Asia-Pacific Investment Report 2021 — Regulation as a Repricing Event

2021 in Asia-Pacific was defined by an event that had no global parallel: a substantial repricing of one of the region's largest markets driven by regulation rather than by monetary conditions.

Measures were introduced across several sectors — platform competition, data handling, financial services offered by technology firms, private education, and gaming — over a compressed period. Each had a stated policy rationale, and the rationales were coherent: competition concerns about dominant platforms, data security, financial stability where lending occurred outside the regulated banking system, household cost pressures, and effects on young people.

Individually, each measure resembled regulatory actions taken elsewhere. Competition authorities in multiple jurisdictions were examining platform market power. Data protection regimes were tightening globally. Non-bank lending was under supervisory attention in many markets.

Collectively, and in the compressed period over which they arrived, the effect was a substantial repricing of an entire category of assets. In some sectors the measures changed the business model fundamentally rather than constraining it at the margin.

The analytical lesson is the durable part, and it is frequently drawn incorrectly. The lesson is not that this market is uninvestable. It is that policy objectives are a variable to be modelled explicitly, and that a market where policy can change business models rapidly requires a different framework from one where regulatory change is slow and heavily litigated.

A second consequence: capital reallocated within the region rather than leaving it. India and Southeast Asia received substantially increased attention as allocators sought regional exposure with different policy characteristics.

Why the measures were hard to anticipate financially

The difficulty investors had is worth examining, because it reveals a genuine limitation in how policy risk is usually assessed.

A financial framework assesses regulation by its cost. What will compliance cost, how much revenue is affected, what is the margin impact. This works when regulation constrains an activity at the margin.

It works poorly when the objective is not economic. Several 2021 measures were directed at social objectives — household costs, effects on young people, data sovereignty — where the economic impact on affected companies was not the primary consideration and in some cases was accepted as a necessary cost.

A measure aimed at a non-economic objective does not stop when the economic cost becomes large. That is what made the outcomes hard to bound from a financial framework: the usual limiting mechanism, in which economic consequences constrain regulatory ambition, was weaker.

The practical requirement this establishes:

• Read policy documents, not market commentary. Stated objectives in official publications were, in several cases, clearer predictors of subsequent measures than financial analysis was.
• Assess whether a business model conflicts with a stated policy objective, not merely whether it complies with current rules. Compliance is a snapshot; conflict is a trajectory.
• Model policy scenarios explicitly, with probabilities, rather than applying an unspecified discount.

A business that complies with every current rule while operating against a stated policy objective is not low-risk. It is a business whose risk has not yet been expressed.

Policy risk as a distribution

The most useful analytical shift from 2021 concerns how policy risk is incorporated into valuation, and it applies to every market rather than to one.

The common approach is a discount. An investor applies a lower multiple to assets in markets perceived as carrying higher policy risk. This is intuitive and it is imprecise in a way that matters.

Why a flat discount misprices in both directions:

• It treats policy risk as a constant when it varies enormously by sector. A business aligned with stated policy priorities may face little risk; one in conflict may face existential risk. A single market-level discount applied to both is wrong twice.
• It does not distinguish direction. Policy can support as well as constrain. Sectors identified as priorities received substantial support, and a discount applied uniformly misses that entirely.
• It obscures the shape of the distribution. A 30% discount could represent a 30% chance of a total loss or certainty of a 30% reduction. These are completely different investments requiring different position sizes.

The better approach is scenario-based:

• Identify the specific policy objectives relevant to the sector, from primary policy documents.
• Assess whether the business model supports or conflicts with each.
• Construct scenarios with explicit probabilities and outcomes.
• Size the position to the downside scenario, not to the expected value — since the downside in a policy-driven repricing can be abrupt and does not permit gradual exit.

This is more work than applying a discount. It is also the only approach that distinguishes between a business the state wants to succeed and one it does not — which, in a market where that distinction determines outcomes, is the entire analysis.

The VIE structure exposed

A structural feature of how several regional companies had been made available to foreign investors received scrutiny it had largely avoided for two decades.

The structure. Foreign ownership was restricted in certain sectors. To permit foreign investment, companies used a variable interest entity arrangement: an offshore holding company, listed abroad, holds contractual rights to the economics of a domestic operating company without owning its equity. Foreign investors buy shares in the offshore entity.

What foreign investors actually hold. Not equity in the operating business, but shares in an offshore entity holding contracts that entitle it to the operating company's economics. The arrangement's enforceability depends on those contracts being honoured and on the arrangement remaining permitted.

The risk had always been present and was widely disclosed — listing documents describe it explicitly, usually at length in the risk factors. It had been treated as theoretical because the structure had operated for two decades without being challenged.

2021 made it concrete, not through the structure being invalidated but through demonstrating that the regulatory environment around it could change materially and quickly. Investors who had treated a disclosed risk as theoretical revised that assessment.

The general lesson is about disclosed risks. A risk that is disclosed, understood and has never materialised tends to be priced as though it will not. That pricing is a judgement about probability, and the judgement is usually based on the absence of past occurrence — which is weak evidence when the structure is young relative to the timescale of the risk.

A disclosed risk that has never materialised is not a small risk. It is an unmeasured one, and the absence of history is not evidence of safety.

Reallocation within the region

The consequence for capital flows was reallocation rather than withdrawal, and the pattern is worth noting.

India received substantially increased allocation. It offered a large market with strong growth, a different policy environment, an improving exit environment, and a technology ecosystem that had matured considerably. The 2021 Indian venture market was among the most active globally.

Southeast Asia received increased attention, though constrained by the fragmentation described in the 2019 report.

Japan and Korea received attention from investors seeking developed-market governance within the region.

The general observation is about how allocators respond to a policy shock: they rarely exit a region entirely, because the underlying growth case that justified the allocation is unchanged. They reallocate within it, toward markets with different policy characteristics.

This has a second-order consequence that is easy to miss. Capital arriving in a market because of conditions elsewhere is less price-sensitive than capital arriving because of conditions in that market. Valuations in the receiving markets rose partly for reasons unrelated to those markets' own fundamentals — the same dynamic the 2015 private equity report describes when institutions allocate out of necessity rather than conviction.

Investors entering India and Southeast Asia in 2021 were, in part, paying a price set by a regulatory event in a different country.

Reading policy documents as investment research

The claim that stated policy objectives predicted 2021 measures better than financial analysis did is unusual enough to deserve development, because it implies a research practice most investors do not have.

Why financial analysis under-performed here. A financial framework assesses regulation by its cost — compliance expense, revenue affected, margin impact — and implicitly assumes that as the economic cost rises, the regulatory ambition moderates. That assumption holds where the objective is economic. Where the objective is social, the economic cost is a price being paid rather than a constraint being respected.

What policy documents provide. Published national plans, sectoral guidance and regulatory consultations state objectives explicitly. They are written to be read, they are usually free, and in several cases the 2021 measures had been foreshadowed in them.

What to look for:

• Stated priorities. Which sectors are named as priorities and which are named as problems. Both are informative and the second is more so.
• Named concerns. Household costs, data security, effects on young people, financial stability outside the banking system. A business model that intersects a named concern is exposed regardless of its compliance status.
• Direction of travel. A policy framework that has been tightening in a direction over several years is likely to continue tightening in that direction.
• The gap between stated objective and current market structure. Where a policy objective is stated and the market has not yet moved toward it, the gap is where measures arrive.

The practical test this produces is different from a compliance assessment: does this business model support or conflict with a stated policy objective? Compliance is a snapshot of current rules. Conflict with an objective is a trajectory, and it predicts future rules.

A business complying with every current rule while operating against a stated policy objective is not low-risk. It is a business whose risk has not yet been expressed — and the objective was published.

This generalises beyond any one market. Every jurisdiction publishes policy intentions, and every regulated business intersects some of them. The practice is cheap, the documents are free, and it is systematically neglected because it does not resemble financial analysis.

What an allocator could act on

Replace the country risk discount with scenarios. A flat discount treats policy risk as constant when it varies enormously by sector and can run in either direction — sectors identified as priorities received substantial support. A discount also obscures the shape: a 30% discount could represent a 30% chance of near-total loss or certainty of a 30% reduction, and those require different position sizes.

Size to the downside branch. In a policy-driven repricing the adjustment is abrupt and does not permit gradual exit. A position sized to expected value is over-sized relative to the branch that would matter.

Treat a disclosed, never-materialised risk as unmeasured rather than small. The VIE structure's risks were described at length in listing documents for two decades. The absence of past occurrence is weak evidence about probability when the structure is young relative to the timescale of the risk.

Read 20-F filings for the issuer's own account of its structure. Free on SEC EDGAR, written under liability, and substantially more precise than secondary characterisations.

Recognise capital arriving for reasons elsewhere. Valuations in India and Southeast Asia rose in 2021 partly because of a regulatory event in a different country. Capital allocated because it must go somewhere is less price-sensitive than capital allocated on an assessment of the destination — and it can reverse for reasons equally unrelated. An investor entering a market on inflows generated elsewhere is paying a price set by conditions they are not exposed to.

What 2021 established for Asia-Pacific

• Regulatory repricing was demonstrated at scale, decoupling one large market from the global monetary cycle.
• Financial frameworks were shown to bound policy risk poorly when the objectives are non-economic.
• Policy risk was reframed as a distribution requiring scenarios, not a discount.
• The VIE structure's disclosed risk became concrete, illustrating that unmaterialised risk is unmeasured rather than small.
• Reallocation within the region raised valuations elsewhere for reasons unrelated to those markets' fundamentals.