Where the Money Goes

Executive Summary

Australia operates one of the largest compulsory savings systems on earth, yet roughly half of its institutionally managed superannuation pool is already deployed outside the country. That fact is not a policy failure in isolation — it is the mathematically predictable outcome of a legal framework that places member returns above all else, applied to a domestic market representing approximately two per cent of world equity indices and concentrated overwhelmingly in banks and resources. The structural tension, however, has sharpened into something more urgent: the same friction that pushes Australian capital offshore is simultaneously deterring foreign sovereign wealth funds from coming the other way. Both sets of actors, operating under entirely different legal obligations, are arriving at the same conclusion about the domestic risk-adjusted return.

The scale of the shift is no longer marginal. Of the $3.4 trillion held in large APRA-regulated funds, approximately half sits offshore — a share that has climbed from roughly 35 per cent a decade ago to more than 50 per cent today. The NAB Super Insights 2026 survey recorded an international allocation of 50.9 per cent in 2025, the first print above 50 per cent. Deutsche Bank's October 2025 review confirmed the offshore share had risen ten to fifteen percentage points over the preceding decade. The Future Fund, under a separate statutory framework, holds approximately 77 per cent of its assets on a physical look-through basis outside Australia. Funds expect 50 to 60 per cent of each new incremental dollar to continue heading abroad. The destinations are no longer abstract: the United States first, with the United Kingdom and India now established as named, policy-backed channels for Australian retirement capital.

The legal architecture leaves trustees no room for sentimentality. Under sections 52 and 62 of the Superannuation Industry (Supervision) Act 1993 (Cth), every regulated trustee must act in the best financial interests of beneficiaries, with retirement income as the paramount objective. APRA's guidance is explicit: the sole purpose test is not aspirational. Civil penalties attach to any contravention of the section 52 covenants, as confirmed by the ATO's legislative analysis. There is no statutory domestic-investment quota, nor any legal mechanism that permits a trustee to overweight Australian assets unless those assets clear the same risk-adjusted return hurdle as their offshore competitors. The duty to members and the failure of the domestic investment environment are therefore not competing explanations for the offshore trend — they are the same problem observed from opposite ends.

The domestic bottleneck compounds the legal imperative. Despite a stated 30-day decision-making target introduced from January 2025, Treasury's own quarterly data recorded a median FIRB processing time of 35 days in Q1 2026, with only 46 per cent of proposals resolved within the window, according to Jones Day's June 2026 analysis. Complex infrastructure and energy acquisitions routinely extend well beyond the statutory period. MinterEllison notes that the 2026 reform package proposes to let the Treasurer rapidly designate new sensitive sectors, leaving investors in critical infrastructure, critical minerals and data centres to expect "dynamic, changing FIRB requirements going forward" — precisely the uncertainty that institutional allocators price with a higher required return, or avoid entirely.

The inbound signal reinforces the outbound one. The Infrastructure Partnerships Australia 2026 Investment Monitor — surveying managers of more than A$530 billion of global infrastructure — found Australia holding steady as a preferred market at 56 per cent, unchanged, while the United States vaulted from 48 per cent to 80 per cent. Macquarie's chief executive told an infrastructure summit in the first week of September 2026 that sovereign wealth funds are directly advising policymakers they will either bypass Australia or demand a higher return because of approvals delays, unexpected outcomes, and the absence of certainty — calling for a "sense of crisis." That is not an advocate's hyperbole; it is a summary of revealed preference by the institutional class Australia most needs to attract.

This dossier examines the mechanics, legal underpinnings, and policy consequences of that trend across eight sections. The central finding is direct: Australia has not become a risky destination; it has become, in the collective judgement of those who must place the next dollar, an insufficiently rewarding one. With the system projected to reach $4.8 trillion and growing at 9.5 per cent annually, the compounding effect of each incremental dollar flowing outward is material. Without reform to approvals architecture, pipeline transparency, and mandate stability, the trajectory will continue — and the reforms announced in May 2026 via the FIRB overhaul are a belated acknowledgement of a price that capital-allocation habits have already begun to pay.

The Scale of the Offshore Shift

The latest official figures establish the system's dimensions precisely. As at 30 June 2026, total superannuation assets reached $4,767.2 billion. Within that pool, $3,411.5 billion — the APRA-regulated institutional book — is managed by trustees who make active allocation decisions on behalf of members. Self-managed super funds account for the remaining $1,107.2 billion, or approximately 23 per cent of the system, and behave very differently from their institutional counterparts when it comes to geography.

The institutional offshore share has crossed a symbolic threshold. The NAB Super Insights 2026 survey recorded an international allocation of 50.9 per cent in 2025 — the first time the figure has exceeded half — rising from 47.8 per cent in 2023. Deutsche Bank's October 2025 review placed the offshore share at approximately 48 per cent as at end-2024 and documented a rise of 10 to 15 percentage points over the preceding decade. The trajectory is consistent across independent sources: institutional super has moved from a predominantly domestic allocation to a predominantly international one within a single generation of fund management.

The contrast with self-managed funds is stark. SMSFs report direct international holdings of just 2 to 2.5 per cent of assets on ATO reporting forms — a figure that looks domestically anchored until one accounts for the treatment of Australian-domiciled global exchange-traded funds. Because those ETFs are registered in Australia, ATO reporting classifies them as domestic assets regardless of the underlying exposure. Correcting for ETF look-through lifts the whole-system offshore estimate to approximately $1.8 to $2.2 trillion, a materially different picture of where Australian retirement savings actually reside.

The sovereign wealth dimension sits at the extreme end of the spectrum. A constructed blend across Australia's full seven-fund sovereign wealth complex — a pool of approximately $356 billion — suggests between 65 and 75 per cent is deployed offshore, though no official consolidated figure exists for the complex as a whole. The Future Fund alone holds approximately 77 per cent of its physical assets outside Australia on a look-through basis, a figure that reflects both the legal mandate to maximise long-term return at acceptable risk and the structural inadequacy of the domestic opportunity set at the scale required.

The forward-looking signal is equally unambiguous. ASFA's March 2026 analysis found that 63 per cent of fund respondents expect to increase international exposure over the next two years, with medium and large funds projecting 50 to 60 per cent of every incremental dollar flowing offshore. Against net contributions of $77.7 billion for the year to June 2026, that projection describes an annual outflow running well into the tens of billions — not a portfolio marginal adjustment, but a structural feature of the system's growth mechanics.

Where that capital lands is now a matter of public record, and the pattern is broadening beyond the United States. The United Kingdom has become the most developed second channel. AustralianSuper, the country’s largest fund with more than A$410 billion under management, has run a London office since 2016, holds approximately £8 billion of UK assets — including a 74 per cent interest in the King’s Cross Estate and a 32 per cent stake in Peel Ports — and in March 2024 committed a further £8 billion of new UK capital by 2030. The British government’s response is the more telling data point: in 2025 it established a dedicated “Supers Unit” within the Office for Investment to court Australian funds, citing £41 billion already invested and a projected £99 billion of deployment over the coming decade. India has moved from frontier allocation to core holding. AustralianSuper’s July 2026 commitment of A$500 million to India’s National Investment and Infrastructure Fund — its second, following A$240 million in 2019 — lifted the fund’s total India exposure to approximately A$3.3 billion, with its chief investment officer citing policy consistency as a key factor. A foreign government building a unit to receive Australian retirement savings, and an Australian fund naming policy consistency abroad as a reason for deploying there, describe from the receiving end the same gap this dossier examines from the domestic side.

What the aggregate figures do not reveal is why the shift has been so sustained and consistent. The answer lies in the legal architecture governing trustees, examined in the following section — but the data establish the precondition clearly: at the scale Australian superannuation now operates, a domestic equity market representing roughly two per cent of global indices and concentrated overwhelmingly in banks and resources cannot absorb the pool without compounding exactly the concentration risk trustees are legally obliged to manage.

The Trustee Imperative: Why the Law Points Outward

The legal architecture governing Australian superannuation trustees rests on two interlocking statutory commands, and neither mentions Australia. Under sections 52 and 62 of the Superannuation Industry (Supervision) Act 1993 (Cth), every trustee of a regulated fund must act in the best financial interests of beneficiaries and maintain the fund solely to provide retirement benefits. APRA's guidance is unambiguous: the sole purpose test, read with the prescribed investment restrictions, ensures the retirement income objective is paramount. There is no statutory domestic-investment quota anywhere in the legislation — not a word of it.

The corollary is uncomfortable for industrial-policy advocates. Directing a trustee to overweight Australian assets collides directly with the sole purpose test and the best-financial-interests duty unless the domestic asset clears the same risk-adjusted return hurdle as its offshore competitor. That is not a matter of discretion; it is statutory compliance enforced by both ASIC and APRA, with civil penalties attaching to any contravention of the section 52 covenants, as the Australian Taxation Office confirms in its guidance on the best financial interests duty. The duty is not aspirational. It is enforceable.

Geography compounds the legal imperative. Australia represents approximately two per cent of world equity markets — a narrow, heavily concentrated base dominated by banks and resources. Diversification beyond that base is therefore a legal obligation, not a preference. A trustee who confined members' capital to the domestic market without clearing the return and diversification covenants in section 52(2)(c) would be in breach, not in compliance. The law, read honestly, points outward by construction.

The Future Fund illustrates the structural ceiling on legislated regard. It sits under a different statute — the Future Fund Act 2006 and the November 2024 Investment Mandate Direction — but the primary duty is materially equivalent: maximise long-term return at acceptable risk. The 2024 mandate instructs the Board to "have regard to" three national priorities — energy transition, housing supply and infrastructure — except where doing so would be inconsistent with the Act or the mandate itself. The practical result is approximately $3.5 billion of new domestic priority commitments against a $290 billion fund, with the overwhelming majority of physical assets remaining offshore. That is regard; it is not a floor. No amount of ministerial instruction changes the arithmetic when the statutory primary duty remains return maximisation. Regard without competitive return is a preference the market consistently overrides.

The section 52 investment covenant reinforces the point with specificity. It requires trustees to give explicit regard to risk, return, diversification and liquidity — each framed as a member-outcome criterion, not a national-interest criterion. APRA administers that obligation as a hard statutory duty, not a soft governance expectation. Funds that have sought to signal domestic commitment through allocation tilts face the same compliance question every investment committee must answer: does this asset clear the hurdle on the same terms as the offshore alternative? If it does not, the tilt is a breach. If it does, no special direction was needed — the trustee would have allocated there anyway.

The absence of a domestic quota is therefore not a gap in the legislative design; it is the design. Parliament constructed a system whose fiduciary core is the member's retirement outcome, not the national current account. The consequence — that approximately half of the institutional book now sits offshore and is projected to rise further — is the mathematically foreseeable result of applying that design to a domestic market whose weight in global indices is roughly two cents in every dollar. Reform of that trajectory must address the domestic return and pipeline environment. Legislating a quota would not change the arithmetic; it would simply expose trustees to a different statutory breach.

The Domestic Bottleneck: Regulatory Friction and Policy Uncertainty

The domestic environment is not failing to attract capital through any single, correctable defect. It is failing through the simultaneous operation of several reinforcing frictions, each of which individually raises the required return for a prospective investor, and which together are sufficient to redirect allocation decisions to competing markets. The most visible of these frictions sits at the regulatory gateway administered by the Foreign Investment Review Board.

Treasury introduced a stated thirty-day decision-making target from January 2025, a reform explicitly designed to signal that Australia was open for business and capable of processing proposals with commercial urgency. The quarterly data do not support that signal. Jones Day's June 2026 analysis of Treasury's own figures recorded a median FIRB processing time of thirty-five days in Q1 2026 — five days beyond the target — with only forty-six per cent of proposals resolved within the thirty-day window. The median itself understates the problem for the investors who matter most. Complex infrastructure acquisitions, energy assets, and anything touching a sensitive sector routinely extend well beyond the statutory period; the thirty-five-day median is pulled down by straightforward residential and commercial transactions that have no bearing on the infrastructure pipeline Australia needs to fill.

The compliance trajectory is moving in the wrong direction. Rather than converging toward full adherence to the thirty-day target as the January 2025 reforms bedded in, the resolution rate has deteriorated across successive quarters, reaching its lowest recorded print in Q1 2026. For institutional allocators conducting forward-looking due diligence on the Australian market, this is not an administrative curiosity — it is evidence that process friction is structural, not residual.

Approval delay compounds with a second, qualitatively distinct problem: policy uncertainty. The 2026 FIRB reform package grants the Treasurer expanded power to designate new sensitive sectors rapidly, without the consultative lead-time that normally precedes legislative change. Investors in critical infrastructure, critical minerals, technology and data centres are now explicitly told to anticipate "dynamic, changing FIRB requirements going forward," in MinterEllison's direct characterisation. A regime that is simultaneously being reformed and expanding its own discretionary reach creates precisely the environment institutional allocators price with a premium — or avoid. Allens judged that no exposure draft legislation had been released as at May 2026, and that finalised legislation was unlikely before late 2026 at the earliest. Investors making commitments in 2026 are therefore pricing against a regulatory framework that is openly in flux and whose final contours are unknown.

Sovereign wealth funds — the counterparties Australia most needs to attract to fill its infrastructure financing gap — are classified as foreign government investors and remain subject to full FIRB screening even for transactions that, in MinterEllison's assessment, have "little, if any, impact on Australia's national interest." The classification is not calibrated to commercial risk; it is categorical. A large pension fund from Canada or Norway seeking to co-invest alongside an Australian superannuation fund in a domestic toll road faces the same mandatory screening process as a state-directed acquirer of a sensitive defence supplier.

Infrastructure Australia's 2026 Annual Budget Statement places the commercial consequence precisely: approval delays are causing investors to move offshore and global supply chains to disengage. That finding, from the Commonwealth's own infrastructure advisory body, closes any interpretive gap between process friction and capital allocation outcomes. The connection is not theoretical. Macquarie's chief executive stated at the September 2026 infrastructure summit that sovereign wealth funds are telling policymakers they will either bypass Australia or demand a higher return because of delays, unexpected outcomes, and the absence of certainty — a convergence of revealed preference by actors operating under no Australian legal constraint whatsoever.

The IPA 2026 Investment Monitor captures the competitive context. While Australia held its position as a preferred infrastructure market at fifty-six per cent — unchanged — the United States vaulted from forty-eight to eighty per cent in a single year. The same competing markets absorb both inbound sovereign capital that might otherwise enter Australia and outbound Australian institutional capital already leaving. Until approvals timelines shorten, policy settings stabilise, and the reform package is legislated with enough certainty to anchor forward-looking due diligence, the required return premium on Australian infrastructure assets will persist — and the allocation math will continue to point outward.

The Sovereign Wealth Signal: Foreign Capital Draws the Same Conclusion

The most arresting evidence in the current debate did not arrive through a portfolio disclosure or a regulatory filing. It arrived through a public warning from the chief executive of Macquarie Group — one of the world's largest infrastructure investors — delivered at an infrastructure summit in the first week of September 2026. Shemara Wikramanayake reported that sovereign wealth funds are telling policymakers directly: they will either not invest in Australia, or they will demand a higher return premium, because of approval delays, unexpected outcomes, and an absence of certainty. Her call for a "sense of crisis" was precise professional testimony, not rhetorical escalation. It summarised the revealed preference of the very class of institutional counterparty that Australia most urgently needs to attract.

The quantitative anchor for that testimony is the Infrastructure Partnerships Australia 2026 Investment Monitor, a survey of owners and managers of more than A$530 billion of global infrastructure assets. Its central finding is precise and uncomfortable. Australia held its position as a preferred infrastructure market at 56 per cent — unchanged from the prior survey. The United States, by contrast, vaulted from 48 per cent to 80 per cent. That is not a marginal reordering at the margins of investor preference. It is a structural reallocation of global infrastructure capital at scale, occurring in exactly those markets — the United States, the United Kingdom, Canada, and Singapore — that simultaneously absorb outbound Australian superannuation dollars.

The significance of that convergence is interpretive as much as quantitative. One reading of Australian super funds' offshore allocation — a reading favoured by those who wish to separate the fiduciary question from the policy question — holds that the outflow is explained entirely by legal obligation. Trustees operating under sections 52 and 62 of the SIS Act must diversify out of a narrow domestic base; offshore allocation is therefore legally compelled rather than commercially motivated. That argument, whatever its partial validity, collapses when placed alongside the IPA data and Wikramanayake's testimony. Foreign sovereign wealth funds operate under no Australian legal constraint whatsoever. They face no SIS Act covenant, no APRA-enforced sole-purpose test, no obligation to look beyond domestic borders. And yet they are drawing an identical conclusion: Australia's risk-adjusted returns, net of approval friction and policy uncertainty, do not clear the hurdle available elsewhere. Two entirely independent sets of actors, subject to entirely different legal architectures, are receiving the same price signal. That is not coincidence; it is market consensus.

The Future Fund's position sharpens the point further. Its Board has been formally directed since November 2024 to "have regard to" national priorities in housing, energy transition, and infrastructure — a direction that carries the weight of a government Investment Mandate. And yet the overwhelming majority of its physical assets remain deployed offshore. The mandate uses the word "regard"; it does not use the word "floor". Where domestic investments cannot clear the return hurdle, regard is precisely what results: consideration, not commitment. The inbound sovereign wealth data confirms that this outcome is rational rather than parochial. If foreign sovereigns, unconstrained by any domestic-preference obligation, are also declining to deploy capital at scale into Australian infrastructure, the problem is not that Australian trustees have adopted an excessively global outlook. The problem is that the domestic opportunity does not price competitively.

"Sovereign wealth funds are saying they will either not invest in Australia or demand a higher return because of delays, unexpected outcomes, and a lack of certainty."
— Shemara Wikramanayake, CEO, Macquarie Group, infrastructure summit, September 2026

What Wikramanayake's warning and the IPA Monitor together establish is that the fiduciary argument and the domestic-conditions argument are not alternative explanations for the capital outflow — they are the same problem observed from different vantage points. Australian trustees send capital abroad because the domestic pipeline is too thin, approvals are too slow, and policy settings are too uncertain to clear the return bar that law requires them to maintain. Foreign sovereigns decline to enter for precisely the same reasons, expressed not in legal language but in the language of required return premiums and portfolio exclusions. Until approvals timelines shorten materially, policy settings stabilise across electoral cycles, and the investable pipeline deepens in both scale and certainty, neither set of actors will change its calculus. The sovereign wealth signal is not a warning about the future; it is a contemporaneous reading of the same conditions already driving the domestic outflow.

The Counter-Case: Diversification, Returns and Member Obligation

The case for offshore allocation does not rest on preference, ideology, or a fondness for global markets. It rests on statute, arithmetic, and the observable performance record. Each element deserves to be stated at full strength, because the counter-case is genuinely compelling — and because understanding precisely where it succeeds, and where it stops, clarifies the policy problem with more precision than either side of the debate typically allows.

The statutory foundation is s52(2)(c) of the Superannuation Industry (Supervision) Act 1993 (Cth), which requires trustees to act in the best financial interests of beneficiaries. APRA administers that obligation as a hard statutory duty, not a soft aspiration, and civil penalties attach to any breach of the section 52 covenants. The investment covenant further requires explicit, documented regard to risk, return, diversification, and liquidity. Retirement income is the paramount objective; geography is irrelevant to the statute unless it serves that objective. A trustee who maintained a domestic tilt that demonstrably underperformed an available offshore alternative would not be acting prudently — they would be in breach.

The arithmetic reinforces the law. A pool of the scale now managed by Australian institutional funds cannot be deployed into a domestic market representing approximately two per cent of world equity indices — and concentrated overwhelmingly in banks and resources — without amplifying exactly the concentration risk the investment covenant requires trustees to manage. As Super Review's 2023 analysis concluded, "a key finding is the continued internationalisation of investment portfolios, which highlights the ongoing challenge for large funds in deploying incoming capital to the domestic market without amplifying concentration risk." Offshore allocation rose not because trustees adopted globalisation as an aesthetic, but because the domestic opportunity set did not scale with the pool. When net contributions run at the pace described in the earlier sections of this analysis, each new increment must find a home that does not crowd existing exposures — and the domestic market cannot absorb flows of that magnitude without distorting pricing and undermining the very return rationale for domestic investment.

The performance record supports the allocation. NAB's 2026 Super Insights survey characterised the shift beyond fifty per cent international as "a commitment to diversification and delivering the best risk-adjusted returns for members," while industry analysts have observed that surging global markets have exposed the performance drag of a heavy domestic bias. The international equity surge of recent years — led by US technology and global infrastructure, with UK real assets and Indian infrastructure increasingly in the mix — has validated the reallocation in realised returns. Trustees who resisted the trend on domestic-preference grounds would have underperformed their peers on a metric APRA watches directly through the annual performance test.

The fiduciary argument is also, in an important sense, self-reinforcing over time. Once a fund has built offshore relationships, infrastructure, currency-hedging capability, and co-investment networks, the marginal cost of the next offshore allocation falls relative to the marginal cost of the next domestic one — particularly if domestic deal flow is thin, approval timelines uncertain, and project terms subject to post-commitment revision. The NAB data and the ASFA survey both reflect not just a snapshot of current allocation but the institutionalisation of offshore capability that has occurred over the past decade.

The counter-case is, therefore, compelling — precisely as far as it goes, which is as far as the member's balance sheet. What it does not answer — what it cannot answer — is whether the domestic pipeline, approvals architecture, and policy stability are adequate to give trustees a genuine domestic option at competitive returns. If they are not, the fiduciary argument for offshore allocation remains wholly intact. But that is itself the indictment. The duty to members and the failure of the domestic environment are not competing explanations for the same phenomenon. They are the same problem observed from opposite ends: trustees are acting lawfully and rationally, and the consequence is capital leaving a country that needs it. The counter-case does not refute the domestic-environment indictment — it mirrors it.

The Policy Gap: What a Coherent Response Would Require

The diagnosis assembled across the preceding sections is clear enough: Australian institutional capital is leaving because the domestic opportunity set does not clear the same risk-adjusted return hurdle as competing markets, and inbound sovereign capital is drawing the identical conclusion independently. The prescription is harder to execute, but it is not opaque. Four interlocking reforms would materially change the calculus — and one constitutional boundary must be stated plainly before any of them is applied.

Approvals consolidation. The problem is not that Australia applies rigorous environmental or national-interest standards; it is that the same standard is administered by multiple agencies across multiple tiers of government, generating compounding friction without additional protection. NSW's December 2025 planning overhaul introduced a single-agency coordination model for development applications — the first systematic attempt by any Australian jurisdiction to collapse sequential referrals into concurrent assessment. The Commonwealth needs an equivalent mechanism for nationally significant infrastructure. Infrastructure Australia's 2026 Budget Statement found explicitly that approval delays are causing investors to redirect capital offshore and causing global supply chains to disengage from Australian projects. That is not an abstract productivity concern; it is a pipeline-destruction mechanism operating in real time. A single Commonwealth coordination point — with statutory timelines binding on all referral agencies simultaneously — is the minimum structural fix at the gateway level.

Pipeline transparency. Trustees placing tens of billions of dollars per quarter cannot make domestic infrastructure allocations against a pipeline that is opaque, subject to re-sequencing between budget cycles, and does not survive elections intact. The UK Government's national infrastructure pipeline was designed precisely to address this problem, providing what its originating documentation described as enhanced "visibility and certainty for investors and the supply chain." Australia has no equivalent mechanism at Commonwealth scale. The consequence is that a trustee who might otherwise favour a domestic asset over an offshore one — all else being equal — cannot construct the forward commitment schedule that large-fund governance and liquidity management require. A credible, legislatively anchored, multi-year infrastructure pipeline, published and updated on a fixed schedule and insulated from routine budget re-prioritisation, would give domestic allocators the forward visibility that offshore markets already supply through listed instruments and established sovereign issuance programmes.

Mandate stability and binding concession frameworks. Shemara Wikramanayake's testimony at the September 2026 infrastructure summit identified changing terms after commitment as one of the three forces driving sovereign capital away. That is precisely the behaviour that raises the required return for all subsequent investors in a jurisdiction, because sophisticated capital prices not only current terms but the probability distribution of future interference. Binding long-term concession frameworks — insulated from inter-election reversals through independent oversight mechanisms or legislated stabilisation clauses — are the structural remedy. Where governments reserve the right to alter material terms, that right should be exercised through transparent, pre-specified compensation mechanisms, not through administrative re-designation or retrospective regulation. The MinterEllison analysis of the 2026 FIRB reform package noted that investors in critical infrastructure and data centres should "expect dynamic, changing FIRB requirements going forward" — precisely the posture that makes binding concession frameworks necessary as a counterweight.

The boundary that none of this can cross. None of these reforms operates, or should operate, as a legislative override of trustee law. The SIS Act's sole-purpose test and best-financial-interests duty are not inconveniences to be worked around; they are the reason the $4.8 trillion system commands member trust and produces retirement income at scale. There is no statutory domestic-investment quota in the existing legislation, and there should not be one. The Future Fund's November 2024 Investment Mandate Direction — instructing the Board to "have regard to" national priorities in energy, housing and infrastructure — has not reversed the offshore trend, because regard is not a return. It is a preference the market is consistently overriding for rational reasons. Legislative direction does not alter that arithmetic; it merely transfers the underperformance cost from the policy architect to the retiree.

The variable that changes the allocation arithmetic is not legislative direction — it is pipeline depth. A trustee facing a deep, visible, stable domestic pipeline of infrastructure assets priced at competitive risk-adjusted returns will allocate domestically, because the fiduciary duty and the commercial incentive will point in the same direction. A trustee facing an opaque, unstable pipeline with uncertain regulatory outcomes will allocate offshore, for exactly the same reason. The four reforms described above are designed to create the former condition: they address the gateway friction that delays or kills deals at origination, the visibility deficit that prevents forward commitment, and the mandate instability that raises the required return ex ante. None of them compels a trustee to accept a submarket return. All of them reduce the probability that the domestic market requires one.

The Clifford Chance assessment of the 2026 FIRB reforms characterised the package as a belated acknowledgement that process friction carries a price — but noted that capital-allocation habits have already formed offshore. Habits formed over a decade of rising international exposure, supported by superior offshore returns and deepening global infrastructure markets, will not reverse in response to marginal process improvements. The reforms required are structural, durable, and cross-jurisdictional. They are also, in principle, achievable — which makes the absence of a coherent Commonwealth programme to deliver them the most consequential policy gap in Australian institutional finance.

Conclusion and Sources

The evidence assembled across this dossier resolves into a single, uncomfortable finding: Australia has become a jurisdiction its own institutional capital does not find sufficiently rewarding, and the capital-allocation habits formed in that environment are now structural. Roughly half of institutional superannuation assets sit offshore. The Future Fund holds approximately three-quarters of its physical assets outside Australia. Funds expect fifty to sixty cents of every new incremental dollar to follow the same path. And foreign sovereign wealth funds — operating under no Australian legal constraint whatsoever — are arriving at the same destination by independent reasoning, telling policymakers directly that they will either skip Australia or demand a higher return premium because of approval delays, shifting regulatory settings, and the absence of certainty.

The honest framing matters here. This is not a story about trustees behaving irresponsibly, or about capital being disloyal. Diversification beyond a market representing approximately two per cent of global equities, heavily concentrated in banks and resources, is a statutory obligation under the Superannuation Industry (Supervision) Act 1993. APRA administers that obligation with civil penalties. No trustee has discretion to overweight Australian assets unless those assets clear the same risk-adjusted return hurdle as their offshore competitors. The offshore shift is therefore the mathematically predictable output of sound governance applied to a domestic environment that has not kept pace.

What transforms the situation from policy inconvenience into structural urgency is the convergence of two independent signals. Australian institutional capital is leaving. Inbound sovereign capital is refusing to fill the gap. Both sets of actors are receiving the same price signal: process friction is high, policy certainty is low, and comparable returns are available elsewhere without either friction. The May 2026 FIRB reform package — promising faster approvals, expanded use of class exemptions, and a new streamlining framework — represents a belated acknowledgement by government that process friction carries a real economic cost, as Clifford Chance noted in its July 2026 analysis. The reforms are genuine, and they are welcome. But as at May 2026 no exposure draft legislation had been released, specialist commentary placed full implementation no earlier than late 2026, and the same package simultaneously proposes to expand ministerial power to designate new sensitive sectors dynamically — a provision that, if exercised, would extend exactly the unpredictability that institutional allocators price as risk.

Capital-allocation habits, once formed, are durable. Offshore deployment teams have been built. Bilateral relationships with foreign regulators and co-investors have been established. Deal pipelines in North America, Europe, and Asia have been populated. Each additional year of domestic friction deepens those relationships and raises the switching cost of returning capital home. The reforms of 2026, if implemented in full and with restraint on new discretionary powers, may stop the deterioration. They will not, by themselves, reverse it. Reversing the trend requires the four structural changes identified in the preceding section: approvals consolidation, a credible national infrastructure pipeline, binding long-term concession frameworks insulated from inter-election revision, and a domestic pipeline deep enough to absorb institutional allocation at scale without amplifying concentration risk. None of those changes overrides trustee law — nor should they. Pipeline depth, not legislative direction, is what changes the arithmetic.

The system continues to grow at a rate that makes complacency expensive. With net contributions running at $77.7 billion in the year to June 2026 and total assets on a trajectory that will comfortably exceed $5 trillion before the decade is out, the volume of capital looking for a home each year is not a rounding error. If domestic conditions do not improve, the next phase of growth will extend the offshore trajectory by sheer force of arithmetic. That outcome is avoidable. It is not yet avoided.

Sources