Rather than slowing as the government had forecast just a couple of months ago, GDP growth re-accelerated in Q3 to 14.4% YoY. That was because of domestic demand rather than net exports, and while that was mostly capex, consumption is growing more than 5%, and the positive output gap widened.
September money and credit data for September are like other indicators for Taiwan: quite strong, but not high enough on their own to think that they suddenly break the CBC's apparently strong desire not to change policy.
Like so much else in Taiwan, labour market dynamics will be determined by the fortunes of the AI cycle. If that turns down, it will be hugely important. A sharp downturn hasn't, however, been the CBC's expectation, but existing wage trends seem inconsistent with the bank's view that inflation slows.
Taiwan's price data still don't suggest the economy is overheating. But pipeline pressure is strong, computer prices are surging, indirect energy prices can't be easily controlled by policy, and underlying inflation still looks firm. That all stands out when real interest rates are still negative.
The CBC last week claimed that even with the policy rate staying at 2%, its monetary stance "remains on the tight side". I am not persuaded, but it is worth studying the bank's argument.
In many advanced economies, governments are drowning in debt. But in Korea and Taiwan they are floating up on the flood in revenue generated by the semiconductor boom. There is so much money that both economies plan looser fiscal policy while also paying back debt.
Consensus was right, and the CBC didn't change rates today. The vote wasn't unanimous, with two members voting for a hike. However, some of the property measures were loosened. To me, this suggests that policy is too loose, and thus that the gradual warming up of inflation will continue.
I expect the CBC to hike. That was also my view 3M ago, and then I was wrong, and the strong consensus is that I'll be wrong now too. Underlying my view is that the CBC can't be confident inflation has peaked. Instead, with real rates negative and the TWD weak, price pressures are likely to persist.
Headline growth rates are slowing, but the government's qualitative assessment remains bullish, there are more signs of life in domestic demand and services, and while a 2026 inflation forecast of a bit over 2% isn't so high, it has been raised four quarters in a row, and is above the policy rate.
Annualised regular wage growth in both manufacturing and services has been above 3% since 2024. In Q2 in manufacturing, it has been near 4%. The risk remains that the export cycle now slows, but TSMC's strong sales and elevated manufacturing overtime don't suggest that is happening yet.
Both headline and core CPI inflation ticked down in July as energy prices eased. However, underlying inflation feels firm, with sequential core failing to break below 2%; import price, PPI inflation and CPI computer price inflation rising to decade highs; and services inflation running at 2.6%.
June's core inflation of 2.3% isn't high in an absolute sense. But outside LNY, it is the highest since the 1990s, and, remarkably, above both the policy rate and 10-year yields. Energy inflation will ease, but with wage growth of 2.7% and rising, and equity prices strong, rates really look too low.
The CBC didn't change policy today. It increased its growth forecast, but didn't sound excited about the outlook. It warned about excesses related to equities, but didn't announce any remedial policies. Inflation is expected to remain below 2%.
Services inflation averaged 0.7% in the 20 years before 2020. In the last five years it has been 2.3%, and is now rising again. Some of that reflects energy prices, with air fares rising 10% YoY in May. But there is also the backdrop of a strong economy, rising stockmarket and rising wage growth.
Today's Q1 GDP data show the economy growing more quickly than any time in over 40 years. The government thinks that continues: in the forecast, downside risk from Iran isn't mentioned, but trickle down from semi to the real economy is. CPI was also revised up, even if it remains (just) below 2%.
My initial interpretation of the Xi-Trump meeting looks wrong. While in the official talks neither side gave much, in a subsequent interview with Fox news, President Trump softened US support for Taiwan. Over the medium-term, that could be significant for politics in Taiwan, and currency valuations.
April trade data show little damage from the Iran war. If the crisis remains limited to energy prices, Taiwan should be somewhat insulated. Energy imports are small relative to chip exports, and while import prices will rise, export prices are now also increasing for the first time in a generation.
Latest video, discussing one of my main themes: that Taiwan macro is more relevant for global investors than it has been for a long time.
Exports in March were strong again. There aren't yet signs of the Iran war derailing the chip cycle, and while energy imports will increase more quickly, the impact on the trade surplus will be limited. The outlook for Taiwan as of now is resilient growth and higher inflation.
I've been arguing that Taiwan is one of the most interesting macro stories in the world right now. I had the chance to lay out some of the elements in this week's Bloomberg Asia Centric podcast.
The surge in the current account surplus of January-September continued into the end of 2025. In Q4, the surplus reached almost 30% of GDP. The government's GDP forecast implies that will be roughly the size of the full-year surplus in 2026. Taiwan needs huge capital outflows to keep the TWD stable.
Taiwan's macro story is fascinating. Rarely has growth in such a rich economy accelerated so much, with reindustrialisation and huge external surpluses, at the same time as rates and the currency barely move. I'd think that something has to give. Will 2026 finally be the year that it does?
Even with data today showing a dip in exports in December, Taiwan's trade surplus last year reached the sort of sky-high levels last seen before the big TWD appreciation of the 1980s. Barring a real dislocation in AI hardware demand, underlying pressure for renewed appreciation will grow in 2026.
There is little growth in exports outside of tech in general, and semiconductors specifically. But the surge in chip exports is big enough to offset all the weakness in other products. The overall trade surplus eased back in November, but the bilateral surplus with the US reached 25% of Taiwan GDP.
In today's GDP release, the government raised estimated growth for Q1, Q2, and Q3. The FY forecast was raised by almost 3ppts to 7.4%. But because of AI, officials remain bullish about the outlook, and so raised the forecast for 2026 as well, And yet, none of this is expected to impact inflation.
