Modern slavery allegation offers first glimpse of the implications of missing social targets

The QIC credit analyst that covers autos is London-based and she sent to us a Reuters article on the evening of Friday 22 July, the day the claims were made. I had a look on the weekend and first thing on Monday asked our investment compliance team to put a restriction on Hyundai bonds so that we couldn’t add to our existing holdings.

QIC has a Monday morning meeting of portfolio managers and credit analysts. I flagged that we would investigate the issue more closely but that given it was likely a modern slavery issue we may need to sell down our holdings. We went through all the available information throughout the course of the day.
QIC has company-wide exclusions, including manufacture of tobacco and controversial weapons, and within the fixed-interest team we also have a dynamic exclusion list. This list is designed to handle transitory risks that might arise while we are holding a bond. At the end of Monday, we added Hyundai bonds to our dynamic exclusion list.

The work we had carried out during the day indicated that these were very serious allegations and, together with Susan Buckley, QIC’s managing director of fixed interest, we agreed to sell our holdings of Hyundai bonds. As a business, QIC will not finance companies that engage in this type of behaviour and the portfolio managers immediately agreed with the recommendation.

At the time we were holding Australian dollar vanilla bonds and US dollar green bonds. We started selling down our positions during the European session on the Monday and were completely out of our positions by the end of the week.

Interestingly, it was harder to sell the US dollar green bonds. We were holding Hyundai’s largest outstanding green-bond line – the US$600 million 1.25 per cent 2026 notes – so it’s noteworthy that it wasn’t particularly liquid.

All the information was publicly available. We found Hyundai’s policies on human rights on its Korean and US websites and found a directive on child labour. This states that “…child labor is prohibited in principle and the company takes measures so that minors’ opportunity for education will not be restricted due to their work”. This is consistent with ILO [International Labour Organization] conventions.

There is a policy on Hyundai’s US website around suppliers and supply chains. At this point, we looked for a management response and found a reference to some shortcomings in a third-party labour-hire firm. This led us to check filings with the US Securities and Exchange Commission. The most recent were from March but they showed that the entity that was subsequently accused of child labour – SMART Alabama – is a manufacturer and consolidated subsidiary of Hyundai Motor Company and is 100 per cent owned by a holding company that is 72.45 per cent Hyundai-owned. This was a pretty clear indication that the issue is within the company’s control – in other words it wasn’t buried deep in the supply chain.

We had to make a quick decision and we were confident in the source of the claims: police allegations, a US government investigation and a global news agency are sufficiently credible. We were not basing our decisions on the claims of an activist report.

Current and former staff have corroborated the allegations, suggesting that it’s not just one child who might have deceptively looked older than they are. The suggestions are there were up to 50 children per shift potentially operating machinery. It’s not a safe work environment and the claims breach Hyundai’s own policies.

As debt investors it is incumbent on us to consider regulatory commitments and clients’ expectations around what their money is financing – and certainly from the number of modern slavery questionnaires we are required to complete this is an issue that is high on investors’ agendas.

We send a monthly report to clients and were able to include this information in the July report. The allegations were published about a week earlier, so it was reasonably timely. We have raised the allegation in meetings with clients and asset consultants, and their reactions were all broadly surprise and shock. Those we have spoken to appreciate the fact that we took quick and decisive action, and that we supported our decision-making framework as best we could.

We are surprised we haven’t heard more about the claim, particularly because we must comply with the Commonwealth government’s Modern Slavery Act which requires a risk-assessment process to be applied to investment portfolios, a company’s own operations, supply chains and so on.

Ordinarily, engagement would be our preferred option but, in this case, our obligations under modern slavery legislation take priority. With a US government investigation ongoing, we judged the likelihood of Hyundai management being able to give us new information as being low and attempts at engagement as futile.

We also took into consideration the management reactions we had seen through public statements from Hyundai and its SMART subsidiary. The attempts at blame-shifting onto temporary-work agencies told us they weren’t as close to the operations as one might expect. Therefore, we concluded it was unlikely that an individual bondholder in Australia would be likely to be able to exert significant influence.

We have engaged with Hyundai twice before, in 2016 and in 2018. We talked about clean energy and carbon-emissions targets during those meetings, and subsequently sat out of bond deals because Hyundai had missed some labour and governance targets that it had set.

However, as we have gone back through our engagement notes, the highly disappointing factor is that we had also asked about labour and governance during those meetings. It’s like we are on a hamster wheel. The same ESG factors come up repeatedly.

On 23 August, a second Hyundai auto-parts plant in Alabama was alleged to be using child labour. We have looked at the consolidated financial statements and cannot find the entity consolidated and so this time it appears to be a supplier. Even so, this constitutes a breach of Hyundai’s supplier code of conduct.

I am surprised that we are not seeing the issue mentioned widely in credit investor updates.
Our first duty is to our clients, both to what we promised we would deliver to them and in helping them to observe regulations. These allegations are probably the first time we have come across slavery claims made directly at an Australian dollar bond issuer.

We have been involved with collaborative investor movements previously: we told another big carmaker on behalf of a client that we couldn't participate in their bond transaction because of cobalt-sourcing concerns. However, we ended up with good engagement with that company, which allowed us better to understand their sourcing policy: we subsequently found out that it was the first time an investor had raised this issue and had sat out of a bond transaction because of it.

At the end of the day, we are making very difficult investment decisions to sit out of these bonds where they are benchmark constituents. We will attempt to engage with Hyundai eventually, because we want to let management know we have exited their bonds until the issue is settled. But, for now, we strongly believe engagement to be futile and not in the best interests of protecting our clients.

We are also looking for any other touchpoints across the rest of the QIC business, because Hyundai is a large conglomerate. We considered whether this could be an issue at other automakers, given we hold bonds in others. Our analysts concluded that most of the other US-based automakers have highly unionised workforces while Hyundai does not, leading us to believe that the issue is likely contained.

Our trading team undertakes constant surveillance of market pricing. It was hard to get any pricing levels on the green bond. We had our marks and we could see bonds on screens, but there wasn’t really any transparency.

It’s not at all what I would have expected. No-one is talking about this from a bondholder’s perspective.

Hyundai has a single framework covering GSS bonds. We owned a US dollar green bond, proceeds from which were to provide loans to consumers to buy specific electric and hybrid vehicles manufactured by Hyundai. The green bond was aligned with the 2018 International Capital Market Association (ICMA) Green Bond Principles (GBP) and had an SPO from DNV GL (DNV), which is on our shortlist of credible SPO providers.

Our process is to look at the issuer as an entity first and the transaction’s structure second, and the structure must meet a series of investment requirements. The label itself is irrelevant. We will not hold on to a green bond because it's a green bond. As this was a claim against the issuer, the structure of the debt didn’t even come into question.

It’s a licence to operate. It’s really as simple as that.

There are very specific assets that continue to warrant the green-bond label. We wouldn't want to discourage an issuer that feels like social responsibility is not within its remit by insisting it has a social framework or targets. In the case of a sustainability-linked bond, targets need to be core to the issuer’s purpose and address the areas where they can have greatest impact. We don’t want to see issuers ignore their key risk areas by opting for softer or easier-to-achieve targets.

No, it doesn’t. We have our own views on who we see as the leading group of SPO providers.

We won’t participate in a transaction if it does not have ICMA or Climate Bonds Initiative alignment, an SPO and also ideally verified ongoing reporting or reporting with assurance. However, the framework comes first because it provides reassurance that a credible group has confirmed that this style of label is acceptable.

Undoubtedly there is a place for the SPO as a central governing process. The market needs a provider of independence from the issuing company. However, I also believe that while it doesn’t need regulation, the SPO sector could benefit from some enhancements. There are a lot of players hanging out their shingle and reputational damage is a risk if they all aren’t deemed to be adding value.

As an investor, legally I can’t rely on an SPO, but it provides comfort some expertise has reviewed the issuer as well. We have access to climate experts and engineers at QIC, but we are finance experts and to some extent we rely on the SPO analyst to provide this technical backstop.

I had wondered if this is the reason why the Australian GSS market has been relatively benign in issuance in the year to date. Are issuers almost too scared to embark on discussions around labelled issuance because they have been led to believe that targets will be retrospectively applied? Market volatility concerns seem to have restricted issuance too.

The credit monitoring within our investment process picked this up. It would have been another week or two before our responsible-investment process did the same.

We are aware that modern slavery legislation in Australia has been a softly-softly approach, and has always been designed to be a multifaceted system where one gets familiar with the idea that modern slavery exists first and that there is a responsibility to be addressing the risk. This is the stage Australian legislation is currently at.

I understand that it was always the intention to revisit the legislation after a few years with a view to putting greater onus on the responsibilities and actions of the investor.

It might be the case that we haven’t seen much action among our peers because carrying out a risk assessment is all the legislation currently requires. At QIC we have taken it upon ourselves to take the next step and we have already carried out what we think legislation will dictate next: removing ourselves from the risk.

I wouldn’t call this a failure of legislation. It is early days and raising awareness of these issues is exactly what the current legislation is designed for.

Data from the Minderoo Foundation show that there are 40 million child slaves globally. Modern slavery is almost impossible to eliminate. In some countries sending a child to work is the only way a family can eat and so one must be realistic – while the ideal outcome is to do as the UN says and ‘stamp it out,’ it’s not always possible. Chances for improvement can come from ensuring access to education and appropriate safe work practices.

Another thing we are thinking about is the role of the index provider and whether it will exclude bonds if the claims are proven. In other jurisdictions, where modern slavery legislation is at the next stage, it may be the case that forcing investors to invest – mandatorily through having to hold the bond in the index – is a breach of the law.

At this stage, if we had carried on holding Hyundai bonds we wouldn’t have broken the law. But if there is future change it will be incumbent on index providers to make decisions about the place of these bonds in their indices.