The business of multifamily lending: Inside Arbor Realty Trust’s books
Arbor’s multifamily lending boomed in a frenzied market. Can it keep distress out of the picture?
If you need a bridge loan to buy 3,200 apartments in Houston, who you gonna call? Jay Gajavelli, the boss of Applesway Investment Group, reached out to Arbor Realty Trust.
Run by Ivan Kaufman, a Long Islander with spiked hair and a thick New York accent, Arbor has solidified itself as a go-to lender for multifamily shops, just like First Republic Bank was known as a mortgage financier for the rich. No other well known, publicly traded lender is so skewed towards multifamily.
“Arbor was able to provide liquidity throughout the pandemic,” Kaufman said earlier this year. “Furthermore, we had the proper liability structures in place, which was driven by the lessons learned from the last crisis.”
Investors spent record amounts of cash buying up apartment buildings in 2021. As an asset class, multifamily has generally been viewed as recession-proof: people will always need a place to live. But many of these investors used floating-rate debt for their purchases, so as interest rates have soared, buyers have started to have difficulty repaying their loans. And as rents have slowed and more supply has cropped up across the U.S., buyers aren’t collecting the income they projected.
Arbor’s bread is short-term, floating-rate loans — called a “bridge” to more permanent financing — perfect for those who want to get in and out of a deal quickly. Its butter is packaging some of these liabilities into collateralized loan obligations, or CLOs, securities then sold off to investors.
However, as rates have risen over the past year, Arbor has become exposed to borrowers who are struggling to pay off these loans. Take Gajavelli’s, for example: Earlier this year, his firm defaulted on the $229 million loan Arbor gave Applesway. In April, Arbor foreclosed on the four apartment complexes.
At the end of June, it reported that about 0.9 percent of its $13.1 billion bridge loan portfolio, around $124 million, was non-performing, according to filings with the U.S. Securities and Exchange Commission. That’s up 1,500 percent from delinquencies reported in December.
And distress is rising on multiple fronts. Across what Arbor calls its agency portfolio — think loans sold to Freddie Mac and Fannie Mae — it reported $343.9 million worth of loans delinquent in June, up 788 percent from the end of December, financial filings show. (The June number, Arbor says, represents just over 1 percent of its roughly $30 billion agency book.)
Delinquencies and trouble within Arbor’s CLO pools are harder to parse. One source said Arbor keeps the data close to the chest and required a nondisclosure agreement to look at one CLO report, unless the person is an active participant in the CLO market.
“The writing is on the wall,” said Dan McNamara, whose firm Polpo Capital shorts commercial mortgage-backed securities. “There will be a significant amount of delinquencies.”
For the loans on its balance sheet, Arbor bears the full impact when a borrower stops paying. That means it holds more risk than other lenders that only specialize in CLO financing, like debt fund MF1 Capital. With a CLO deal, the lender generally holds up to 30 percent of the securities.
Long Island strong
Looking at Arbor’s balance sheet, it’s natural to assume the firm’s beginnings would be in Texas or Florida, where most of its loans go now. But the firm’s roots are on Long Island, in the suburban village of Westbury, about 30 miles east of New York City.
Kaufman was in his first semester of law school at Hofstra University in 1983 when he and his father, Morris Kaufman, a data processor born in the Bronx, came up with a plan to capitalize on the exodus from the city to the suburbs. The movers needed mortgages, and the Kaufmans decided to provide them.
The two formed American Mortgage Banking — characterized as a “competitive lender” by the New York Times in 1986, when the firm offered a 15-year mortgage with a 9.75 percent interest rate (the average mortgage rate was around 10.18 percent).
Five years later, the elder Kaufman had died, but his son was hooked on mortgages — and wanted to grow quickly. In 1991, when American Mortgage Banking was originating about $550 million in loans a year, Kaufman acquired 13 of General Electric Capital’s mortgage-origination offices and renamed the company Arbor National Mortgage. (To stay true to its etymology, Arbor also started a program of giving a tree to every client that closed a loan.)
In 1993, Kaufman wanted more than just residential. He opened up a new commercial lending subsidiary at the company.
“The writing is on the wall. There will be a significant amount of delinquencies.”
But Arbor began to struggle. The U.S. Federal Reserve hiked interest rates from about 3 percent in 1993 to 6.56 percent by the end of 1994 and the mortgage industry retrenched. BankAmerica swooped in and bought Arbor National Mortgage for $119 million — a deal “worth far less than the market price of Arbor’s stock,” Reuters reported at the time.
Kaufman, without his residential business, pivoted solely to commercial lending and investing. And where better to go than the Big Apple, where commercial real estate prices were starting to soar.
In the early 2000s, Arbor bought stakes in 450 West 33rd Street — an office tower Brookfield has since redeveloped as Five Manhattan West — and 1107 Broadway.
The firm also had its eye on a new product: the collateralized debt obligation, or CDO. Banks would issue loans, pool them and sell interest in the portfolios as securities to investors. Arbor put together its first CDO in 2004, four years before the economy melted down.
“They were very risky,” said one real estate finance attorney. “Then the crisis happened and a lot of investors pulled back.”
Arbor cooled it on CDO issuance, along with the rest of the market. But after a few years, investors seemed to have an appetite for asset-backed securities again.
Enter the collateralized loan obligation. New name, but, attorneys say, more conservative.
While CDOs could have a range of debt types in one pool — junior, riskier debt or second-lien mortgage loans — CLOs were limited to strictly first-lien mortgage loans, seen as the safest for investors.
As Arbor started boosting its CLO expertise, it naturally started boosting its multifamily loan portfolio.
“The product lends itself really well to transitional assets,” said the attorney familiar with CDOs and CLOs. “These are loans on properties that are in some form of rehabilitation,” like apartments that are undergoing renovations.
Lenders hold the power
In 2013, about two-thirds of the assets tied to Arbor’s loan portfolio were multifamily, according to financial filings. By 2019, multifamily made up 80 percent of Arbor’s loans.
Part of that growth was because Arbor supercharged another product: on-book bridge loans.
Arbor offers one-to-three-year bridge loans for up to $100 million, marketing materials show.
Those bridge lines are funded one of two ways: with repurchase (repo) lines, or with the company’s own cash on hand.
“They don’t have a source of cash like a bank does from deposit, or like a life insurance company does from premiums on life insurance policies,” said one attorney familiar with Arbor. “They need to get the capital from somewhere.”
Repo lines are essentially corporate credit lines. In this case, Arbor borrows money from a bank, though it’s structured as a sale and repurchase agreement. The bank agrees to fund the line for a specific purpose — it has to go towards financing a multifamily property, for example — so each repo line has collateral.
But repo lines come with their own risks. Economists, including Yale’s Gary Gorton, have argued that a so-called “run on repo lines” contributed to the 2008 financial crisis. Banks wrote down the value of subprime-housing loans, which were collateral for many repo lines, and issued what’s known as a margin call. That forced financial institutions to cough up more cash to pay down the lines.
“The lenders under a repurchase facility could decide, ‘Okay, there’s been a change in the value of these loans or the value of the property,’” said an attorney familiar with repurchase lines.
In the case of Arbor, its repo line lenders could decide that multifamily properties aren’t worth what they were when they provided the line. In April, Green Street estimated that apartment building values were down 20 percent from their 2021 peak.
“There’s always an issue and concern of margin calls,” Kaufman said on an earnings call in May. “Knock on wood, so far we’ve been in really good shape.”
To ensure it could meet margin calls, Arbor has to boost its cash on hand and keep it there. It’s clearly trying to.
“They are a serial issuer of stock,” one analyst said.
With a repo line, Arbor is at the beck and call of its lenders’ discretion, too. Banks could decide they no longer want to finance multifamily properties, and many have.
“Having a lender say, ‘I know it meets all the bells and whistles, but we just don’t like financing hospitality, or [assets in] Wichita or Kansas City anymore,’” the attorney added. “Or if multifamily is the case, ‘We don’t want to finance anymore, we have too much exposure at the bank across all of our platforms.’”
Delinquency threat
At the end of the second quarter, analysts noted that three new non-performing loans had popped up on Arbor’s balance sheet, totaling about $116 million.
Arbor did not give specifics, but said the loans were made to three different borrowers . Two are tied to Houston buildings, the other in Atlanta, Kaufman said on the firm’s May earnings call.
Kaufman laid the blame on bad borrowers.
“When you see stress in the portfolio like we’re seeing, it’s a fact that the sponsors are not executing along their plan,” Kaufman said. “Elevated interest rates do put stress on these assets.”
Bad operators, Kaufman added, “run into payment issues and often they are late on the payment, trying to raise additional capital and try to get them in a proper position.”
But Arbor dramatically grew its business in 2021, when so many new multifamily players entered the market, trying to capitalize on cheap debt. The firm handed out about $16.1 billion in loans in 2021, up 76 percent from $9.1 billion a year prior, which Arbor called “record volume.”
Arbor has handed out at least one loan to Tides Equities, according to Morningstar data, an aggressive syndicator that grew a $6.5 billion portfolio on floating-rate debt in a low-rate era. Tides is now struggling to meet debt payments and has warned investors that capital calls may be coming.
Arbor has options when a borrower fails to make payments on a loan. But it depends on what type of debt it is.
For a troubled loan in a CLO group, Arbor can cough up the cash to buy out the loan — an amount equal to what’s owed.
If many loans within a single CLO pool fall into trouble, Arbor would have to cough up a lot of cash to buy them out.
Arbor can decide to let the delinquent CLO sit in the group. But that route forces more tough decisions, according to attorneys. A special servicer would be tasked by figuring out whether a sale or foreclosure is the best alternative. If the property sells for less than the original loan, “that’s a loss to the deal,” one attorney added.
The scenario would not be ideal for Arbor, as it holds nearly a third of securities in a CLO, and most often, the riskier below-investment-grade stuff — “skin in the game,” the source said.
“If a significant number of those deals go bad, it could potentially work its way up into the investment-grade securities,” the attorney said.
Gajavelli at Applesway couldn’t even hold out until maturity to default on Arbor’s loan on 3,200 units. The firm defaulted about a year before the debt was due, according to Texas property records.
But Arbor seems convinced someone can execute a business plan on the apartments. After Arbor foreclosed on the properties and New York-based investment firm Fundamental Partners acquired the units, Arbor gave the firm a new loan on the same portfolio.
For Arbor, it’s the problematic borrowers who are to blame.
Correction: An earlier version of this story misstated the percentage of Arbor’s $13.1 billion bridge loan portfolio that was non-performing. It is 0.9 percent.