Smart-money record · SMART
Decision-support · not advice
Decision-support · not advice
Temasek just hit a record — S$518 billion. But its 20-year return was 6.8%. Is that good, or bad? The only honest way to know is to compare it — to Warren Buffett, and to the most boring investment on Earth. The answer isn't what you'd guess.
Here's the reveal most people get wrong. Everyone "knows" Warren Buffett compounds at ~20% — but that's his 60-year record since 1965. Over the last 20 years, Berkshire returned about 11% a year, and actually slightly trailed the S&P 500, which returned roughly 10–11% with dividends reinvested.
Temasek's 20-year return was 6.8% in Singapore dollars, 8% in US dollars. Either way, it sits below both Buffett and a simple index fund over the same two decades. And this isn't a fluke of one fund: over 15-year windows, roughly 93% of large-cap active funds underperform the index. Beating the market, it turns out, is brutally hard — even for the best in the world.
Judging Temasek purely on return misses what it's built for. Read it fairly — both sides are true.
Temasek is a defensive, diversified sovereign fund built to preserve a nation's wealth across generations — not to beat the S&P. At 6.8%, a portfolio doubles roughly every decade, and Temasek's did exactly that. Resilient, lower-volatility, mandate-appropriate.
As pure return, it trailed a basic index fund over 20 years — even in USD (8%). A stronger Singapore dollar shaved ~2 points, and the 20-year figure includes a one-time valuation uplift from a 2016 accounting change. Measured against the index, the giant lagged.
These are Temasek's Q1 2026 US purchases, straight from its public filings — the kind of "smart money" moves Spiking tracks. Note the tilt: AI and semiconductors, alternative-asset managers, and infrastructure — matching Temasek's stated push into AI and core-plus infrastructure.
Copying a giant's holdings feels like a shortcut. It isn't a magic edge — and the reasons matter.
13F filings are disclosed up to 45 days after quarter-end. What you see is where the money was, not where it is now.
You see US-listed long positions — not what they sold, not their timing, not their entry price, and not why they bought.
These 16 names are a fraction of a ~S$518 billion global portfolio spanning private companies, funds, credit and unlisted assets you can't buy.
Remember the scoreboard: even Temasek's entire portfolio trailed a simple index fund over 20 years. Copying a slice of it carries no guarantee of doing better.
The value in tracking smart money was never blind imitation. It's context. Seeing that a sovereign fund is tilting hard into AI, semiconductors and infrastructure tells you where the world's most careful capital thinks the next decade is heading — a starting point for your own research, not a shopping list. The edge is understanding why the giants move, then doing your own work — because even they can't reliably beat a basic index by following anyone.
This report does not tell you to buy, sell, or hold any of these stocks, and contains no price target. Listing what an institution bought is not a recommendation that you buy it too.
Return figures are approximate 20-year annualised totals to roughly 2026 and vary by source, exact window and currency; Temasek's are its own reported TSR (6.8% SGD / 8% USD), Berkshire's and the S&P's are drawn from public performance data (~11% and ~10–11%). Individual stocks carry concentrated risk and can fall sharply. Verify current figures before acting.
Educational, decision-support material only — not financial, investment, legal, or tax advice. Do your own research and consider a licensed professional.
Join the free live masterclass and learn to read the disclosed record for yourself — the institutional flows, the context behind the trades, and how to turn "smart money" data into your own informed decisions. No hype. No price targets. Just the method.
Save my free seat →