Sustainability assurance on the JSE is rewarded by investors, but its quality is hard to see

Introduction

Listed South African companies publish a lot more than a set of financial statements these days. Most of the large ones put out an integrated report, and a growing share of them pay someone outside the company to give an opinion on the sustainability or non-financial parts of it. Mushwana, Chikutuma and Erin (2024) looked at 100 companies on the Johannesburg Stock Exchange (JSE) over 2018 to 2020 and found more of them assuring their integrated reports as the period went on. That is a real change in what accountants and assurance providers are asked to do, and it fits a wider pattern in South African reporting (Herbert and Graham, 2021). It also raises a plain accounting question, which is whether anyone on the market side actually values it.

The research that tries to answer this belongs to the value relevance tradition, where the test is whether an item of reported information is associated with share prices or firm value. This paper argues that the JSE evidence points to a market reward for having sustainability or integrated-report assurance (Jeriji and Nasfi, 2022; Thompson et al., 2022), but that the same evidence tells us very little about the quality of the assurance being rewarded. Provider type, tenure, scope and the role of the audit committee are all thinly studied, and where they have been studied the results are mixed (Velte, 2025; Ruiz-Barbadillo and Martinez-Ferrero, 2022). The paper first sets out the value relevance findings, then asks why assurance might matter, then looks at what the evidence cannot see, and finally considers governance inside the company and what regulators and audit committees should do about it.

What the JSE value relevance studies find

The clearest finding is a positive one. Thompson et al. (2022) used a fixed-effect panel model on JSE-listed firms and reported that sustainability reporting is positively related to firm value, and that assurance adds explanatory power on top of that. Jeriji and Nasfi (2022) reached a similar answer from a different angle. They measured firm value with Tobin’s Q for 88 JSE companies and 83 French companies on the SBF120 index between 2007 and 2018, and found a positive association between assurance and value. Their reading was that the rules behind assurance (Grenelle II in France, King III in South Africa) changed how investors judged firm performance. Two studies, two designs, one direction, and that does count for something.

The related integrated reporting literature on the JSE leans the same way, though it measures other things. Ngcobo and Sibanda (2021) studied JSE mining firms and found that higher integrated reporting scores went with a lower cost of equity and smaller analyst forecast errors. They note that integrated reporting has been expected of JSE-listed companies since 2010. Mushwana, Chikutuma and Erin (2024) added a link between assurance of the integrated report and investment decision-making, using their 100-company sample. Put together, these studies suggest that the South African market has been rewarding more complete reporting and, separately, the assurance placed around it. Still, I would not read the results as a clean price for assurance. The samples are small, the periods differ, and in at least two cases the outcome is something other than share price, so the studies are not all asking one question.

There is also the matter of what ‘mandatory’ means here. Jeriji and Nasfi (2022) treat assurance in both countries as mandatory because of the regulations they name, and that framing drives their interpretation. King III is a governance code and not a statute, and Ngcobo and Sibanda (2021) describe integrated reporting as a JSE expectation from 2010, which is a rather different thing from a law with a penalty attached. Because their sample pools the French and South African companies, the abstract does not tell us whether the South African firms alone drive the result, and a reader should be careful about that. I think it is fair to say the finding is encouraging for the JSE and not quite as firm as the headline suggests.

Before going further it helps to be clear about how much weight a value relevance result can carry. These studies show association. They do not show that buying assurance causes a higher valuation, because the companies that buy it are probably the ones with better reporting habits, bigger sustainability teams and more attention from analysts to begin with. Thompson et al. (2022) used fixed effects, which deals with some of that by looking at changes within the same firm over time, but it cannot remove everything. Jeriji and Nasfi (2022) have the same issue with Tobin’s Q, which reflects growth prospects and intangible assets and not only the way investors react to a reporting choice. So the safest wording is that assured firms tend to be valued more highly, and that the reason is still contested.

Sample coverage is another limit. The samples are small and tilted towards big names, with 100 companies in Mushwana, Chikutuma and Erin (2024), the top 100 in Ruziwa, Mvunabandi and Nomlala (2025), and a mining-only group in Ngcobo and Sibanda (2021). Large firms attract far more analyst and investor attention, which suggests that a smaller listed company might see a very different payoff, or none at all. That is only a guess until someone tests it. Nobody has done so on the JSE as far as I can tell, so claims that every listed firm should expect a premium go beyond what these samples can support, and the same caution applies to the cost of capital results.

Why assurance might earn a reward

The usual explanation is credibility. Management writes the sustainability numbers, investors cannot check them, and an outside opinion narrows that gap. Jeriji and Nasfi (2022) offer two reasons for the reward they found: assurance strengthens a firm’s legitimacy and reputation, and it reduces information asymmetry between managers and shareholders. Those are different mechanisms and the data do not separate them. A firm that wants to look legitimate and a firm that wants to reduce what its managers hide could both buy assurance, and both would show up with higher value in a regression. Herbert and Graham (2021), reading South African integrated reports from 2011 and 2015 through legitimacy theory, found that companies applied the reporting Framework’s principles quite consistently whatever their size, which fits the idea that much reporting behaviour is about being seen to conform.

The better evidence on credibility comes from studies that measure the quality of the report itself. Donkor, Djajadikerta and Mat Roni (2021) used content analysis of JSE integrated reports to score the quality of combined assurance, and then paired that with Bloomberg data on sustainability and financial reporting quality. Their results show strong associations between combined assurance quality and both integrated and sustainability reporting quality. The picture for financial reporting was less tidy, because combined assurance related differently to their two earnings management measures. That mixed result matters. It may be that assurance helps most where the information was weakest and least checked to begin with, namely the non-financial material, and that it does not simply lift everything. Erin (2025) adds that, across 301 South African companies in nine sectors, integrated reporting practices were affected by assurance and corporate governance together, which suggests the two work as a pair and not as alternatives.

What the evidence cannot see

The first gap is who provides the assurance. Thompson et al. (2022) report that investors do not appear to separate assurance from Big 4 firms from assurance by specialist consultancies. A cheerful reading is that the market simply wants an opinion from somebody. A less cheerful one is that investors cannot tell a thorough engagement from a thin one, so provider brand tells them nothing. Velte (2025) reviewed 52 quantitative studies on the choice of sustainability assurance provider and found that results were either too varied or too scarce to show clear patterns in the drivers or in the effect on firm value. So the JSE result is not an oddity. It sits inside an international literature that has not settled whether the provider matters at all.

The second gap is independence over time. Ruiz-Barbadillo and Martinez-Ferrero (2022), using international listed firms from 2007 to 2016, found that assurance statements with more informative content were linked to lower information asymmetry, but that this benefit was held back when the relationship between company and assurance provider had run on for too long. Nothing like that has been tested properly for the JSE, as far as I could find, although Thompson et al. (2022) did look at provider type. The assurance statements that JSE companies publish usually name the provider, so tenure looks like something local researchers could reconstruct from public documents. Until someone does, it is an open question whether the long-tenure problem seen in the international sample also shows up here.

The third gap is the way studies code assurance. Most of the JSE work treats assurance as a yes or no variable, or as a score from the report. That hides what was actually assured, to what level, and against which standard. Donkor, Djajadikerta and Mat Roni (2021) moved beyond a simple dummy by scoring the quality of combined assurance, and that is probably why they could say something about credibility. Mushwana, Chikutuma and Erin (2024) were more limited, and they say so themselves, noting that a single-country sample constrains what can be concluded. If the market reward turns out to be a reward for the label, then companies have a reason to buy the cheapest label available, which would not be good for the people relying on the reports.

Audit committees and the governance chain

Inside the company, the audit committee is the body that is meant to hold this together. Ruziwa, Mvunabandi and Nomlala (2025) examined the top 100 JSE-listed corporations over 2017 to 2021 and found that audit committee independence was positively related to the quality of environmental, social and governance reporting, while meeting frequency and the average age of members showed no relationship. They used the Ernst and Young Excellence in Integrated Reporting Awards as the quality measure, which is an imperfect proxy, and the authors frame their results as evidence for the value of internal assurance. Still, the independence result is the interesting one. It suggests that who sits on the committee counts for more than how often they meet, which is a point regulators and boards sometimes forget when they ask for more meetings, and it matches the governance emphasis in Erin (2025).

Erin (2025) supports the governance link from another direction. Using board and audit committee variables together with an assurance variable for 301 South African companies, the study reports that integrated reporting practices respond to the combination of governance and external assurance. Read next to Donkor, Djajadikerta and Mat Roni (2021), the picture is of layers. Management produces the information, internal functions and the audit committee review it, and external providers give an opinion on top. Combined assurance is basically the idea that these layers should be coordinated so that gaps and overlaps are visible. The difficulty for the researcher is that the market sees only the outer layer, the assurance statement, and cannot observe whether the audit committee actually challenged anything beforehand.

That is why the audit committee matters to the argument of this paper. If investors reward assurance without being able to judge its quality, then the committee is the only party placed to judge it on their behalf. It chooses or recommends the provider, sees the scope, and can decide whether a long-serving provider is still independent in any real sense. Ruiz-Barbadillo and Martinez-Ferrero (2022) show why that last decision is not trivial, and Velte (2025) shows that the choice between provider types has no clear evidence behind it. In practice, then, committees are making a high-stakes choice with thin guidance. That is a governance weakness in its own right, whatever the market premium turns out to be.

What follows for regulators, committees and researchers

For regulators and the JSE, the evidence supports asking for more information about assurance and not necessarily more assurance. A listed company could be expected to say who provided the assurance, for how many years, on which parts of the report, at what level of comfort and under which standard. None of that looks costly to disclose, though I have not seen it costed. It would give investors something to price other than the presence of a statement, and it would give researchers variables they currently have to construct by hand (Donkor, Djajadikerta and Mat Roni, 2021; Velte, 2025). It would also help to settle whether South Africa’s experience, which Jeriji and Nasfi (2022) and Thompson et al. (2022) treat as a reward story, holds up once quality is observable.

For audit committees, the practical step is to treat sustainability assurance as something to be reviewed and not merely received. The committee should record why it chose a particular provider, how it judged independence over time, and what the assurance covered and left out. Ruziwa, Mvunabandi and Nomlala (2025) suggest that independence on the committee is where the reporting benefit appears, so the composition of the committee should be taken as seriously as the choice of provider. For researchers, there is plenty to do. Studies could separate the South African firms in cross-country samples, test tenure on the JSE, and move from yes or no coding to scope and level. The disclosure behaviour that Herbert and Graham (2021) describe suggests the raw material is there in the reports, and Ngcobo and Sibanda (2021) show the JSE is a workable setting for testing market effects.

Conclusion

The JSE evidence is reasonably consistent that investors respond well to assured sustainability and integrated reporting. Thompson et al. (2022), Jeriji and Nasfi (2022) and Mushwana, Chikutuma and Erin (2024) each point that way, with different samples and different outcome measures, and Ngcobo and Sibanda (2021) find a related benefit for integrated reporting in the mining sector. That is worth knowing, and it suggests companies are not wasting money.

What the same evidence does not do is show that the reward goes to good assurance. Provider type appears not to register with the market, tenure has been tested internationally but, as far as I could find, not yet locally, and most studies code assurance too crudely to judge scope or level (Velte, 2025; Ruiz-Barbadillo and Martinez-Ferrero, 2022). The audit committee, whose independence seems to matter for reporting quality (Ruziwa, Mvunabandi and Nomlala, 2025), is the one party able to look behind the label. My position is that the market reward is real enough to justify the practice, but fragile enough that disclosure about the assurance itself should now be the priority, before the label becomes the whole product.

Reference list

Donkor, A., Djajadikerta, H.G. and Mat Roni, S. (2021) Impacts of combined assurance on integrated, sustainability and financial reporting qualities: evidence from listed companies in South Africa. International Journal of Auditing, 25(2), pp. 475-507. doi:https://doi.org/10.1111/ijau.12229

Erin, O.A. (2025) Corporate governance, external assurance and integrated reporting practices: empirical evidence from South Africa. Meditari Accountancy Research, 33(7), pp. 280-312. doi:https://doi.org/10.1108/medar-02-2024-2341

Herbert, S. and Graham, M. (2021) Applying legitimacy theory to understand sustainability reporting behaviour within South African integrated reports. South African Journal of Accounting Research, 36(2), pp. 147-169. doi:https://doi.org/10.1080/10291954.2021.1918481

Jeriji, M. and Nasfi, A. (2022) The value relevance of mandatory sustainability reporting assurance. South African Journal of Accounting Research, 37(2), pp. 122-138. doi:https://doi.org/10.1080/10291954.2022.2148887

Mushwana, P., Chikutuma, C. and Erin, O. (2024) Does integrated reporting assurance impact investment decision-making? Insight from 100 listed firms in South Africa. Cogent Business and Management, 11(1), a2369213. doi:https://doi.org/10.1080/23311975.2024.2369213

Ngcobo, B.N. and Sibanda, M. (2021) Effects of integrated reporting on the cost of capital and analysts’ forecasts errors: a study of Johannesburg Stock Exchange listed mining firms. Journal of Economic and Financial Sciences, 14(1), a675. doi:https://doi.org/10.4102/jef.v14i1.675

Ruiz-Barbadillo, E. and Martinez-Ferrero, J. (2022) Is tenure important in the sustainability assurance context? An empirical analysis of the informative value of assurance statements. Meditari Accountancy Research, 31(5), pp. 1483-1517. doi:https://doi.org/10.1108/medar-01-2022-1580

Ruziwa, R.M., Mvunabandi, J.D. and Nomlala, B.C. (2025) Audit committee characteristics and environmental, social, and governance reporting quality: an analysis of top 100 JSE-listed corporations. Accounting and Financial Control, 6(1), pp. 38-52. doi:https://doi.org/10.21511/afc.06(1).2025.04

Thompson, E.K., Ashimwe, O., Buertey, S. and Kim, S.-Y. (2022) The value relevance of sustainability reporting: does assurance and the type of assurer matter? Sustainability Accounting, Management and Policy Journal, 13(4), pp. 858-877. doi:https://doi.org/10.1108/sampj-08-2021-0329

Velte, P. (2025) Determinants of the selection of sustainability assurance providers and consequences for firm value: a review of empirical research. Meditari Accountancy Research, 33(7), pp. 443-465. doi:https://doi.org/10.1108/medar-07-2024-2575