How smaller companies can help tackle Asia’s concentration risk problem
Asia’s growth story extends well beyond its largest companies, with smaller businesses offering a wider range of opportunities.

Duration: 5 Mins
Date: 07 Sept 2026
Let’s first be clear about what concentration risk is. Broadly speaking, it’s a phenomenon that occurs when an investment portfolio – or even an entire financial system – relies too heavily on a single asset, sector or region.
The so-called “Magnificent Seven” technology stocks have been concentration risk’s poster children for several years. Together, they represent a potentially unhealthy proportion of numerous US indexes.
For example, Nvidia alone accounts for around 7% of the S&P 500. This goes a long way towards explaining why the company’s quarterly earnings calls are routinely preceded by several weeks of fevered speculation and not a little trepidation.
The S&P 500 tends to soar if Nvidia does well. But if Nvidia stumbles – say, by falling short of analysts’ forecasts – the benchmark instead tends to take an outsized hit, with passive investors often unpleasantly surprised by the abruptness and magnitude of their losses.
Amid ever-fluctuating sentiment over the sustainability of the AI revolution, such events are becoming more frequent. As a result, more investors are recognising the merits of diversification.
Yet it’s vital to appreciate that diversification comes in many forms, some of which are likely to prove more effective than others. To underline this point, let’s consider what would happen if a passive investor were to start tracking some of the major indexes in Asia.
Sure enough, this could help offset the concentration risk arising from a narrow focus on just a handful of US tech behemoths. However, the fact is that many Asian benchmarks are vulnerable to the very same threat – and in some cases the dangers are even more extreme.
The shock that tore through Asian markets at the end of July demonstrated as much. Both Taiwan’s Taiex and South Korea’s KOSPI plunged dramatically in the face of a sudden flurry of AI-related negative dynamics.
Why? These indexes are even more weighted towards technology companies than the S&P 500. A single chipmaker, TSMC, is responsible for over half of the Taiex’s overall market capitalisation, while two others, Samsung Electronics and SK Hynix, comprise more than half of the KOSPI’s.
We might therefore reasonably infer that relentlessly piling into the most popular names, wherever in the world they might be, can be perilous. By extension, we might also reasonably infer that smaller companies can be attractive from a diversification perspective.
Real growth engines and real diversification
Over 50% of global growth comes from Asia. Believe it or not, this figure can’t be uniquely attributed to a tiny array of trillion-dollar tech businesses – regardless of how potent their performance might be from time to time.
Many of the real engines of growth can instead be found at the opposite end of the market-capitalisation spectrum. This is where our fund seeks out the region’s hidden gems.
We draw on a combination of in-depth research and direct engagement to identify small-cap and mid-cap businesses that can outperform over the long term. Consequently, our fund bears scant resemblance even to its benchmark index – less still to the lopsided likes of the Taiex and the KOSPI.
It’s right to acknowledge that some of the companies we favour are linked to the AI boom. They can be seen as “picks-and-shovels” stocks, which is to say they rank among the behind-the-scenes enablers of the far-reaching transformation now under way.
Take South Korea’s Hyundai Marine Solution. We originally invested in the business when it was dedicated purely to repairing and refitting ships – a specialism sufficient to generate 50,000 purchase orders and 80,000 deliveries a year.
Now, though, Hyundai Marine Solution looks likely to carve out a useful sideline in maintaining engines that have been reconfigured to power what are known as FDCs – floating data centres. We must confess that we didn’t see this development coming.
This is a remarkable snapshot of the second-order – and even third-order – impacts of artificial intelligence’s spread. In light of such pervasiveness, investors could be forgiven for assuming there’s no other game in town.
Yet there are still many businesses that have no significant connection to AI. Our holdings include China’s Zhejiang Shuanghuan Driveline, a brand-agnostic maker of gear components; Century Pacific Food, a leader in branded food products in the Philippines; and Aegis Logistics, India’s main importer and handler of liquified petroleum gas.
We would argue that variety of this kind represents genuine diversification. In our view, it’s not enough merely to diversify geographically – particularly if doing so simply adds to concentration risk.
True diversification is practised across multiple dimensions, including market capitalisation. Its fundamental purpose is to spread and reduce risk, not to duplicate and heighten it.
In our experience, smaller companies have a big role to play in this respect. They may not lay claim to countless column inches, but they can prove eminently capable of stealing the headlines from their larger counterparts over the long run.
Companies selected for illustrative purposes only to demonstrate the investment management style described herein and not as an investment recommendation or indication of future performance.
Important information
Risk factors you should consider prior to investing:
- The value of investments, and the income from them, can go down as well as up and investors may get back less than the amount invested.
- Past performance is not a guide to future results.
Other important information:
Issued by abrdn Fund Managers Limited, registered in England and Wales (740118) at 280 Bishopsgate, London EC2M 4AG. The company is authorised and regulated by the Financial Conduct Authority in the UK.
Find out more at aberdeeninvestments.com/aas or by registering for updates. You can also follow us on Facebook and LinkedIn.





