Governance: The Quiet Driver

BEN LEVY, ESG Analyst

Why ‘G’ remains the most reliable engine of long-term value

Ask investors what “ESG” means to them, and the conversation drifts quickly to carbon footprints, supply chains, diversity targets – the environmental (E) and the social (S). Ask them what actually moves the needle on returns, and the answer changes. In a 2024 survey of 47 major institutional investors and asset owners by Stanford Graduate School of Business¹, 68% named governance as the primary consideration shaping their ESG-related decisions. Environmental factors trailed at 23%, and Social factors barely registered at 2%.

Source: Stanford GSB.

We have spent decades at AVI making this argument through where we allocate capital and how we vote: the businesses worth owning for the long run are, almost without exception, well-governed ones. This is the case for why governance (G) has never stopped being the driver, even as the spotlight moved elsewhere.

The Evidence: Governance as the Value Driver

Governance is arguably the most quantifiable of the three ESG pillars. Where E and S cases can often rest on longer-term or harder-to-isolate outcomes such as decarbonisation across a global supply chain, G is the one where the evidence has had the most time to accumulate. This link between governance and value has been tested repeatedly, and the data continues to confirm it.

MSCI research shows that in the U.S., companies that displayed governance leadership consistently outperformed governance laggards between 2015 and 2023, with an excess annualized return of 2.7% over the full nine-year study (26.3% cumulative)². That is nearly a decade of compounding evidence that boards which hold management accountable, separate oversight from execution, and protect minority shareholders appear to run better businesses.

The mechanism behind this is straightforward. MSCI’s study of 4,319 global issuers from 2015-2024 found that companies in the top ESG-rating quintile financed themselves at an average cost of capital of 6.8%, against 7.9% for the bottom quintile – a full percentage point, and statistically significant at the 99% confidence level³. Markets, in other words, already appear to price governance quality into the cost of every pound a company raises. Good governance is not a nice-to-have sitting alongside the balance sheet; it is baked into it.

This is echoed at the institutional level. The OECD’s 2025 Corporate Governance Factbook finds 76% of the 52 jurisdictions it tracks now require or encourage the separation of CEO and board chair roles, up from just 44% a decade earlier⁴. Regulators, like markets, have concluded that governance architecture is not incidental to performance; it is a precondition for it. It is the same logic that has underpinned AVI’s engagement for decades: find where governance lags value and close the gap.

Governance Reform is Not a Straight Line

Progress on governance reform is never uniform. Some jurisdictions push ahead with real conviction, others quietly retreat, and the overall picture looks less like a steady march forward than a patchwork of gains in one market offset by losses in another. Japan is the clearest recent case of the former. Since the Corporate Governance Code was introduced over a decade ago, and revised in July 2026, the share of Prime-listed companies with at least one-third independent directors has risen from 6.4% in 2014 to 98.8% in 2025⁵. Cross-shareholdings have plummeted from 50% of total market capitalization in 1990 to just 11% today⁶. Progress, but not completion: cross-shareholding unwind remains slower than the headline figures suggest, and disclosure of what companies do with the proceeds is still inconsistent.

Korea offers an even sharper before-and-after. In the decade before the Corporate Value-Up Program launched in 2024, the KOSPI grew just 35%, against 179% for the S&P 500 and 155% for the Nikkei 225⁷. In response to persistent demands from foreign and domestic investors to address the issue, the government-initiated actions to improve capital efficiency and align the market closer with global corporate governance best practices. Since the programme began, buybacks have risen from KRW 8.2 trillion (2023) to KRW 20.1 trillion (2025), and cancellations have quadrupled to KRW 21.4 trillion⁸. The Korea Value-up Index rose 89.4% in 2025 alone, outpacing the broader KOSPI by nearly 14 percentage points.

Set this against the picture in more “mature” markets, and the straight-line narrative breaks down entirely. The UK is home to some of the world’s oldest governance codes, yet lost 88 companies to delisting in 2024, the highest since the financial crisis, against just 18 new listings, despite efforts by the UK government, regulators and the LSE to boost the City’s attractiveness by reforming market rules⁹. This doesn’t reflect a governance failure, but rather two structural headwinds that have swamped the UK’s governance advantage: nearly a decade of Brexit-related uncertainty that cut London’s share of European IPO proceeds from 40% to 30%¹⁰, and a near-total retreat by UK pension funds from domestic equities, whose ownership of the UK stock market has fallen from 39% in 2000 to just 4% in 2023¹¹. Good governance did not cause this exodus, and cannot reverse it alone; which is why hands-on engagement matters more in a market like this, not less.

Perhaps the starkest evidence of reform’s non-linearity comes from dual-class shares. Dual-class structures show why markets eventually catch up with governance risk, even when it takes years. A 2024 study spanning 40 years of IPO’s found that dual-class firms trade at a 13.9% valuation premium in their first year on the market (Figure 3), but that premium fully reverses into a discount within seven to nine years, as the gap between insider voting power and cash-flow rights becomes impossible to ignore¹².

None of this should surprise a serious observer of corporate governance. Reform is never granted; it is negotiated, market by market, board by board, and it can just as easily be legislated away as it was legislated in. That is precisely why active, informed engagement matters more, not less, in markets where the direction of travel looks settled.

Why Engagement, Not Exit

Governance can regress even in well-regulated markets. The real question is what an investor does about it. A conventional response is to exit – sell out of the underperformer, reinvest where governance is stronger, and leave the valuation gap for the next owner to solve. That does not close a governance discount; it simply changes whose problem it is. We would rather be the ones who close it, and increasingly, the evidence backs that approach.

Institutional investors overwhelmingly say engagement shapes outcomes: 85% report that engagement significantly influences their voting decisions, and 52% say they are demanding more direct access to board members than they were a few years ago¹³. But support must be earned, not demanded; 87% of respondents say they back proposed change at a company only when the financial case is sound and the plan is credible – not because the request was persistent.

Former AVI investee company Hipgnosis Songs Fund is a clear example of what this looks like in practice. Our concerns about the fund’s financial controls, transparency, and the effectiveness of its board had built for some time but came to a head in 2023 when management proposed selling some of their music catalogues to a related party at a steep discount to its carrying value. Rather than sell out, we published a public letter urging shareholders to vote against the proposals, and shareholders responded decisively – 83% voted against continuation, and 84% rejected the sale. Two directors resigned before the AGM and the chairman lost his own re-election vote. We then worked with other shareholders to install two new independent directors, one of whom became Chairman. Within months, a competitive process delivered a bid from Blackstone at a 48% premium to the undisturbed share price¹⁴.

Aker is another AVI example that illustrates something related but different: what concentrated ownership can achieve, and where its limits lie. We held a position in the Norwegian industrial holding company from 2008 to 2025, drawn by the alignment of interest with a controlling shareholder that consistently thinks in decades rather than quarters – and has sagaciously allocated capital to the benefit of all shareholders. That kind of patient capital allocation is exactly what we look for in a controlling shareholder, and it has delivered strong returns over the holding period. But it is not a guarantee against misjudgement: Aker Horizons has faced a far harder road than the market expected when it launched in 2020, a point its own Chairman has acknowledged directly¹⁵. We take that as validation rather than contradiction of our approach – even well-run, long-term-minded companies benefit from shareholders who keep asking questions, because governance is a standard that has to be maintained, not a box that gets ticked once.

This distinction matters more than it might first appear, as a literal reading of governance scorecards would flag much of our own portfolio at AVI as a problem. Holding companies, by design, concentrate control in a single family or founding shareholder; and thus we likely carry a higher proportion of dual-class structures than the broader market. On a rigid checklist that prioritises board independence and one-share-one-vote, these companies would score poorly. We do not see that as a contradiction of our own argument, we see it as a persistent governance misdiagnosis in the market.

Concentration of control is not, on its own, a governance risk; the absence of accountability is. Telling the two apart, identifying which controlling shareholders behave like Aker’s and which behave like Hipgnosis’s former board, is a skillset that has built at AVI through case-by-case engagement. It also has an evidence base of its own: Credit Suisse’s Family 1000 research finds that family- and founder-controlled businesses have outperformed the broader global universe of non-family-owned companies at an annual average of 404 basis points since 2006¹⁶.

The broader trend points the same way. 2025 was a third consecutive record year for global shareholder campaigns pushing for change¹⁷, with capital-allocation demands now appearing in 31% of them, up from a 19% five-year average – and European AGM seasons are seeing rising opposition to remuneration and share-issuance resolutions, up from 30.7% to 37.9% year-on-year on pay alone¹⁸. Engagement, in short, is not a relic of stewardship codes, it is becoming the market’s default response to governance that falls short.

What Good Looks Like Going Forward

If the last decade has taught us anything, it’s that governance progress is measurable, and worth measuring precisely because it can stall or reverse. A few markers are worth watching closely.

Board independence remains the clearest structural signal, and the shift has been global rather than confined to any single market. Across the 52 jurisdictions the OECD tracks, the share requiring or encouraging separation of the CEO and board chair roles has grown from 44% to 76% since 2014, and board-level responsibility for risk oversight is now written into rules or codes in 92% of jurisdictions, up from 62% a decade ago¹⁹. Disclosure has kept pace with structure: 88% of jurisdictions now require or recommend that companies disclose director qualifications, up from just 61% in 2014. What was once a patchwork of local practice has converged into something close to a global norm.

Capital discipline is the second marker. Global dividends hit a record $1.75 trillion in 2024, while buybacks – up nearly 182% since 2012 versus a 54% rise in dividends – have gone from just over half of dividend value to 94% of it²⁰. Set alongside the Korea and Japan figures from earlier in this piece, the pattern is global: boards are being held to a higher standard on what they do with shareholders’ capital, not just how they govern themselves.

The third marker is simpler than it sounds: how a board treats deals that benefit an insider at other shareholders’ expense. Related-party transactions are the clearest test. The OECD finds 94% of the 52 jurisdictions it tracks now require immediate disclosure of these deals, up from 50% in 2016, and 87% require board approval, up from 54% a decade ago. That is a lens we often return to in our own engagement work across the globe.

None of these markers guarantee an outcome. But together, they describe what we look for before we can call a company well-governed: a board that answers to shareholders, capital allocated with discipline, and minorities protected as a matter of structure rather than goodwill.

Closing

Return, for a moment, to the survey we opened with, where 68% of institutional investors named governance as their primary ESG consideration, against 23% for the environment and 2% for social factors²¹. That gap is not really a verdict on E and S which ask you to price in value that has not yet arrived. G asks something more fundamental, that the people running a business are accountable to the people who own it. There is nothing fashionable about that, it has simply always been correct.

Decades of engagement have taught us that this accountability is never permanently secured. It is built, market by market, board by board, and it must be defended just as deliberately as it was won – in Tokyo and Seoul as much as here in London. That, in the end, is the quiet driver.

Scan to visit our Insights page: