The number of triple-A buyers that would buy at market tights included US banks (even the bank that arranged the deal would sometimes backstop the tranche), insurance companies, large asset managers, CLO ETFs, regional Japanese banks and European accounts. One source stated that this year has been similar to 2021 in terms of how many triple-A CLO tranche buyers there are.

So much for that old refrain ‘there aren’t enough triple-A investors out there’.

Among the regional Japanese buyers are Chugoku, Joyo, Shinsei, Shinkin, sources state. Insurance companies such as Athene, MetLife, Meiji Yasuda, Security Benefit, and Symetra have all been active, and large asset managers such as Axa IM, PGIM (on behalf of Prudential) and Pimco (specifically for short-dated paper) too.

Most US banks that are in the CLO market as arrangers, now have pockets of capital to invest into CLO triple-As, according to several sources, with some like Bank of America, BNP Paribas and Societe Generale preferring to buy triple-A bonds in loan-only form. Banks like loan CLO tranches as the capital charges for buying these are less punitive. One US bank that is not yet buying triple-As again is Wells Fargo, according to sources.

Got there eventually

Not wanting to miss out on all the fun, Japan Post and Nochu acquiesced and are now buying at benchmark of SOFR+135bps, sources say. Nochu was the anchor triple-A investor in Barings CLO 2024-IV on 19 July, according to 9fin sources, which was the first time the bank went to S+135bps. Nochu has since been the anchor investor at S+135bps for CVC Credit Partners and Blackstone Credit in August.

Meanwhile, Japan Post was the anchor triple-A investor in Nuveen’s new issue Symphony 45, sources say, on 29 July. The two anchor investors were telling US CLO managers that S+135bps may be as tight as they could go, as both have highlighted that the strong US dollar means foreign investments are less attractive.

But that was before a global sell-off was triggered in early August that weakened the US dollar and made foreign trades more attractive for Japanese investors.

The exchange rate to start the year was 142 JPY to 1 USD. On July 10 that was up to 161 JPY to 1 USD, before it dipped down to 145 JPY to 1 USD after the sell-off in August.

Not only was syndicating triple-As a way to move triple-A spreads tighter, sources say that dealing with smaller accounts is easier than banks such as Nochu and Japan Post. They state that the two anchor buyers have more stringent document requests in the CLOs they invest in, compared to other triple-A buyers.

The presence of Nochu and Japan Post in primary CLOs has a meaningful impact, even if they are matching benchmark levels rather than moving the market tighter.

When it comes to par subordination, for instance, Nochu and Japan Post have higher requests compared to other triple-A investors. Nochu likes the tranche to have 38.5% par subordination and Japan Post prefers a 40% attachment point, sources say. Both banks also require two rating agencies to rate the tranche they are buying.

Nochu and Japan Post also do not lock in a level when they commit to a deal. Instead they commit to taking the whole tranche and then negotiate the triple-A spread level. Oftentimes on marketing emails, the triple-A tranche will be ‘pre-placed’ while the manager markets the rest of the capital structure. A level is then finalized before the deal prices, sources say.

This can lead to “intense negotiations” between the CLO manager, arranger and Nochu/Japan Post, according to sources. During the period earlier this year when both banks were stuck at S+150bps, the “stand-off” between the parties was “longer than usual, and heated”, one 9fin source stated.

This does not mean that the US CLO market has turned its back on Nochu and Japan Post, however, as some managers have continued to price deals with one of the anchors involved in order to stay on the banks’ list of preferred US CLO managers, sources say. Both banks were active in Q1 and Q2 this year, but not at the tights and not as often as in years past.

Also, while the US CLO market does not necessarily need the two banks at the moment, these conditions will not last forever and if the market reverts to how it was in 2022 and 2023, when US banks and domestic buyers took a step back, Nochu and Japan Post will be key for issuance, according to several sources.

Ripping tighter

Traditional triple-A anchor buyers were buying at the start of the year at SOFR+150bps and kept triple-A spreads at or near the level for months, even as the rest of the CLO capital structure was tightening. At the time, only two firms were at the benchmark level with Oak Hill Advisors and BlackRock issuing five-year reinvestment CLO triple-As at S+148bps. Oak Hill priced its new issue in January and BlackRock in February.

These were the only two managers to go inside S+150bps until GoldenTree Asset Management priced GoldenTree US CLO 20 on 30 April via Bank of America (the lead arranger), Morgan Stanley and Wells Fargo. The triple-As priced at S+145bps and sources say that BofA took a significant portion of the $307.5m tranche.

The triple-A benchmark for five-year reinvestment CLOs began to edge tighter from there. On 23 May, Elmwood Asset Management priced its triple-As at S+143bps on its new issue Elmwood 30, before Allstate Investment Management Company and Nomura reset Aimco CLO 10 and printed its triple-A tranche at S+141bps. Both managers syndicated their triple-As, according to 9fin sources.

AGL and Palmer Square Capital Management then tightened the benchmark further to S+140bps in late May, before the benchmark settled at S+135bps in July. Allstate Investment Management Company has now brought the US CLO triple-A benchmark down to S+134bps with its reset of Aimco CLO 11 on 1 August.

Before Allstate priced at S+134bps, 20 US CLOs from 14 different managers priced at S+135bps as the market found itself at a standstill, similar to how it was stuck at S+150bps earlier in the year.

Triple-A spreads started ripping tighter for a variety of reasons, but two of the main ones are that there has not been a cut in interest rates this year, and CLO triple-A prepayment rates for post-reinvestment deals are very high, sources tell 9fin. Without a cut in rates, SOFR continues to be over 5%, making floating-rate paper very attractive. Even at S+135bps, CLO triple-As are yielding over 6%.

Meanwhile, the combination of so much of the CLO market being out of reinvestment, and prepayment rates high, CLO triple-A investors were flush with cash and needed to replace those bonds that were paying down. Therefore they reupped into the CLO primary market and were willing to go tighter than traditional anchor investors, according to people familiar with the matter.

ETFs making waves

Another factor for why syndicated triple-As have become more prevalent this year has been the rise of CLO ETFs, sources say. CLO ETFs inflows this year have been gargantuan with Janus Henderson’s triple-A CLO ETF (JAAA) now over $11.5bn after starting the year at $5.3bn.

Triple-A CLO ETFs were a natural landing spot for junior triple-A tranches, according to sources, but as money continued to flow in to the funds, Janus had to up its investments and started buying $100m tickets of senior triple-A tranches also. The largest CLO position in JAAA from a deal that priced in 2024 is $140m of the senior triple-A tranche of Sycamore Tree 2023-3 which was reset in April, the tranche is $336m in total size.

Triple-A CLO ETFs are less sensitive to spreads compared to traditional investors, sources say, as the funds are “all-in yield buyers” rather than “spread buyers”. Meaning that CLO ETFs take the SOFR rate into account, rather than just the spread. With SOFR above 5%, CLO triple-As are yielding almost 7%, one of the best yields for triple-A paper in the credit markets.

Our customers receive this content ahead of the crowd — find out more about what we do for structured credit.