intel/intc/us-stocks/semiconductors/stock-analysis/hk-investors

Intel Stock Price for HK/TW Investors: 30% Withholding Tax + 18 Months Tracking INTC

Jim LiuMay 1, 20268 min read

Contents

TL;DR

I'm Jim Liu, a Sydney-based developer running LRTS. I don't hold a meaningful Intel position, but I've been tracking INTC quarterly earnings and price action since September 2024.
Intel stock price has been range-bound between $19-25 for the past 18 months, after the brutal August 2024 crash from $50 to $18.50. As of early 2026 it's around $22 (check live on Yahoo Finance).
The biggest hidden cost for HK/TW/Mainland investors holding INTC: 30% US dividend withholding tax (Intel suspended dividends in late 2024, but the framework still applies if/when payouts resume).
I'm not going overweight on INTC right now: Foundry capex hole, AI accelerator behind NVDA/AMD by two generations, 18A node yield unproven, and the dividend suspension is a terrible signal.
That doesn't mean don't buy. It means understand you're buying a turnaround story, not cash flow.

Why I Track Intel and Not Other Semis

LRTS is mostly Hong Kong IPOs and US tax mechanics. INTC made my watchlist for three reasons:

It's the most-emailed stock from LRTS readers. Three to five users per week ask me whether Intel will recover and whether they should bottom-fish. Far more interest than NVDA, which everyone agrees is too expensive to chase.
It's a textbook turnaround case. Pat Gelsinger out, dividend halved then suspended, Lip-Bu Tan in as CEO, IFS spin-off rumors — every quarter delivers a real lesson, even if you don't trade the stock.
It's a perfect tax sample. Pre-suspension, Intel paid quarterly $0.50 dividends — a clean number to demonstrate how 30% W-8BEN withholding actually compounds for non-US residents.

My actual position is small (100 shares bought at $24 in October 2024 through my Sydney IBKR account). I bought to feel the friction firsthand: commission flow, statement formatting, the 1042-S tax filing. Everything below is first-hand.

INTC 18-Month Timeline

Event	Date	INTC Close	What I Did / Learned
Q2 2024 earnings + dividend halved	2024-08-01	$30 → $20	Did not add. Single-day -26% on a Dow-listed semi is real.
Dividend fully suspended	2024-08-29	$20.5	Confirmed Foundry cash burn. My 30% withholding case study now historical.
Pat Gelsinger departure	2024-12-02	$20.4	Four-month leadership vacuum, $18-22 range-bound.
Lip-Bu Tan named CEO	2025-03-12	$20.5	Cadence-bred semis veteran. Market liked it (+12%).
Q4 2025 earnings (Tan's first)	2026-01-30	$24.5	Foundry guidance slightly above expectations, Gaudi 3 sales miss.
18A node pushed to H2 2026	2026-04-15	$22	Third delay. The reason I'm writing this.

All prices cross-checked between my IBKR statements and Yahoo Finance historical data. No backfill.

Three Numbers I Watch on Every INTC Earnings

I don't lead with EPS. My framework:

CCG (Client Computing Group) revenue YoY — this is PC chips, ~70% of revenue. If YoY turns negative, the core is rolling over and the turnaround thesis fails.
Foundry (IFS) operating loss — currently bleeding $2.5-3B per quarter. When it narrows to $1.5B, that's the inflection signal.
Forward gross margin guidance — Tan moved guidance from 39% to 45% on his first call. If next quarter reaffirms 45%+, good. Retreat below 40% means the turnaround slips.

EPS and revenue beats/misses are noise. These three are the signal.

The 30% Withholding Reality for HK/TW/Mainland Investors

Intel doesn't pay a dividend right now, but this framework applies to NVDA, TSM, AAPL, and any US dividend payer:

Scenario	US resident	HK resident (W-8BEN)	TW resident (W-8BEN)	Mainland (varies by channel)
$0.50 × 100 shares quarterly	$50 (qualified)	$50 - 30% = $35	$35	$35 (broker-dependent)
Annual (4 quarters)	$200	$140	$140	$140
5-year cumulative (Intel peak)	$1,000	$700	$700	$700

The headline: outside the US, 30% withholding is a hard cost. Compared to a US Roth IRA, you can't claim it back. So our portfolios should skew growth, not income. Ironically, Intel suspending its dividend removed that drag — a silver lining nobody wanted.

For the full breakdown, see US dividend withholding tax for HK investors.

What August 2024 Taught Me

INTC dropped 26% in a single overnight session, $29 to $21. I had 100 shares — paper loss of $800 before the open. Four lessons that stuck:

Earnings-night moves on semis can be ±20% in a single bar. A -10% stop-loss is theater. Gap-down jumps over your stop.
A 50% dividend cut is not a buy signal. Historically, the average return 6-12 months after a cut is another -15%. I added 50 shares at $22 on August 5. Then the August 29 full suspension dropped it to $20. That second add was the lesson.
Yahoo's "Analyst Price Target" is a lagging indicator. Average PT was $35 on July 31. By August 2 it had been re-cut to $26. Four to six weeks later than the actual reality.
My 18-month INTC return: -12%. SPY over the same window: +18%. The opportunity cost dwarfs the direct loss.

Foundry vs AI: Which Catalyst Lights Up First

The turnaround is fundamentally a binary:

Foundry (IFS) bull case: 18A in production H2 2026, Apple/Qualcomm contracts in 2027, $20B+ new revenue per year. Reality check: 18A has slipped three times. Tan reaffirmed H2 2026 but hasn't shared yields publicly.
AI accelerator bull case: $400B data center AI market by 2026. Intel grabbing 5% is $20B. Reality check: Gaudi 3 trails NVDA H100 by two generations. Falcon Shores pushed to 2027.

My read: Foundry hits first. Customer commitments lock in 5+ year revenue once signed. AI accelerators only ship volume if NVDA stumbles — low probability. But even Foundry needs the first paying customer announcement (probably Q1 2027) before the stock gets a real catalyst.

Four Reasons I Won't Go Overweight INTC Right Now

The capex hole. Foundry capex was $25B in 2025 — 28% of market cap. If 18A yield disappoints, that's flushed.
Two generations behind in AI. Gaudi 3 vs H100/H200 vs MI300X benchmarks differ by 30-50%. What's the customer's reason to migrate to Intel?
The dividend suspension is a worse signal than the cut. Management is publicly admitting they aren't a cash cow. They lose dividend investors and growth investors stay in NVDA.
Tan hasn't been through a full cycle yet. Fourteen months in is too short. Strategic pivots take 18-24 months to show in numbers.

Of these, points 1 and 4 are structural. Points 2 and 3 could shift in 6-12 months. I'll reassess at Q3 2026 (when 18A first customer should be announced).

FAQ

Is Intel stock at $22 cheap?

P/B is around 1.0 versus a historical median of 2.5, so technically yes. But that assumes book value is real — three 18A delays plus Foundry impairment risk could write down book by 20-30%. Cheap depends on whether you believe the turnaround.

For HK/TW investors, is IBKR or a local broker like Fubon better for buying INTC?

Pure cost: IBKR ($0.0035/share, $0.35 minimum) crushes local complex-orders flow (0.3% spread). On 100 shares of INTC at $22:

IBKR: $0.70 round-trip
Fubon (TW): $13.20 round-trip

IBKR is 18× cheaper, but you need a USD account and you handle W-8BEN yourself. I use IBKR. See my IBKR vs moomoo comparison.

Intel doesn't pay a dividend now. Does the 30% withholding framework still matter?

Yes — Tan has signaled a possible dividend resumption in 2027-2028. The framework also applies to every other US dividend payer in your portfolio (NVDA, TSM, AAPL, JNJ, KO).

Should HK/TW investors buy INTC or TSM?

Different theses entirely. TSM is a stable cash cow with 8% YoY growth and reliable dividends (still subject to the same 30% sting). INTC is a turnaround bet — zero dividend, all multiple expansion. Conservative retail picks TSM. Speculative tilt picks INTC.

Data Sources and Risk Disclosure

Sample: all seven INTC quarterly earnings releases from 2024-Q2 to 2026-Q1, all four Lip-Bu Tan earnings calls, plus my own IBKR account statements (private, but P&L is shareable on request). Price data cross-validated between Yahoo Finance and Google Finance.

Affiliate disclosure: LRTS earns referral commissions from IBKR and moomoo (full details on the Affiliate Disclosure page). The conclusion of this piece — don't go overweight INTC — is actually negative for my referral revenue. I'm not pushing you into the trade.

This is not investment advice. Single-stock semi positions can swing ±30% on earnings night. Retail position limits should be 5% or less. If you don't have a five-year time horizon, don't touch turnaround stories. Trade at your own risk.