The following short-sell analsys is from a recent issue of my short seller’s newsletter. To learn more about this follow this link: Short Seller’s Journal information
ABR loans money to speculators and operators who acquire CRE buildings – primarily multi-family and office buildings – either as “value” plays or rehab plays. It also underwrites GSE single-family and FHA multi-family mortgages but this is a tiny percentage of its operations. The loans that ABR both holds directly and that it stuffs into the CRE CLOs it sponsors are becoming distressed at an alarmingly rapid rate.
Given the increasing distress in CRE CLO’s and other risky debt instruments held by ABR and for which the number of non-performing loans (NPLs) are starting to pile up, I believe that ABR will collapse possibly within the next 18-24 months, depending on how quickly ABR’s borrowers can no longer make debt service payments and the degree to which multi-family and office buildings continue to lose value. It is highly likely that the stated book value of most of the real estate that collateralized the mortgages underwritten by ABR over the last several years, particularly when interest rates were near zero, is worth well less than the outstanding loan amounts. ABR has stuffed about 60% of this paper into CRE CLOs and it holds around 40%.
The bulk of ABR’s loan portfolio consists of risky bridge loan financing and CRE CLO loan obligations. Most if not all of the loans could labeled as “the dregs of the industry.” Per a report by Banco Santander, the percentage of loans in Arbor CLO’s that went delinquent in Q4 doubled from Q3 to 16.5%. According to industry data, this is about 2.5 times the delinquency rate for the CRE CLO market at large. Santander’s analyst expects the delinquency rate on ABR’s loans to continue rising this year. Of ABR’s $12.8 billion loan portfolio, about $7.3 billion is represented by loans used in CLO financing structures.
But it gets worse. According to specialized research firm, Viceroy Research (@viceroy-research/viceroyresearch.com), based on data on ABR’s CLO book provided by CreditIQ, through the end of February 20% of the loans structured into ABR’s CLOs are now delinquent. The delinquency rate thus has risen just two months into 2024 from 16.5% to 20%. Also per Viceroy’s analysis of the data, the number of loans with a greater than 30-day delinquency rate has risen from 7 loans in September 2023 to 50 this month (March 2024).
In addition, the principal loan amount of ABR’s NPLs soared from 2022 to 2023. A loan is designated as non-performing when it is 90+ days past due. Per the table below from ABR’s 2023 10-K, the unpaid balance (UPB) of NPL’s jumped from $7.7 million in 2022 to $274.1 million in 2023:
With respect to the latter, ABR has been swapping some of the bad loans from the CLO trusts with the performing loans that the Company holds. But this is merely a short term fix. In going through the 2023 10-K, I spotted several areas of accounting treatment that raise red flags. The number of accounting red flags I noticed in reading through the footnotes of the 10-K gives me cause to believe that the $274.1 million is fraudulently understated.
Furthermore, in the face of the rapidly rising NPLs, management has been using extend and pretend gimmicks, like loan extensions and shell-game refinancings to defer the eventual default and foreclosure of many of these loans. As Viceroy Research notes, per data available from the trustee of ABR’s six CLOs (U.S. Bancorp), the average net operating income to debt service (debt service coverage ratio) is below 1. This means that the combined cash flow of each borrower with loans held by these CLOs is negative (table produced by viceroyresearch.com):
As mentioned above, for now ABR has been able to plug the cash flow holes in its CLOs by drawing down on the reserve in each trust, extending loan maturities and swapping performing loans held by ABR for NPLs in the CLO trusts.
For now, ABR is generating positive GAAP income. The weighted average interest rate on the mortgages it underwrites is 8.42%. But keep in mind that the average yield on the 10yr Treasury between 2021 and present is roughly 3%. The spread between the 10yr and the weighted average interest rate on ABR mortgages is 5.42%. This would imply a triple-C bond rating, which means this paper has a high probability of defaulting. Viceroy Research believes that every mortgage on ABR’s books, including the CLO mortgages, will go bust.
ABR produced $400.5mm in net income in 2023. But in my opinion this is because it is under-reserving for credit losses by a substantial amount. Part of this understatement is attributable to ABR overstating the amount of recovery (the amount of proceeds from selling after foreclosing on defaulted mortgages net of the the loan amount). In my opinion ABR is committing fraud in this regard.
ABR’s stated book value is $3.2 billion. If just 26% of ABR’s loans are wiped out, its book value goes negative. Keep in mind that 20% of ABR’s outstanding loan balance is delinquent and that rate of delinquency is rising rapidly. Furthermore, based on the recent building sale data, the commercial real estate values have started to head south quickly. The influx of multi-family units (detailed a few issues ago) will further exacerbate the rising distress rate of multi-family real estate.
The all-time high in ABR’s stock is $20. It hit $4 during the worst period of the covid ordeal. The state of distress in CRE is similar if not worse to what it was in 2020:
The caveat on shorting the shares outright is that, for now, ABR pays a 13% dividend. But if the loan performance data continues to deteriorate, that dividend won’t last much longer. Also, the short interest is 40%. I think the best way to express the view that ABR will hit the wall within the next 18 months is with October $10 puts. I also think August 2025 $5 puts are interesting. Recall that NYCB plunged from $10 to $3 after it reported bad numbers and cut its dividend. If ABR is forced to cut the dividend, the stock will trade below $5 quickly.