Last month, I sat across from Kenny, a software engineer in his early thirties, based in Singapore, someone who reads financial news, has a brokerage account, and thinks carefully about his money. I asked him about his portfolio. He listed: Apple, Microsoft, Google, Nvidia, and Amazon.

I asked about Singapore stocks. He paused. “I don’t really look at those,” he said. “I just feel like I understand tech companies better.”

He uses an iPhone. He uses Google Maps. He watched the Netflix documentary about Enron. He follows Elon Musk on X.

He does not follow Singtel’s earnings calls.

He was not investing in what he understood. He was investing in what felt familiar. Those are not the same thing.

The data behind the anecdote

We work on an AI-native portfolio intelligence platform. Over a recent two-month period, we analysed 82 anonymised retail portfolios submitted by investors across Singapore and Vietnam. What we found was not what conventional financial theory would predict.

The three most frequently held securities across our sample were Apple (AAPL), Microsoft (MSFT), and JPMorgan Chase (JPM), appearing in 32.9 per cent, 32.9 per cent, and 30.5 per cent of portfolios respectively. Gold (GLD) and long-duration US Treasuries (TLT) each appeared in 29.3 per cent of portfolios.

Not a single SGX-listed security appeared in the top 30 most commonly held positions.

Read that again. In a sample where the majority of users are Singapore-domiciled retail investors, no Singapore-listed stock was commonly enough held to crack the top 30.

Behavioural finance has a well-established concept called home bias, the tendency of investors to overweight domestic stocks relative to the theoretically optimal global portfolio. French and Poterba documented it in 1991. It has been replicated in virtually every market studied since. The academic consensus is that investors buy what is local, familiar, and proximate.

Our data suggests something has changed, or at least, something is changing at the leading edge of digital investor behaviour in Southeast Asia. These investors are not exhibiting home bias toward Singapore. They are exhibiting a different bias entirely: anchoring to the US mega-cap companies whose products they use every single day.

We call it reverse home bias. And it carries risks that standard suitability frameworks were not designed to catch.

Also Read: Tried-and-tested marketing strategies for startups across all stages in Singapore

The mechanism: You invest in your ecosystem, not your address

This is not simply the observation that technology has lowered the barriers to international investing, though that is true. The question is not whether you can buy Apple from a Singapore brokerage; the question is why one in three retail investors in our sample have chosen to.

The answer, I suspect, is cognitive availability. Apple is not a foreign stock to someone in Singapore. It is the company that made the phone in their pocket, the laptop on their desk, and the watch on their wrist. It appears in their social media feeds, in the financial content they consume on YouTube and TikTok, and in the investment discussions on Reddit and Seedly. JPMorgan appears daily in financial news. Microsoft is their workplace operating system.

This is the availability heuristic, a concept from Tversky and Kahneman’s foundational work on cognitive bias, operating across national borders. What you can easily imagine tends to feel safer. What saturates your attention feels like information, even when it is not.

The Singapore investor who holds Apple is not making an informed bet on AAPL’s earnings trajectory relative to its valuation. They are making a bet that feels safe because they cannot imagine a world without iPhones.

Why this is a problem worth naming

A portfolio concentrated in US mega-cap technology and financial stocks is not a balanced, internationally diversified portfolio. It is concentrated exposure to: US equity market risk, Nasdaq sector concentration, US Federal Reserve interest rate sensitivity, and USD/SGD currency risk.

None of these risk factors appears on a standard retail investor suitability questionnaire. Brokers ask whether you are growth-oriented or conservative. They do not ask: Does your portfolio move in lockstep with Nasdaq? Are you exposed to a single country’s monetary policy? Do you hold any asset that is genuinely uncorrelated with US equities?

Also Read: Singapore lands OpenAI’s first lab outside the US with US$225M commitment

The scoring system of DNA Score, a composite behavioural risk metric computed from portfolio position data, flagged meaningful risk differentiation in the sample. Portfolios in the Speculative Investor archetype, which held high-momentum narrative stocks like MicroStrategy (MSTR) and Palantir (PLTR), scored a mean of 53 out of 100. The more diversified archetypes scored in the mid-80s.

The investors with the low scores were not taking conscious speculative positions. They were following communities, chasing stories they had absorbed on financial social media, and concentrating on names that felt exciting and familiar in equal measure.

The most dangerous portfolio is the one that feels safe and is not.

What the AI era changes, and does not

There is reason to think reverse home bias will intensify, not diminish, as AI-assisted investing goes mainstream. When a retail investor in Singapore asks ChatGPT which stocks to consider, the names most represented in the AI’s training data are overwhelmingly US large-caps. When TikTok’s finance creators in the region discuss their portfolios, they discuss Apple, Nvidia, and Tesla, not Keppel or ComfortDelGro.

The infrastructure of financial information has globalised faster than the infrastructure of financial advice has localised. The result is that millions of first-generation retail investors in Southeast Asia are being guided by content optimised for engagement rather than advice optimised for their specific risk profile, currency exposure, and financial goals.

This is not an argument against holding US equities. It is an argument for holding them consciously, knowing why you own them, what risks they carry, and whether your overall portfolio is as balanced as it feels.

Also Read: RIE2030’s hidden flaw: The one capability Singapore’s startups are missing

What you can do right now

Run your portfolio through a behavioural diagnostic. Not the risk tolerance questionnaire your broker sent you when you signed up; those are designed to satisfy regulatory minimums, not to give you genuine insight. A real diagnostic looks at what you actually hold, computes your concentration, identifies your factor tilts, and tells you which behavioural biases are embedded in your current positions.

A true behavioural finance system processes your portfolio and returns a DNA Score, a breakdown of seven behavioural bias flags, and a regime-aligned action plan. It requires no broker credentials, no passwords, no transaction data, just your positions.

And the question is worth asking: when you look at your portfolio, are you seeing a strategy, or are you seeing a reflection of your screen time?

The post Why Singapore investors hold more Apple than Singtel, and why it should worry you appeared first on e27.

When we run citation checks for clients at ESBO Ltd, my agency, the sources assistants quote for buying questions are rarely brand websites or big media. They are creators: a YouTube comparison from someone who tested four products on camera, or a forum thread with real numbers in it. The people being quoted are, almost without exception, not being paid for it.

That is a pricing failure, and I doubt it lasts long.

The machines quote creators now

The scale surprised me when the data landed. In January, Adweek pulled together citation research from four independent firms, 6.1 million AI citations in all, and found that YouTube had overtaken Reddit as the most quoted social platform in AI answers. YouTube now appears in roughly 16 per cent of assistant responses, and its share of social citations more than doubled between August and December 2025 while Reddit’s halved.

The why matters more than the ranking. Models reach for creator content on buying questions because it looks like evidence. Brand pages make claims. A creator video shows the thing working, with a transcript, specific numbers and an audience arguing in the comments. When an assistant has to answer which accounting tool suits a small Philippine agency, somebody’s tested comparison beats a feature page.

The region built its house on the unquoted platform

Here is the uncomfortable part for Southeast Asia. This region’s creator economy runs on TikTok to a degree no other market matches, with over 150 million active users regionally, and Momentum Works found content commerce drove 32 per cent of Southeast Asian e-commerce GMV in 2025. As a selling engine, it works. For the answer layer, it barely exists: the same Adweek data set has AI systems citing YouTube about 50 times more often than TikTok and 18 times more often than Instagram. Short vertical video with thin transcripts is close to invisible to the machines assembling recommendations.

Also Read: The creator economy is distribution, not marketing. Most Asian businesses are still scaling it like a campaign

The reach economics stay brutal at the same time. TikTok’s Creator Rewards Programme was still not live in a single Southeast Asian market as of mid-2026, and creators with mainly domestic audiences earn effective payout rates measured in cents per thousand views. A creator in Manila or Jakarta needs a multiple of the audience a London creator needs to earn the same platform payout. So the region’s creators are optimised for a currency that pays them worst, on the platform the answer layer reads least.

Pricing the quote

Influence stopped being a follower count a while ago; citation data just makes the replacement measurable. A reviewer with 30,000 subscribers whose comparison video gets quoted whenever assistants answer a category question carries more commercial weight than a lifestyle account with three million followers and zero citations. The tracking tools to see this per brand and per market already exist, and what brands can measure, they eventually price.

In practice that means paying creators for durable presence in the content machines quote, rather than for a burst of reach. A two-year-old tested review that assistants keep citing is media that never stops running. One warning belongs here: sponsorship has to be disclosed, and undisclosed astroturf is the fastest way to lose the audience and the citations together, since models lean on community validation. The value comes from genuine testing with a sponsor attached, never from a script.

Also Read: Why reach doesn’t equal credibility in the creator economy

For founders reading this from the brand side, I argued in June that startups need public proof before they scale. Creator citations are becoming part of that proof layer, and in this region they are still remarkably cheap to earn honestly.

Building citability

Four habits separate quoted creators from merely famous ones.

Own a narrow set of questions. “Best payroll tools for Indonesian SMEs” gets answered by machines thousands of times a month, and somebody’s testing will be the source. Pick the questions and become that somebody.

Go where transcripts live. Long-form video with clean captions, plus a blog or newsletter carrying the same findings in text. Machines read words. Give them words.

Publish numbers. “It felt fast” does not survive summarisation. “It processed 500 invoices in 40 minutes” does, and gets attributed.

Keep your name consistent across platforms, so authority accumulates to one entity instead of scattering across handles the models cannot connect.

For a decade this region’s creators have been paid for attention. The machines now pay attention to something else, and the first creators and brands to price it will look early for about a year, then look obvious.

—

The post Paid to be quoted: The creator revenue line Southeast Asia hasn’t priced yet appeared first on e27.

Bitcoin rose 0.82 per cent in 24 hours to US$76,318.25. Ethereum gained 0.80 per cent to US$2,418.94. The total crypto market cap increased one per cent, and the broader crypto market rose 0.99 per cent. These numbers point to a single conclusion. A relief bounce tied to the Federal Reserve lifted the entire asset class.

No coin-specific catalyst appeared in the data. The primary force came from the central bank. Bitcoin slightly underperformed that broad rise even as it gained. Ethereum tracked its larger peer almost exactly. This synchronised move indicates that the market is currently driven by macro headlines rather than project-level news.

The Fed raised rates by 25 basis points on September 16 to a target range of 3.75 per cent to 4.00 per cent. Market participants had widely anticipated this unanimous decision. The confirmation removed near-term uncertainty. Risk assets responded with a modest rally. The two largest digital assets moved in lockstep with that broader tide.

Bitcoin’s 90-day correlation with gold recently hit a multi-year high. That detail matters. It shows the leading cryptocurrency now trades more like a macro asset than a speculative tech bet. Ethereum remains highly sensitive to central bank cues and Bitcoin’s direction in the short term. The move has less to do with each network’s fundamentals and more with a market-wide sigh of relief.

This is a beta trade, not a fundamental repricing. A priced-in event often produces this kind of reaction. Traders sell the rumour and buy the fact. The fact here was a rate hike that no longer surprised anyone. The market had already absorbed the news before the Fed spoke, so the actual announcement simply cleared the air.

Also Read: The CLARITY Act vote could send crypto to US$2.73T or crash it to US$2.6T

Supporting data in derivatives markets adds nuance. Bitcoin open interest fell 3.1 per cent. Liquidations dropped 65.79 per cent. That decline in forced selling suggests a calmer backdrop. Bitcoin dominance stayed elevated near 58.85 per cent. Capital has not rotated aggressively into riskier altcoins.

Instead, it remains defensive. Ethereum told a slightly different story. Average perpetual funding rates rose 40.74 per cent over 24 hours to +0.0053 per cent. Some derivatives traders leaned cautiously bullish. The absolute rate stayed far from extreme levels. Ethereum also benefited from its place in the Layer 1 narrative, which posted a 0.99 per cent sector gain.

Risk capital is rotating toward large-cap blockchain platforms, but it is doing so selectively. Bitcoin still leads. Ethereum follows. That relationship defines the current market structure. The lack of a leverage washout and the sustained dominance of the largest asset create a stable floor, but they also limit upside momentum. When capital stays defensive, rallies tend to be measured and shallow rather than explosive.

Institutional flows provide the most important test. U.S. spot Bitcoin ETFs recorded US$450 million in outflows on September 15. That figure shows hesitation among institutional investors. A return to net inflows would confirm renewed demand. Until then, price stability rests more on reduced selling pressure than on a fresh wave of buying.

Ethereum faces a similar question. The daily ETF flow report will show whether spot Ethereum ETF flows turn positive in the next 24 to 48 hours. Positive flows would confirm a return of institutional interest. Sustained outflows could pressure the support zone. The bounce then looks technical rather than durable.

This flow data matters more than any single derivative metric because it reflects real capital allocation from large investors. Without that capital, the rally depends on short-term traders and macro sentiment. That foundation is thin and can crack quickly if the next data release or policy comment shifts the mood.

Also Read: The Fed is the real crypto story, Bitcoin and Ethereum are just following

Technical levels define the near-term battlefield. Bitcoin trades just above the US$75,000 support level, which has held for weeks. If the largest asset holds above US$75,000, a retest of US$78,189 resistance becomes possible. A break below US$75,000 would shift focus to the next support near US$74,000.

Ethereum consolidates between support at US$2,350-US$2,400 and resistance at US$2,500-US$2,600. Its 4-hour RSI sits at 53.17, a neutral reading. A daily close above US$2,500 would signal a breakout attempt. A break below US$2,350 would risk a deeper correction toward US$2,200.

The market is in a wait-and-see mode. It balances relief from the Fed against lingering regulatory uncertainty from the failed CLARITY Act. That legislative setback removed a potential positive catalyst and left the market without a clear regulatory path forward. Without that path, institutional investors may continue to hesitate, and that hesitation shows up in ETF flows.

In my view, the synchronised price action tells a story of a market where macro forces set the tone but internal dynamism remains weak. The Fed-induced relief rally is welcome. It is also fragile. It is a pause, not a pivot.

The path forward depends on two developments. One is that ETF flows must reverse from negative to positive. That shift would provide fresh institutional demand. The other is that both assets need convincing technical breaks above resistance. Bitcoin must reclaim and hold above US$78,189. Ethereum must close above US$2,500. Without those confirmations, the crypto complex remains vulnerable to the next macro shock or regulatory headline.

The high correlation with gold and Bitcoin’s persistent dominance show that capital seeks the safest harbours within the asset class during uncertainty. Until capital rotates more clearly into Ethereum and beyond, the recovery remains a beta-chasing exercise rather than a genuine broad-based bull market.

The next 24 to 48 hours of ETF flow data will offer the primary real test of whether this relief rally has legs. I would watch the US$75,000 level for Bitcoin and the US$2,350 level for Ethereum as the lines that separate consolidation from correction. I would also watch funding rates for signs of overheating. A sharp reversal there could trigger a squeeze and undermine the calm that currently supports prices. For now, the market has bought itself time, but it has not earned a new trend.

—

The post Why did Bitcoin and Ethereum move in near-perfect lockstep after the Fed rate hike? appeared first on e27.

Picture the scene. The lights dim. A drone shot. A cello. Employees smiling at the camera. Then the line: We are changing the world.

The film was beautifully made. The investor picked up his phone ten seconds in.

I didn’t set out to start a company over that. But a founder I was coaching kept running into the same wall — and eventually asked me a question I couldn’t ignore.

His startup was heading to CES. He needed a film. He’d already worked with media agencies, spent real money, and wasn’t satisfied. The results were professional. Polished. Something was missing underneath. He asked me: How should we actually do this?

The problem wasn’t the marketing firm or the production company. It was that they showed up too early.

Before anyone wrote the copy or picked up a camera, someone needed to ask the questions an investor would ask:

Why would an investor care? Where would they attack this business? What changes the investment case — and what can wait?

Answer those first, and the agency has something real to work with. Skip them, and even a beautiful film is built on the wrong foundation.

That gap — between investment logic and production — sat between professions. Nobody owned it.

So I jumped in. Not with a business plan. The founder had a problem; I thought I could solve it. That decision was closer to just do it than anything I’d call strategy.

Also Read: Taiwan bets on Gen Z founders to move beyond its chip-supplier image

I began with the investor’s questions. What is this business worth paying attention to? Where’s the evidence? Why now? Then I approached it as a journalist — strip away the company’s own language, find what makes an outsider stop.

Only then did I think about the film.

The first real test came before CES. I was working with a startup entering the Korea Ministry of SMEs and Startups’ Global IR competition — 92 companies in the field, every one of them with a deck, a pitch, a story they believed in.

What I focused on wasn’t the slides. It was the sequence of recognition: what does an investor see first, and does it make them want to see the next thing?

The startup won. Grand Prize, out of 92.

I didn’t think much of it at the time. One competition. Maybe the company was simply strong.

Then came CES 2025.

This time the scope was wider: pitch deck, investor film, and a piece examining the company’s technology with the rigour of business journalism rather than the language of a brochure. Three formats working in sequence — 90 seconds earns attention, the article builds conviction, the deck closes the argument.

The startup went on to win a CES Innovation Award — and raise funding.

That’s when I stopped thinking of this as pitch coaching.

Maybe this wasn’t a better way to make a pitch. Maybe there was an entire category missing between investment logic and production.

Founders know their companies better than anyone. That becomes a liability when they have to explain them to someone who doesn’t.

When investors don’t respond, the instinct is to add more. Another slide. More market data. A longer technical explanation. A 20-page deck becomes 30, then 40. The assumption: if the investor has enough information, eventually they’ll understand.

Also Read: Why Beyond Border thinks visas are now part of the founder playbook

But nobody gives you time to explain.

Investors spend an average of 2 minutes 14 seconds on a first-pass deck review, according to DocSend analytics, 2024–2025.

The investor doesn’t owe a founder 40 minutes of attention. The founder has to earn the next minute.

Recognition happens before a paragraph is finished. Analysis comes after — but only if recognition happened first. Every slide you add before earning that moment is a petition to a decision that hasn’t started yet.

The question stopped being: How do I explain everything?

It became: What does an investor need to recognise first?

The purpose of 90 seconds isn’t to replace the next 60 minutes. It’s to earn them.

I became fairly ruthless about this. If I can’t make the investment case in 90 seconds or on one page, I don’t make the pitch longer. I go back to the business. Because sometimes the problem isn’t the story. You may not have a fundable business idea yet.

That’s why I came to see compression not as an editing technique, but as a stress test. And it’s what separates what AN Lab does from video production — or storytelling.

What I hadn’t expected was how cleanly three decades of apparently disconnected work converged on this one problem.

Thirty years in investment banking and finance taught me to look past the product and find the investment logic underneath it. Writing as a guest columnist for international business media taught me to cut through complexity and find the story that matters to an outsider. I discovered video as an extraordinary compression tool — data, numbers and moving images can communicate in seconds what takes pages to explain. Working with a documentary filmmaker whose work includes BBC and CNN commissions showed me something else: how powerfully film can reveal the human conviction behind a business. And AI became a creative partner — a way to show what a camera can’t capture, what doesn’t exist yet, what would otherwise be impossible to film.

For years, these looked like separate chapters. Only when I was sitting with that founder’s problem did they resolve into one toolkit.

AN Lab is that bet: that investment logic, journalistic compression, and film grammar belong in the same room — applied to the same 90 seconds.

Most founders preparing to fundraise ask: What should we put in the deck?

After two experiments and one pattern I couldn’t unsee, I think there’s a more useful question.

What does an investor need to recognise — before they owe you another minute?

—

The post Nobody gives you time to explain. That’s the real fundraising problem appeared first on e27.

Every cycle, people say that crypto is dead. By retail, and by funding. Attention moves on to the next bubble, this time it is AI. However, when Y Combinator published its list of Biggest Startup Opportunities of 2026, crypto kept its place alongside AI, healthcare, defence, enterprise software, and climate technology.

What surprised me was not that crypto is on the list. Rather, it was the particular formulation of crypto opportunities that caught my eye: YC did not ask the startups to design the next Layer 1, to build another memecoin or NFT marketplace, or to develop a yield farming protocol. Instead, they pointed to the need for stablecoin financial services, crypto infrastructure, institutional crypto products, tokenised assets, and agentic commerce.

This is the first sign that crypto is slowly evolving from an innovation layer to an enabler of other innovations, and, therefore, moving from an industry to an infrastructure. This is not the first sign, either, if you have been paying attention to the broader ecosystem.

Recently, Stripe announced Stablecoin Financial Accounts, a product that allows businesses in more than 100 countries to hold and transfer digital dollars around the world, bypassing the traditional banking system to a large extent. Meanwhile, Visa continues to develop stablecoin settlement and tokenised asset initiatives, and PayPal has expanded the use of its PYUSD stablecoin beyond Ethereum and blockchain-based payments.

The most notable trends in crypto adoption as financial infrastructure are also visible in Southeast Asia. GCash, the largest digital wallet in the Philippines with over 94 million registered users, has partnered with Ava Labs to tokenise EURC, USDC, and USDT on Avalanche in GCrypto. This makes it possible for everyday users to make payments in digital dollars via the most popular local app, instead of buying crypto on centralised exchanges.

Also Read: How AI and blockchain could make commerce decisions more accountable

By contrast, the competition between different Layer 1s may well be turning into a race to enable the adoption of digital assets as financial instruments by banks, payment processors, asset managers, governments, and enterprises. BlackRock’s BUIDL, the world’s largest tokenised money market fund, has also joined the ecosystem.

The same can be said for Stellar, which has embarked on a long-term journey to enable cross-border value transfers and displace traditional financial infrastructure in emerging markets and remittance corridors. The recent partnership with MoneyGram and UNHCR, as well as the introduction of Paxos Global Dollar (USDG), are all examples of this.

The broader financial services industry is also undergoing a similar transition. OKX and Standard Chartered Bank recently announced the launch of a collateral mirroring programme, which allows institutional customers to use tokenised money market funds and crypto assets as collateral in OKX’s custody under a regulated framework. The growing consensus within traditional finance is that digital assets will become an unavoidable part of the financial ecosystem.

However, the integration of crypto into traditional finance is much more nuanced and complex than many in the crypto community have cared to admit. DeFi is undergoing a similar transformation at the protocol level, as evidenced by Aave’s recent decision to sunset several smaller Layer 2 markets in favour of a more focused approach.

Also Read: The real status of blockchain gaming in Southeast Asia: Not hype, not dead — just growing up

In other words, Aave has opted for quality over quantity by shifting its resources to more liquid and relevant chains and products. Once again, we see the signs of a maturing ecosystem that is beginning to move away from the narrative of limitless possibilities and multiple ecosystems to the pursuit of efficiency and pragmatism.

In many ways, the evolution of crypto as an infrastructure layer has already begun. And, ironically, it may well be the payment apps, enterprise software, payroll processors, and AI agents that enable the greatest number of daily crypto transactions around the world.

The end-user will hardly distinguish between a transaction settled on the XRP Ledger, Ethereum, as long as it is cheap, seamless, quick, and available 24/7. Ripple continues working with banks and central banks on CBDC pilots through its CBDC Platform. Hedera is being used by organisations exploring tokenisation and digital identity. The industry’s competitive edge is gradually shifting away from speculation toward financial infrastructure.

It is no wonder that YC keeps believing that crypto is one of the major startup opportunities for the next decade. The next major leap for crypto may be driven not by crypto natives but by traditional payment companies that are looking to disrupt the financial system with better UX and more attractive yield opportunities. Perhaps one day, we will look back on this period of programmable money experiments as a brief episode of adolescence, when people talked a lot about crypto but used it even more.

—

Editor’s note: e27 aims to foster thought leadership by publishing views from the community. You can also share your perspective by submitting an article, video, podcast, or infographic.

The views expressed in this article are those of the author and do not necessarily reflect the official policy or position of e27.

Join us on WhatsApp, Instagram, Facebook, X, and LinkedIn to stay connected.

The post Crypto is dead? Apparently not, says Y Combinator – Blockchain is still worth building appeared first on e27.