Rising Expectations for Removal of Government Bond Tax Exemption; New Treasury Yields May Rise 10-20BP
nashnova research
The PBOC's Q3 report flagged how tax exemptions distort government bond pricing, fuelling expectations that the corporate-income-tax break on government bond interest will be scrapped; Huatai Securities estimates new-issue 10Y CGB yields could rise 10–20 BP, forcing banks, insurers and mutual funds to rethink allocation.
What did the PBOC say, and why is the market nervous?
The PBOC's Q3 2026 monetary-policy report stated explicitly that CGB tax exemptions let investors accept lower yields and that interest-rate comparisons must "restore the tax factor." This means → the central bank is laying the groundwork for removal, not conducting an academic exercise.
The historical precedent sharpens the signal — the Q1 2025 report flagged the VAT exemption issue, and roughly 12 weeks later the VAT exemption was abolished. In plain terms = when the PBOC "names" a tax arrangement in a policy report, it is usually a countdown to action.
The market is already front-running: from Nov 11–13, 10Y CDB yields rose 7.5 BP cumulatively, the CDB-CGB spread widened from 5.2 BP to 11.0 BP, and the 30Y CGB future fell more than 0.5% on Nov 16.
Which goes first — local government bonds or CGBs?
The two bond types rest on different legal bases, which dictates the pace. CGB tax exemption is written into the Enterprise Income Tax Law — removal requires NPC Standing Committee legislation. Local government bond exemption is a departmental regulation — it can be adjusted with State Council approval alone, no law change needed.
Huatai's base case: "Grandfather existing bonds, start with local government bonds, CGBs follow after the law is amended." This means → legacy bonds will likely keep their exemption; only new issues get taxed.
Key dates to watch: whether the Ministry of Finance announces local-bond adjustments in coming months; whether the NPC Standing Committee conducts a second reading of the tax-administration-law amendment in late December 2026; and whether the April–May 2027 legislative work plan includes amending the Enterprise Income Tax Law.
How much will yields actually rise?
Huatai frames two extremes: a theoretical ceiling of ~54 BP — the rate increase needed if bank proprietary desks demand unchanged after-tax returns — and a floor of ~10 BP — implied by how far the CDB-CGB spread already compresses the effective tax rate, with new CGBs converging toward CDB levels.
The 2025 VAT grandfathering experience offers a reference: the 10Y CGB new-vs-old spread averaged about 4.8 BP per month, closer to the 3% tax bracket for asset-management products. In plain terms = new taxable bonds are priced mainly by trading desks, while the heaviest-taxed bank proprietary desks prefer to hold tax-exempt legacy bonds — so the actual increase is far below the theoretical ceiling.
Huatai's central estimate: new-issue 10Y CGB yields rise ~10–20 BP; 30Y CGBs rise ~15–25 BP. Guosheng Securities, using a different method — spreading ~RMB 102.3 bn in first-year tax across ~RMB 112 tn of affected bonds — arrives at an impact of ~9.1 BP/year. Both houses converge at the lower bound.
Will the government actually collect more revenue?
Huatai estimates that under grandfathering, year-one net fiscal revenue gain is about RMB 21 bn, growing to ~RMB 190.4 bn by year ten. But there is a tipping point: if yields rise more than ~40 BP, extra interest costs exceed extra tax revenue and net fiscal gain turns negative.
Local governments face an even sharper bind: they pay the interest on local bonds, yet the new corporate income tax accrues mostly to the central government and to headquarters cities. Huatai calculates the local-level breakeven at no more than 16 BP — beyond that, locals are net losers. This reflects a hidden redistribution: interest costs flow from less-developed regions toward wealthier ones.
How are banks, insurers and mutual funds affected?
Banks bear the brunt — they hold 70.6% of interbank CGBs and 71.7% of local government bonds. At a 10Y CGB yield of 1.61% and a large-bank funding cost of ~1.3%, the after-tax net carry on new bonds drops from ~55 BP to ~17 BP, a fall of ~38 BP. This means → banks will cling to tax-exempt legacy bonds, demand higher rates at primary auctions, and pull back from ultra-long government debt.
Insurers face a vanishing cushion: the after-tax spread of new 30Y local government bonds over the 2.0% traditional-life pricing cap narrows from ~58 BP to ~6 BP. But capital constraints under C-ROSS II mean insurers are more likely to raise required buy yields than to shift en masse into other assets.
Mutual funds may be the biggest winner. After exemption is removed, a bank holding CGBs through a mutual fund earns ~42 BP more after tax than holding directly. In plain terms = the tax distortion migrates from "exempt vs taxable bonds" to "direct investment vs fund investment" — whether fund-dividend tax exemptions get adjusted next may be the next policy shoe to drop.
What about policy-bank bonds, credits and railway bonds?
Policy-bank bonds and credit bonds reacted more violently in this episode. Guosheng's explanation: these instruments are paying back the debt of excessively compressed spreads — once taxation expectations push government new-issue rates higher, the already razor-thin spreads cannot absorb the repricing from an anchor shift.
Railway bonds are the exception: their interest is taxed at half the corporate rate. Once government bond exemptions are removed, railway bonds flip from a tax disadvantage to a tax advantage relative to CGBs. This reflects how the policy change reshapes not just absolute yields but the relative-value pecking order across bond types.
The ultimate magnitude hinges on two checkpoints: the timing gap of local-bond policy implementation and the legislative progress on CGB law amendment — together they will set the pace of the market's repricing.
市场有风险,内容仅供研究参考,不构成投资建议。
