According to CNA, the New Taiwan Dollar (TWD) stood at 32.326 against the US dollar at midday on July 17, 2026, a decline of 7.7 cents from the previous close (E1). The session opened at 32.25, which also marked the day's high, while the low touched 32.338 (E3, E4). The spread between the day's high (32.25) and low (32.338) was roughly 8.8 cents, framing the midday print of 32.326 near the weaker end of the morning's range.
CNA reported midday turnover in the Taipei foreign exchange market at US$1.07 billion (E2). No separate breakdown of buy-side versus sell-side volume was provided in the report.
Central bank (中央銀行) records cited in a December 11, 2017 press release show that currency-versus-dollar performance for the TWD and the Korean won (KRW) has diverged across different periods. From the start of 2017 through the end of September that year, the KRW appreciated 5.4% against the US dollar while the TWD appreciated more, at 6.5% (E8). But from end-September through December 8, 2017, the pattern reversed: the KRW rose 4.8% while the TWD gained only 1% (E9).
The same central bank release also referenced the 2008-09 global financial crisis as a historical benchmark: the TWD fell from a December 31, 2007 close of 32.443 to a crisis-period low of 35.174, a maximum depreciation of 7.8%, while the KRW fell from 936.1 to 1,570.3 over the same window, a maximum depreciation of 40.4% (E6, E7).
| Period | NT Dollar move | Korean Won move | Source |
|---|---|---|---|
| Jan–Sep 2017 | +6.5% | +5.4% | E8 (cbc.gov.tw) |
| End-Sep–Dec 8, 2017 | +1% | +4.8% | E9 (cbc.gov.tw) |
| 2007-end to 2009 crisis low | -7.8% | -40.4% | E6, E7 (cbc.gov.tw) |
Note: these historical figures come from a 2017 central bank statement and predate the July 17, 2026 midday quote reported by CNA (E1); they are cited here only as background on past TWD-KRW divergence, not as directly comparable to the current session.
In a press release dated December 11, 2017, the central bank (中央銀行) addressed a foreign media report alleging a "mysterious late-session drop" in the NT dollar, stating that "this report is purely speculative and does not match the facts" (E5, E12). As a specific example, the central bank cited November 22, 2017, when the TWD moved from 30.012 to a close of 30.020 in the final 15 minutes of trading — a depreciation of just 0.8 cents, or 0.027% — noting that this price change sat squarely within the range seen across other periods of that day's session (E10).
The same central bank release linked diverging TWD and KRW performance in late 2017 to differing foreign investor behavior: between end-September and December 8, 2017, foreign investors were net buyers of Korean equities to the tune of US$2.36 billion, while net-selling Taiwanese equities worth US$530 million (E9). Separately, the central bank noted that domestic banks' foreign-currency deposits had risen by NT$477.1 billion (approximately US$15.6 billion) as of end-October 2017 compared with end-2016 (E11).
The cited evidence shows two distinct time frames: a same-day midday snapshot from July 17, 2026 (32.326, down 7.7 cents, on US$1.07 billion in turnover, per CNA) and a separate set of central bank figures from 2017 addressing a different controversy over late-session TWD pricing. Within that 2017 dataset, the central bank's own numbers show the TWD-KRW relationship shifting between periods — the TWD outpacing the KRW in early 2017 but lagging in the following months, a divergence the central bank tied to opposite foreign fund flows into Korean versus Taiwanese equities (E8, E9). The central bank's explicit denial of "mysterious" late-session manipulation (E12) was backed by a specific numerical example — an 0.027% move in the final 15 minutes on November 22, 2017 (E10) — illustrating how the bank used granular percentage data to counter a qualitative allegation.
According to CNA, midday turnover in the Taipei foreign exchange market was US$1.07 billion (E2).
No. In a December 11, 2017 press release, the central bank called the report 'purely speculative and not matching the facts,' citing a November 22, 2017 case where the final 15 minutes saw only a 0.8-cent, 0.027% move (E10, E12).
As of 22 July 2026, standard API pricing (input/output per 1M tokens) and Artificial Analysis capability scores for major LLMs: OpenAI GPT-5.6 Sol ($5/$30, 59) and Anthropic Claude Opus 4.8 ($5/$25, 56) lead on capability; xAI Grok 4.5 and Claude Sonnet 5 sit in the value band; the DeepSeek V4 family prices output under $1 for high-volume use. Pricing changes often — verify on official pages.
According to a report by CNA (Central News Agency), the U.S. Commerce Department confirmed that Chris Fall, director of the Center for AI Standards and Innovation (CAISI), resigned after only three months in the role. The department did not disclose a reason, and a UDN report notes the departure marks the latest personnel shift in the Trump administration's AI policy.
According to Inside.com.tw, Taiwan's Legislative Yuan Transportation Committee has passed a preliminary review of an amendment to the Telecommunications Management Act that would exempt satellite operators from foreign-ownership caps under national-security review. Per CNYES, Chunghwa Telecom (中華電信) chairman Jie Zhicheng said the company would immediately move to bring in Starlink once the law passes, while Digital Affairs Minister Lin Yi-ching noted Starlink has told the government Taiwan's market potential is limited.