C1Q26 Strategy Update
We are sharing summary of our C1Q26 investor letter:
Dear Investors:
‘Chaos’ is often a necessary condition for ‘rationality’ to emerge, but its arrival is usually unadvertised within securities. Since C2025 wasn’t anywhere close to being as chaotic as we had anticipated, we note continued value disconnects within headline Indian equities. Accordingly, we see significant room for these price disconnects to unravel before compelling values re-emerge. Dalal street’s unyielding optimism can get capitalized at unreasonable levels as long as somewhat reasonable underlying growth can be sold to India’s momentum chasing, notoriously fickle, and often blinkered individual investor and family office capital. As growth stumbles, that capital is unlikely to support valuations that are near impossible to justify to any reasonable investor that isn’t restrained by capital controls - We note increasingly apparent fault lines as retail trading’s influence continues to simmer down[1], ticker tape reactions to headline earnings get more discerning (please click here for our work on reactions to headline prints), and heady subscriptions to public offerings cool off[2]. It is well documented that Minerva India Under-served has historically disengaged as domestic liquidity chases street’s most-peddled narratives. The converse is as true. It’s pointless to argue against the fact that Dalal street’s perennial half-full narratives would eventually lose their influence over India’s largely ‘green’ and entitled investor crowd. As this liquidity continues to recede[3] and uncertainties begin to get priced in more effectively, we fully expect to extract our pound of proverbial flesh.
Perseverance with finding wisdom is always a struggle. The only way to minimize/eliminate the associated boredom/pain is to keep seeking the elusive truth vs. trying to make perfect sense of every outcome within an otherwise imperfect environment. Those outcomes in our work are inevitably tied to the asset class we invest in (which are smaller Indian small caps in our case). Our primary job is to establish a favorable asymmetric relationship with that asset class over complete cycles, vs. merely riding on the class itself. The latter could always be more effectively executed through passive vehicles.
In C2025, Minerva India Under-served strategy was up +3.4% (net; in INR), vs. -6.6% decline (gross; in INR) in BSE Smallcap, and +6.4% gain (gross; in INR) in BSE 500. Minerva India Under-served Fund (Tattva series USD), which tracks our onshore strategy, was down -2.38% (net; in USD) for the year. We note that INR lost -4.78% during the year. Please note that, unlike our pre-tax onshore INR returns, reported performance of Minerva India Under-served Fund (Tattva series USD) is net of not only taxes paid on realized gains, but also net of tax accruals on unrealized gains. Headline numbers are therefore not comparable between structures. Over nearly 15 years of deployment in India, Minerva India Under-served strategy was up +17.0% annualized (net; in INR terms), vs. +12.8% and +11.5% annualized gains (gross; in INR terms) in BSE Smallcap and BSE 500 respectively. Importantly, this decade-plus alpha wasn’t driven by skewed asset-class driven prints of liquidity-surge years in 2014, 2017, 2021, or 2023.
Within names that we still held at the end of the year, our best performing names for the calendar were an ‘Auto parts’ name (+53% in INR, following on a +41% gain in 2024) and a ‘Spirits’ name (+23% in INR, following a flattish 2024). Our two worst performing positions were both tail names and collectively accounted for under 4% of the book at the beginning of the year – a Building materials name (down -39%, and a Media name (down -35%).
We ended the year with about 29% liquidity at hand, vs. close to 40% at the beginning of 2025. Within the 9 positions we held at the end of the year (which was incidentally the lowest number of holdings in our book in almost 14 years), we had added exposure in 8 positions over the course of the year. Our heaviest additions during the year were predictably in our top-3 holdings – an Auto parts name (exposure expanded by >60% through multiple additions; stock ended about flattish for the year), a Forestry name (exposure expanded by >30% through multiple additions; stock was down about -14% for the year), and an Apparel name (exposure expanded by nearly 40% through two additions; stock was up +9% for the year).
Elevated valuation differentials only partly explain the material underperformance for Indian headline names vs. emerging peers. Poor absolute growth, delivery against sub-par expectations, and muted employment trends are equally evident. As things stand, we do not expect remarkable underlying growth to alter this in the near term, while headline valuations (still) remain undeniably disconnected.
1. We don’t share Dalal street’s somewhat sanguine view on an earnings re-surge for headline India. We neither see a favorable hiring environment, nor any meaningful improvement in capacity additions to suggest any significant acceleration in headline India earnings. In our October investor letter, we had noted the impact of ongoing IT layoffs in India on discretionary consumption, with material weakness in the tech-cluster[4] across discretionary categories such as luxury watches, or premium apparel and footwear.
We would be surprised if Nifty50 (ex-financials) ends up doing anything over low double-digit[5] earnings growth for F2027E (Mar), vs. the current +17% predictably cheerful street consensus as we pen our year-end update.
2. Headline India’s earnings revisions have been decidedly poor, despite muted expectations. Not only was India considerably more expensive at the beginning of last year (and still is!), but MSCI India’s C2025E earnings revisions have also trailed most of the EM peers over the course of the year. While Latin American markets such as Brazil and Mexico reported slightly less worse cuts, they still ended up with absolute growth numbers ahead of India. Given that these markets began last year at rather palatable multiples, they ended up between 25-30% higher in local currencies (and even higher in USD). As we see it, it’s somewhat striking that headline India even eked out a positive return[6] in C2025 despite unremarkable earnings expectations[7] getting capitalized at near hysteric multiples to begin the year.
3. Quality job creation remains woeful. F2027E (Mar) doesn’t indicate any meaningful turnaround either. Within industries where India could have rightfully competed, ASEAN nations and, to some extent, Bangladesh[8] (for apparel) have remained the top draw for supply chain reconfigurations. India’s continued indifference/feet-dragging to Chinese inward investment has played its part. Ownership issues, security clearances, local sourcing requirements in certain sectors, and sometimes arbitrary snail-paced case-by-case approvals[9] have often discouraged Chinese businesses. Be it electronics components in Vietnam, apparel manufacturing in Bangladesh or Cambodia, or automotive supplies in Thailand, the fact that Chinese businesses themselves have significantly influenced eventual destinations of such supply chain shifts meant that India hasn’t been as large a beneficiary[10]. While majority of supply chain reconfiguration opportunity might arguably have already been capitalized upon, we believe that India’s lag in capital flows can still be at least partly fixed if economic considerations could trump geopolitical posturing. Regardless, even elsewhere, India simply isn’t creating enough quality employment within manufacturing.
Services meanwhile paint a somewhat mixed picture – While GCCs continue to proliferate (and these often eliminate roles at Indian IT vendors[11]), entry-level hiring within IT services has seen sharp cuts. Meanwhile, emerging roles such as data annotation work to train machine learning models[12], for instance, barely pay equivalent of what a full-time food delivery job (often mischaracterized as a ‘gig’ in India) makes, and is simply not the kind of work that is sustainable. Elsewhere, for all the hoopla, data centers simply don’t create enough employment, and yet continue to get automated. In our view, India cannot create materially more than meager 6,000 jobs at such data centers over the next 5-7 years.
If low utilization levels[13], muted industrial approvals data[14] and planned capex[15] of India’s biggest manufacturing companies is anything to go by, ability of domestic manufacturing to drive job creation in F2027E (Mar) doesn’t appear encouraging either. Until recently, even though white-collar job creation was muted, we were at least seeing steady growth in requirements for contract labor. Our last two contractor checks[16] however have shown among the lowest growth readings since we began conducting these checks in 2022.
Dalal street seems to have discovered what has eluded security markets since eternity – ‘Certainty’. Over the last 5-6 years, street has flipped a cardinal rule in investing - Apparently, ‘here and now’ growth in India wasn’t just good to have. Instead, it was a brazen entitlement, and heady growth was getting capitalized at extremely elevated multiples. The apparent certainty with which heady growth has been getting priced in would inevitably end up with a thud. We charted Price/Book and ROAs for 70+ sub-industries (using same components) within BSE 500 at 3 different points – 1. December-end 2007 (around the time that Nifty50 peaked before GFC), 2. December 2010 (before losing a quarter of its value the following year), and 3. December-end 2025. The median component in this exercise began this year at 4.4x book, or 5.1x if one were to exclude Financials. Nearly 45% of sub-industries were trading at >6x book. That compares against significantly lower 18% of sub-industries in December 2010 and 32% of sub-industries in December 2007. For context, Minerva India Under-served’s median holding began this year at 2.1x book[17]. Median operating return of these names meanwhile has barely changed[18]. This is despite the fact that operating returns for most sub-industries saw anywhere between 50-200 bps benefit since the 2019 tax cuts. The argument that elevated valuations are somehow justifiable because headline India’s core operating returns have moved favorably is flimsy.
We disagree with postulations that suggest that headline India would likely entirely capture underlying earnings growth over the next few years. We fully expect India to de-rate from here, and meaningfully offset earnings growth. We would ascribe a material probability to a >20% cut to headline India’s multiples. That it does or doesn’t de-rate this year however is not a fundamental debate.
We remain extremely concerned not only about asset bubbles, but also elevated associated vulnerabilities to overall consumption within India’s top decile[19]. As Indian equities entered, what ended up, relatively speaking, a rather shallow correction about a year ago, we had noted a materially adverse impact on the discretionary consumer within high equity ownership cities[20]. A deeper and more sustained decline could likely test what we believe remains a highly green and entitled investor base in India.
We certainly do not share street’s view/hope on stickiness of the liquidity spigot either - We note that systematic investment plan (SIP) cancellations at Indian equity funds spiked every time over the last 5 years when trailing returns turned somewhat unpleasant – In particular, note the cancellation spikes when trailing returns reported double digit index declines a year ago.
Headline United States looks as poor. For S&P 500, buck shouldn’t stop at ‘Technology’. With underlying profitability for Technology running at nearly 4x[21] rest of S&P 500, and incomparable revenue growth (mid-teens in C2025 vs. nearly nothing for the rest of the benchmark), it’s somewhat bizarre that Information Technology has suddenly[22] become the posterchild for listed equity froth narratives within S&P 500. While a reasonable skeptic should rightly question the runway of AI infrastructure spends once architecture backbones are in place, the debate at some point also needs to shift to the remarkably mediocre ~2/3rd of the benchmark that has gone largely unnoticed. This part is replete with names that are more dependent on the home turf that is now struggling to move the needle on employment, is more dependent on a consumer that is likely to see its real wages flatten/decline this year, and is likely to be a relatively smaller beneficiary of potential USD weakness. Yet, as we write this quarter-end update, this unremarkable 2/3rd of the benchmark inexplicably trades at nearly 21x C2026E, especially when USD is unlikely to be favorable. Unless the growth wheels[23] entirely come off for Technology or one is more focused on trading off of earnings prints, we believe that at least as apparent risks lurk elsewhere within S&P 500. To be clear, we aren’t in the least suggesting that C2026E estimates for benchmark Technology names aren’t loaded with downgrade risks. Instead, we are simply stating that the rest of the benchmark components would need baffling upgrades to growth[24] and operating returns to justify trading where they began the year. That preposterous acceleration and highly disconnected valuations of S&P 500’s non-tech components lack any reasonable foundation at this time. There isn’t an escape valve. We retain our view that headline names in both United States and India should continue to materially drag behind emerging Asia (and that this trade has significant legs).
To access the entire letter, please click here.
[1] Having accounted for a lion share of daily trading volume for nearly all of the period between late C2019 and September 2024, retail share of daily trading volume in India has continued to come off over the last year, reaching low 40s recently.
[2] Average retail oversubscription in C2025 was under 25x, vs. >34x in C2024 (Source: Prime Database)
[3] Minerva India Under-served has handily outpaced all benchmarks off of prior small-cap peaks (beginning in early 2011 and late 2017), and observation of current drawdown off of September 2024’s peak has been somewhat similar.
[4] 50K layoffs had been announced until September end. We considered the following within India’s tech cluster - Bangalore, Hyderabad, Pune, Gurgaon, and Noida. For contrast, the following were classified under the ‘non-tech’ cluster - Lucknow, Bhopal, Ahmedabad, Indore, Bhubaneswar, and Goa.
[5] This compares against Nifty50 (ex-financials) annualized earnings growth of about 12% (excluding the 2019 tax cut tailwind) between F2014-F2024.
[6] In USD terms, MSCI India was barely positive for the year. Also, while MSCI Thailand and MSCI Indonesia reported declines in local currencies, we note that Thai Baht and Malaysian Ringgit were the best performing currencies in Asia last year (both up 10%).
[7] Street was pegging C2025E earnings growth for MSCI India at 16% to begin last year. It is now unlikely to materially cross 7%.
[8] New supply lines have opened from Brazil and West Africa, alongside emergence of new local accessories’ facilities. Renewed focus on sustainability provided an incremental edge, as EU and NA clients increasingly sought out sustainability focused ‘super vendors’. The latter is particularly notable because Bangladesh had historically been perceived to carry relatively high risks in terms of social and environmental compliance.
[9] Likes of Luxshare, BYD, and GWM had all flagged bureaucratic delays before canceling planned investments
[10] ASEAN nations however continued to report solid inward FDI inflows from China last year
[11] Large clients such as Target Corp have fully internalized not only software development but also maintenance work (by bringing contracted resources on payroll). Our discussions suggest that many others such as Ameriprise, Nike, and Best Buy also intend to go down this path over the next few years. Meanwhile, headline Indian IT services names trade at >30% premium over the likes of Gartner and Accenture.
[12] Expected to employ nearly a million people in India by C2030E. India has only trailed Pakistan in creating these jobs over the past 5 years (Source: Revelio Labs).
[13] RBI’s quarterly surveys suggest a low-mid 70s utilization. This somewhat overstates utilization because OBICUS surveys use GVAs as weights and therefore have higher weight of ‘Materials’, which show better utilization levels. Industrial utilization levels are worse in our view.
[14] 6-month rolling federal approvals data on projects (Industrial, Mining, Infrastructure, and Power) have confirmed yoy declines throughout whole of last year.
[15] Planned F2027E (Mar) aggregate capex for manufacturing names within Nifty 100 is expected to grow by barely 3%, which is a sharp deceleration vs. >20% growth in F2026E (Mar). Excluding ‘electricity generation’ capex, expected growth trickles down further to 2%.
[16] Conducted every 6 weeks
[17] Against 2.2x for MSCI Emerging Markets (with 22% weight of ‘Financials’) and 3.7x for MSCI India (with 29% weight of ‘Financials’)
[18] Median ROA of 6.5% now, vs. 7.6% and 5.7% in Dec 2010 and Dec 2007 respectively. Compared against Dec 2007, little over half of BSE 500 sub-industries report higher operating returns today.
[19] Estimated to account for about 2/3rd of India’s overall discretionary consumption
[20] We had segregated some of India’s biggest cities into ‘high equity ownership’ and ‘low equity ownership’ buckets, and evaluated deceleration (or not) in January and February within consumption in certain premium categories (luxury watches, apparel, footwear, hotels, and smartphones). We had noticed significantly sharper deceleration within high equity ownership cities across all categories.
[21] C2025E operating margin of about 26% for S&P 500 Information Technology, vs. barely 6% for the rest of the benchmark
[22] Having exploded for most of last year, flows within AI and Robotic funds have plateaued since early November (Source: EPFR). Meanwhile, short interest within benchmark Technology names also declined towards the end of the year. There is certainly ample room for short interest to build up this year, while fund flows continue to cool off.
[23] CIO surveys are currently suggesting that C2026E spends should outpace the high single digit growth of C2025, with particular focus on software, networking, and storage. Nonetheless, C2026E expectations remain scary high, with nearly +40% earnings growth penciled for the overall Technology sector within S&P 500.
[24] Excluding Technology, S&P 500’s earnings have grown at <7% between 2013-2025, and <5% if one were to adjust for below the operating line drivers (buybacks and 2017 tax cut).