Israel's figures allow a rough but interesting estimate. Bank of Israel estimates residents hold approximately $1B in crypto; voluntary disclosure only captured $50M — about 5% penetration. If this ratio holds in Taiwan where crypto holdings are far larger, the potential undeclared tax exposure could be substantial. Behaviorally, a '5% disclosure rate' means 95% of holders judged 'non-disclosure risk < disclosure cost.' This assessment may be rational before CARF, but won't hold afterward.
From the perspective of 'what makes a tax amnesty program actually succeed,' Israel's case is a textbook negative example. Successful voluntary disclosure programs need: procedural certainty (how taxes are calculated after disclosure must be clearly specified, not vague 'case by case'); anonymity protection (at minimum until taxes are paid, disclosure doesn't trigger further investigation); clear exemption boundaries ('if I disclose X, I'm definitively not pursued for Y'). If Taiwan's FSC and Ministry of Finance launch similar programs in the future, the design of these three conditions will determine success or failure.
Three specific low-cost actions Taiwan investors should take now: First, build a transaction record spreadsheet — each time you buy/sell USDY/OUSG/BENJI, record date, token count, USDC amount, TWD exchange rate. Record redemption values. This spreadsheet will be your most important tax document after CARF takes effect. Second, December 31 annual snapshot — record all tokenized asset positions' daily NAV and TWD equivalent as annual overseas income calculation basis. Third, choose platforms with good record export functions — Ondo Finance and Franklin Templeton have relatively complete transaction history export features; keeping these records is far easier than reconstructing them later.
CARF's global spread is faster than most crypto holders expect. As of 2026, over 50 countries have committed to CARF implementation, including: all EU member states (starting 2027); UK (2027); Japan, South Korea, Singapore (2027-2028); Australia, Canada (2026-2027). Taiwan isn't on the OECD member list, but Taiwan's major VASPs (BitoEX, MAX) may need to comply with CARF for users in OECD member countries. More importantly, Taiwan's FSC VASP management regulations are building a similar local framework. The direction is certain; only the timeline is uncertain.
In August 2025, Israel's Tax Authority launched a voluntary disclosure program for crypto assets: report unreported holdings before the deadline, receive immunity from criminal prosecution. Official expectation: $1 billion in tax revenue.
Final result: 58 filers, approximately $50M total — a 20x shortfall from expectations.
Israeli tax attorney Iftach Simhony identified the problem directly: 'In the crypto field, the lack of anonymous reporting channels is especially severe. When the process can't provide certainty or anonymity in the first phase, voluntary disclosure incentives are naturally weakened.'
OECD's CARF requires exchanges to automatically share user transaction data with tax authorities starting 2027-2028 — just like CRS for bank accounts. Once CARF takes effect, passive non-reporting strategies face fundamental risk.
For Taiwan investors holding USDY, OUSG, or BENJI: building complete transaction records now (purchase date, amount, USDC/TWD exchange rate) costs far less than reconstructing them under pressure later.