Capital Structure and Taxes: What Happens When You (Also) Subsidize Equity?

Summary

"Capital Structure and Taxes: What Happens When You (Also) Subsidize Equity?" by Frédéric Panier, Francisco Pérez-González, and Pablo Villanueva (working paper, December 2015; presented at the BIS Research Network) is one of the best-identified evaluations of the debt-bias-reducing effect of an allowance for corporate equity. It uses Belgium's notional interest deduction (NID) — the "hard" full-stock ACE introduced in 2006 — as a natural experiment and finds that removing the tax code's discrimination against equity measurably de-levered Belgian firms.

Its headline is stated plainly in the abstract: "This paper shows that capital structure significantly responds to changing tax incentives."[1] That matters because, as the authors note, the prior empirical record was weak — they quote Myers (1984), "I know of no study clearly demonstrating that a firm's tax status has predictable, material effects on its debt policy," and Graham (2008), "there is no known study that documents tax-related time series effects in debt usage."[1] The Belgian NID gives them the clean shock the earlier literature lacked.

The Natural Experiment

The NID is "an explicit equity deduction introduced in 2006 with the objective of reducing the tax-driven distortions that favor the use of debt financing."[1] Mechanically, it "allows firms to deduct from their taxable income a notional charge equal to the product of the book value of equity times a benchmark interest rate based on historical long-term government bonds."[1] Because the deduction applies to the whole equity stock regardless of the source of new financing, "firms' marginal financing decisions are provided with a significant tax deduction regardless of their source of financing"[1] — precisely the rent-only design motion the wiki tracks: stop penalizing the normal return to capital, and the artificial premium on debt disappears.

Identification exploits the 2006 timing and, crucially, a control group: "The results are robust to using data from neighboring countries as a control group, as well as relying on a battery of tests aimed at isolating the effect of other potential confounding variables."[1] The data come from the National Bank of Belgium.

What It Establishes

The authors summarize "four" main findings:[1]

1. "The NID led to a significant increase in the share of equity in the capital structure."
2. "Both incumbent and new firms increase their equity ratios after the introduction of the NID."
3. "The largest responses to these changing tax incentives are found among large and new firms."
4. "The increase in equity ratios is explained by higher equity levels and not by a reduction in other liabilities."

The fourth point is the sharpest: firms did not merely reshuffle their balance sheets — they raised more equity, exactly the financial-stability improvement the debt-bias critique targets. The paper's overall conclusion generalizes the lesson: "the evidence demonstrates that tax policies designed to encourage the use of equity financing are likely to lead to more capitalized firms."[1]

Why It Matters Here

This is the leverage/debt-bias result the rent-targeting-taxes-reduce-debt-bias outcome rests on, delivered by a well-identified quasi-experiment on the strongest ("hard" full-stock) form of the ACE. It sits alongside Branzoli & Caiumi (2020) on Italy's incremental ACE: two different countries, two different ACE designs, the same direction — leverage falls when the tax subsidy to debt is removed. Together they make the debt-bias effect the best-replicated result in the ACE literature. The Belgian evidence is also cited in the allowance for corporate equity concept page (alongside Princen 2012 and aus dem Moore 2014).

Honest Limits

• A financing result, not an investment result. The paper identifies the effect on capital structure (equity share, leverage), which is the debt-bias margin. It does not establish that the NID raised real investment — that margin is the contested one in the ACE literature (Hebous & Ruf 2017 find no production-investment effect for multinationals; Konings, Lecocq & Merlevede 2022 find positive affiliate effects). This page supports the financing half of the outcome claim, not the real-investment half.
• Working-paper status. The version cited is the December 2015 working paper (BIS Research Network, 2016); confirm figures against the latest circulated draft.
• The "hard" ACE is expensive. Belgium's full-stock NID is exactly the fiscally costly design later narrowed (2018) and repealed (2023): its leverage effect is real, but the revenue cost and the multinational "double-dip" arbitrage are part of the same record (see the ACE concept page). The cleaner cost-effectiveness belongs to Italy's incremental design.

Bears On

• Benefit (supports the debt-bias claim): Rent-targeting corporate taxes reduce debt bias without penalizing marginal investment — supplies the quasi-experimental evidence that a full-stock ACE (Belgium's NID) raised the equity share via higher equity, the financing-side confirmation that a rent-only base removes the leverage subsidy.
• Concept: Allowance for Corporate Equity — the instrument this study evaluates.

See Also

• Rent-targeting corporate taxes reduce debt bias — the outcome this supports
• Allowance for Corporate Equity — the design
• Branzoli & Caiumi (2020): Italy's incremental ACE — the replication at lower revenue cost
• Hebous & Ruf (2017) — the contested real-investment margin
• Cash-Flow Tax · Economic Rent

Sources

1. Frédéric Panier, Francisco Pérez-González & Pablo Villanueva (2015), "Capital Structure and Taxes: What Happens When You (Also) Subsidize Equity?", working paper, December 2015 (presented at the BIS Research Network conference, 2016) — used for the "capital structure significantly responds to changing tax incentives" headline, the NID description (explicit equity deduction; notional charge = book equity × long-term government-bond rate; deduction "regardless of their source of financing"), the neighboring-country control-group robustness, the four main findings (equity share rose; incumbents and new firms; largest among large and new firms; driven by higher equity not lower liabilities), the "more capitalized firms" conclusion, and the Myers (1984)/Graham (2008) statements of the prior evidence gap (all quotes ≤50 words, verified against the BIS-hosted full-text PDF this session). Free PDF (BIS) · Presentation slides (BIS)