Our fine little CRE CLO business has exploded over the past couple of years, hasn’t it? Last year, around this time, I recklessly predicted for my friends at Commercial Mortgage Alert that we might hit $30 billion of issuance in 2021. I was the outlier…by a lot. Well, it looks like we’ll finish the year closer to $45 billion and it’s clear that in the first quarter of 2022 we will be on fire. Now, of course, I also thought that we’d have a fantastic year in early 2020 and then that pesky little bug changed our lives, so please consider my on-fire prediction subject to caveats, limitations and restrictions including, but not limited to, disease, dictators and the possibility that the Fed is making a colossal mistake.
All else being equal, the CRE CLO business will continue to grow and I don’t really see the appetite for this technology receding any time soon. Could it? Sure. Annoying black swans aside, if the curve radically changes shape and creates outsized demand for fixed rate product, the CRE CLO business, as it has grown up in the past couple of years, will see challenges. But more on this later.
Let’s set the table before I veer off into the land of non-recourse, shameless speculation. As we know it today, a CRE CLO is a leverage vehicle. This distinguishes it from the moving company model of the conduit CMBS business. Typically, in a CRE CLO, the sponsor retains 15-20% of the bottom of the capital stack. In large measure, tax transparency is achieved by hanging the vehicle under a public or private REIT (there are alternate approaches). Distinguishing this technology from the other most common form of leverage in the CRE space, the repo, the CRE CLO is non-marked-to-market, essentially non-defaultable, non-recourse, provides the sponsor with more management flexibility and provides something close to match-term financing for the underlying assets.
The managed CRE CLO (and frankly, most nominally static structures include a replenishment feature which the market endearingly doesn’t view as a managed structure) is characterized by an ability to add collateral to the structure after the closing. In virtually all deals today, the manager (or the preferred) has the ability to significantly modify underlying financial assets in a manner consistent with best portfolio asset management practice unimpeded by the servicing standard.
As the borrower community has increasingly become reliant (fixated?) on floating rate loans, this has become, for obvious reasons, a hugely popular method of turning 4% coupons into 12% yields for the non-bank lending community. (I think that’s my new job description… 4% coupons into 12% yields.) And now that the average weighted coupon on CRE CLO liabilities has come down to levels competitive and in some cases inside repo financing, what’s not to love?
Evolving investor views on the CRE CLO (and perhaps fading memories of the late, lamented CDO) has led to a concomitant increase in the appetite for CRE CLO paper in investor cadres. This has accelerated the growth of the CRE CLO market, as the “push” of the borrower needs and the “pull” of investor interest have grown apace. As we have said several times in the past, the CRE CLO offers the absolute best alignment between sponsor and investor in the entire structured finance marketplace. A factor not to be overlooked.
If the current arc of the growth of this technology continues, we’ll surely see big and bigger numbers put up over the next several years. My January prediction of $65 billion for 2022 (see caveats above) is not outlandish (maybe wrong, but not grotesque).
Largely, this has been a vehicle for the non-bank lenders. The non-bank lending sector is likely to continue to be at the heart of the demand for this technology, and non-bank lenders are multiplying like Tribbles (obscure cultural reference). That, by itself, means both more issuers, more active issuers and more deals.
But we’re going to see other entrants. Lifecos, major asset managers and even bank holding companies are likely to turn to this technology. As a lifeco, you simply can’t keep making 3% CRE loans when your liability structure is 7%-ish. Same for GIC and annuity sellers. We’re already seeing a variety of Separate Accounts managed by major insurers and asset managers looking at the technology and we expect that trend to grow. I mentioned bank holding companies. While banks themselves hardly need to turn to this relatively complex and expensive species of leverage, bank holding companies with non-bank subsidiaries do and we expect to see that represent a significant increase in demand for the technology in the coming year.
However, it strikes me that the CRE CLO technology can, with a little imagination, be repurposed in ways that will turbocharge its growth in 2022 and the out years.
First, let’s talk geography. The CRE CLO has largely been a Made-in-the-USA product for the last several years. We’ve now seen our first, albeit small, European CRE CLO and we expect to see real growth in non-US geographies over the next couple of years. There’re certainly some small-ball, courses for horses, adjustment to be made to the US technology to accommodate non-US legal systems, but on early days’ evidence, non-US CRE CLOs will continue to look remarkably like their US cousin, just as non-US corporate CLOs look like US CLOs. We expect these transactions to execute under either US or UK law.
Just as in the US, the CRE CLO is a compelling solution to the fact that coupons remain suppressed and the non-bank marketplace can’t attract investors with unlevered coupons. As European governments have been increasingly heavy-handed in their regulation of the prudentially regulated banking sector, and that’s only going to get worse, we will see the non-bank sector continuing to grow and therefore a concomitant growth in the need for the CRE CLO technology to obtain adequate leverage. Might we see $10 billion in non-US product in 2022? Sure.
If the paradynamic CRE CLO is a broadly distributed 144A offering, we are seeing and will continue to see the growth of private transactions. Using the same CRE CLO technology with or without a rating, transactions have and will continue to get done where sponsors will have nailed down major investors for a big hunk of the investment grade securities in advance. These transactions, often unrated, are done on streamlined disclosure documentation and big-boy letters from the investors. (As your lawyer, note this is a declaration against interest.) Usually, these transactions have simpler capital stacks and therefore some attendant pricing inefficiency, but these poor-man versions of the broadly distributed CRE CLO will grow.
While cost savings can sometimes be the driver of these transactions, they can also be driven by the inability of a sponsor to get to an adequate HERF score because what it does is make larger high-quality loans and just can’t accumulate sufficient note count for a regular way securitization. These can be terrific deals for investors, just likely unrated.
Another variation on the theme is the so-called direct issuance structure in which an issuer issues securities directly into the capital markets without the intermediation of a depositor and SPE issuer. In this case, the lender becomes the issuer. While this type of structure can be unwieldy, particularly when dealing with state mortgage licensure issues, it can provide a significant benefit. As there is no transfer of the financial asset, the transaction can be structured without risk retention. Is this a commercial mortgage SPAC?
Let’s not forget the possibility of this technology migrating to other asset classes. It seems to me that it won’t be long before we see someone endeavor to finance construction and development loans through the technology. We already see future funding, often construction light in many, many deals so, is this such a large leap? Construction loans are very different, representing a much higher risk profile than either fully stabilized bridge loans, or even bridge loans with material future funding, but, with eyes wide open, and the right pricing structure AND the requisite management/sponsor competence (not everyone who makes loans should even think about making a construction loan), there’s nothing to prevent this from being done, and it will be.
I hate to even mention the potential applicability to sub-debt, as I can envision townspeople with torches and pitchforks waiting for me on my return home from the office, but really? Think about the HRR market 5 years out from when risk retention first slouched into Bethlehem. There’s a ton of horizontal risk retention assets on the balance sheets of many and monetizing this asset is a more than twice a week conversation in our office. An issuer MOA can be constructed and “fully recourse” debt can be issued. Is that a CRE CLO? Or is that a corporate CLO? Does the distinction really matter? In either case, it’s a pool of financial assets supporting notes sold directly into the capital markets.
How about multiple sponsors securitizations? One of the biggest problems in the CRE CLO space to date has been a race to acquire enough product to meet the ratings agencies’ (and probably investors’), diversity and HERF fascination before term and yield maintenance protection on the underlying assets burn off. If multiple lenders could co-sponsor a structure, that would go a long way to solving that rat-wheel sort of problem. While it hasn’t yet been done, it surely will be. Multiple sponsors could sell into a single structure, agreeing to an evaluation of the collateral up front (much like happens in your run-of-the-mill conduit securitization MOA), the sponsors could own shares of an upstream REIT which holds the issuer in proportion to those valuations, be individually responsible for a rep or warrant breaches, but share the economics of the leverage in proportion to their relative value share. The accumulation vehicle can be the risk retention holder. Some challenges there? Surely, but could they be overcome? Equally certainly.
This has been done and we expect this subsector of structured finance to grow in the next year as well.
What about the Holy Grail of ameliorating the need for a pre-issue warehouse (or 2…or 3) with a CRE CLO which is all ramp, or virtually all ramp, issuing variable funding notes? This can also be done by merging a traditional repo with a capital market transaction by securitizing the buy side of the repo with variable funding notes? That’s sort of a CRE CLO isn’t it?
In this type of transaction, the issuer would fund its aggregating directly into the capital markets. How? One way is to establish, pre-closing, bond levels which are agreed between the sponsor and investor. The sponsor would then fund up from first dollars through the structure based on those established attachment points. At some specified future point, the sponsor would undertake to obtain actual ratings and defined attachment points and then the future advances would be designed to bring the structure into compliance with those ratings levels. We’ve seen this done already in a private transaction (and frankly, this might only be a structure that could be accommodated in a private transaction) and we expect to see this type of structure grow in attractiveness in the marketplace as we proceed through 2022.
How about this fascination with floating rate paper? There’s no structural reason this can’t be used to lever fixed rate or a combination of floating rate and fixed rate paper. To date, the compelling reason not to do so is pricing and the real absence of a portfolio demand for longer tenured fixed rate assets, but this could change. If the numbers work, the technology is easily adapted to non-bridge loan product and we expect this to happen.
While I’ve run out of creativity for the moment, once you break down the silos around the various products in the securitization space, you can begin to see real opportunities. Our business tends to conservatism; it tends to cause people to stay in their lane, but there’s money to be made by breaking down those silos and exporting technology from one sub-market to others.
Stand by. It’s going to be a really interesting year.