Business

Business — Know What You Own

Figures converted from INR at historical FX rates — see data/company.json.fx_rates. Ratios, margins, and multiples are unitless and unchanged.

Shakti Pumps is two businesses sharing one factory: a government-tender solar-pump specialist (~70% of FY25 revenue, 24% peak EBITDA) selling subsidised irrigation systems to state DISCOMs under PM-KUSUM, and a stainless-steel submersible-pump exporter (17% of FY25, growing) selling pumps in 100+ countries. The first business prints margins that do not belong to a pump company; the second is what a normal pump business should look like. The mistake the market keeps making — in both directions — is to confuse one with the other.

The 24% EBITDA margin in FY25 is what happens when a tender deadline floods volume through fixed-cost capacity; the 9–10% margin in Q3-Q4 FY26 is the post-deadline air pocket. Neither is the true business. What decides what this company is worth in 2030 is whether the 2.2 GW captive solar-cell plant turns the tender business from a cyclical print into a structurally higher-margin engine.

1. How This Business Actually Works

Shakti is a vertically-integrated OEM that wins L1 reverse auctions floated by state DISCOMs, ships a complete solar-pump system (pump + motor + controller + mounting structure + panels + 5-year service), and gets paid in tranches by the state government over 200–250 days. That sentence contains every economic lever in the model.

The business is closer to a government-channel project EPC than a consumer pump brand. Three implications follow. First, working capital is the second income statement: 152 days of receivables in FY25, 178 in FY24, ~190 days at end-Q2 FY26 — every revenue print is a bet on state-government cash flow, not on end-customer demand. Second, operating leverage is brutal in both directions: the same factory that printed 24% EBITDA in FY25 printed 11% in Q3 FY26 once volumes fell ~15% YoY, because the cost base barely moved. Third, the moat is not the pump — it is the bundle: MNRE 1A accreditation + DCR-compliant cell supply + 400+ service centres + 15 patents + execution track record on 71,500+ pumps installed in FY25 alone. New entrants can buy a pump; they cannot buy that.

The retail/export side runs on a cleaner economic engine: sell a stainless-steel submersible to a dealer or distributor, receive payment in 60–90 days, book a ~12–18% EBITDA margin without subsidy distortion. This is the "real" pump business, growing 24.8% CAGR FY21–25 in exports, now 100+ countries, with Africa 43% / Middle East 25% / US 15% of FY25 exports. It is also the part of the business that does not need PM-KUSUM 2.0 to exist.

2. The Playing Field

The peer set is two sub-industries pretending to be one. The right comparable today is Oswal Pumps; the right second comparable is Shakti to itself across cycles. KSB and Kirloskar Brothers are different businesses on multiples that do not transfer.

Source: screener.in 8 May 2026; market caps converted at 2026-05-09 spot rate (₹1 = $0.01058). Shakti DSO is FY25 print, others are most recent disclosed annual.

The chart shows the puzzle: the two highest-margin operators (Oswal and Roto) trade at low multiples; KSB sits at the highest, with Kirloskar Brothers in the middle. This is not market irrationality. It is the market saying Oswal's 29% margin is at risk of being a one-cycle subsidy artifact and Roto's specialised industrial niche won't scale, while KSB is a durable franchise whose 14% margin is the steady-state and Kirloskar's project mix gets a smaller premium. Shakti sits in the middle on every axis — which is exactly where the market cannot decide what kind of business this is.

What the peer set teaches:

• Oswal is the right read for the tender business. It captures ~38% of PM-KUSUM share to Shakti's ~25%, prints ~29% op margin and ~88% ROE on a smaller revenue base, and trades at ~13× TTM earnings — the steep discount reflects exactly the cycle skepticism Shakti faces. If you believe Oswal's print is repeatable, you should believe Shakti's tender business at a comparable cycle multiple.
• KSB and Kirloskar are not relevant comparables for tender economics. Their 14% margin is a floor on what a fully diversified, channel-distributed industrial pump franchise earns through a cycle. Shakti's 15–16% steady-state guide implicitly maps to this band, not to the FY25 24%.
• WPIL is the cleanest read on water-infrastructure project lumpiness — 16% margin, high working capital, large EPC orders. Shakti's solar-pump tender business actually behaves more like WPIL than like KSB.
• Roto Pumps is the niche export precedent. 22% margins on small scale through differentiation, no subsidy exposure. This is what Shakti's export business could grow into if it doubles from $51M (FY25) to $115M+.

The deeper takeaway: Shakti's "right" multiple depends entirely on revenue mix. A 70/30 tender/non-tender Shakti deserves an Oswal-style cycle-discounted multiple in the low-to-mid teens on peak earnings. A 50/50 Shakti deserves a blended mid-cycle 20–25× depending on the diversification mix. A Shakti that has structurally captured DCR cell economics deserves more. Investors arguing about whether Shakti is "expensive at 26x" or "cheap at 16x peak FY25 earnings" are arguing about the wrong thing — they should be arguing about the mix.

3. Is This Business Cyclical?

Yes, and on a different cycle than every comparable. The cycle is subsidy-disbursement and election-timing, not GDP or housing. The cycle hits revenue, margin, and working capital at the same time — there is nowhere for the income statement to hide.

The cycle hits in three places, in order:

First, the order book. State governments pause new tenders during election model code or when budgets tighten, and the order book is the canary. Shakti carried $194M at end-FY25 (≈6 months execution), $145M by Nov 2025 (≈4 months), reflecting GST 2.0 disruption holding up new state awards in Q2-Q3 FY26.

Second, working capital. When state cash flow tightens, retention payments slip. Shakti's debtor days went from 178 (FY24) to 152 (FY25, the peak year) and back up to roughly 190 in Q2 FY26 — receivables hit $185M against quarterly revenue of $75M, an explicit indicator the FY25 collection improvement was cyclical, not structural. Total debt rose from $10M (FY24) to $20M (FY25) to $52M (FY26) almost entirely to fund the working-capital balloon.

Third, margin. The two-quarter sequence Q3 FY26 (10.7% op margin) and Q4 FY26 (9.7%) is the canonical post-deadline air pocket: revenue did not collapse but margin halved because lower-volume quarters cannot absorb the fixed cost base, and management deliberately chose to chase Maharashtra Magel-Tyala scheme volume at a lower price point. A scheme deadline (March 2026) front-loads revenue and back-loads margin compression. Shakti is mid-process through that exact dynamic.

4. The Metrics That Actually Matter

Forget P/E for a moment. The four metrics that explain Shakti's value creation and failure are order book run-rate, receivable days, captive integration depth, and the gap between print EBITDA and management's 15–16% steady-state guide.

Two of these are unusual enough to dwell on.

Receivable days are a leading indicator of earnings quality, not a lagging one. The cleanest read on whether Shakti's tender business is structurally profitable or just timing-aided is whether DSO trends below 130 over a full year while revenue grows. FY25 hit 152 on a 84% revenue print — directionally good but still 30+ days from the company's own 120 target. Until DSO sits below 130 for four consecutive quarters, every margin print should be discounted.

The integration capture ratio matters more than the order book. PM-KUSUM tender prices fall every cycle (the L1 mechanism guarantees it). Shakti's only defence is reducing the cost stack faster than the price falls. Each integration layer captured (motor in-house in 2010s, controller in early 2020s, structure now, DCR cell from FY27) is permanent margin floor — independent of how the next tender prices. The 2.2 GW SESL plant is not a growth project; it is a margin-defence project.

5. What Is This Business Worth?

This is one economic engine, not a sum-of-parts. Shakti has no listed subsidiaries, no investment portfolio, no holding-company discount, no separable real estate. The right valuation lens is normalised earnings power × backward-integration premium, with cycle position determining where in the band you anchor.

The company is best valued not at FY25 peak earnings nor FY26 trough earnings, but at a mid-cycle revenue and margin assumption that respects management's own 15–16% steady-state guidance. The premium or discount to that base case rests on three observable drivers.

A back-of-envelope frame, not a price target: Shakti at $5.82 trades at ~26x FY26 EPS ($0.22) and ~16x FY25 peak EPS ($0.40). At a defensible mid-cycle revenue of $320M × 16% op margin × 75% (interest + tax) ≈ $38M net income or ~$0.31/EPS, the stock is at ~19x mid-cycle earnings. That is a premium to Oswal's ~13× TTM multiple but a meaningful discount to KSB's 55× and Kirloskar Brothers' 33× — the discount the market is charging for tender concentration and working-capital risk relative to diversified industrial-pump peers, while still pricing more cycle-stability than Oswal earns. The thesis is therefore not that the stock is mispriced today; it is that the through-cycle earnings power is mis-estimated. If DCR cell capture and export mix shift the steady-state to 17–18% margin and $375–425M revenue by FY28, the implied EPS approaches $0.43 and the same 19x multiple gets you to a very different stock. If neither happens, the current multiple is the multiple, and the only return is earnings growth tracking GDP+.

6. What I'd Tell a Young Analyst

Watch four things and stop reading everything else.

Three things the market is most likely getting wrong, in order:

(1) Treating FY25 24% EBITDA as the new normal. Management itself guides 15–16%. The Q3-Q4 FY26 prints below 11% confirm operating leverage works in reverse just as fast. Anyone modelling 22%+ steady-state is anchoring on the cycle peak.

(2) Treating Shakti like KSB or Kirloskar. It is not. They are diversified industrial-pump franchises with multi-year capex spec engagements; Shakti is a tender-bid solar specialist with a growing exports tail. The right comparable is Oswal, and the right second comparable is Shakti to itself.

(3) Underweighting the DCR cell plant. This is the most important strategic move in the company's history. It either turns Shakti into a structurally higher-margin tender winner or it does not. The market is pricing it as if it is just another factory. It is not.

What would change the thesis decisively:

• Up: PM-KUSUM 2.0 with larger central outlay + DCR plant operational at full ramp + DSO sustained below 130 + non-tender mix above 35%.
• Down: Scheme renewal stalls + state payment cycles stretch beyond 200 days + DCR plant slip beyond H2 FY27 + L1 tender pricing erodes 200+ bps.

If the data 24 months from now lands on the first set, this is a materially bigger company than today's multiple implies. If it lands on the second, this is a cyclical microcap that printed a once-in-a-decade subsidy windfall and reverted to its 2017–2024 character. Both outcomes are observable from the four signals above. Track them — the multiple debate resolves itself.