Rashi Peripherals — the middleman of Indian tech hardware, riding a price wave

The Short Version

Rashi Peripherals is a distributor — the company that buys computer hardware from global brands (processors, graphics cards, laptops, networking gear) and sells it onward to thousands of Indian resellers, shops and businesses. It’s a high-volume, wafer-thin-margin business: out of every ₹100 of sales it keeps only about ₹2–3 as operating profit. In the year just ended (FY26) it did ₹15,827 crore of revenue (up 15%) and a record ₹282 crore of profit (up 35%). A big part of that growth, though, isn’t more boxes shipped — it’s that the boxes got more expensive: a global shortage has sent memory-chip (RAM) prices up two-to-three times, which inflates the rupee value of everything Rashi sells. The honest reading is a well-run middleman with a strengthening balance sheet, genuinely exposed to the AI-and-data-centre hardware boom, but whose recent “record” growth is partly a price wave that could recede. At about 13 times earnings, the market isn’t paying up for it.

What This Company Actually Does

Picture the journey of a graphics card from, say, an American chip company to a gamer’s PC in Pune. The brand doesn’t sell directly to every Indian shop — it appoints a distributor to import the product, hold the stock, extend credit to retailers, deliver it, and handle warranties and repairs. Rashi is one of India’s largest such distributors, doing this for a long list of global technology brands across two buckets:

• Personal Computing & Devices (PCDP) — the consumer-facing stuff: processors, graphics cards, components, laptops, mobiles, peripherals. Higher volume, more tied to the consumer PC cycle.
• Enterprise & Specialty (ES) — networking, storage, UPS and enterprise gear sold to businesses and data-centre buyers. Stickier and slightly better-margin.

The economics are the key to understanding everything: a distributor earns a tiny markup on enormous turnover. Operating margins of 2–3% aren’t a weakness — they’re the entire model. What actually makes or breaks a distributor isn’t margin, it’s two other things: how much cash gets tied up in inventory and in money owed by resellers (the “working-capital cycle”), and the interest cost of financing all that stock. So the right things to watch here are debt, receivables and cash flow — not the headline margin.

Rashi IPO’d in February 2024. The founding family owns ~64% (and has been quietly adding to its stake — a small sign of confidence), domestic mutual funds have built a ~17% position, and foreign investors are essentially absent. The stock has roughly doubled off its low but, at ~13× earnings and ~1.8× book value, remains modestly priced — appropriate for a thin-margin, cyclical business.

The Long Game

For a distributor, the long-term health check is unglamorous: steady volume growth, a tightening cash cycle, falling reliance on debt, and a rising credit rating (because cheaper borrowing directly lifts the thin bottom line). On all four, Rashi has been moving the right way. The IPO money was used to cut debt; the company’s credit rating was upgraded into the “AA club” during FY26 (which lowers its interest cost); operating cash flow turned positive; and a long-stuck large receivable (from a data-centre customer, Yotta) was fully collected. Management’s stated aspiration is to compound revenue at about 20% a year.

The genuinely interesting long-term angle is where the hardware is heading. Two structural waves run straight through Rashi’s product shelf: the AI and data-centre build-out (which needs graphics cards, servers, storage, networking — much of what Rashi distributes; it is, notably, the oldest Indian distributor for NVIDIA, though it’s careful to say not exclusive), and the steady upgrade cycle in enterprise IT and premium consumer devices. Management has signalled a planned re-entry into AI data-centre project work in FY27, citing a sales funnel of ₹20,000–25,000 crore — but pointedly only “subject to ROI,” a discipline that matters in a business where chasing low-margin volume can destroy more value than it creates.

The reason to give the story time: it’s a scale-and-discipline game. Rashi has quadrupled revenue over seven years, and the moat — relationships with global brands, a national reseller network, and the working capital to fund it — compounds slowly. The reason for realism: distribution will never be a high-margin business, and a chunk of FY26’s growth is a price phenomenon, not a volume one.

The Year, Told Simply

The thread through FY26: volumes grew respectably, but the dramatic-looking revenue and profit jumps owe a lot to a component-price surge — and management was refreshingly clear about that.

First half (the June and September quarters). A steady start, then acceleration. By the September quarter, revenue was up ~12% (about 20% excluding lumpy project work), the balance sheet was visibly healthier — debt down, cash flow positive, the big Yotta receivable collected, and the credit-rating upgrade in hand. Management guided to ~15% core growth and reiterated the thin ~1.5% net margin reality, while flagging persistent memory and chipset shortages and a clarification worth noting: it’s NVIDIA’s oldest Indian distributor, not its exclusive one (a useful check against over-hyping the AI-GPU angle).

Third quarter (reported February) — the price wave shows up. Revenue jumped 43% and profit doubled. But management was candid that roughly half of that came from price, not volume — memory (RAM) prices had risen two-to-three times, and a weaker rupee inflated the cost of imported goods, both of which pump up reported revenue. Unit volumes were guided to be roughly flat quarter-on-quarter even as the rupee value kept climbing. The genuine improvements that quarter were real, though: the credit-rating upgrade and positive nine-month operating cash flow.

Fourth quarter and the year (reported May). A record finish — Q4 revenue up 51% — closing FY26 at ₹15,827 crore (+15%) and ₹282 crore of profit (+35%), explicitly described as riding a component-pricing “super cycle.” Management reaffirmed its ~20% growth aspiration, flagged that more price increases (~20%) were likely but also warned of possible consumer-demand softness in the second half, and laid out the planned FY27 re-entry into AI data-centre projects against that ₹20,000–25,000 crore funnel — but only where the returns justify it.

The pattern a long-term investor should read: the business is being run well — debt managed, credit improving, cash flow turning, the founder adding to his stake. But the reported growth is flattered by a component-price spike that, like all price waves, can reverse — and when it does, revenue growth will look weaker even if the underlying volumes are fine. Read Rashi on volumes and cash flow, not on headline revenue.

What a Patient Investor Would Watch

On a known multi-year clock. The ~20% volume-growth aspiration — and whether actual unit growth holds up once the component-price wave fades. The FY27 re-entry into AI data-centre projects (the ₹20,000–25,000 crore funnel), pursued with ROI discipline. The continued march of the enterprise/data-centre mix, which is stickier and slightly better-margin than consumer. And the credit rating and cash cycle, which directly drive the thin bottom line.

What could genuinely matter. The price wave cuts both ways: rising memory/chip prices have inflated revenue and may have delivered some inventory gains, but a sharp reversal would deflate revenue growth and could leave the company holding high-cost stock — the classic distributor risk. Working capital is the perennial pressure point: a thin-margin distributor lives or dies on how fast it turns inventory and collects from resellers, so any creep in debtor days or debt is worth watching closely (borrowings have crept back to ₹991 crore as the business scaled). Consumer-PC demand is cyclical, and management itself flagged possible second-half softness. And the AI/GPU angle, while real, should be sized soberly — Rashi is one distributor among several, not the exclusive gatekeeper.

The simple test for next year. Did volumes grow (not just rupee revenue inflated by component prices)? Did the cash cycle stay tight and debt stay controlled? Did the AI data-centre re-entry actually happen — and at decent returns? Did the credit rating hold or improve? Four questions, all answerable from next year’s filings — and notably, for a cheap, thin-margin distributor, the bar the market has set is low, so steady execution is what it needs, not heroics.

The Four Checks

1. Quality and moat. A well-run business in a structurally thin-moat industry. What Rashi has is scale and relationships: 82 brand partnerships built over three decades (including being NVIDIA’s oldest Indian distributor — though pointedly not its exclusive one), a network reaching 10,000-plus resellers in 700-plus towns through 55 branches, and the balance sheet to fund the inventory and credit that the whole machine runs on. That’s real — a newcomer can’t replicate it quickly — but it isn’t durable pricing power. Brands can and do appoint multiple distributors, operating margins of 2–3% are the permanent ceiling, and the only defences are cost discipline and being too useful to drop. Call it a competent middleman with stickiness, not a moat in any strong sense.

2. Returns on incremental capital and runway. Every incremental rupee here goes into working capital — stock on shelves and credit to resellers — and what it earns is moderate: ROCE of 16.8% as of the June 2026 snapshot, but the trend is the honest part — 25% in FY22, compressed to 14% in FY24–25, recovered to 17% in FY26 on the component-price super-cycle. ROE sits at 14.7%, with the three-year average a lower 13.3%. The runway is genuinely long — management cites a ₹1.5-lakh-crore industry, a 20% CAGR aspiration, and a ₹20,000–25,000 crore data-centre funnel — but the cash conversion cycle has stretched from 48 to 60 days, meaning growth keeps swallowing capital: operating cash flow was negative four years running before FY26’s first positive turn (+₹114 crore). A long runway at mid-teens returns, with the returns lately flattered by a price wave.

3. Capital allocation for the stage. Mostly sensible, judged by actions. The IPO money went where it should — borrowings cut from ₹1,082 crore to ₹700 crore — though they’ve since crept back to ₹991 crore against a ~₹2,025 crore net worth, which is tolerable for the model. The most telling decision was restraint: in FY26 management consciously skipped large data-centre projects because the regular business was already growing 31%, and frames the FY27 re-entry as “subject to ROI” — exactly the discipline a thin-margin distributor needs. Dividends are token (5% payout) and there are no buybacks, but that’s defensible when every spare rupee funds working capital at 17% ROCE; the promoter family adding from 63.4% to 64.0% is a quiet vote of the same conviction. The quibble is the doubling of working-capital days since FY22 — growth has been bought with tied-up cash, and FY26 is the first year the engine returned any.

4. Price. As of the June 2026 snapshot the stock trades at ₹539 — near its 52-week high of ₹575, roughly double the low — at 12.8 times earnings and about 1.8 times book. For a profitable, deleveraged distributor compounding revenue at 15%-plus, that multiple is undemanding on its face. The asterisk is the denominator: FY26’s record ₹282 crore profit rides a memory-price super-cycle that management itself says contributed 20-odd points of second-half growth and may run only another year or two. Pay 13 times peak-wave earnings and the true multiple on normalised, volume-driven earnings is somewhat higher. Fair-to-cheap if volumes hold and cash flow stays positive; merely full if the price wave recedes and takes the record year with it.

Sources

• Concall transcripts (4): Q1 FY26 (Aug 2025), Q2 FY26 (Nov 10, 2025), Q3 FY26 (Feb 4, 2026), Q4 FY26 + full-year (May 2026) — BSE filings, converted to markdown. (One quarter’s transcript needed a raw-text fallback as the PDF parser underperformed; facts were cross-checked where the text was rough.)
• Annual reports (2): FY24 (Rashi’s first as a listed company, post-IPO) and FY25 sections — governance- and segment-heavy.
• Screener.in snapshot: quarterly and annual tables, ratios, shareholding — fetched 2026-06-06 (logged-out session).
• Research files: vault/Sources/Earnings/Rashi Peripherals Ltd/ — raw transcripts, AR sections, snapshot, per-document digests (not published).