Up 8% when small caps were bleeding. How?

Summary: The small-cap benchmark is down 1.5 per cent over the last two years. Aditya Khemani's fund is up close to 8 per cent over the same period. The decisions that produced that gap were made two and a half years ago.

When markets turn choppy, most small-cap funds feel it first and hardest. Aditya Khemani’s doesn’t. As Head of Equity at Invesco Mutual Fund, Khemani has built one of the best-performing small-cap funds in the country. And notably, a large part of that alpha came in the last two years, when the small-cap benchmark was down roughly 1.5 per cent while his fund was up close to 8 per cent.

With over 20 years of experience and a framework built around quality, risk management and bottom-up stock picking, he has shown that small cap need not mean reckless. In this conversation, he explains how he thinks about portfolio construction, position sizing, and manages valuation risk in a richly priced market.

We’d like to understand a little more about how your investing style has helped you navigate the last two years of volatile markets, and in turn generate the alpha that makes Invesco India Smallcap one of the top funds in its category. For context, since September 2024, the small-cap benchmark is down about 1.5 per cent, while your fund is up close to 8 per cent in that time. What are the two or three key decisions that led to close to 10 per cent outperformance in such a volatile market?

The first thing is the quality of the portfolio. We keep talking about risk management, and for us as fund managers, it’s the most important thing apart from returns. If you make the quality of the portfolio good, it stands out in bad times — because in good times, most portfolios do well, but it’s in bad times that risk management really comes to the fore. So the first thing is maintaining a high quality of the portfolio.

Second, two and a half years ago, I decided that we should not be running a very small-market-cap fund. So we purposefully reduced exposure to stocks below Rs 7,000–8,000 crore in our portfolio. Over the last 18 months, a lot of the pain has been in that part of the market, so staying away from the smaller-market-cap names has helped us. Today, less than $1 billion market-cap names make up just 8–9 per cent of the portfolio.

Third — and this isn’t specific to the small-cap fund — two and a half years back, after four to five years of a rally, a lot of spaces had become very expensive, and investors weren’t differentiating between good and bad businesses; everything was getting a high valuation. At that point, we reduced exposure to some of the very overheated sectors, including industrials. Most industrial companies two and a half years back were trading at 40–50 P/E — these are cyclical companies. So when even a slight bit of earnings or growth deterioration happened over the last two years, there was a knock in terms of the multiple investors were willing to pay. So navigating the overheated parts of the market slightly better also helped us.

These are things we want to do consistently, because small cap tends to be riskier than other parts of the market, so it’s on us as fund managers to manage that risk. The common notion is that because you’re taking the highest risk in small caps, the return expectation should also be high — but that hasn’t proven true historically. If you look at the last 10 years, the small-cap index has given roughly 14–15 per cent CAGR, and mid caps have given about 17 per cent CAGR. So it’s not the case that small cap will be the best-performing space just because you’re taking more risk. That’s why risk management becomes very important, especially when the market is overheated — and this has helped us not just in the small-cap fund, but across our funds, in navigating the last two years.

Investors generally expect a small-cap fund to have sector overweights and underweights that are much starker relative to the benchmark. When we looked at your fund, we found that over the last three years or so, you’ve been roughly 9 per cent underweight in Materials, and about 4 per cent overweight in Consumer Discretionary, 3 per cent in Technology, and about 2 per cent in Financials. What’s your approach to constructing a small-cap portfolio, and how does that translate into these sector-level outcomes?

First, look at the benchmark. The small-cap index is actually a very poor construct for growth. Unlike most other benchmarks, where Financials is the largest sector, here it isn’t — Industrials is the largest, with about a 19–20 per cent weight. Second, some sectors like Healthcare have a much higher weight in this index than in, say, the large-cap index, where Healthcare is only about 5 per cent. So the sector composition of the small- and mid-cap indices is very different, and a lot of these sectors are sectors of the future with much higher growth potential. That’s my starting point — look at how the index is constructed, then think about how you can do better.

For me, four sectors are critical for any portfolio I build: Financials, Consumption, Healthcare, and Industrials. Broadly, each of these would be 20–25 per cent of the portfolio, and the balance 10–20 per cent — would come from allocations to sectors like Materials and Real Estate.

Within each sector, though, there are very different business models. Take Healthcare — the traditional part is pharma producers, which is what we’ve largely seen over the last 10–15 years, but there’s also a services piece: hospitals and CDMOs (contract development and manufacturing organisations), which is the non-traditional side. We have much more exposure to hospitals and CDMOs. Within Consumption, we don’t have too many traditional consumer staples companies — we have more exposure to quick commerce, and convenience is a theme we play. Aviation, even though it’s classified as industrials, we look at as a consumption stock. Within Financials, banking might look homogeneous, but there are sub-sectors like capital market players — exchanges, wealth managers, asset management companies — and we’ve historically put more weight there than on the lending side.

So the construction within each sector is very different from how the index is built. Our strength has been more bottom-up stock picking. Sectors like Industrials and Healthcare are very heterogeneous, and even within Financials, non-banking models differ significantly from banking. We’ve taken a fairly non-traditional approach to portfolio construction rather than relying on what’s worked over the last 10 years.

On Materials specifically — it’s a large sector in the small-cap index, around 15 per cent, and again very heterogeneous: cement, chemicals, fertiliser, oil and gas companies. We’ve chosen lower exposure there because we tend to favour companies that are more stable and less cyclical. A lot of companies within Materials have very volatile earnings — a metal company’s earnings, for instance, depend heavily on metal prices, which are influenced by what’s happening in China and government duty measures. We’ve chosen to avoid areas like that, which is why we’re largely underweight in metals.

So sector by sector, we decide where we want to be overweight from a top-down view, and then make the call based on the bottom-up work we do on individual stocks.

When we do in-house performance attribution for funds, we usually find that winners and losers largely cancel out, with a few select stocks driving the real alpha. In your fund over the last three years, we found one financial-sector stock that’s gone up roughly 10x, and a recently listed technology-sector company that’s up three to four times in under a year. Both were big winners, but with seemingly different position-sizing strategies. What was your initial thesis on these picks, how much do you attribute to that thesis playing out versus luck, and how do you generally decide position sizing?

Position sizing is a function of several things. First, we look at the index construct — where a particular theme or sub-sector sits within the index, and what its weight is. That’s the starting point, because at the end of the day any fund manager wants to beat the benchmark, so you need to know clearly what you’re targeting to beat.

Second, it’s a function of your learnings — meeting people, meeting companies, understanding what’s happening in the environment — which leads you to themes you like for the future. For example, hospitals is a theme we’ve played; financialisation is a theme we play; convenience is a big theme we’re playing. You write down the themes you want to play, then look at which stocks within the index belong to that theme. Through a process of elimination, you narrow that down — say, from eight stocks in a “financialisation” theme to two or three. Then, given the index weight for that theme, you decide how much you want to be overweight, and distribute that weight across the stocks you’ve chosen, based on valuation, management quality, and so on.

That’s both an art and a science. For example, CDMO was a space we were very positive on, and there was one company we liked far more than the others, so we gave it a disproportionate share of that allocation. Hospitals, on the other hand, had several good options across small, mid, and large caps, so that allocation got spread across multiple names.

For a small-cap portfolio specifically, I mentally cap the largest single holding at around 5 per cent, and while there’s no hard floor, I like taking a minimum of about 1.5 per cent in a stock — below that, even if it does well, it won’t meaningfully move the portfolio’s return, so it’s not worth holding. That’s why it’s a fairly tight portfolio — around 65 names today, but the bottom 15 names make up hardly 5 per cent of the portfolio. 

So about 50 names make up 90–95 per cent of it, and it’s a carefully selected list built through top-down theme identification and bottom-up stock selection.

Some sectors are more heterogeneous than others — Pharma/Healthcare, Consumption, and Financials together make up about 60 per cent of the portfolio, and adding Industrials (16–17 per cent) takes it to about 80 per cent. Over the years we’ve also developed a framework around cyclicals — cyclical investing requires being a good trader in terms of entries and exits, which honestly isn’t our strength, so we’ve chosen to largely avoid highly cyclical names and instead focus on more stable compounders. That’s also helped us over the last two years.

As a style, we don’t trade very often — we want to buy into a theme and stay with it unless the investment thesis breaks, whether that’s at the sector, theme, or stock level. Beyond that, valuation and relative attractiveness can prompt some changes, but it’s a fairly dynamic process overall. That said, if you compared our fact sheet from six months ago to today, you wouldn’t see too many changes.

Looking at the top holdings in your fund — or really any stock with more than 2 per cent allocation — most of them appear, at face value, richly valued, or you could say they have fairly high expectations priced in. What’s your strategy for picking these high-growth stocks, and how do you manage the valuation risk, given that a missed guidance or execution delay could trigger a de-rating? And how do you decide when to enter and exit such stocks?

This is a question we could go on and on about, but let’s start with what the risk actually is when you buy a stock. A one- or two-quarter miss is really a risk for a very short-term investor. When we look at companies, we don’t focus too much on what’s going to happen over the next year or two — there will be earnings misses, because companies operate in different circumstances, and it’s not realistic to expect a perfectly linear trajectory. There have been so many disruptions over the last five, six, seven years — we’re going through one right now too — so we try to be reasonable about what a company can realistically deliver, and we’re fairly liberal in how we judge a company based on one quarter’s results.

On picking companies — the small-cap space is very heterogeneous, and there are different approaches. One is to buy a niche company that dominates a small total addressable market — a large fish in a small pond. The other is to buy a small fish in a large pond. I’ve always preferred the second, because a large fish in a small pond will eventually run out of growth, whereas a small fish in a large pond, if it executes well, can become much larger.

When we evaluate a business, we look at two things: the business itself, and the promoter and management. The best combination is a good business run by good people, and the weight you give a stock depends on how that combination, along with valuation, stacks up. Promoter and management quality matters enormously to us — a good top management can pivot even an average business into a good one over time. There’s also a difference between the “zero-to-one” journey of a company and the “one-to-ten” journey — the skill set needed to scale up is very different from the skill set needed to start something, and in most cases, promoters alone can’t manage that scale-up. You need good professional management, a strong second line, the right culture, and importantly, “fire in the belly.”

A lot of these promoters are already worth Rs 10,000–15,000 crore, and at that point it would be natural for them to want to take things easier — but our journey with the stock is only starting, so that hunger has to still be there.

That’s why we never start with valuation. If you look at a company trading at 50 times forward earnings, your mind immediately judges it as expensive. Instead, we first assess the business, management, and everything else on its own merit, decide what we think a fair price is, and only then look at the one-year forward P/E — and even that alone isn’t sufficient. Because these are relatively small companies — Rs 10,000–15,000 crore market cap — absolute market cap becomes an important lens: is this a small fish in a large pond that could become very large? Terminal value matters a lot for these companies, even though it’s hard to know exactly what will happen five or ten years out. A lot of these companies’ financials — working capital cycles, margins — aren’t yet optimised, and as they scale, you’d expect those to improve. So looking only at one-year forward P/E doesn’t do justice to all of that.

Ownership structure also matters — an owner-operator business needs a different lens than an owner-owned business run by professional management, or one backed by private equity. In a typical one-and-a-half-hour management meeting, an hour might go into assessing the second line of leadership — what they’re doing, what’s happening. One question we often ask is how many holidays the promoter takes in a year — if the promoter never takes a holiday, that suggests the company can’t run without him; the more holidays he takes, the more it suggests the business can operate independently. There are a lot of nuances that go into this — investing can be as simple or as complicated as you make it. It’s a fairly comprehensive assessment, rather than simply looking at a one- or two-year forward P/E and deciding whether to buy. 

Also read: What does it take to top the category?