Climate reporting, financial stability, artificial intelligence and research impact all meet at the intersection of business and public policy. Each area raises questions about how organisations assess risk, make decisions and prepare for change. Leaders need to connect regulation, finance, and technology and research could help them make these connections, according to Christian Leuz, Charles F. Pohl Distinguished Service Professor of Accounting and Finance at the University of Chicago’s Booth School of Business.

In a recent interview with Richard Holden, UNSW Sydney Vice-Chancellor’s Professor and Chief Societal Economist, Prof. Leuz spoke about climate disclosure rules in Australia and the United States, the lessons of 2008 and Silicon Valley Bank, the effect of AI on business education, and how researchers could make their work useful to practitioners.

Prof. Holden: What do you make of the way the United States and Australia approach corporate climate disclosure?

Prof. Leuz: The current administration in the United States opposes rules on climate disclosures and on emissions. The previous administration proposed a climate rule that was within the confines of the US reporting system. That system, like the Australian one, is based on materiality: a company discloses to the market information that is potentially material to its shareholders. The system does not separate financial from non-financial information; any material information is potentially relevant. The climate rule was trying to clarify what that means for climate reporting. It was not expanding reporting beyond information deemed financially material to investors.

The Securities and Exchange Commission (SEC) issued the rule during the Biden administration, and courts immediately challenged it. The Biden administration withdrew it, and with the change of administration, the SEC will likely rescind it. I find the pushback odd because the rule is consistent with the system. I can understand pushback against the European Union’s approach, which uses a double materiality system. Companies there have to report information relevant to shareholders, as well as their impacts on society and a range of other stakeholders. That is a substantial extension of the reporting system.

Learn more: Can accounting fix global climate disclosure challenges?

Australia is interesting because, with few exceptions, it sticks to materiality for investors. Climate is a new space, and what counts as financially material in climate and energy is not obvious. Australia uses the rules set by the International Sustainability Standards Board. One standard, S1, sets general principles and another covers climate. Australia adopted these only for climate, not for other areas of sustainability, and is introducing them in stages, with large companies going first and assurance rules coming later. That’s smart, as it allows us to walk down the learning and cost curves. The rest of the world can look at Australia as an example of how to introduce such a system. In that sense, there is more to learn from Australia than from the United States, and Australia is currently leading the way in this space.

Prof. Holden: How does materiality work in practice? If a CEO discovers a health condition, for example, would a company have to disclose it?

Prof. Leuz: It could trigger an 8-K reporting obligation, which is an ad-hoc disclosure outside the regular reporting intervals. The test is whether the information is material to the decisions of hypothetical investors or shareholders. It is not straightforward, because interpretation is involved. Standards exist to clarify what the general disclosure requirement means in specific cases.

Prof. Holden: The lack of transparency about risks in mortgage-backed securities was a problem in 2008. Dodd-Frank and other measures followed, and the Trump administration is pushing to deregulate. How worried are you about a corner of financial markets (perhaps private credit) where visibility is limited and the next crisis could start?

Prof. Leuz: I start with 2008, because I do not think we learned the right lessons regarding reporting. The narrative at the time was that fair value accounting, also called mark-to-market accounting, contributed to the crisis, and many people still believe that. The evidence goes in the opposite direction. We held a conference in Chicago to examine what went wrong in the financial system, as investigators do in an autopsy after a plane crash. Two co-authors and I looked at the accounting and regulatory side. We find that banks delayed recognising losses and did not take prompt corrective action. In our research, when regulatory capital is shielded from losses on marketable securities, banks take less action to correct their balance sheets and exposures.

"When regulatory capital is shielded from losses on marketable securities, banks take less action to correct their balance sheets and exposures"

CHRISTIAN LEUZ

I think the same misunderstanding led to the Silicon Valley Bank debacle/failure, which held significant amounts of Treasury securities. Treasury values fell as interest rates rose, which in principle ate into the bank’s economic capital. Because regulatory capital was shielded from these losses and the securities were not marked to market, the bank took little action. If the losses had hit regulatory capital on a mark-to-market basis, I expect boards and regulators would have reacted sooner, that is, to each interest rate rise, probably more gradually and not in a zero-to-100 way. Given the accounting treatment that shielded these assets from write-downs, banks and their boards could look the other way. They even moved more securities into this held-to-maturity category, where banks do not take the losses. To some extent, that made matters worse.

The lesson in both cases is that leverage combined with short-term funding usually brings down banks. That holds for a traditional bank and for a shadow bank. New shadow banks are created to disintermediate the banking system, often in the belief that the rules about leverage and short-term funding do not apply outside banking. In banking, we recognised the problem and created rules and guarantees. Banks receive deposit guarantees because guarantees prevent runs, and the quid pro quo is regulation so that banks do not gamble given the guarantee. Many shadow banks have no such arrangement, and are therefore prone to runs and sometimes even still rely on traditional banks.

Private credit is an example. Lending to private companies has moved away from banks and comes from business development companies. These companies still rely on banks, through revolving credit lines, when they need to pay dividends or meet redemptions. The banking system remains at the centre, which can create risks and even a meltdown, as it can in other areas where disintermediation occurs. We need to apply the principles learned in banking to other areas as the financial system disintermediates.

Subscribe to BusinessThink for the latest research, analysis and insights from UNSW Business School

Prof. Holden: As a Professor of Accounting and Finance, what do you think the future of business education looks like?

Prof. Leuz: I don’t have a crystal ball, but I expect dramatic changes. For example, I recently taught an executive MBA class on financial statement analysis and valuation. The final project is a deep dive into a company and its valuation, including a discounted cash flow model. I gave Claude the final project without further instructions or prompts, and there was little back-and-forth. My teaching assistant graded the Claude answer alongside students’ answers, and it landed around the students’ median.

Interestingly, Claude lost points on the topics I cover in class. Of course, once you tell Claude where to look, it can easily improve the result. If I give Claude my class notes, it will do better. But the key point is: to evaluate the Claude solution, a person still needs to know the concepts. For instance, in valuation, much changes as the forecast moves from the explicit forecast horizon to the terminal value. A person needs the concepts from the class to know that.

I borrow a point from Luis Garicano’s blog: a job is not a task. AI is good at tasks, but a job is a bundle of tasks and it is embedded in organisations. If the Claude solution is good, you still have to go to your boss or attend a meeting and explain why it is good. If your boss asks whether the solution can be trusted, someone needs to evaluate it and vouch for it. That requires concepts, so what we do in business school remains important.

"The key point is: to evaluate the Claude solution, a person still needs to know the concepts"

CHRISTIAN LEUZ

Delivery will also change. Claude can produce slides for me or provide the examples I need. It could also create an introductory course, so students skip some very basic stuff and then come better prepared to mine and other classes. I can see that happening, but I believe students still need much of what we teach.

Prof. Holden: Australian universities push academics, including junior ones, to produce research with societal impact, and not only Google Scholar citations or publications in FT50 journals. Some scholars ask what that means for work such as time-series econometrics. What do you tell them?

Prof. Leuz: My answer is mixed. Society needs some people to do basic research. The time-series econometrician might be in that group, as might someone who studies quantum mechanics. That said, Most universities around the world receive public funding. Thus, it is reasonable for society to ask for accountability and for impact as a return on the money it spends and on the working lives it affords researchers. Impact is a way of giving back for the jobs researchers hold.

But it is also important to distinguish between junior and senior academics. Junior academics should probably focus on their research first and chase tenure. At a later point, the balance needs to tilt thinking about how research benefits society. For instance, it could tilt towards translating research.

Evidence-based medicine offers an example. This idea is considered one of the top 10 achievements in medicine alongside the discovery of antibiotics. Its place on that list comes from its effect on practice. This began as a small idea from researchers at McMaster University in Canada and became a concerted effort across the field. The effort covers methods, including randomised controlled trials and Cochrane reviews, which aggregate research in a given area to answer a practical question, such as what to do for a patient with certain symptoms. Hundreds of researchers contribute to these reviews. There is a manual that sets out how to aggregate research, how much weight to give evidence depending on how it is produced, and each review comes with an interpretation for a layperson.

Learn more: How AI is changing work and boosting economic productivity

In economics, finance and accounting, we do nothing comparable, although these fields are applied in some respects. Individually, we explain research to regulators, and we probably have some impact through those conversations, but that is one-off. What is missing is the infrastructure in between, which would aggregate the body of research and communicate it to practitioners or society. We can think of a board member as akin to a physician with a patient presenting symptoms and a place to find the answer. So translation is very important and would highlight the value of research.

I would say evidence-based medicine is harder because of the political economy. Preferences and political views matter more in economics and public policy. Medicine, in principle, is less political, although COVID clearly showed it is not free of politics. We cannot copy the medical model one-for-one, but we can take several lessons from it. But to be clear, I would not place the onus on junior academics; rather, I would place it on the leaders in the field and ask them to act collectively.