September’s Stock-Bond Rout: Diversification “Failed”, Costing Each Member HK$5,060 — Is the Alternative-Assets Lifeboat Arriving in Time?

On 7 October 2026, MPF Ratings published its September MPF performance survey: a single-month loss of 1.47%, or roughly HK$5,060 per member — the third monthly loss this year. The third quarter still managed +0.5% (HK$1,667 per member, a third consecutive positive quarter), and year-to-date stood at +6.2% (HK$20,161 per member), with total assets of HK$1.684 trillion, the third-highest on record.

But the numbers are not the story; the chairman’s commentary is. Francis Chung, chairman of MPF Ratings, stated publicly that in September, equities and bonds recorded losses simultaneously and that, with US Treasury yields at multi-decade highs, inflation worries and growth-slowdown risks, “traditional diversification strategies may appear ineffective”. Rising stock-bond correlation, he added, supports the MPFA’s mid-2025 policy pivot to “selectively allow alternative assets into Hong Kong’s mandatory retirement system”.

In plain terms: the first rule of MPF discipline taught over the past two decades — diversify across stocks and bonds — fails in a rate-hiking regime, and it fails precisely when you need it most. This is not the first demonstration: in 2022, the Manulife Global Bond Fund lost 13.57% for the full year.

Through the data: all three legs broke in September

GUM’s statistics, as of 18 September, show all three asset classes retreating together:

The MPFA’s provisional data to end-September adds the longer lens: over the past 12 months, bond funds averaged -1.5%, dead last across all categories, against +10.3% for equity funds and +8.5% for mixed-asset funds. The bond “stabiliser” has been running in reverse for a year.

And the 2022 precedent is starker still:

Note the middle row: the Age 65 Plus Fund is the terminus of DIS automatic de-risking — every DIS member ends up 100% in it after age 64, roughly 20% equities and 80% bonds, the system’s designed “safe harbour”. In 2022 it fell 13.21%, nearly as badly as a pure bond fund. When stock-bond correlation flips positive, so-called diversification is just two legs breaking together.

The compounding toll: one hole, three recovery clocks

Scale a 2022-style bond shock to today’s system: 13.57% × the average balance of HK$338,950 ≈ HK$46,000 vaporised per member; system-wide, roughly HK$228.5 billion. September’s 1.47% (HK$5,060 per member) is about one month’s contribution (HK$5,000) vanishing — a milder rehearsal.

The real toll is the recovery clock. After a 13.57% drawdown, a 15.7% gain is needed just to get back to even:

Same hole, slowest recovery in the “safe” options — 11 years in the conservative fund, nearly a full contribution cycle. This is asymmetric risk disclosure in action: product documents label the Age 65 Plus Fund and conservative funds “low risk”, yet nowhere do they state that in a hiking regime they fall alongside equities — and climb back slowest afterwards.

The structural problem runs deeper. DIS’s statutory de-risking mechanism shifts roughly 6.7 percentage points a year into the 80%-bond Age 65 Plus Fund from age 50. That fund has returned only about 2.5% annualised since its 2017 launch (MPFA statistical digest, to 30 September 2025; annualised inflation over the period was 1.8%). De-risking is not risk-avoidance — the system pushes you into a vehicle that falls in sync with equities during hiking cycles and delivers a real return of barely 0.6–0.7 percentage points.

The fix: read the menu, then watch the lifeboat

First, read the menu honestly. The MPF fund menu has no genuine third leg: beyond equities, bonds and cash (the conservative fund), all “diversification” is a stock-bond weighting game. Mixed-asset funds fell 1.2% in September, between equities’ -1.5% and fixed income’s -0.2% — a “cushion” of just 0.3 percentage points, and only because the bond leg fell a mere 0.2% that month. In 2022, when the bond leg fell 13.57%, the cushion became an amplifier.

The lifeboat is being launched, slowly. The policy timeline:

• Mid-2025: the MPFA’s policy pivot to selectively allow alternative assets (confirmed by Francis Chung, 7 October 2026)
• 10 March 2026: active ETFs admitted as a permitted investment class, capped at 10% of a fund’s net asset value
• 7 July 2026: gold ETFs moved from case-by-case approval to class-based approval (three conditions: SFC-authorised, SEHK-listed, physical non-derivative), cap kept at 10% of NAV
• 8 September 2026: the Financial Services Development Council’s “Roadmap 2.0” report proposed dedicated arrangements to widen MPF participation in alternative assets and infrastructure, noting the 10% cap “lacks economies of scale”; the MPFA responded it would prudently assess liquidity and transparency risks
• 16 September 2026: Policy Address package — removal of the aggregate investment cap on eligible index-tracking ETFs, with the ITCIS regulatory-framework review to complete by year-end

A member’s practical defences (within T+1/T+2 forward pricing — no day-trading):

1. Do not confuse “de-risking” with “risk avoidance”. Two more FOMC meetings remain in Q4 (27–28 October, 8–9 December), and Barclays is calling 6% on the 30-year Treasury. Avoid large switches around rate-decision weeks — switch instructions execute at unknown forward prices, and the blind window is a timing tax.
2. Use the TVC back door to keep the growth engine outside the de-risking machine. A TVC account can hold the Core Accumulation Fund directly, free of statutory de-risking.
3. The conservative fund is the menu’s only uncorrelated shelter — price it honestly. Ten-year annualised returns of roughly 1.2–1.5%, against August core inflation of 1.9%, mean a real return of about -0.6% a year. It is a car park, not a harbour.
4. Track policy delivery; never pay active prices for the word “diversified”. When the first constituent fund holding gold ETFs or alternative assets appears, check two things: its fund expense ratio, and the actual allocation under the 10% cap. The FSDC has already said the quiet part out loud — the 10% cap cannot deliver economies of scale, so fee accountability is the next battlefield.

Two accountability questions for the system: will the FSDC’s proposed raising of the 10% cap land? And will the DIS reference portfolio (60% FTSE MPF All-World Index + 37% FTSE MPF World Government Bond Index (HKD hedged) + 3% prescribed savings rate — a pure stock-bond recipe) admit alternative assets in its comprehensive review? September’s HK$5,060 tuition fee is paid. The next time stocks and bonds fall together, the menu should offer more than two legs.