G
G 6.5Price and target zones价格与目标区间
In the buy zone在买入区间内
What the company does公司做什么
No company description is on file. Run python .mcp/profiles.py to get one.
Investment thesis投资逻辑 from the watchlist来自关注列表
⭐ IN ZONE $30-36 — ANALYZED 8/11, first file ever. ACCUMULATE. Passes this sleeve's test cleanly: a specific problem with a metric that closes the gap. The problem is that the market prices Genpact for the terminal decline of seat-based BPO at 7.9x forward / 8.0x FCF (12.6% FCF yield); the metric is revenue growth against average headcount growth, and it has now been positive for two quarters. 🔍 Q2 2026: revenue +7.1% on average headcount −2.6% — productivity index 1.04. Through FY25 it ran the other way (revenue/employee $36.3k→$34.8k). In the same fortnight Infosys grew 2.4% and CUT guidance. That divergence is the only hard test in this cohort → [[pattern-headcount-revenue-divergence-tests-ai-deflation]].
What's working
- ATS $363M +24.1% (27% of revenue), FY26 ATS target raised to ≥25%; largest-ever quarterly bookings, six $50M+ deals in Q2 vs 6 in all of H1'25; agentic TCV >$200M in 2025 → >$1B targeted 2026, over half from new clients
- 3yr fundamentals are genuinely good, not an artifact: operating income +36.7% on +16.2% revenue (real leverage, unlike AKAM or DBX), FCF +23% CAGR $389M→$724M, OCF conversion 81%→107% of operating income, gross margin 35.2%→36.5% and rising — the one quantitative signal that separates moat erosion from moat intact, pointing the right way. ROIC 13.9%, ROE 22.4%. Net debt $688M = 0.95x one year's FCF, interest covered 10.3x
- ~7.2% shareholder yield: nine straight dividend raises (~13%/yr, FY26 $0.75) at a 17.5% FCF payout — the dividend survives a 75% FCF cut — plus ~5%/yr buyback, and they bought 1.20M shares at $38.15 in Jan-Feb, cheaper as it fell
- Client concentration beats the "GE captive" reputation badly: top-5 15.2%, top-10 23.4%, no client >10% of receivables, GE no longer disclosed at all. ⚖️ Graham IV $34.13 against a $34.09 tape — a four-cent gap. The only name in the 8/11 batch where Graham lands on the market price, and it is Graham's natural habitat. Nothing is being paid for the future. 🔍 Reverse-DCF: EV $6.63B on $724M FCF prices −2 to −3% FCF decline in perpetuity, in a business compounding FCF at 23%. That gap is the trade
Red flags
- The bear case is contractual, not narrative — the 10-K states "some of our MSAs and SOWs require certain minimum productivity benefits to be passed on to our clients." AI efficiency can be revenue-negative by contract. Management is candid about it, which is worth something against Infosys refusing to quantify
- Core Business Services is 73% of revenue growing 1.9% and DECELERATING from 3.8%. ATS cannot carry the company if the core turns negative
- 24% growth in 27% of the business bought only 25bps of FY26 margin — if agentic work is not structurally higher-margin the mix shift is a treadmill
- Every insider is a seller with zero open-market buys; Kalra is 13 sells / 0 buys in five years — the loudest negative and the hardest to explain at 7.9x. ⚖️ Nalanda (Pulak Prasad) added 1.02M shares in Feb — a concentrated quality buyer, and the more informed of the two groups
- Industry BPO ACV −14% in 2025; WNS taken out by Capgemini at $3.3B; JPM Underweight, Citi Neutral, ~−30% YTD
Watch
- Two quarters is not a trend — that is exactly why 6.5 and not 8.
- FY23 net income $631M sits at operating income $627M — a tax item. Use operating income and FCF for the trend, never net income. FV $39-48 (central $43). Entry $30-36, in zone now. Trim 12x fwd.
- Sleeve rule applies — exit when the gap closes.
- Cluster discipline: G / INFY / DBX / AKAM are one trade in four costumes, all cheap on the same AI-displacement repricing, all resolving on the same question. Own G from that cluster, not four. Also correlated with the held ACN / CTSH / EPAM / IBM — see Concentration Map.
Would upgrade on
- Upgrade: four consecutive quarters of the divergence → 7.5 any insider open-market purchase a take-private bid.
⭐ 在 $30-36 买入区间内 — 8/11 已分析，史上第一份文件。 ACCUMULATE（逐步买入）。 ** Genpact，业务流程外包与数据/AI 服务。 它干净地通过了这个板块的检验标准：一个具体的问题，加上一个能把估值差距关上的指标。 问题是市场按「按人头计价的 BPO 模式终将消亡」来定价，给出 7.9x 前瞻市盈率 / 8.0x 自由现金流（12.6% 自由现金流收益率）；指标是收入增速与平均员工人数增速的对比，而它已经连续两个季度为正。 🔍 2026 年 Q2：收入 +7.1%，平均员工人数 −2.6% — 管理层的「生产力指数」为 1.04。 整个 FY25 期间这个数字是反着走的（人均收入 $36.3k→$34.8k）。 在同一个两周内，Infosys 收入只增长 2.4% 并且下调了指引。 这个背离是整个行业里唯一的硬性检验 → [[pattern-headcount-revenue-divergence-tests-ai-deflation]]。 ✅ ATS 板块 $363M，+24.1%（占收入 27%），FY26 的 ATS 目标上调至 ≥25%；史上最大单季预订额，Q2 有六笔 $50M+ 的合同，而 2025 上半年全期只有 6 笔；智能体（agentic）合同总价值 2025 年超过 $200M，2026 年目标 >$1B，其中一半以上来自新客户。 ✅ 三年基本面是真的好，不是会计假象：营业利润 +36.7%，而收入只增长 +16.2%（真正的经营杠杆，不像 AKAM 或 DBX），自由现金流年复合增速 +23%，从 $389M 到 $724M，经营现金流对营业利润的转化率从 81% 升到 107%，毛利率 35.2%→36.5% 并在上升 — 这是区分护城河侵蚀与护城河完好的唯一量化信号，而它指向正确的方向。 ROIC 13.9%，ROE 22.4%。 净负债 $688M，等于 0.95 年的自由现金流，利息覆盖率 10.3x。 ✅ 约 7.2% 的股东回报率：连续九年提高股息（年均约 13%，FY26 为 $0.75），派息只占自由现金流的 17.5% — 就算自由现金流砍掉 75%，股息依然安全 — 外加每年约 5% 的回购，而且他们在 1-2 月以 $38.15 买入了 120 万股，股价越跌买得越便宜。 ✅ 客户集中度远好过它「GE 附属公司」的名声：前五大客户占 15.2%，前十大占 23.4%，没有任何客户超过应收账款的 10%，GE 已经完全不再被单独披露。 ⚖️ 格雷厄姆内在价值 $34.13，对上 $34.09 的股价 — 相差四美分。 这是 8/11 这批股票中唯一一个格雷厄姆模型正好落在市场价上的名字，而且这正是该模型最擅长的场景。 市场对未来一分钱都没有支付。 🔍 反向 DCF：企业价值 $6.63B 对上 $724M 自由现金流，隐含永续每年 −2% 到 −3% 的自由现金流衰退，而这家公司的自由现金流正以 23% 的速度复合增长。 这个差距就是这笔交易。 🚩 看空理由是写进合同的，不只是叙事 — 10-K 明文写着「我们的部分 MSA 和 SOW 要求把一定的最低生产力收益让渡给客户」。 AI 带来的效率可能按合同就是收入负面的。 管理层对此直言不讳，相比 Infosys 拒绝量化，这一点是有价值的。 🚩 核心业务服务（CBS）占收入 73%，增速只有 1.9%，而且从 3.8% 在减速。 如果核心转负，ATS 撑不住整家公司。 🚩 占收入 27% 的业务增长 24%，却只买来 25 个基点的 FY26 利润率扩张 — 如果智能体业务在结构上并不是更高毛利，这个业务结构转换就是一台跑步机。 🚩 所有内部人都在卖出，没有任何公开市场买入；Kalra 五年内 13 次卖出、0 次买入 — 这是最响亮的负面信号，也是在 7.9x 估值下最难解释的一点。 ⚖️ Nalanda（Pulak Prasad）在 2 月增持了 102 万股 — 一个集中持股、看重质量的长期买家，在两群人里是更了解情况的那一方。 🚩 行业 BPO 年度合同价值 2025 年 −14%；WNS 被 Capgemini 以 $3.3B 收购；摩根大通给「减持」，花旗给「中性」，年初至今约 −30%。 ⚠️ 两个季度不构成趋势 — 这正是为什么信心分是 6.5 而不是 8。 ⚠️ FY23 净利润 $631M 几乎等于营业利润 $627M — 那是一个税务项目。 判断趋势请用营业利润和自由现金流，绝不要用净利润。 合理价值 $39-48（中枢 $43）。 买入区间 $30-36，现在就在区间内。 减持 12x 前瞻市盈率。 ⚠️ 板块规则适用 — 估值差距关上就退出。 ⚠️ 集群纪律：G / INFY / DBX / AKAM 是同一笔交易的四套衣服，全都因为同一个「AI 取代」重新定价而便宜，全都取决于同一个问题。 从这个集群里只拿 G，不要拿四个。 同时它也与已持有的 ACN / CTSH / EPAM / IBM 相关 — 见集中度地图。 **
Competitive advantages竞争优势 from the thesis and the reports来自投资逻辑和报告
- Margin direction is the one quantitative signal that separates moat erosion from moat intact, and here it points the right way.
- But the 10-K states MSAs "in most cases do not specify pricing terms or obligate the client to purchase a particular amount" — contract length overstates lock-in .
- The agentic pivot and the rebrand are his.
- Moat replies that the 10-K contains a clause requiring productivity give-backs on some contracts, and that every digitised process makes the "context moat" more portable.
- Fundamentals owns the last eight quarters; Moat owns the terminal question.