The Mansion at Glen Cove on Long Island’s Gold Coast, with its landscaped gardens, a wellness spa and a Georgian-style home that once served as Herbert Hoover’s summer White House, was an unlikely setting for a conference about solving world poverty. But here, in the spring of 2018, representatives from the Rockefeller Foundation, Morgan Stanley and Apollo Global Management gathered to talk about climate change and inequality.

During one session of the Global Impact Investing Network annual meeting, former JPMorgan Chase & Co. investment banker Desiree Fixler proposed using Wall Street financial engineering to benefit some of the world’s poorest. A structured-credit specialist, Fixler described a plan to attract institutional investments by pooling microfinance loans into a tradable financial product. It would provide the poor with resources to help themselves, she said, while delivering investors a reasonable return. Fixler’s idea was so well-received she was invited to discuss it with a top official of the US Overseas Private Investment Corp., a government agency known as OPIC whose mission was doling out taxpayer funds to encourage development.

Fixler’s idea hit the market in July 2019 — a $175 million collateralized loan obligation, or CLO, that channeled money to 26 microfinance and small-business lenders in 18 countries, including Botswana, Indonesia and Peru. OPIC put in $131 million. But Fixler’s rosy predictions worked out better for the bankers, lawyers, consultants and middlemen who make up the microfinance industry’s infrastructure than it did for many of the borrowers. At least three companies that received funding have been accused of abusive sales and debt-collection practices. Others charged annualized interest rates as high as 120%.

Although the financial firms and development banks Fixler worked with told investors they would weed out all but the most socially responsible companies, the vetting was largely outsourced to the Smart Campaign, an industry-funded initiative that rarely censured lenders publicly for engaging in predatory collection practices or charging excessive interest rates. One firm that lost its Smart accreditation received funding from the CLO anyway.

While predictions that microlending would alleviate poverty were downgraded years ago to the more modest goal of financial inclusion for the unbanked, even that has proved largely unattainable. Instead, the drive for profit and the billions of dollars in investments have transformed the industry into an increasingly commercialized one, where borrowers’ poverty is compounded by debtors’ prisons, coerced land sales and sometimes even suicide, according to a Bloomberg News investigation.

Most microlenders don’t engage in such practices, but even those that charge reasonable interest rates and strive to help the poor have produced little evidence of a long-term positive impact, according to recent studies. Alternative methods such as credit unions and village savings associations that provide a more effective path out of poverty but aren’t as profitable have received less attention from investors. Despite microfinance’s flaws, development banks and foundations continue to pump record amounts into the industry — more than $50 billion of committed funds in 2020, industry data show.

Fixler’s CLO was just a rivulet in that flood. She had helped create one of the first microfinance CLOs in 2006, a $100 million deal arranged by Morgan Stanley and investment manager BlueOrchard Finance Ltd. It was the same year that Bangladeshi economist Muhammad Yunus won the Nobel Peace Prize for his work lending small sums to poor women to boost development. But optimism about building a trillion-dollar industry soured during the financial crisis. Fixler set out to jumpstart the movement in 2018 with her new investment vehicle.

“We wanted to take this niche market, generally funded by the public sector, and make it mainstream,” says Fixler, who got into impact investing after reading a book by Yunus. “Where else are you going to get the scale and the power? You must tap the capital markets. You have to bring in institutional money.”

A CLO is a pool of loans sold in tranches with varying interest rates depending on how much risk investors are willing to take. It’s similar to a collateralized debt obligation — the instrument that helped spark the financial meltdown in 2008 — but it packages corporate loans instead of subprime mortgages. In Fixler’s deal, the target returns would flow from the payments made by shop owners in Cambodia and farmers in Peru.

The firm Fixler was then consulting for, Swiss impact investor responsAbility, hired JPMorgan to help structure the deal. ResponsAbility would serve as the middleman. It chose the lenders, distributed funds, collected money and disbursed interest payments to investors, including OPIC. RepsonsAbility was promised as much as $4.9 million in upfront and servicing fees, while OPIC would be paid $393,000 in return for its investment, according to the offering document. JPMorgan’s fees weren’t disclosed, and the bank declined to comment about how much money it made. But a spokesperson for the bank said microfinance organizations “broadly speaking do a good job of lifting up communities and creating opportunities.” If there are occasions where borrowers are treated unfairly, the spokesperson said, “the correct regulatory authorities should address them and we would certainly be supportive of that.”

Because microfinance companies are scattered around the globe and borrowers are sometimes in remote villages, conducting due diligence is complex and expensive. Consumer protection laws in many of the more than 90 countries where microfinance is prevalent are patchy or nonexistent. For all its pledges to empower the vulnerable at the base of the economic pyramid, there are few clear standards regarding two important issues: What interest rates are predatory and what profit margins are exploitative?

Investors leave the bulk of the screening to investment firms like responsAbility. The CLO illustrates how in many cases those middlemen rely more on each other’s reputations than their own hands-on research. In this case, responsAbility relied mostly on the Smart Campaign to vet recipients. Although Smart’s website listed companies it no longer certified, it said some may have allowed their certification to lapse and didn’t specify why or when. One person with direct knowledge of the matter said they could remember only one firm in Southeast Asia being decertified for cause by Smart in its 11-year history, before it was disbanded in 2020: Cambodia’s Prasac Microfinance Institution Plc.

About $20 million raised through the CLO went to microlenders in Cambodia, including Prasac after it was decertified, more than any country except India, according to a prospectus for the offering. Cambodia is one of the most indebted places in the world. The average Prasac loan soared more than fivefold in the 10 years ending in 2018 to almost $5,000, three times the country’s average household income, industry data show. About one in five Cambodian adults now has a microfinance loan. For years academics and human rights groups have been warning about the crushing debt burden that has led to families to ration food, take their kids out of school so they can work and sell their land.

Prasac, one of Cambodia’s biggest and most profitable microfinance institutions, received $8.5 million, even though the Cambodian League for the Promotion and Defense of Human Rights, or Licadho, has documented scores of cases of borrowers, including Prasac customers, being coerced to sell land to repay their debts. Bloomberg News also interviewed several Prasac customers who described high-pressure tactics used by the firm’s loan officers.

Suy Sokna, a 32-year-old storekeeper in central Cambodia, is one of them. She says she borrowed $8,000 from Prasac to buy a motorbike and build a shop where she could sell groceries, vegetables and pork. That was about 10 years ago, she says, long before the CLO, and over the next decade she kept topping up the loan until, by 2020, it had ballooned to $27,000. Sokna, who can barely read or write, signed her contracts with a thumbprint.

For years she made the $600-$700 monthly payments, but she says she fell behind when her access to fresh pork dried up during the pandemic. Groups of loan officers on motorbikes began showing up, pressuring her to sell the land. She did last July, using the $30,000 to repay the loan and other debts. She says the property, a present from her parents on her wedding day in 2008, was worth $50,000, but she had to sell quickly as Prasac had threatened to take her to court. That meant closing the store and losing her home. She now lives with her husband and four children in a one-room house owned by her sister-in-law with no toilet or running water.

“They told me that they would get my land sold by themselves if I failed to pay,” says Sokna, as she fried bananas at a nearby wooden shack, wearing pink pants and a bright red fake Gucci shirt. “I am very frightened of facing court. I couldn’t sleep because of that.”

Prasac didn’t respond to emails and phone calls seeking comment.

Another microlender funded by the CLO, Edpyme Alternativa in Peru, charged annualized interest rates as high as 120% at the time, according to its website. One-third of its customers earn less than $5.50 a day. Alternativa 19 del Sur in Chiapas, Mexico, received $5 million, even though its website advertises annual interest rates of 98% for some small loans — about 30% higher than the average for microfinance lenders in Mexico.

Edpyme stopped charging interest rates of 120% in June 2021, and those small loans with daily payments accounted for less than 0.1% of its portfolio, General Manager Fernando Bautista wrote in a text message. The lender now charges an average interest rate of 45%, he added. Executives at Alternativa 19 del Sur didn’t respond to emails and phone messages.

Impact investors say they strive to achieve what they call a double bottom line of profitability and social good. Gauging the first is relatively straightforward. Lenders seeking funds from the CLO made their financial data available to prove they were solvent. OPIC, which had almost 20 years experience in microfinance and a rigorous screening process, reviewed the books to assess credit risk. Its decision to invest in the CLO was viewed by other investors as a seal of approval.

Evaluating social impact presents a greater challenge. Before investing, OPIC would send a questionnaire asking if a microlender is engaged in prohibited behavior, including child labor, environmental violations or human trafficking, and if it has connections to organizations on sanctions and terror watch lists. But its screening for consumer protection is less rigorous. A spokeswoman for the US International Development Finance Corp., known as DFC, which took over OPIC’s operations in 2019, said the agency relied largely on the Smart Campaign for that. She said the agency’s analysts visited only two of the lenders that received funds from the CLO — one in Kosovo, the other in Ecuador. None of the lenders was rejected.

Another investor in the CLO was Calvert Impact Capital, whose vetting process has been hailed as a model for socially minded firms. Calvert helped impact investing go mainstream in 1995 by launching a fund that allows investors to buy stakes for as little as $20. Today, it manages almost  $500 million in assets, about 20% of which finances microlenders around the world.

Calvert requires analysts to search financial records and public documents and conduct interviews in affected communities. But Caitlin Rosser, Calvert’s director of impact management, has said those interviews don’t usually include the people the investments are intended to reach. “We are so many layers removed from the kind of end communities that are ultimately benefiting from the capital,” she said at a June 2020 seminar. Instead, she said, Calvert relies on fund managers to assess the social impact of the financial products they are pitching.

Jennifer Pryce, Calvert’s president and chief executive officer, declined to comment about the screening process or whether the company was aware before investing of the high interest rates and aggressive collection tactics employed by some lenders that got funding from the CLO.

Like Calvert, other investors in the CLO reviewed lenders mostly from their home offices, what some call “desk underwriting.” That left much of the vetting to responsAbility, a Zurich-based firm founded in 2003 that manages about $3.6 billion and has more than 200 employees who can make on-the-ground assessments.

ResponsAbility says on its website that its multi-layered screening process starts by making sure a firm targets only the neediest low-income clients. The next step is an analysis of a company’s financials. That’s followed by an onsite review that looks at everything from risk management to client protection and involves meeting branch employees and borrowers. ResponsAbility’s co-head of financial inclusion debt, Thomas Mueller, says only about 500 of the approximately 10,000 microfinance firms meet his company’s standards.

But responsAbility selected three Cambodian lenders — Prasac, LOLC Cambodia and Kredit Microfinance — that were subsequently linked to coerced land sales by Licadho, the Cambodian human rights organization. “Whatever due diligence these firms did, it clearly failed to capture the human rights violations that these investors are now complicit in,” says Naly Pilorge, Licadho’s outreach director. “The human rights abuses we are raising, such as coerced land sales, do not appear on financial sheets — they are the product of a sector that operates in a highly corrupt environment without any consumer protection.”

Martin Heimes, responsAbility’s co-head of financial inclusion debt, says his firm has a good relationship with Prasac and LOLC Cambodia and considers them conscientious lenders that prioritize consumer protection. He says both were certified by Smart, which developed and published a list of six client protection principles, including avoiding over-indebtedness and having appropriate collection practices. “If they are Smart certified,” Heimes says, “then this is a shortcut for us, then we don’t need to look too deeply into this topic.”

It apparently came as news to Heimes that Smart had decertified Prasac before the CLO was marketed. He initially said the microlender was certified at the time and only lost its accreditation later because Covid travel restrictions made it impossible for consultants to assess its operations. Heimes, who helped construct the CLO, later said Prasac had been stripped of its accreditation in 2018 mainly for allowing too many customers to refinance loans early by borrowing ever-larger amounts.

The Cambodia Microfinance Association, an industry regulatory body, stipulated that such high-risk refinancings should be less than 5% of a lender’s portfolio. Prasac had breached that ceiling, according to Heimes, though it wasn’t clear by how much. ResponsAbility says on its website that the 5% limit is an important “countermeasure against market overheating.” But Heimes says the firm assessed the breach as “ultimately acceptable” and considered Prasac still certified. He says responsAbility was assured by Cambodia’s central bank and the Cambodian Microfinance Association that the largest microlenders were not involved in pressuring delinquent borrowers to sell their land. “They are also monitoring the behavior of their staff looking for any bad apples that might not be behaving well,” he says, adding that LOLC Cambodia and Prasac also denied being involved.

LOLC Cambodia said in an emailed statement that its agents do not pressure delinquent borrowers to sell land or homes to repay debts and that such behavior is prohibited by the firm’s code of ethics. The lender said it has not received any complaints from customers alleging such practices.

Fixler says she played no role in choosing which firms were included in the CLO. Her intention was to help improve poor people’s lives by opening up the microfinance industry to institutional investors to help bridge a funding gap. Fixler became an impact investing celebrity in 2021 after blowing the whistle on greenwashing at DWS Group, Deutsche Bank AG’s asset management unit, where she was head of sustainability until she was fired that April. She alleged in an interview with the Wall Street Journal in August that DWS had exaggerated how many of the firm’s assets had been screened for ESG criteria, claims now being investigated by authorities in the US and Germany.

Fixler’s back in New York working as a consultant for companies looking to burnish their ESG credentials. Although she isn’t working on any impact investing deals, she says she’s open to doing so again.

While Sokna and other Cambodians lost their livelihoods during the pandemic, Prasac reported a $155.5 million profit in 2021, a 43% rise over the previous year. Six of the eight largest Cambodian microlenders, including Prasac and LOLC Cambodia, also reported record profit that year. Prasac and LOLC have posted returns on equity of more than 27% in most of the past seven years, almost double the rate many in the industry view as the border between a fair profit and exploitation.

“If someone is doing good and making a good return out of it, I don’t think that’s a bad thing,” says responsAbility’s Mueller. “You’ve hit the sweet spot maybe. We’re not judging them on their profits.”

Mueller says microfinance has had a positive impact in Cambodia and that responsAbility has deals with both lenders in the pipeline. But, he says, his fund is “acutely aware” of the risks of over-indebtedness and unfair client treatment.

Last year responsAbility sold a $177.5 million “social bond” marketed as contributing to United Nations sustainable development goals in a post-Covid world. The Swedish International Development Agency and Danske Bank A/S were investors. Prasac and LOLC Cambodia received funds from that deal too. ResponsAbility has started work on a follow-up deal that Heimes and Mueller say they hope will appeal to a broader range of investors, including hedge funds and insurance companies.

As for Sokna, she has only recently returned to work after giving birth at the end of January. She says she’s concerned about her family’s future. The man who bought her house and land has put it on the market for $50,000 — $20,000 more than she sold it for — but she has no hope of raising that much money to buy it back. She used all her savings to repay Prasac and now has only $75 to her name. “I need to spend more for my children’s study,” she says. “I am very worried.”

— With Michael O’Boyle, Sinduja Rangarajan, and Christopher Cannon.

This article was provided by Bloomberg News.