A-Portfolio

The company in one breath

Accelevation Holdings, ticker ACCV on the Nasdaq, designs, builds, and installs the power gear, steel frames, cooling pieces, and wiring that go inside data centers. Customers pay Accelevation to replace a pile of separate vendors with one company that can design the job, make the parts, and install them.

The company grew from less than $3 million of revenue in 2021 to $447.8 million in 2025. That means a small Ohio manufacturer turned into a serious data-center supplier in four years. Good for a young industrial company would look like proving that speed came from repeat customer demand, not one heroic sprint or a shopping spree.

Why now

The stock closed Thursday at $16.59, according to Yahoo Finance at 4:00 p.m. Eastern on October 1, 2026. That is 7.8% below the $18 IPO price and 17.1% below the original $20 bottom of the marketing range. In plain English, institutions rejected the first price, then public buyers pushed it lower. Good would look like the stock finding a floor while orders, cash collection, and margins keep improving.

The business growth is wild. First-half revenue reached $437.5 million, up 175.8% from one year earlier. That means Accelevation sold almost as much in six months as it sold during all of 2025. Good for a physical manufacturer would usually look like growth above 20%, so this cleared the bar by a ridiculous distance.

The catch is who got paid. The IPO sold 30 million shares at $18, but Accelevation issued only 10 million of them. Existing owners sold the other 20 million, so two-thirds of the deal put cash in old holders' pockets rather than new machines on the factory floor. Good would look like most IPO proceeds funding growth or repairing the balance sheet. Here, the company received about $170.6 million before its own offering expenses, and all of that was earmarked for debt repayment.

The numbers, translated

• Backlog was $1.11 billion on June 30. Backlog is work customers have ordered but Accelevation has not yet counted as revenue, and this amount equals about 1.27 times the latest six-month sales pace stretched over a year. Good would look above one year of revenue with firm purchase orders, broad customers, and healthy margins, so the size passes while the quality still needs inspection.
• The six-month book-to-bill ratio was 2.5 times. Book-to-bill compares new orders with completed sales, so Accelevation booked about $2.50 of future work for every $1.00 it delivered. Good would look above 1.0, because orders then refill the pipeline faster than shipments empty it. This is excellent demand, assuming customers do not delay or cancel.
• Gross margin was 26.6%, down from 30.4%. Gross margin is what remained after materials and production labor, before office costs, interest, and taxes. Good for a custom equipment maker would look above 30% and stable, so growth came with weaker economics.
• Operating income was $35.3 million, equal to 8.1% of revenue. Operating income is profit from the business before interest and taxes. Good would look above 10 cents per sales dollar, so Accelevation got close but did not clear the line.
• Net income was $19.3 million, equal to 4.4% of revenue. This is the profit left after every reported expense. Good would look above 8 cents per sales dollar, so the company is profitable but still thin-skinned.
• Free cash flow was negative $15.8 million for the first half. Free cash flow is cash from operations after equipment spending, the money owners can actually keep or reinvest. Good would look positive while backlog grows, so accounting profit has not yet turned into cash.
• Pro forma debt was $479.9 million after the IPO, versus $19.3 million of cash. Pro forma means the filing shows the balance sheet as if the IPO and debt repayment had already happened. Good would look like debt below two times annual adjusted operating profit and cash comfortably covering working-capital swings. Instead, debt is about 3.5 times the latest adjusted operating-profit pace and cash is skinny.
• The top two customers produced 46.6% of first-half revenue. That means almost half the sales depended on two buyers. Good would look below 25%, because one delayed data-center campus should not wreck the year.
• At $16.59, the fully exchanged equity value is about $3.73 billion and enterprise value is about $4.19 billion. Equity value prices all 225 million shares that could exist after old owners exchange their private units, while enterprise value adds debt and subtracts cash. Good value for this mix of speed, debt, and weak cash flow would look below 4 times annualized sales and below 25 times annualized adjusted operating profit. Today the stock costs about 4.8 times sales and 30.7 times adjusted operating profit, so the IPO discount did not create a bargain.

What I like

The company sells picks and shovels for the data-center rush. It does not need to guess which artificial-intelligence model wins. It needs customers to keep building rooms full of power-hungry computers. Good would look like cloud and AI spending turning into signed orders, and the $1.11 billion backlog says that is happening.

The integrated model has teeth. Accelevation designs, manufactures, and installs the gear, which can reduce vendor handoffs and field mistakes. Management estimates its modular approach can improve field-installation productivity by as much as 84% for certain jobs. That means customers may get a data hall running sooner. Good would look like repeat orders, stable pricing, and gross margin rising above 30% as factories get busier.

The founders still operate the business. Michael and Shawn Rubiera built it from roughly 25 employees after two small acquisitions into a company with more than 1,700 employees. That means the people who found the opening still have skin in execution. Good founder leadership would look like protecting customer trust and cash flow after private equity and public-market pressure arrive.

The order machine is not living on old products. Management says more than 80% of backlog came from products designed and launched within the prior year. That means Accelevation is adapting to fast-changing power and cooling demands instead of selling yesterday's catalog. Good would look like those products reaching planned margins without warranty surprises or cost overruns.

What worries me

The private-equity fingerprints are all over the capital structure. Olympus Partners financed distributions that reached about $303.4 million by July, while the company carried heavy debt and later used IPO proceeds to repay part of it. That means public investors arrived after the sponsor had already pulled a large pile of cash from the table. Good would look like debt falling through operating cash, no new sponsor distributions, and capital spending beating financial engineering for once.

Cash flow is the ugly part. Accelevation reported profit but used $4.2 million of operating cash and $15.8 million of free cash in the first half as receivables, contract assets, and inventory grew. That means sales moved faster than customer payments. Good would look like operating cash turning positive and at least 8 cents of free cash from each revenue dollar once growth settles.

Contract accounting can bite. The company recognizes much of its revenue as projects progress, using estimates of the total cost to finish. First-half estimate changes cut revenue by $9.4 million, and loss provisions on contracts reached $16.3 million. That means one bad cost estimate can reverse profit already celebrated. Good would look like loss provisions below 2% of revenue and no large downward catch-up adjustments.

Control is lopsided. Olympus controls about 85% of the voting power, while public IPO buyers hold roughly 13%. That means regular shareholders can own the stock and still have almost no say at the table. Good would look like sponsor control falling below 50% without dumping a wall of shares into the market.

Then there is the tax receivable agreement. Accelevation promises old owners 85% of certain future tax savings, with a possible long-run payment estimate of $589.8 million under the filing's all-exchange example. That means public owners may fund tax payments to pre-IPO holders for years. Good would look like actual payments staying small relative to free cash flow. I do not know that yet, and neither does anyone outside the company.

What would make this a buy

My position size is zero. This is research and commentary, never a buy signal.

The price door opens from $13 to $14, an analytical range based on the October 1 close and the final prospectus. At $13, the enterprise value would be about 3.9 times the latest sales pace. That means the price would finally compensate me for sponsor control, customer concentration, debt, and weak cash conversion. Good would look like that range arriving without a customer loss or backlog collapse.

The proof door matters more. I want two public quarters with positive operating cash flow, gross margin back above 30%, top-two customer concentration below 40%, and net debt below three times the latest annual adjusted operating-profit pace. That combination would show the company can turn orders into cash while loosening the grip of its biggest customers and lenders.

I would pay above $14 only if backlog stays above one year of sales, the book-to-bill ratio stays above 1.2, and free-cash-flow margin reaches 5%. That would mean the growth story graduated from prospectus math to bank-account proof.

What makes me walk away

I walk if backlog falls below 0.8 times annual revenue while the data-center buildout stays healthy. That would mean Accelevation lost share or customers stopped trusting its delivery. Good would look above one times annual revenue through a normal cycle.

I walk if gross margin stays below 25% for two quarters or new contract losses exceed 3% of revenue. That would mean speed and customization are creating rework instead of an advantage. Good would look above 30% gross margin with contract losses below 1% of sales.

I walk if free cash flow stays negative through the next four reported quarters. That would mean the company keeps booking profit without collecting enough cash. Good would look like full-year free cash flow above 5% of revenue.

I walk if the sponsor sells aggressively after the 180-day lockup while leverage stays above three times adjusted operating profit. A lockup temporarily blocks insider sales after an IPO. Good would look like debt falling first and sponsor exits arriving slowly, not a fire hose pointed at the public market.

The verdict

Score: 67 out of 100. Bucket: watch. Accelevation earned points for founder-led execution, explosive growth, a $1.11 billion backlog, a 2.5 times book-to-bill ratio, fast product development, domestic manufacturing, and a real role in the data-center supply chain.

It lost points for negative free cash flow, shrinking gross margin, almost $480 million of post-IPO debt, heavy customer concentration, 85% sponsor voting control, a secondary-heavy IPO, a tax deal that favors old owners, and a valuation that still charges for clean execution. A candidate starts at 80. This is thirteen points short. The factory is exciting. The balance sheet showed up wearing somebody else's bar tab.

The DBTech read

Backlog is work a company expects to turn into future revenue, but it is not cash and it is not a guarantee. Accelevation's $1.11 billion backlog gives it visibility, which means management can see more of the road ahead. Good backlog comes from firm, profitable orders spread across many customers. Weak backlog contains soft commitments, cancellation risk, or jobs priced too cheaply. Read backlog like a restaurant reservation book, a full dining room means nothing if half the tables cancel and the other half ordered the loss-leading special.

Sources and as-of date

• Price: $16.59 at 4:00 p.m. Eastern on October 1, 2026, the latest completed regular-session close, and $16.40 at 6:27 a.m. Eastern on October 2, 2026, the latest premarket print retrieved shortly after 7:00 a.m. Eastern, Yahoo Finance. Good valuation work uses the liquid closing price, so every valuation calculation uses $16.59.
• Business, financial statements, backlog, margins, customer concentration, debt, cash flow, contract accounting, ownership, tax receivable agreement, and final IPO terms: Accelevation final prospectus, filed October 1, 2026. Good primary research reads the final prospectus because the preliminary pitch did not contain the final $18 price.
• IPO closing and governance documents: Accelevation Form 8-K, filed October 1, 2026. Good deal work separates shares sold by the company from shares sold by existing owners.
• Company positioning and offering announcement: Accelevation investor relations and pricing release, dated September 22 and September 29, 2026. Good research uses management's claims to understand the pitch, not to prove the pitch.
• Independent IPO context: IPO Scoop and Renaissance Capital, retrieved October 2, 2026. Good independent checks confirmed the below-range pricing and sponsor-heavy sale.
• Community research covered September 2 through October 2, 2026. The focused evidence found one on-topic Reddit thread, two relevant YouTube analyses, and one GitHub listing-tracker item. Good evidence would include customers, electricians, and data-center operators with direct project experience, so I treated the thin online sample as a warning about limited public proof, not as proof against the company.
• IPO calendar: Nasdaq (NDAQ) data retrieved shortly after 7:00 a.m. Eastern on October 2, 2026 showed no confirmed upcoming U.S. IPO entries. Good calendar work leaves the section empty rather than promoting an unconfirmed listing.
• Earnings calendar: Earnings Whispers data retrieved October 2, 2026 showed no featured name, prior feature, or identified A-Portfolio watch-list company reporting during the next full week. Good calendar work skips unrelated reports instead of stuffing the page.
• Calculations use Decimal arithmetic, the $16.585 October 1 regular-session close, 225 million fully exchanged shares, $479.884 million of pro forma debt, $19.289 million of pro forma cash, and first-half revenue and adjusted operating-profit pace from the final prospectus. Research and commentary only, never a buy signal.