Investing/Nov 14 edition/Reporting & analysis
StocksMarketsMacro

StocksEquities, earnings & valuations

Goldman Sachs Forecasts 19% Return for Asia ex-Japan Stocks in 2027, Cuts Taiwan and China A-Shares to Market Weight

Goldman Sachs expects the MSCI Asia ex-Japan Index to return 19% in dollar terms in 2027, driven by earnings growth. It is also moving its overweight calls away from AI-heavy Taiwan and China A-shares and toward India and China H-shares.

From the Bloomberg Markets article of November 14, 2026 and cited sources: Goldman Sachs expects the MSCI Asia ex-Japan Index to return 19% in US dollar terms in 2027, mainly from earnings growth. [1]

SOURCES[1] Bloomberg: Goldman Sachs Sees 19% Return on Asia ex-Japan… bloomberg.com

AI-assisted, not human-edited. Methodology Corrections Who owns this site

KEY FACTS
What happened
Goldman Sachs expects the MSCI Asia ex-Japan Index to return 19% in US dollar terms in 2027, mainly from earnings growth. [1]
Original source
Bloomberg Markets article, independent perspective
Brief by
AT Pulse research desk AI-assisted; publisher and contact on the desk page
When
November 14, 2026 article published
Evidence
5 sources across 5 sites

Reach: not measured yet

THE CORE IDEAS3 TAKEAWAYS
01

Goldman Sachs expects the MSCI Asia ex-Japan Index to return 19% in US dollar terms in 2027, mainly from earnings growth. The bank also warns that uncertainty around AI could make markets volatile. [1]

02

Goldman cut Taiwan and China A-shares to market weight and raised India and China H-shares to overweight. Korea and Japan stay overweight. A month earlier, Taiwan and China A-shares had been among its preferred markets. The bank is spreading its AI exposure more widely rather than abandoning it. [1] [3]

03

Regional earnings depend heavily on AI hardware. In October, Goldman expected information technology to provide 47% of 2026 regional earnings, with memory chips making up about 70% of that. Separately, its global team says AI investment as a share of GDP could peak in early 2027, and lists an early AI peak as a downside risk. [3] [4] [5]

WHY IT MATTERS

Goldman still expects earnings-led gains but has cut its two most AI-exposed overweights.

Read the full assessment

Implication: AI capital spending and memory-chip prices now drive Asian equity forecasts directly, and the shift points to spreading AI exposure rather than pulling back from it.

Sources are collected, so I'm writing the dossier now.

Executive brief

Goldman Sachs expects the MSCI Asia ex-Japan Index to return 19% in dollar terms in 2027, driven by earnings growth. In the same report it also warned that uncertainty over AI could make markets volatile (Bloomberg). The bigger change is in its country calls. A month earlier, Goldman was overweight Taiwan and China A-shares, the markets most exposed to AI. It has now cut both to market weight and raised India and China H-shares to overweight. In other words, the bank still expects gains but is spreading its AI exposure across more markets. No independent reporting has repeated the 19% figure.

What changed and event timeline

  1. "Stick with Asia's winners."

    Goldman expected mid-teen returns for Asia Pacific ex-Japan in the second half of 2026. It forecast earnings growth of 60% for 2026 and 22% for 2027, was overweight Korea, Taiwan, Japan and China A-shares, and highlighted the memory-chip supercycle ().

  2. Forecasts raised

    Goldman forecast a 26% dollar return over 12 months, with an index target of 1,120. Its earnings-growth forecasts rose to 75% for 2026, 25% for 2027 and 14% for 2028. It stayed overweight North Asia ().

  3. Global strategy warns AI spending may peak

    Goldman expects S&P 500 earnings growth to slow from 36% to 11% and says AI investment as a share of GDP could peak in early 2027 ().

  4. Macro outlook

    Goldman forecasts 2.8% global GDP growth and AI capital spending rising to $1.8 trillion from $1.1 trillion. An early peak in the AI cycle is one of three downside risks it names ().

  5. Asia ex-Japan, 19% for 2027

    India and China H-shares were raised to overweight. China A-shares and Taiwan were cut to market weight.

Capabilities and access

  • What it is: a sell-side equity strategy forecast, not a product.
  • Index naming: the earlier reports refer to MSCI AC Asia Pacific ex-Japan, which includes Australia (CNBC, Crypto Briefing).
  • Access: the full Goldman research is for clients.
Read the full section
  • What it is: a sell-side equity strategy forecast, not a product. The target is the MSCI Asia ex-Japan Index, with a 19% total return in USD for calendar 2027 (Bloomberg).
  • Index naming: the earlier reports refer to MSCI AC Asia Pacific ex-Japan, which includes Australia (CNBC, Crypto Briefing). The reviewed excerpt doesn't show whether the 19% and 26% figures refer to the same index.
  • Access: the full Goldman research is for clients. The news coverage only summarizes it.

Technical analysis for researchers and developers

  • How the forecast is built: return is treated as earnings growth plus any change in valuation multiples.
  • Concentration: in October, information technology was expected to make up 47% of regional earnings in 2026, and memory chips about 70% of IT earnings (Crypto Briefing).
  • Implied valuation view: if 2027 earnings growth near 25% still holds, a 19% return means Goldman expects valuation multiples to shrink slightly.
Read the full section
  • How the forecast is built: return is treated as earnings growth plus any change in valuation multiples. Goldman's June note said about 80% of the region's 2026 gains came from earnings growth or upward revisions to forecasts (CNBC).
  • Concentration: in October, information technology was expected to make up 47% of regional earnings in 2026, and memory chips about 70% of IT earnings (Crypto Briefing). That makes the earnings base very sensitive to memory pricing and hyperscaler capital spending.
  • Implied valuation view: if 2027 earnings growth near 25% still holds, a 19% return means Goldman expects valuation multiples to shrink slightly. This is an inference; the reviewed sources don't confirm the November earnings figure.
  • Reproducibility: the model inputs, exchange-rate assumptions and dividend treatment are not public.

Claims and evidence

  • 19% 2027 return and the country rotation: these appear only in Bloomberg's account. No other reporting corroborates them.
  • 75%/25%/14% earnings path: documented for October (Crypto Briefing). It's unclear whether the November note keeps these numbers.
  • AI spending peak in early 2027: this is Goldman's own forecast (Yahoo Finance). All of these figures are the bank's projections, not observed data.
Read the full section
  • 19% 2027 return and the country rotation: these appear only in Bloomberg's account. No other reporting corroborates them.
  • 75%/25%/14% earnings path: documented for October (Crypto Briefing). It's unclear whether the November note keeps these numbers.
  • AI spending peak in early 2027: this is Goldman's own forecast (Yahoo Finance). All of these figures are the bank's projections, not observed data.

Context and prior work

Goldman's Asia outlook has moved in steps through 2026. Now it puts 2027 at 19% and has dropped two of its main AI-linked overweights (CNBC, Crypto Briefing). The India upgrade follows its October work on 42 Indian companies that supply AI infrastructure, such as power, data centres and semiconductors.

Read the full section

Goldman's Asia outlook has moved in steps through 2026. In June it expected mid-teen returns. In October it raised its 12-month forecast to 26%. Now it puts 2027 at 19% and has dropped two of its main AI-linked overweights (CNBC, Crypto Briefing). The India upgrade follows its October work on 42 Indian companies that supply AI infrastructure, such as power, data centres and semiconductors. Those stocks were up about 60% in 2026, while the Nifty fell about 12% (Crypto Briefing).

Limitations, safety and contested findings

  • Only a feed excerpt of the Bloomberg article was available.
  • Goldman's own risk scenarios run against the base case. Its global team also expects stocks to rise but with high volatility (Yahoo Finance).
  • Recent levels and year-to-date performance for the index could not be confirmed from MSCI data.
Read the full section
  • Only a feed excerpt of the Bloomberg article was available. Strategist names, index levels and the reasoning behind each country change are not in the reviewed sources.
  • Goldman's own risk scenarios run against the base case. They include an early AI peak, oil at $132 a barrel if the Iran conflict escalates, and stress in French government bonds (BigGo Finance). Its global team also expects stocks to rise but with high volatility (Yahoo Finance).
  • Recent levels and year-to-date performance for the index could not be confirmed from MSCI data.

Business and practitioner implications

  • For AI-industry readers: a major bank now treats AI spending as a key macro variable.
  • Supply chain: Goldman's lower weightings on Taiwan and China A-shares show the bank sees more risk in how concentrated the AI hardware supply chain is.
  • India: the move points to investor interest shifting toward the infrastructure that supports AI, such as power and data centres.
Read the full section
  • For AI-industry readers: a major bank now treats AI spending as a key macro variable. Hyperscaler capital-spending guidance and memory pricing affect Asian equity earnings forecasts directly.
  • Supply chain: Goldman's lower weightings on Taiwan and China A-shares show the bank sees more risk in how concentrated the AI hardware supply chain is. It isn't predicting a collapse in demand.
  • India: the move points to investor interest shifting toward the infrastructure that supports AI, such as power and data centres.
  • This describes Goldman's published views and is not investment advice.
FOLLOW THE EVIDENCE

The source trail.

Sources (5)
A LITTLE LESS NOISE. A LOT MORE CONTEXT.

Stay curious.
Follow the evidence.

Independent perspectives, the original sources, and room for the questions that don't have easy answers.

How we build the brief
Connect with us

Find us where you already read.