Richtech Robotics
This is the archived full research. Read the story
A tiny Las Vegas maker of serving, delivery and cleaning robots, sitting on far more cash than its shares are worth, and run by the family that controls the votes.
The scores
Growth2/4 →
Revenue up 9.6% in nine months, but FY2025's $5.05M was below FY2023's $8.76M and product sales fell 45%.
Business quality1/4 ↓
No moat: cheaper Chinese robots lead shipments, contract makers build for rivals, and gross margin halved in two years.
Financial position4/4 ↑
$339.6M of cash and investments, no debt, and operations using about $18M a year before interest income.
Execution record1/4 ↓
Three years restated, two late reports, a Nasdaq notice, a lawsuit over a release, and revenue below 2023's.
Per share capture1/4 ↓
Shares more than tripled in under three years, a plan that can add 18% a year, and a founder with voting control.
Potential2/4 →
About 67,000 hospitality and cleaning robots sell a year, but Richtech's signed orders are one 200-robot deal and an undisclosed dealer rollout.
Risk profile2/4 →
Survival is not at risk, but the cash can leak through spending, new shares and a controlling founder's choices.
Valuation4/4 ↑
The shares trade at 0.45 times book value (what it owns minus what it owes), so the price pays nothing for the business.
Evidence quality2/4 →
The cash is filed, but the accounts were restated, controls are weak, customers are unnamed and there is no earnings call.
If it works 0.3× · if it fails you keep about 100%.
The Verdict
We rate Richtech Robotics Neutral, with low conviction. The business is weak: sales of about $5M a year have shrunk since 2023, the profit on each sale has halved, and the latest accounts had to be restated. The price is cheap, because the shares are worth about half the cash in the bank, so buyers pay nothing for the robots. We do not call that a bargain, because the founder's family controls the votes, a share plan can add shares every year, and nobody can make the company hand the cash back. The bet is safe but pays little: holders keep the cash if the robots fail.
The long caseIf rising wages keep pushing restaurants, stores and dealers to rent robots, Richtech's cleaning and delivery robots grow from a small base while its cash pays for the wait.
How we weighed it
The price against our value
In its favour
No fair value is drawn because the business makes no lasting profit; the shares trade at about half the cash and investments, so the business is priced below zero.
The business
Against it
Weak
A weak business: cheaper Chinese rivals lead shipments, the robots are built by contract makers, and gross margin halved.
Growth against what the price assumes
Against it
Sales are below their level of three years ago and rest on one large cleaning robot order and an undisclosed dealer rollout.
What could go wrong
Against it
Restated accounts, weak controls, a class action, a founder with voting control and a plan that can add shares every year.
Against competitors
Against it
Pudu, Keenon and an LG-backed Bear offer similar robots, often for less, at far larger volumes.
How its group is behaving
Cuts both ways
Robot and AI shares swing together on headlines, which lifted this one twice and dropped it twice.
The latest news and report
Cuts both ways
A buyback and a large cleaning-robot order arrived, but the latest reports came late after a restatement.
The evidence under this - 6 claims
Cash has been spent since June on operations and possibly buybacks, so the September figure will be lower than $339.6M.
The figures sit under a material weakness in financial controls not yet remediated, and three earlier years of statements had to be restated.
FY2025's rise came partly from one-off orders for older delivery robots, which the 10-K/A says do not reflect the long-term trend.
Management says rentals spread the robot's cost over the contract, so the quarterly margin swings with the mix rather than with pricing.
The board may add fewer shares than the 18% ceiling, and the 2026 awards were far below it.
The company
Richtech sells and rents service robots: Matradee carries food to tables, Titan hauls parts around car dealerships, DUST-E scrubs floors in stores, and ADAM is a two-armed robot that makes coffee and drinks. It also sells robot training data and is building a wheeled humanoid, Dex.
Its buyers are restaurants, hotels, casinos, retailers and dealers short of staff. Without it they hire people or buy a similar robot from Pudu, Keenon or Bear, usually for less. Most robots are made for it by contract makers in China.
The evidence under this - 3 claims
How it makes money
Mostly by renting. In the nine months to June 2026 it earned $1.5M renting robots for events, $1.1M from robots on monthly contracts, $0.9M from selling robots and $0.5M from other work. A rental pays a little each month, so sales grow slowly even when robots go out, and customers can stop renting when the contract ends.
The evidence under this - 3 claims
Product sales fell 45% in the nine months, so growth came from rentals that recognise revenue slowly; the analysts' one estimate put Q3 at $1.55M against $1.37M reported.
The dealer group is unnamed and rental revenue of $0.69M across 55 contracts is about $12,600 a contract a year.
FY2025's rise came partly from one-off orders for older delivery robots, which the 10-K/A says do not reflect the long-term trend.
Where the money comes from, stream by stream
| Stream | Share of sales (nine months to June 2026) | How it earns it |
|---|---|---|
| Event rentals | 38% | Robots hired out with staff for trade shows, parties and promotions. |
| Robots on monthly contracts | 27% | Serving, delivery and cleaning robots rented with service, paid monthly. |
| Product sales | 22% | Robots and parts sold outright. |
| Other, including the cafes | 12% | Food and drink sold at Alphamax cafes and other services. |
Who runs it
Founder Wayne Huang has run Richtech since 2016 with his brother Michael as finance chief. Thanks to shares with ten votes each, Wayne holds about 52% of the votes and Michael about 14%, so outside holders cannot change the board.
Pay comes mostly as share awards, given every year, not tied to profit. No insider bought shares in the market last year.
The evidence under this - 3 claims
Why it wins, and who could beat it
Why customers come back
There is no clear edge. Richtech says it is an early brand with a broad range, but rivals sell similar robots, often cheaper, and the profit on each sale fell from about two thirds in 2023 to about a third in 2025.
What would show an edge is customers paying more each month for its service while that profit holds above half. Rising pay for hospitality workers helps every robot maker, not this one alone.
The evidence under this - 4 claims
Management says rentals spread the robot's cost over the contract, so the quarterly margin swings with the mix rather than with pricing.
The figure is Keenon's own release of a paid study, not IDC's publication.
Wage growth of about 4% a year is steady rather than accelerating, and hospitality robot sales fell 11% in 2024 regardless.
The threat
Pudu and Keenon, two Chinese makers, ship far more serving and delivery robots; Keenon says it holds about a fifth of the world market. LG Electronics now controls Bear Robotics, maker of the Servi serving robot, and Tennant sells robotic floor scrubbers through a large sales force.
The evidence under this - 4 claims
The figure is Keenon's own release of a paid study, not IDC's publication.
The rivals: TNC, and 4 more that are private, in-house or listed abroad.
Who else does this, rival by rival
| Ticker | Rival | Where it competes | How far along | Priced against it |
|---|---|---|---|---|
| Private | Pudu RoboticsPriced against it: Private | Service and industrial robots, rented, sold and run | On sale | Private |
| Private | Keenon RoboticsPriced against it: Private | Service and industrial robots, rented, sold and run | On sale | Private |
| Bear Robotics, controlled by LG ElectronicsPriced against it: Listed abroad | Service and industrial robots, rented, sold and run | On sale | Listed abroad | |
| TNC | Tennant autonomous scrubbers and AvidbotsPriced against it: One segment only | Service and industrial robots, rented, sold and run | On sale | One segment only |
| Private | Miso RoboticsPriced against it: Private | Service and industrial robots, rented, sold and run | On sale | Private |
How big it could get
How big this could get
The market is real but modest. The International Federation of Robotics counted almost 200,000 professional service robots sold in 2024: over 42,000 for hospitality, down 11%, and over 25,000 for cleaning, up 34%.
Richtech's sales are roughly one robot in every 300 of those, if a robot earns about $25,000 a year; the price per robot is not disclosed, so that share is a rough guess.
The evidence under this - 3 claims
The IFR counts a sample of 294 suppliers and does not project to the whole industry.
Event rentals and restaurant sales are not robots sold, so the comparison overstates the robot count.
The dealer group is unnamed and rental revenue of $0.69M across 55 contracts is about $12,600 a contract a year.
Who has signed
A national retailer signed a two-year service agreement in 2025 and in September 2026 agreed to deploy 200 DUST-E cleaning robots over five years; the value was not disclosed. A top-five US car dealer group approved Titan robots across its stores. Neither sets a firm amount the buyer must pay.
The evidence under this - 2 claims
The retailer is not named and no price is given; 200 robots over five years is about 40 a year.
The dealer group is unnamed and rental revenue of $0.69M across 55 contracts is about $12,600 a contract a year.
The results
The numbers
Sales grew 19% in the year to Sep 2025, faster than the −52% in the year before.
Operating margin is −341%, down from −6.2% in the year to Sep 2022.
Running costs take 377¢ of each sales dollar, up from 72¢ in the year to Sep 2022.
Shareholders' money lost 34% a year, down from −17% in the year to Sep 2022.
It used $14.1M of cash in the latest year, so the $252M it holds lasts about 17.9 years at that pace.
It holds $252M of cash against $118K of debt, $252M more cash than it owes.
The share count rose 205% in 3.0 years, and pay in shares costs 74% of sales.
All the numbers
Filed figures from the company's 10-K and 10-Q, read 2026-11-17. Fiscal years end in September.
| Sep 2022 | Sep 2023 | Sep 2024 | Sep 2025 | |
|---|---|---|---|---|
| Size and growth | ||||
| Sales | $6.05M | $8.76M | $4.24M | $5.04M |
| Sales growth | - | +45% | −52% | +19% |
| Profit | ||||
| Gross margin | 65% | 69% | 64% | 36% |
| Operating margin | −6.2% | 3.3% | −183% | −341% |
| Net margin | −8.4% | −3.9% | −32% | −973% |
| What growth costs, per sales dollar | ||||
| Research and development | 29% | 23% | 62% | 65% |
| Sales and marketing | 4.9% | 2.7% | 31% | 15% |
| General and admin | 37% | 40% | 154% | 297% |
| Pay in shares | - | - | - | 74% |
| Capital spending | - | - | 20% | 102% |
| Cash | ||||
| Cash from operations | −$2.65M | −$2.90M | −$4.94M | −$8.91M |
| Free cash flow | - | - | −$5.79M | −$14.1M |
| Balance sheet | ||||
| Cash and short-term investments | $327K | $433K | $30.5M | $252M |
| Debt | - | $845K | $53.0K | $118K |
| Net cash | - | −$412K | $30.5M | $252M |
| Returns and shares | ||||
| Shares (diluted) | 40.0M | 62.2M | 70.0M | 122M |
What growth costs
Growth is costly and has not paid back yet. Spending on staff, research and share awards runs several times sales, the profit on each sale has narrowed as cheaper rentals replaced sales, and the new building adds cost before it adds revenue.
The evidence under this - 3 claims
Management says rentals spread the robot's cost over the contract, so the quarterly margin swings with the mix rather than with pricing.
The last report (2 Sep)
Quarter ended June 2026, filed figures.
In the quarter to June 2026, the third of its fiscal year, sales rose 16.7% to $1.37M and the profit on each sale recovered to 58%, as rentals grew and product sales shrank. The operating loss was large because of share awards and a software write-off.
Interest on the cash, $9.8M over nine months, earned far more than the robots did, which kept the cash used by the business close to zero. The report itself came late, after the restatement.
What it did to the thesisIt changed little. Rentals grew and the profit on each sale recovered, which supports the idea of a service business, but sales are still tiny, the interest on the cash did the earning, and the late, restated accounts kept the doubt about the numbers alive.
The quarter in detail
Sales growth from the previous quarter, filed figures. The company has restated its own history, so a quarter either side of the restatement may describe a different business; n/c marks one that does not compare.
Earnings per share $-0.04 against $-0.05 expected; $-0.04 as filed, with no adjustments.
The evidence under this - 4 claims
Product sales fell 45% in the nine months, so growth came from rentals that recognise revenue slowly; the analysts' one estimate put Q3 at $1.55M against $1.37M reported.
Management says rentals spread the robot's cost over the contract, so the quarterly margin swings with the mix rather than with pricing.
The figures sit under a material weakness in financial controls not yet remediated, and three earlier years of statements had to be restated.
What management did with the money
In nine months it sold about $105M of new shares, at prices up to $4.55, then spent $21M of it on a building in Las Vegas. In August, with the shares below the cash per share, the board approved buying back up to $12M of shares. A share plan can add up to 18% more shares every year until 2034.
Insiders in 2026Q1: 0 purchases and 3 sales, from their Form 4 filings.
The evidence under this - 4 claims
No purchase under it has been filed yet: the 396,724 treasury shares at June were withheld for employees' taxes, not bought.
The board may add fewer shares than the 18% ceiling, and the 2026 awards were far below it.
The price, and where we disagree
In short. Cheap: the shares cost about half the cash the company holds, so the price pays nothing for the robot business.
The shares fell 77% in a year to $0.74, below what the cash alone is worth per share, and a quarter of the shares that trade are sold short. The board answered with a buyback, and a retailer ordered 200 cleaning robots.
The next test is the annual report due by 29 December 2026: how much cash is left, whether any shares were bought back, and whether the accounts are now filed on time.
The evidence under this - 5 claims
No purchase under it has been filed yet: the 396,724 treasury shares at June were withheld for employees' taxes, not bought.
The retailer is not named and no price is given; 200 robots over five years is about 40 a year.
How we value it
No fair value: the one analyst forecasts a loss every year to 2029 and the business has no lasting operating profit after stock pay, so a value built on earnings would rest on a guess; the cash and the bet price it instead.
How we read the multiple
At $0.74 the 225M shares are worth about $166M. The company holds $340M of cash and investments and owes almost nothing, so a buyer pays about half the cash and gets the robots for nothing.
The cash also lasts: before interest, the business used about $18M a year, so at that pace it would take well over a decade to spend. The discount is about trust and control, not about running out.
Against its own history. It trades far below its usual price against its net worth, because the money raised this year is now nearly all it owns and buyers doubt they will see it.
The evidence under this - 4 claims
Cash has been spent since June on operations and possibly buybacks, so the September figure will be lower than $339.6M.
The figures sit under a material weakness in financial controls not yet remediated, and three earlier years of statements had to be restated.
The building, the restaurants and the rental fleet are spent through capital spending, which ran $21.7M in the same nine months.
Priced against its rivals
None of the listed rivals above sells the same thing to the same buyers as a large part of what it does. A price comparison needs at least two, so the price is judged against its own history instead.
Also checked and set aside: Tesla (a substitute, not the same product).
Private or foreign rivals that cannot be priced: Bear Robotics, Inc., Pudu Technology Inc., Avidbots Corp, Miso Robotics Inc., Cafe X Technologies, Inc., Figure AI Inc.
The bet: what it pays if it works
This is a bet on the cash more than the robots. If the retailer's cleaning robots, the dealer rollout and the cafes all land, sales could triple, but at a normal price for such a business that adds only about a quarter of today's market value.
If the robots fail, holders keep the cash, which in the worst case we can see is still worth more than the whole company today. So the bet is safe and pays little.
| Revenue in two to three years | Dollars of yearly revenue needed | At the stated multiple |
|---|---|---|
| $8.43M | about 8.4M at 100% kept | 0.2× |
| $15.00M (the big win) | about 15.0M at 100% kept | 0.3× |
| $27.10M | about 27.1M at 100% kept | 0.5× |
To pay 2× (a Fair bet) it would need about $110.98M of revenue (about 111.0M dollars of yearly revenue); to pay 5× (a Cheap bet), about $277.44M of revenue (about 277.4M dollars of yearly revenue), at the same multiple.
The steps it has to take
- Done A national retailer orders 200 cleaning robots over five years
- Done A $12M buyback authorised while the price sits below cash
- Not started The FY2026 annual report filed on time with controls fixed due 2026-12
- Not started Quarterly revenue above $2M due 2027-12
- Not started Quarterly revenue above $3M with gross margin over 50% due 2028-09
The detail behind the label
The evidence under this - 4 claims
The figures sit under a material weakness in financial controls not yet remediated, and three earlier years of statements had to be restated.
The retailer is not named and no price is given; 200 robots over five years is about 40 a year.
The dealer group is unnamed and rental revenue of $0.69M across 55 contracts is about $12,600 a contract a year.
Where we disagree with the market
The cash will be spent or diluted away, so it is worth half its face value.
The cash is real and burns slowly, but owners cannot force it back, so cheap can stay cheap.
The market treats the cash as half gone: spent on more ventures, handed out as share awards, or diluted by new share sales, as happened twice this year. A quarter of the shares that trade are sold short.
We think the cash is mostly real and burns slowly, which is the case for the shares. But the family controls the votes, so the discount can last as long as they choose.
What analysts say, and the evidence
The evidence under this - 5 claims
Cash has been spent since June on operations and possibly buybacks, so the September figure will be lower than $339.6M.
The board may add fewer shares than the 18% ceiling, and the 2026 awards were far below it.
The building, the restaurants and the rental fleet are spent through capital spending, which ran $21.7M in the same nine months.
What could go wrong, and how we would know
The risks
- More shares The share plan can add up to 18% more shares each year, and the company sold shares twice this year below today's cash value.
- No say for holders The founder's family holds about two thirds of the votes, so the board need not return cash.
- Accounts in doubt Three years of accounts were restated and controls are still weak, so the cash figure itself rests on trust.
- Lawsuits A class action over a January release about Microsoft, and three suits against the directors, are open.
- Cheaper rivals Chinese makers ship far more robots for less, which keeps prices and the profit on each sale low.
The evidence under this - 6 claims
The board may add fewer shares than the 18% ceiling, and the 2026 awards were far below it.
The restated revenue moved by 6.6% in one quarter only; the errors were mostly in how warrants and awards were valued, not in sales or cash.
The company disputes the claims, and the class period is only three days long, which limits the damages a court could award.
What we still do not know
- Whether the Dex humanoid and the data services earn any money, or stay a story.
- Whether the Alphamax robot cafes are run to make a profit or to show off ADAM.
- How much of the $12M buyback is actually used before it expires in August 2027.
The evidence under this - 4 claims
Neither the Q2 nor the Q3 FY2026 10-Q reports Dex sales or a data services contract.
No purchase under it has been filed yet: the 396,724 treasury shares at June were withheld for employees' taxes, not bought.
What would change our view
- Neutral→AttractiveThe cash starts coming back
It needs all of:
- ○The company buys back the full $12M of shares it approved.checked at the next results, due by 29 Dec
- ○The FY2026 annual report is filed on time and says the weak controls are fixed.watched as it happens
- Neutral→UnattractiveThe cash leaks away
It needs any of:
- ○Cash and investments fall below $250M while sales stay under $3M a quarter.checked at the next results, due by 29 Dec
- ○The company sells new shares or buys a business while the shares trade below cash.watched as it happens
- Neutral→UnattractiveA price that pays for the business
It needs:
- ○The market value rises to the company's book value, what it owns minus what it owes, so the price starts paying for the robots.about $1.64 today, 122% above the last close
We would also be wrong if
- Another restatement or late report after the FY2026 annual report.
- Quarterly sales pass $3M with the profit on each sale above half.
What is coming
- 29 December 2026 FY2026 results and annual report: Cash at 30 September, how much of the buyback was used, and whether the controls weakness is fixed.
Every dated event ahead
- 29 December 2026 FY2026 results and annual report: Cash at 30 September, how much of the buyback was used, and whether the controls weakness is fixed.
- 17 February 2027 Share plan top-up on 1 November and the Q1 FY2027 report: How many new shares the plan added, and whether revenue from the retailer's cleaning robots starts.
- 21 August 2027 Buyback authorisation expires: Whether the board spent the $12M while the shares sat below cash.
Since we wrote this
- 17 NovThesis writtenNeutral, low convictionWritten with the shares at $0.74, about half the cash per share, a month before the FY2026 annual report.
Sources 17
Sources - 17
Every claim above cites one of these, or states the reasoning that derived it. The quote is frozen at the date it was read: a link can rot, and an agent fetch returns a reading of a page rather than the page. Backs counts the whole claim ledger, so it can name a claim this page does not surface.
- S01primarycompany filing2026-08-07the business and its three pillars, the named competitors, suppliers and staff, the MSAs, revenue by type, the restated statements, the dual-class shares and the earlier Nasdaq price noticebacksC01C02C04C06C08C19C20C21C27C31C32C35
- S02primarycompany filing2026-08-19revenue by type, costs, the impairment, investment income, cash and investments, the share issues and private placement, the evergreen plan, the lawsuits, the material weakness and the Alphamax restaurantsbacksC01C03C04C05C06C07C09C10C11C12C14C15C30C31C32C36
- S03primarycompany filing2026-06-12the audit committee's finding that FY2024, FY2025 and four quarters must be restated, and whybacksC09
- S04primarycompany filing2026-05-28Nasdaq's notice that the Q2 FY2026 10-Q was late, and the compliance timetablebacksC10
- S05primarycompany filing2026-06-03the $21.18M Las Vegas building bought with cash on hand, and what it is forbacksC12
- S06primarycompany filing2026-08-25the $12M share repurchase authorisation to August 2027 and the executives' healthcare stipendbacksC13
- S07primarycompany filing2026-09-21the national retailer's statement of work for 200 DUST-E cleaning robots over five yearsbacksC19
- S08primarycompany filing2026-09-08the share count by class, votes per share, beneficial ownership and voting power, the officers and their paybacksC06C08C28
- S09secondaryvendor data2026-11-16Research pack, frozen beside this thesiscontent/RR/research-pack.jsonthe close, the price history and its split, short interest, the analyst panel and estimates, and the insider rosterbacksC16C17C18C29
- S10primaryindustry data2025-10-07International Federation of Robotics, 'Service Robots See Global Growth Boom' (World Robotics 2025) International Federation of Roboticsunits of professional service robots sold worldwide by category, and the growth of rented fleets, from the industry's statistics body
The total number of service robots sold for professional use reached almost 200,000 units in 2024, marking a 9% increase. ... Hospitality robots remain second place with more than 42,000 sold - down 11%. ... The market for professional cleaning robots, in third place, grew by 34% to more than 25,000 units sold. ... The robot-as-a-service fleet (RaaS) has grown impressively by 31%.
backsC22 - S11primarycompany release2025-07-21Keenon Robotics, 'KEENON Robotics Continues Global Lead in Commercial Service Robot Market, Securing Triple No.1 Rankings, IDC Reports' Keenon Roboticsa Chinese rival's share of commercial service robot shipments, citing IDC
KEENON Robotics claimed the top spot globally with a 22.7% shipment share ... KEENON Robotics leading the sector at a 29.8% market share
backsC23 - S12primarycompany release2025-01-24LG Electronics, 'LG Acquires Majority Stake in Bear Robotics to Bolster Robotics Capabilities' LG Electronicsthat the maker of the Servi serving robot, named in Richtech's 10-K, is now controlled by LG
LG will hold a controlling 51 percent stake in Bear Robotics ... an initial investment of USD 60 million in March 2024
backsC24 - S13primarycompany filing2026-02-24Tennant Company Form 10-K, year ended 31 December 2025 Tennant Companya listed floor-care maker selling autonomous scrubbers, and its stake in Brain CorpbacksC25
- S14primarygovernment data2026-11-07U.S. Bureau of Labor Statistics, Average Hourly Earnings of All Employees, Leisure and Hospitality (CES7000000003), via FRED U.S. Bureau of Labor Statisticshourly pay in restaurants and hotels, the cost a service robot is bought to save: $23.01 in October 2025 and $23.90 in October 2026
Average Hourly Earnings of All Employees, Leisure and Hospitality
backsC26 - S15primarycompany filing2026-11-13the November 2026 and December 2025 share awards to officers and directors, and the former president's salesbacksC29
- S16primarycompany release2026-01-27the release the class action is about: a joint engineering effort with Microsoft's AI Co-Innovation Labs on the ADAM robot; read through a reprint, since GlobeNewswire refuses this repo's fetcher
LAS VEGAS, Jan. 27, 2026 (GLOBE NEWSWIRE) -- Richtech Robotics Inc. (Nasdaq: RR)
backsC11C36 - S17primarycompany filing2026-01-20the annual report as first filed: the same business description, risk factors and competitors the 10-K/A repeats, and the figures the restatement later changedbacksC01