Financebotresearch desk研究台

Evergreen Compounders长期复利公司

Manhattan Associates, Inc.

MANH 8

Software - Application · 11.6B market cap · site

Price and target zones价格与目标区间

$199.65
$180–195 entry买入$229.74+ full value估值上限 · 42× fwd × $5.4752wk52周 $119–$227-12% off the high距高点

What the company does公司做什么

Manhattan Associates makes computer software for warehouses, stores, and delivery networks. The software tells workers where to put each item, which order to pick next, and how to ship it out. Large companies such as retailers and manufacturers use it to run their supply chains — the steps that move goods from a factory to a customer's door. Most customers now rent the software over the internet, called "the cloud", instead of installing it on their own computers. Manhattan also sends staff to set up the software and train workers to use it.Manhattan Associates为仓库、门店和配送网络制作计算机软件。这套软件告诉工人把每一件商品放在哪里、接下来先拣哪个订单、以及怎么把货发出去。像零售商和制造商这样的大公司用它来管理供应链——也就是把货物从工厂运到顾客家门口的那些环节。现在大多数客户改成通过互联网租用这套软件,也就是所谓的“云”,而不是安装在自己的计算机上。Manhattan还派人去安装软件、培训工人使用。

Earns from营收来源 most money comes from recurring cloud subscription fees, plus one-time fees for setup and training.大部分钱来自反复收取的云订阅费,再加上一次性的安装和培训费。

Read the full description查看完整介绍

Manhattan Associates, Inc. develops, sells, deploys, services, and maintains software solutions to manage supply chains, inventory, and omni-channel operations. It offers warehouse management solution for managing goods and information across the distribution centers; Manhattan Active Warehouse Management, a cloud native and version less application for the associate; and transportation management solution for helping shippers navigate their way through the demands and meet customer service expectations at the lowest possible freight costs; Manhattan SCALE, a portfolio of logistics execution solution; and Manhattan Active Omni, which offers order management, store inventory and fulfillment, call center, POS, and customer engagement tools for enterprises and stores. The company also provides demand forecasting and replenishment, allocation, and unified business planning; technology platform including Manhattan Active Platform solutions, a cloud-native product designed to provide version-less product access; maintenance services, which offers on-premises software licensees with software upgrades for additional or improved functionality and technological advances; and professional services, such as solutions planning and implementation, and related consulting services. In addition, it provides training and change management services; and resells computer hardware, radio frequency terminal networks, radio frequency identification chip readers, bar code printers and scanners, and other peripherals. The company offers products through direct sales personnel and partnership agreements with various organizations. It serves retail, consumer goods, food and grocery, logistics service providers, industrial and wholesale, high technology and electronics, life sciences, and government industries. The company operates in the Americas, Europe, the Middle East, Africa, and the Asia Pacific. Manhattan Associates, Inc. was founded in 1990 and is headquartered in Atlanta, Georgia.

Revenue breakdown营收构成 share of revenue营收占比

47%Services服务
38%Cloud subscriptions云订阅
14%Maintenance and hardware维护与硬件
1%Software license软件许可
Pick a slice for detail点击某一部分查看详情
Services服务47% · ~$508.3M

Manhattan's own staff and partners install the software at a customer's site, move the customer's records into the new system, and train warehouse and store workers to use it. This work is a one-time project for each customer, so it does not repeat like a subscription fee. It also builds the close ties that keep a customer on Manhattan's software for years afterward.Manhattan自己的员工和合作伙伴在客户现场安装软件,将客户的记录迁移到新系统中,并培训仓库和门店员工使用软件。这项工作对每个客户来说都是一次性项目,不像订阅费那样重复发生。它也建立了紧密的联系,使客户在之后的多年里继续使用Manhattan的软件。

  • Implementation and consulting实施与咨询 — experts install the software and move a customer's existing records into the new system.专家安装软件,把客户现有的记录迁移到新系统里。
  • Training and change management培训与变革管理 — teaches warehouse and store staff how to use the new software.教仓库和门店的员工如何使用新软件。
Cloud subscriptions云订阅38% · $408.1M

This is Manhattan's core growth engine. Customers rent the company's supply-chain software over the internet and pay a recurring fee, instead of buying and running it on their own computers. The fee renews every year, so this revenue is predictable and grows as more customers move from older on-site systems to the cloud.这是Manhattan的核心增长引擎。客户通过互联网租用公司的供应链软件并支付定期费用,而不是自行购买和运行软件。这笔费用每年续约,因此这部分收入可预测,并随着更多客户从老旧的本地系统转向云端而增长。

  • Manhattan Active Warehouse Management — tells warehouse workers where to put each item and which order to pick and pack next.告诉仓库工人把每件物品放在哪里,以及接下来该拣选和打包哪一个订单。
  • Manhattan Active Omni — links a store's shelf stock with online orders, so a retailer can ship an order from the nearest store instead of a distant warehouse.把门店货架上的库存和网上订单连起来,这样零售商就能从最近的门店发货,而不用从很远的仓库发。
  • Manhattan Active Transportation Management — picks delivery routes and shipping companies to move goods at the lowest cost.挑选配送路线和承运公司,以最低成本把货物运出去。
Maintenance and hardware维护与硬件14% · ~$151.4M

Customers who still run Manhattan's software on their own computers, instead of over the internet, pay a yearly fee for updates and technical support. Manhattan also resells barcode scanners and other warehouse equipment that works with its software. This part of the business is shrinking as more customers move to the cloud.仍在自己的计算机上运行Manhattan软件、而非通过互联网使用软件的客户,需支付年费以获得更新和技术支持。Manhattan还转售条码扫描仪和其他与其软件配合使用的仓库设备。随着越来越多客户转向云端,这部分业务正在萎缩。

  • Software maintenance软件维护 — a yearly fee for software updates and technical help on systems installed at the customer's own site.针对安装在客户自己场地上的系统,收取每年的软件更新和技术支持费用。
  • Hardware resale硬件转售 — barcode scanners, handheld computers, and RFID readers (radio tags that track items), sold to work with the software.条形码扫描器、手持电脑和 RFID 读取器(追踪物品的无线标签),配套软件一起出售。
Software license软件许可1% · ~$10.8M

A customer pays once to install Manhattan's software on its own computers, instead of renting it over the internet. Almost every new customer now chooses the cloud version instead, so this one-time payment has nearly disappeared from Manhattan's revenue mix.客户一次性付费,将Manhattan的软件安装在自己的计算机上,而不是通过互联网租用。现在几乎所有新客户都选择云端版本,因此这笔一次性付款在Manhattan的收入构成中已几乎消失。

  • Perpetual license永久许可 — a one-time payment to install the software on the customer's own computers, instead of renting it online.一次性付款,把软件安装在客户自己的电脑上,而不是在线租用。

Company facts公司资料

Founded成立时间
1990
Headquarters总部
Atlanta, Georgia
Employees员工人数
4,100
Industry行业
Software - Application
Country国家
United States

Investment thesis投资逻辑 from the watchlist来自关注列表

⏳ DEFER to the 10/27 print — RE-ANALYZED 9/28, conviction 8.5→8.0. Wide/stable moat, net cash $132M, all Structural rows CARRIED — the cut is evidence freshness, not the business: the four claims that earned the 8.0→8.5 raise on 7/29 (bookings, RPO, Active Agents, win rate) are UNTESTED for a 2nd-3rd straight pass — two months, two passes, zero re-tests. Conviction must reflect the volatility of our own analysis, so it takes the haircut. WATCH was also too active a verb — it implies a considered "no" on the merits, when the truth is there is no call to make before Oct 27: a binary print in 29 days, a guide-punished season, one trigger 24bp from firing, and five rows untestable until then. [8.0] + DEFER is the consistent pair; neither number alone was the error.

What's working

  • Offsets that held the cut to one notch: FY26 consensus was revised UP $0.14 to $5.47 by five analysts inside the 11% drawdown; the margin drag is sized and identifiable growth spend — S&M +53.7% yoy while R&D was flat; and the predicted heavy insider selling at $220+ never appeared. FV $195-250 → $180-230 (mid $222 → $205) = 33-42x FY26 $5.47 / 29-37x FY27 $6.14; spot $199.63 = 36.5x FY26, 53rd percentile. Every input that moved, moved in the stock's favour — the band fell only because the assumption did (the 8/27 36-46x range was set into a rate-cut-hope tape; the Fed hiked 9/16). Corroborated from outside: SWS narrative FV $180 = exactly the new floor, DA Davidson $210 sits inside the band. Corrected EV $11.51B vs vendor $11.88B — three passes, three errors, two different signs: the field is unreliable in both directions, not "stale low." Ladder: 40% @ $180-195 35% @ $162-180 25% @ $140-160. Trim 42x fwd($5.47).

Watch

  • RETRACTED — and it was rendering wrong on the site for 32 days. The 46x fwd trim's valuation judgment (~$252 ceiling) was fine; the encoding was wrong. site.py computes EPS = price / forwardPE, and Yahoo's forwardPE (32.38) divides FY2027 EPS of $6.1432, not the FY26 guide-mid of $5.47 the baseline wrote against. The site rendered $284 against an intended $252 — +12.6%, silently. Cause: [[pitfall-multiple-trim-inherits-the-broken-vendor-field]], written 2026-08-12 — fifteen days before the baseline. The baseline caught the wrong-fiscal-year trap for its multiples table and missed it for the trim.
  • The $150-210 "zone" was never a zone — it was a frontmatter flattening artifact: the baseline's real structure was a three-tranche ladder ($195-210 / $175-195 / $150-170) collapsed into the union of all three. A 40%-wide band reports "in zone" at essentially any price — MANH has read as in-zone-and-green at $206, $210 and $223 against a verdict that said do not initiate. This is the mirror image of [[pitfall-stale-entry-zone-suppresses-a-name]]: that note records a zone set beyond reach suppressing a name; this is a zone set too wide silently promoting one. → [[pitfall-frontmatter-flattens-a-tranche-ladder-into-a-false-zone]] Fixed three ways: the ladder was re-derived from scratch, the frontmatter now carries T1 only ($180-195) with width cut 40% → 8.3%, and at $199.63 the name is correctly out of zone. 📉 Operating margin recomputed from the quarterly statements: 27.09 → 27.49 → 24.78 → 23.01 → 22.24% — four straight declines, 24bp above the <22% break trigger. The baseline called Q1's 23.0% "still the last data point on record" when Q2 was already filed. (Ex the $8.26M restructuring charge Q2 is 25.02%; the basis is pinned to GAAP as-reported, the conservative read — the baseline never stated a basis.) 📉 FY27 consensus $6.1432 = +12.3% on FY26 — the first quantified test of the "2027 inflection toward the 26% cloud rate," and it undershoots.
  • Caveats you should hold: the EDGAR full-text sweep returned HTTP 503, so "no 8-K / no Form 4 since Aug 11" is an aggregator-only negative finding — and it is currently counted as a positive force in the conviction table; re-verify at Q3. Today's tape is a thin print (~187k shares vs ~798k average), and the close is quoted at both $199.04 and $199.63. 🔬 Best unused question for Q3: billings four-quarter-average growth of ~6.5% against revenue +9.3% and RPO +23% — three different stories, gap unexplained. The source is an algorithmic screen so it moved nothing this pass, but if it verifies it is more sensitive than four of the five break triggers. Q3 FY26 print ~Oct 27 (date confirmed, after close).

⏳ 推迟到 10/27 财报——2026-09-28 重新分析,信心 8.5→8.0。 ** 宽阔且稳定的护城河,净现金 $132M,所有结构性主张都保留——这次下调是证据的新鲜度问题,不是业务问题:7/29 那次把信心从 8.0 提到 8.5 的四条主张(订单、RPO、Active Agents、中标率)已经连续第二到第三轮未经检验——两个月,两轮分析,零次复核。 ** 信心必须反映我们自己分析的波动性,所以它要挨这一刀。 「WATCH」这个动词也太主动了——它暗示这是一个基于基本面深思熟虑后的「不」,而事实是在 10 月 27 日之前根本没有可做的判断:29 天后是一个二元的财报,正处在一个「惩罚指引」的季节,一个触发条件距离触发只差 24bp,还有五条主张在财报前无法检验。 [8.0] + DEFER 才是自洽的组合;单独看哪个数字都不是错误所在。 ** ❌ 已撤回——而且它在网站上错误渲染了 32 天。 ** 46x fwd 这个减持位的估值判断(约 $252 的上限)本身没问题;错的是编码方式。 ** site.py 计算的是 EPS = 价格 / forwardPE,而雅虎的 forwardPE(32.38)除的是 FY2027 的每股收益 $6.1432,不是基准报告所依据的 FY26 指引中值 $5.47。 网站渲染出了 $284,而本意是 $252——高了 12.6%,而且是悄无声息的。 ** 原因:[[pitfall-multiple-trim-inherits-the-broken-vendor-field]],写于 2026-08-12——比基准报告早十五天。 ** 基准报告在它的估值倍数表里抓住了「错误财年」这个陷阱,却在减持位上漏掉了。 * ⚠️ 那个 $150-210 的「区间」从来就不是一个区间——它是一个 frontmatter 压平产生的假象:基准报告真正的结构是一个三段式阶梯($195-210 / $175-195 / $150-170),被压缩成了三段的并集。 这正是 [[pitfall-stale-entry-zone-suppresses-a-name]] 的镜像:那条笔记记录的是区间定得够不着从而压制了一个名字;这一条是区间定得太宽从而悄悄地推销了一个名字。 ** → [[pitfall-frontmatter-flattens-a-tranche-ladder-into-a-false-zone]] 已从三个方面修复: 阶梯从零重新推导,frontmatter 现在只装 T1($180-195),宽度从 40% 收到 8.3%,而在 $199.63,这个名字正确地显示为不在区间内。 📉 用季度报表重新计算的营业利润率:27.09 → 27.49 → 24.78 → 23.01 → 22.24%——连续四次下滑,距离 <22% 的触发条件只剩 24bp。 ** 基准报告把 Q1 的 23.0% 称为「仍然是记录在案的最后一个数据点」,而当时 Q2 已经报送了。 ** (剔除 $8.26M 的重组费用,Q2 是 25.02%;口径锁定为 GAAP 报告值,即更保守的读法——基准报告从未说明过口径。 ) 📉 FY27 一致预期 $6.1432 = 在 FY26 基础上 +12.3%——这是对「2027 年向 26% 的云业务增速拐头」这一说法的第一次量化检验,而它没达标。 ** ✅ 把下调控制在一档之内的对冲因素: FY26 一致预期被五位分析师在那 11% 的回撤期间上调了 $0.14 到 $5.47;利润率的拖累是有明确规模、可识别的增长投入——销售与市场费用同比 +53.7%,而研发费用持平;以及被预测会在 $220+ 出现的大量内部人抛售根本没有出现。 ** 合理价值 $195-250 → $180-230(中位数 $222 → $205)= FY26 的 $5.47 的 33-42 倍 / FY27 的 $6.14 的 29-37 倍;现价 $199.63 = FY26 的 36.5 倍,位于第 53 百分位。 每一个发生变化的输入,都是朝对股票有利的方向变的——区间下移只是因为假设下移了(8/27 那个 36-46 倍的区间是在一个期待降息的市场环境里设定的;而美联储在 9/16 加息了)。 从外部交叉验证:SWS 的叙事合理价值 $180 恰好等于新的下限,DA Davidson 的 $210 落在区间之内。 修正后的企业价值 $11.51B,对比供应商的 $11.88B——三轮分析,三次错误,而且是两个不同的方向:这个字段在两个方向上都不可靠,不是「偏低且陈旧」。 ** 阶梯:40% @ $180-195 35% @ $162-180 25% @ $140-160。 ** 减持 42x 前瞻(以 FY26 的 $5.47 为基准,约 $230)。 ** ⚠️ 需要留意的两条: EDGAR 全文检索返回了 HTTP 503,所以「自 8 月 11 日以来没有 8-K / 没有 Form 4」只是一个来自聚合器的否定性结论——而它目前在信心表里被当作一个正面因素计入;Q3 时需要重新核实。 今天的成交是一笔清淡的换手(约 18.7 万股,对比约 79.8 万股的均量)。 🔬 *Q3 最值得问、但本轮没用上的问题:订单额四季度平均增速约 6.5%,对比收入 +9.3% 和 RPO +23%——三个不同的故事,差距无法解释。 来源是一个算法筛选器,所以本轮没有据此改动任何结论,但如果它被证实,它比五个触发条件中的四个都更敏感。 **

Historical results历史业绩 past fiscal years — these values do not change过去财年——这些数值不会变动

Profit margins利润率 the share of revenue that stays as profit营收中留存为利润的比例

Gross profit毛利润
56%
Operating profit营业利润
25%
Net profit净利润
19%

of every $1 of revenue每 $1 营收中的占比

Return on equity净资产收益率
96%
Revenue growth营收增长
+9%

Competitors竞争对手

Blue YonderOracleSAP

Why customers stay客户为什么留下

once a warehouse trains its staff and connects its scanners and other systems to Manhattan's software, switching to a rival is slow and risky, so most customers renew instead of leaving.一旦一个仓库培训好了员工,把扫描枪和其他系统都连到了Manhattan的软件上,换到对手那里就又慢又有风险,所以大多数客户选择续约而不是离开。

Recent news近期新闻 last 7 days近7天

Sources来源