Four kinds of responsible investing, and what DWS tells a buyer

In September 2023 the US securities regulator (the SEC) announced that DWS Investment Management Americas, a Deutsche Bank subsidiary, had agreed to pay $25m across two cases, $19m of it for misleading statements about ESG (US Securities and Exchange Commission, 2023). The buyer's question is: what does a "responsible" label say the fund actually does, and how can I check? Four strategies give a way to frame it.

What are the four kinds of responsible investing?

According to the library's table, drawn from Louche and Lydenberg (2010), there are four strategies (Louche and Lydenberg, 2010). Exclusion avoids firms or sectors seen as harmful; the usual limitation is narrow impact. Inclusion actively puts capital into socially beneficial activity; the limitation is concentration risk if the portfolio is narrow. Relative selection picks ESG leaders in each sector; the limitation is that it still invests in controversial sectors and there is no precise relative ESG yardstick. Shareholder engagement means dialogue, voting and filing shareholder resolutions; the library lists no limitation for this one. The limitations are the library's notes, not quotations from the original chapter.

The library's framework page also records a point worth remembering: "ESG integration" and "thematic investing" are sometimes mentioned alongside these, but they are not among the four. That is the library's note, not a quotation from the original chapter; the chapter sits behind a paywall and I could not read it, so I do not know what the authors wrote about those two ideas.

One point about the figure's layout: the four columns are ways of acting, not grades of virtue. Nothing in the library's table ranks one above another, and in my reading a fund can combine them, for instance by excluding a sector and also voting at the shareholder meetings of companies it holds. I mention it because a label such as "responsible" often hides which of the four is meant, and the table is a way to ask.

How far can each kind be checked?

This test is my proposal. It asks: if a fund claims this kind, what document lets a buyer check? The "can check" and "hard to check" columns are my reasoning, not conclusions of Louche, Lydenberg or the SEC.

The last row is in brackets because it is not one of the framework's four. The first two rows allow fairly direct checking; the last is where the DWS case sits. "Hard to check" does not mean the claim is false. It means the buyer relies more on trust and less on documents, and that is a different position from holding a list in one hand and a portfolio in the other.

What did the SEC say about DWS?

The SEC release of 25 September 2023 says DWS agreed to pay $25m in total, without admitting or denying the findings: $6m for anti-money-laundering failures and $19m for materially misleading statements about its controls for incorporating ESG factors into research and investment recommendations for "ESG integrated" products (US Securities and Exchange Commission, 2023). The SEC found that from August 2018 to late 2021 DWS failed to implement parts of its own global ESG integration policy and failed to adopt policies ensuring its public statements were accurate (US Securities and Exchange Commission, 2023). For the ESG part, the violations cited are sections 206(2) and 206(4) of the Investment Advisers Act (US Securities and Exchange Commission, 2023).

An SEC official said DWS advertised that ESG was in its "DNA" (US Securities and Exchange Commission, 2023). The release does not say what DWS earned from ESG products, and I did not read the two attached orders. The post When is profit justified? uses the same case to ask about profit and information gaps; here I only ask which kind of claim it was.

Why is ESG integration harder to check?

In my reading, it is because the promise concerns how work is done inside the firm, not portfolio outcomes. A fund says it built ESG into research and recommendations: the buyer cannot see the research, nor whether each recommendation weighed ESG. If a fund does not follow its own policy, the trace sits in process records rather than in the portfolio, so a holdings table would not show it; the SEC concluded that DWS did not do as much as it said (US Securities and Exchange Commission, 2023).

This is my reading: the SEC describes DWS's products as "ESG integrated", not as one of the four kinds, so the case tells us about process claims more than about exclusion or selection. I have no basis for saying which of the four, if any, DWS used, or how other DWS funds operate. The information gap is wide: the seller knows the process and the buyer reads a brochure; see also greenwashing.

What should a fund buyer ask?

The five questions below are my own assembly from the table; no source proposes this list. A hypothetical example: a small company with VND 5bn of spare cash wants to put it in "Fund A", an open-ended fund (invented name) carrying an "ESG" label.

1. Which kind does Fund A say it is: exclusion, inclusion, relative selection or engagement, or does it only say "ESG integration"?
2. If exclusion: where is the exclusion list, and are holdings published well enough to check against it?
3. If relative selection: who scores, by what method, and can I see it?
4. If engagement: are voting records and resolutions filed published?
5. If only ESG integration: what written policy does the fund have, and does any independent party confirm it is followed?

This approach suits you when you are about to move a large sum into a fund and can ask for documents. It does not suit you if you buy through a retail channel that answers no questions, or if your goal is return rather than ESG. An evasive answer, or no document, is information too, though not proof of anything: a fund may simply not publish what you ask for. Compare answers from two or three funds rather than judging one in isolation, and keep the replies, because a vague answer in writing is easier to revisit than one remembered from a phone call. The post Sustainability marketing is not CSR looks at the seller's side and environmental claims.

Going deeper: what did I read and not read?

Louche and Lydenberg (2010) is chapter 21 of a Wiley book (pages 393 to 417). It sits behind a paywall and I could not read it; the description of the four kinds, their limitations and the note on ESG integration and thematic investing all come from the site's library summary, so I give no page numbers. The library is also inconsistent in one place: the framework's question lists "screening, integration, shareholder engagement or impact investing" while its table and note list exclusion, inclusion, relative selection and engagement. This post follows the table and note.

SEC (2023) is press release 2023-194; I read the whole of it through an Internet Archive capture because the original page blocks automated access, and I did not read the two attached orders. The "can check or not" table and the five questions are mine. Fund A is hypothetical. The post does not assess the quality of ESG investing or of any other fund.