is valuing the operating business at roughly $95-100M EV despite renewed double-digit growth, seven consecutive quarters of GAAP profitability, active buybacks, and a liquid-balance-sheet anchor. The setup is not clean quality: China/VIE governance, sub-scale conversational AI margins, and low trading liquidity keep the discount deserved. Base fair value is $6.65 per ADS, but the stricter net-current-asset floor is materially below the headline cash floor.
FY2027E revenue is estimated from Q1 2026 revenue $37.7M, Q2 $40.4M, Q3 guidance $41-42M, and a Q4 step-up estimate near $44M, then normalized to $190M for 2027. Sensitivity: +/-0.5x EV/Revenue moves FV by about +/-$1.13/ADS, or roughly 17% of base FV. The $6.65 base is 2.5% below the latest named BofA $6.80 target dated 2026-08-14 and 33% above Benzinga's one-analyst $5 feed, so target dispersion is an explicit risk. Not investment advice.
| Component | Assumption | USD/share |
|---|---|---|
| Core RTE + AI platform EV | $190M FY2027E revenue x 1.70x EV/Rev = $323M EV / 83.8M ADS | +3.85 |
| Net financial cash | ($361.7M cash/deposits/products - $100.3M borrowings) / 83.8M ADS | +3.12 |
| China/VIE governance haircut | -$29M risk reserve for VIE/PRC discount / 83.8M ADS | -0.35 |
| SBC, FX and ADS liquidity reserve | -$15M reserve for dilution/FX/liquidity friction / 83.8M ADS | -0.18 |
| Conversational AI option value | 20% probability x $92M incremental platform value / 83.8M ADS | +0.22 |
| FV base case | 3.85 + 3.12 - 0.35 - 0.18 + 0.22 = $6.66, rounded to $6.65 | 6.65 |
Insider/governance check: no material open-market insider sales were found in the latest Form 4 screens; recent filings are director RSU vesting/exercise events. The more relevant insider signal is CEO Tony Zhao's June 1, 2026 stated plan to buy up to $20M of ADSs or ordinary shares with personal funds over the following 12 months.
| Item | FY2023 | FY2024 | FY2025 | H1 2026 | Guidance / FY2026E |
|---|---|---|---|---|---|
| Total revenue ($M) | 141.5 | 133.3 | 141.1 | 78.2 | ~163-165 |
| Revenue growth | -11.9% | -5.9% | +5.9% | +15.8% | ~+16% |
| Gross margin | 63.2% | 64.1% | 66.4% | 63.5% | 63-65% |
| Net income ($M) | N/D | -53.5 | ~8.0 | 3.3 | positive |
| Operating cash flow ($M) | N/D | N/D | positive in Q4 | 3.6 | low positive |
| Cash/deposits/products ($M) | 216.7 | 269.7 | 374.9 | 361.7 | cash-rich |
| Metric | Q2 2025 | Q3 2025 | Q4 2025 | Q1 2026 | Q2 2026 |
|---|---|---|---|---|---|
| Revenue ($M) | 34.3 | 35.4 | 38.2 | 37.7 | 40.4 |
| Gross margin % | 66.8% | 66.0% | 65.1% | 63.4% | 63.7% |
| Net income ($M) | 1.5 | 2.7 | 4.9 | 1.1 | 2.2 |
| Cash/deposits/products ($M) | 377.3 | 374.3 | 374.9 | 366.1 | 361.7 |
Business model - global RTE platform plus China Shengwang
Agora global platform ~$90-95M FY2026E (55-58% rev) ramping Santa Clara-led international RTE and conversational AI business. Growth is driven by live shopping, financial services, customer service, and AI agent workloads. Main asset: low-latency infrastructure and developer APIs. Shengwang China ~$70-73M FY2026E (42-45% rev) discounted Shanghai-based China RTE PaaS brand. The business returned to growth in recent quarters but carries macro, regulation, currency and VIE-structure risk that justifies a lower multiple. Conversational AI early revenue, option value early scale Agent Studio and voice AI agents target call-center and workflow automation use cases. Adoption signals are encouraging, but unit economics remain early and gross margin is not yet back to 2025 levels.
Legal, regulatory and risk analysis
SWOT analysis
- +Cash/deposits/products of $361.7M exceed the quoted market cap at $4.06.
- +Seven consecutive quarters of GAAP profitability through Q2 2026.
- +Q2 revenue growth accelerated to 18.0% with DBNRR back above 100%.
- +Repurchase program has materially reduced ADS-equivalent share count.
- −Operating profit remains thin; cash interest helps reported net income.
- −Gross margin declined as AI products are still sub-scale.
- −China/Shengwang exposure limits multiple expansion.
- −Low trading volume can keep the discount persistent.
- →Q3 2026 guidance provides a near-term proof point for re-acceleration.
- →Conversational AI agents could reopen higher-growth infrastructure use cases.
- →Continued buybacks below book value can compound per-share value.
- →A small improvement in peer confidence can move EV/Revenue materially.
- !PRC policy, data rules or VIE sentiment can widen the discount quickly.
- !AI workloads may carry lower gross margin than legacy RTE workloads.
- !Buybacks consume cash if growth quality deteriorates.
- !Analyst support is thin and target feeds conflict.
Summary by assessment area
- Large liquid balance, low financial debt, and accretive buybacks.
- Strict downside floor is lower than headline cash, so use $2.61-$3.12 as the risk band.
- Revenue is growing again, but margin mix must improve.
- Q3 2026 is the next clean checkpoint for the AI/RTE thesis.
- Base FV $6.65 offers +64% upside from $4.06.
- VIE discount and low liquidity keep this a dislocation, not a simple compounder.
Sources: Agora investor relations Q2 2026 release and SEC 6-K dated 2026-08-13; Agora Q1 2026 release dated 2026-05-26; Agora FY2025/Q4 release dated 2026-03-02; Agora Q3 2025 release dated 2025-11-19; SEC 20-F FY2025; StockAnalysis market cap and financial overview; Investing.com historical price page; MarketBeat and Finshort short-interest data; PriceTargets, Benzinga, ChartMill and TipRanks analyst target feeds; Agora management share purchase plan dated 2026-06-01. Market data used for valuation: API close $4.06 on 2026-09-11, market cap about $340M, 52-week range $3.14-$5.495, 83.8M ADS equivalents at 2026-06-30, short interest 0.61% as of 2026-08-31. This document is for informational purposes only and does not constitute financial or investment advice.