Nifty valuations near post-Covid lows. Alchemy Capital’s Alok Agarwal explains what investors should buy now

India’s benchmark Nifty 50 is entering a valuation reset, with its forward price-to-earnings multiple falling to 17.4 times from 21.5 times two years ago. Alchemy Capital Management Deputy CIO Alok Agarwal says the index is now close to its cheapest level since the Covid pandemic, despite delivering no returns over the period.

Agarwal expects the next phase of gains to come from beyond large caps, with earnings growth emerging in mid- and small-cap stocks and investment opportunities expanding across AI infrastructure, healthcare, manufacturing and selected defence companies.

Edited excerpts from a chat:

Why hasn't the Nifty really become cheap despite giving no returns in the last two years? Were we that overvalued in September 2024?

It has become relatively cheaper, in our view. It may not feel cheap primarily due to no returns and slow earnings growth. Two years ago, in September 2024, the Nifty50 was trading at a one-year forward P/E of 21.5x vs 17.4x now in September 2026. During this period, the Nifty50 has corrected by 11% while earnings kept growing (albeit at a subdued pace). So the multiple did the falling. Current valuations are close to the lowest levels we have seen in the post-Covid era.

Were we overvalued in 2024? Stretched, yes. The one-year forward P/E was barely 8% below its all-time highs and well above historical averages. All this happened while earnings growth slowed for nine straight quarters and remained in single digits. Prices have run ahead of profits, and the last two years have been a period of corrections.

Here's the thing. People anchor "cheap" to Covid-crash levels of 11.5x (one-year forward P/E). In our view, that's a panic price, not a fair value. What we have now is a reasonable one, which is where long-term returns usually start.

The Nifty has underperformed several pockets of the broader market this year, while many mid- and small-cap companies have continued to deliver earnings growth. Is this structural broadening or a late-cycle rotation?

It appears to be structurally broadening, in our view, and earnings trends support this. A late-cycle rotation is often money chasing stories because the leaders are tired. This time, the numbers are doing the work. So far in 2026, the Nifty50 is down about 10%, while the Nifty Midcap 150 Index is up 3%, the Nifty Smallcap 250 Index is up 10%, and the Nifty Microcap 250 Index is up 19%. Behind that, Q1FY27 profits grew 23% for the Nifty Midcap 150 Index, and 31% for the Nifty Smallcap 250 Index, against 21% for the Nifty50. There's a quieter reason too. About 40% of the mid- and small-cap index constituents have changed over two years. The indices keep swapping in growth. The Nifty50 has barely changed (may have changed by less than 10% in the last two years).

But broadening isn't a free pass for everything with a small market cap. Higher growth, and visibility of the same, are generally being rewarded. Moreover, manufacturing has been in focus – led by cyclicals, which have been staging an earnings-led comeback after being overlooked for over a decade between the Global Financial Crisis and Covid.

FPIs have again turned sellers, with rising global bond yields, a weaker rupee, elevated crude prices and geopolitical risks affecting flows. What needs to change for foreign investors to return meaningfully?

FPIs have pulled out about USD 25 billion from Indian equities so far in 2026. That's already more than all of 2025. But look at the drivers. The US 10-year bond yield is around 5%, Brent crude oil has been near USD 100 per barrel, and the rupee has touched INR 96/USD. That's a crude-and-dollar story, not an India story, in our view.

Three triggers could cause FPIs to return: US bond yields coming down sharply, crude settling back much lower than current, and the rupee stabilising. A combination of these could be enough to support a return of flows.

And remember how quickly they came back in mid-June, July and August 2026, adding about USD 7 billion in two and a half months. Historically, foreign money hasn’t come back gradually. It tends to show up all at once, usually when the valuation gap has already closed. Indian valuations look attractive and are already helping improve the risk-reward equation. Moreover, India has underperformed global markets and emerging markets by a record margin over the last two years – making the risk-reward equation tilt in favour of India.

What are your expectations from Q2FY27 earnings? Some think trouble could begin Q3 onwards as the high base effect kicks in.

Q2FY27 earnings are likely to be solid, although likely not as healthy as Q1FY27. Q1FY27 was a genuine positive surprise. Large-cap profits grew 21% against expectations of 14%. Demand appears steady enough to support the top line through Q2FY27. The pain point is energy, in our view. Crude near $100 per barrel squeezes margins, and that's where most of the downgrades are concentrated. The bulk of the high energy costs were absorbed by the oil marketing companies in Q1FY27.

The base effect worry for H2FY27 is fair. Growth may cool from 20%-plus. But cooling from 22% to the mid-teens is normalisation, not trouble. The market seems to confuse a slowing growth rate with a falling profit number all the time.

Revenue growth appears to be holding up. That's the real signal. Margins can face temporary headwinds.

The AI investment cycle is creating demand for power, cooling, transmission, equipment and data-centre infrastructure. Where are the most investible opportunities?

It’s the picks and shovels, in our view, and that's where we've been for a while. Every AI rack ultimately creates a power requirement. India's data-centre capacity was about 1.6 GW in mid-2026, and JLL expects it to reach roughly 6 GW by 2029. That's close to four times the power requirement that has to be generated, transmitted, stepped down, backed up and cooled.

So, the investible chain runs through power generation, transmission and distribution, transformers and switchgear, cables, and thermal management. These businesses have order books, pricing power and a customer base far wider than AI. Manufacturing, railways and urbanisation all need the same equipment. On similar lines, all this power requires silver and copper – both of which are facing deficits.

Is India likely to emerge primarily as a data-centre destination, a supplier to the global AI infrastructure ecosystem, or both?

Both, but likely in sequence. The data centre story is already underway. Hyperscalers have committed over USD 50 billion to self-built Indian capacity through 2029. Add a 20-year tax holiday and low-cost data consumption, and India is positioned as a natural home for compute that serves Indian users.

The supplier story is the bigger prize and it's slower. India may not manufacture the most advanced chips any time soon. But the physical layer of AI is another matter: transformers, cables, electrical equipment, precision components and, eventually, assembly and packaging. In these areas, India may benefit from the China+1 sentiment that's already working in electronics.

In our view, the destination story gets the headlines, but the supplier story builds the compounders.

Within healthcare, where do you see the best medium-term risk-reward: hospitals, domestic pharma, CDMO, diagnostics or medical devices?

CDMO (Contract Development and Manufacturing Organisation) first, hospitals a close second. We believe CDMO has the cleanest tailwind in healthcare: global innovators want to reduce their dependence on China, and India has the chemistry depth to absorb that work. Contracts are sticky and multi-year, and each successful molecule compounds the relationship. That's the kind of growth we find attractive.

Hospitals benefit from the structural demand in the sector, with a real bed shortage, an ageing population and rising income levels. The catch is valuation. A lot of the good news is priced in, so execution on new beds matters more than the theme.

Domestic pharma appears steady, but relatively less compelling. Diagnostics has turned into a price war. Medical devices may have the longest runway, but the sector is still at an early stage and fragmented. For now, it remains an area to watch.

Defence has become a major structural theme, but valuations are demanding. Is this a bad time to start picking defence stocks?

It may not be a bad time to start, but selectivity is critical. The Nifty India Defence Index trades at about 57 times trailing earnings. Over three years, earnings have grown about 24% a year and prices have risen by about 43% a year. That gap is the problem. The multiple has been doing a lot of the lifting.

The theme itself is real: indigenisation, export ambition, and budgets that don't reverse with election cycles. But a great theme at full price just means lower returns ahead.

The world spends about 2.5-3% of GDP on defence. This could rise with an increase in geopolitical conflicts, the US removing security support and the outcome of the recent US-Iran war. During the Cold War period, this number was at 6% of GDP. We wouldn’t be surprised to see defence spending inching back towards 6% of GDP levels over the next few years. This can increase the total addressable market significantly.

In our view, the runway is long, selecting the right opportunities at the right time could make the journey fulfilling. A staggered approach may be prudent rather than going all in. The focus could be on businesses where orders are translating into execution and cash flows, not just order-book headlines. And look beyond the obvious names towards electronics, components and the supply chain, where the growth is similar and the crowding is thinner.

What percentage of a long-term equity portfolio should reasonably be allocated to midcaps and smallcaps, given the possibility of a sharp drawdown?

That is the classic balance. In India, we are seeing the emergence of newer sectors and the revival of the cyclicals after a long hiatus. The bulk of these opportunities have been in the broader markets. Hence, it is important to be adequately present in the broader markets. But the allocation should be sized according to the drawdown an investor can sit through, not the return one is hoping for.

For a long-term investor with a horizon of seven years or more, who has lived through a bear market without selling at the bottom, an allocation of 35–50% in mid and small caps may be considered reasonable.

A portfolio's biggest risk may be investor psychology, not the asset class. Smallcap drawdowns have historically been deep and long, sometimes lasting years rather than months. The allocation only works if one is still holding when the recovery comes.

Every runner knows this. The pace you can hold for 42 kilometres is slower than the pace you can hold for five. Pick the pace that gets you to the finish.

(Disclaimer: Recommendations, suggestions, views, and opinions given by experts are their own. These do not represent the views of the Economic Times)