Why do chickens hatch eggs faster than banks settle loans?
An introduction to broadly syndicated loans and the loan trade settlement process
Chicken eggs hatch after 20 days of incubation. According to the LSTA, U.S. par broadly syndicated loans settle 28 days after they are traded. The loans technically settle in T+20 business days, but the trading industry is fussier about working on weekends than chickens are.
Intro to Broadly Syndicated Loans
Feel free to skip ahead to “What is trade settlement?” if you already have a nuanced understanding of the broadly syndicated loan market
Semantics matter
In finance there are two types of definitions - dictionary definitions and context dependent definitions. For the sake of brand consistency, I’ll refer to the latter as Wall Street Definitions. Oftentimes, you’ll come across two words with the same meaning: credit/debt, issuer/borrower, or financial sponsor/private equity firm. This type of Wall Street definition is pretty straightforward. You can easily google the fancy sounding synonyms and pick the vocab up pretty quickly.
The trickier type of Wall Street definition is when a single word has multiple meanings. These definitions are nuanced and all about context.
The canonical example would be the definition of risk. If you talk to quants, economists, macro traders, portfolio theorists, and academics, they will define risk as the probability an investment’s returns differ from the expected return. This includes both positive and negative surprises. These professionals use volatility as a synonym for risk, and measure it quantitatively as the standard deviation between the asset’s returns and the broader market’s returns.
If you talk to literally anyone else in finance, they will define risk as the chance you lose money on an investment. They hardly ever use the word risk in reference to the chance of outperformance.
I studied economics in college and learned this semantic divergence the hard way. I was laughed at in an interview after proudly proclaiming the definition of risk is “the probability that an asset’s actual returns differ from its expected return as measured by its standard deviation.”
Throughout this post and in future posts, I plan to highlight differences between the dictionary and Wall Street definitions of certain words. For the Wall Street vets among my readership, this will probably be elementary. However, I think these distinctions are important and often unintuitive.
What is a loan and what’s the loan market?
The dictionary definition of a loan is a financial transaction in which a lender provides money to a borrower and the borrower repays this amount plus interest over a specified period of time. This is probably obvious since loans are all around us - student loans, auto loans, mortgages, home equity lines, etc. Without added context, the “loan market” could conceivably reference markets for any type of debt instrument.
On Wall Street, however, when people discuss the “loan market,” they are typically referring to a subsegment of the corporate credit market. The corporate credit market can have multiple definitions, but it usually means lending to corporations in the corporate bond market, the private credit market, and the broadly syndicated loan (BSL) market. The loan market could in theory reference both the private credit and BSL markets since both traffic in term loans, however, the “loan market” usually just refers to the BSL market.
Some people will loop the BSL market in with bonds and call the combination “public credit.” I think that’s a misnomer since BSL investors need to get permission from BSL borrowers in order to invest. They even need to sign NDAs to get access to the borrower’s financials in the first place. That doesn’t sound very public to me.
The BSL market deals in the issuance of term loans where a borrower receives a lump sum of cash up front, and repays the loan over a specified period of time. These term loans are floating rate which means the interest rate is recalculated throughout the life of the loan as the current base rate (formerly LIBOR, now SOFR) plus a credit spread (i.e., SOFR + 300 basis points). This contrasts with the bond market where bonds are typically fixed rate (i.e., an unchanging 9% per year). When interest rates are rising, floating rate borrowers pay more in interest. Additionally, term loans are usually the most senior debt capital and secured by the assets of the company. This just means that in a bankruptcy, term loan lenders get paid before the bondholders and equity investors. The loan market can also include revolvers, but unless otherwise stated, I’ll be focused on term loans in this piece.
Everything in the loan market is governed by a credit agreement which is just a legal contract negotiated between two or more parties. I wrote that term loans are usually floating rate, senior secured, and funded up front, but in practice all of those features are up to negotiation and these instruments are quite bespoke.
The broadly syndicated loan market is intermediated by investment banks. If your college roommate went into LevFin at J.P. Morgan, he spent his time structuring these loans. When a company needs to borrow $400m+ to undertake an acquisition, the company will call up the LevFin team at an investment bank and ask for a quote. The bank commits to providing the requested financing within some sort of pricing range (perhaps SOFR + 250-350 bps), and then calls up everyone it knows to take a piece. The bank will lend $40m (or less) of the $400m and the rest is “syndicated” to clients including CLO managers (75%+ of the market), credit hedge funds, and other asset managers. This is called the primary market and issuance averaged $865bn annually over the last 10 years per Refinitiv. After the these loans are issued in the primary market, they’re traded in the secondary market. Annual trading activity in 2020 was almost $800bn, up from just over $400bn in 2010. In aggregate, there is ~$1.2tn of broadly syndication loans outstanding.
What does it mean to “trade” a loan?
This piece is about loan trading - where an existing lender sells its stake to a new lender (the buyer). I’ve found the concept of “purchasing a loan” is unintuitive to people outside of finance. The following might help:
Imagine you buy a house with a mortgage from Bank of America. The house costs $1m and BofA lends you $700k to finance the purchase. You agree to repay the mortgage in monthly increments over 30 years. BofA issued your loan and is therefore your lender, but they might decide to sell that loan to a different bank. If they sell the loan the JP Morgan, JP Morgan becomes your lender. You now wire your monthly mortgage payment to JPM instead of BofA. On the flipside, if this loan was an undrawn home equity line instead of a mortgage, JPM is now on the hook to wire you money when you make future draw requests.
Circling back to broadly syndicated loan trading, secondary trades in this market settle 20 business days after the trade date (T+20). This means that if you agree to sell a loan on July 10th, it won’t settle until August 7th. Stocks, by comparison, settle in T+2 business days. People think Bitcoin is slow because transactions take 1 hour…
What is trade settlement?
There are two main components in the settlement of any trade: Cash and Accounting. If I sell you a share of stock in my startup for $100, you need to wire $100 to my bank account (cash), and I need to update my accounting records to show you as a new owner of the stock (accounting).
In the broadly syndicated loan world, the process flows roughly as follows:
Prospective buyer needs to get the borrower’s consent to make a purchase. Borrowers in the loan market can actually blacklist lenders they don’t like.
The seller (the bank, a.k.a. the dealer) needs to onboard the buyer. This involves running KYC for the buyer’s various legal entities, gathering the contact details, and confirming the wiring information. This is surprisingly painful in practice.
The bank agrees to sell the buyer $X of ABC loan at a specific price, usually via instant message in Bloomberg.
The bank’s operations personnel and the buyer’s operations personnel need to agree on the details of the trade.
The bank and the buyer need to sign off on 3 key documents: Trade Confirmation, Assignment Agreement (literally assign the legal lending obligation to the buyer), and Funding Memo.
The buyer needs to wire the money to the bank.
The bank needs to update its own books and records for the buyer’s purchase. This formally makes the buyer a “lender of record.”
This process sounds (sort of) simple but a lot can go wrong. The following fictional trade should help shed some light on the people, processes, and technologies involved in broadly syndicated loan trade settlement.
Walkthrough the loan trade lifecycle
Introducing our cast of characters
To set the stage, let me introduce our cast of fictional characters and financial institutions named with the help of ChatGPT:
Samantha Hawthorne, Capital Markets, Blackpebble Capital
Samantha works for a $50bn AUM private equity firm that recently financed its $10bn LBO of Vertexify using $4bn of broadly syndicated debt. As a member of the PE firm’s capital markets team, she serves as the internal LevFin function and helps the LBO deal professionals find debt financing for their transactions.
Derek Montgomery, CFO, Vertexify
Derek is the CFO for Vertexify, a $10bn SaaS company recently acquired by Blackpebble. As the most senior finance professional at the firm, he oversees the credit agreements and manages the company’s debt load which includes a Term Loan A (TLA), Term Loan B First Lien (TLB 1L), and a Term Loan B Second Lien (TLB 2L).
Raj Patel, Credit Analyst, Yellowstone Asset Management
Raj works for a $4bn AUM credit-focused hedge fund and is a sector specialist in technology. Specifically, he covers 20-30 large technology companies that have outstanding broadly syndicated loans. He’s familiar with the narratives surrounding these companies and has a rough sense for the prices at which he would like to buy or sell the various loans. In the following example, Raj will be buying a loan in the secondary market.
Ben Reynolds, Loan Trader, Yellowstone Asset Management
Ben works with Raj but does not focus on any specific sectors, follow the company-specific narratives, or make decisions to buy or sell a given loan. He is very plugged into the macroeconomy, keeps a pulse on inflows/outflows in the asset class, and has relationships at all of the investment banks. His main job is to figure out how to buy a loan efficiently when Raj wants to buy, and sell a loan efficiently when Raj wants to sell.
Nate Thompson, Loan Operations, Yellowstone Asset Management
Operations includes the systems, processes, and documents required to make the loan trade actually happen. Nate handles everything outside of the investment decision making, and trade execution. In practice, his role will be filled by separate people across middle office and back office functions (a distinction we’ll save for another day). But for simplicity today, Nate will handle all of Yellowstone’s operations.
Isabelle Rousseau, CLO Portfolio Manager, Atlas Investment Group
Isabelle works at a $20bn AUM multi-asset manager that happens to run a CLO vehicle. Without going into the details, a CLO is a type of structured product that many credit managers employ. CLOs are the main participant in the broadly syndicated loan market and account for 75%+ of the volume. Her job is similar to Raj’s, except she’s a bit more senior and focuses on the overall portfolio of holdings, rather than the specific companies and loans. In the following example, she will be selling a loan in the secondary market.
Max Harrison, Loan Trader, Atlas Investment Group
Max does the same thing as Ben at Yellowstone, but on behalf of Isabelle at Atlas.
Avery Mitchell, Loan Sales & Trading, JP Salomon
Avery is a loan sales and trading professional at a global investment bank. Most banks have a person doing loan sales and a separate person doing loan trading, but for the sake of simplicity, Avery will handle both functions. Avery’s job is to enable her bank’s clients to buy loans that they want to buy, and sell loans that they want to sell. She’s (mostly) not placing directional bets on loans to trade up or down, but rather publishing prices at which she will both buy and sell a loan in order to make a small spread. Each morning Avery sends a list of 2x2s to her clients for the loans she trades. A 2x2 for the Citrix Term Loan B (“TLB”) might say “92-94”. This means she’s willing to buy $2m of the Citrix TLB for 92 cents per dollar of principal, and willing to sell $2m of the Citrix TLB for 94 cents per dollar of principal. Avery trades Citrix loans because her bank’s LevFin team underwrote the original Citrix TLB primary issuance and serves as that loan’s administrative agent.
Sales & Trading is similar to CarMax. CarMax doesn’t have a long-term thesis on the 2020 Toyota Camry, but it will buy any good condition 2020 Toyota Camry for $18k and sell any good condition 2020 Toyota Camry for $24k.
Brad Sullivan, Loan Agency, JP Salomon
Brad is another combo character. He handles both loan operations, and admin agency at JP Salomon. Admin agency has two main functions: (1) maintain the ledger for lenders of record when secondary trades happen, and (2) make sure the right cash flows are distributed to the right people at the right time. Term loans typically have regular interest payments that vary depending on the current Secured Overnight Financing Rate (SOFR). The admin agent calculates a given period’s interest payment, collects that money from the borrower, and then distributes that money to the various lenders recorded on the ledger. It sounds simple in theory. It’s chaos in practice.
An example trade
Friday, July 7, 4:00pm: Delays in software implementation timelines throughout Q1 and Q2 have negatively impacted Vertexify’s cash flow and brought down debt service coverage ratios. Derek (CFO, Vertexify) submits the final Q2 numbers to S&P Global, his company’s credit ratings agency, and uploads the financial package to the lender data room hosted by FIS’ SyndTrak. Data rooms are basically Google Drive folders with compliance features suited to financial institutions. Other data room vendors serving this market include IntraLinks and S&P’s DebtDomain.
Monday, July 10, 7:00am: Due to Vertexify’s worsening liquidity profile, S&P downgrades Vertexify’s TLB 2L from B to CCC.
Monday, July 10, 7:10am: Isabelle (CLO PM, Atlas), wakes up to an automated email notification - “S&P downgrades Vertexify TLB 2L to CCC.” Shit. She calls Max (loan trader, Atlas), “S&P downgraded our Vertexify credits. Sell our position.” This doesn’t come to a surprise as both Isabelle and Max have been monitoring Vertexify over the past few weeks. As a CLO, they’re only allows to have 7.5% of their portfolio invested in CCC-rated credits. They were already bumping up against this threshold and were worried Vertexify would be downgraded. She logs into a dashboard on Allvue’s Black Mountain, a popular order and portfolio management software system for CLOs, and sees that Atlas’ portfolio surpassed the CCC limit. It took her 18 months and $1m in consulting fees to implement Black Mountain, but at times like this she’s glad she did. Until they offload this position, coupon payments to their CLO’s equity tranche will be halted.
Monday, July 10, 7:12am: Max shoots Avery (S&T, JP Salomon) an instant message via Bloomberg chat, the preferred communication channel for the loan trading industry. “Vertexify 2L downgraded. Sell at market.” Avery is already at the office and has been since 6am. She’s obviously aware of the Vertexify downgrade and has received sell orders from 9 CLO clients in the last 12 minutes. She’s sitting in a conference room with 8 other members from JP Salomon’s loan S&T team and announces “Another axe. Atlas needs out of Vertexify.” Axe is short for ‘axe to grind,’ trading parlance for a position the trading desk wants to buy or sell. Avery’s bank is going to receive a lot of volume today for Vertexify loan trading because they are the administrative agent for the loan. In fact, the admin agent on any given loan receives 67%+ of that loan’s secondary trading volume. They hold the keys to the books and records, and trade settlement is impossible without their cooperation.
Monday, July 10, 7:20am: After discussing with her team, Avery sends out her morning trading run with the Vertexify TLB 2L 2x2 at 82-84. “This will garner some attention,” she thinks. Her previous run was at 92-94. It’s not uncommon for pricing to gap down 10 points when a credit gets downgraded to CCC. Most of the investors are CLOs, and they’re all rushing to the exit. She posts her revised pricing on Bloomberg for her loan trading clients. On the client’s Bloomberg portal, they’ll see what looks like a black, yellow, and green email inbox with timestamped 2x2s from traders at the various banks (i.e., “07:20; Run; JPS; 82.000 / 84.000; 2M x 2M; Mitchell A; VRTX TL B 2L”).
Monday, July 10, 7:30am: Raj (credit analyst, Yellowstone) checks Bloomberg and sees Vertexify’s TLB 2L gapped down. He’s not surprised. Friday’s results were pretty brutal so he had expected S&P to downgrade the credit. He’s been covering the company since Blackpebble’s acquisition and took a look at JP Salomon’s primary issuance, but decided to pass at the time. He’s glad he had taken the earlier look because he’s already under NDA and therefore has access to Vertexify’s private financials. If he hadn’t, he would need to reach out to his coverage representative at JP Salomon to get access to the data room. That would be painful and slow him down. Over the weekend, he downloaded the latest financials from FinDox, a compliant data room aggregation portal owned by Reorg Research, and updated his financial model. He also spent some time with the credit agreement and read through some summary reports from two credit research portals - Xtract Research and Covenant Review. He feels there’s an attractive risk/reward if they can get exposure below 85 cents. His fund is pretty large, so he wants to build at least a $20m position. He messages Ben (loan trader, Yellowstone), “See if we can pick up $20m of Vertexify TLB 2L under 85 cents.”
Monday, July 10, 7:31am: Ben pings Avery “hey can u get us 20m of vertexify at 84.” Avery responds, “1l or 2l??” Last week she failed to clarify which lien a client wanted and the confusion kicked out settlement by a full week. She wasn’t making that mistake again. “2L,” Ben responds. Avery replies, “Best I can do is 4m at 84 and 4m at 85. maybe 10-12 more at 86.” Ben is not surprised. Loans, even actively traded names, rarely trade in blocks bigger than $2-5m at a time. It often takes a while to put together a scaled position. “Fine. We’ll do 4m at 84 and 4m at 85. Hold off on the rest.” Avery confirms, “Deal. Yellowstone to buy $8m of Vertexify TLB 2L at 84.50.” Interestingly, this counts as a legally binding voice trade under New York law. Ben screenshots his chat with Avery and sends it via email to Raj and Nate (loan operations, Yellowstone).
Monday, July 10, 11:00am: Nate starts to go through loan trades from this morning and enters Ben’s Vertexify trade into his trade capture system. Nate’s firm built trade capture internally and integrates it to a back-office data and accounting platform called SS&C Geneva. They also use S&P’s WSO suite to manage their loan portfolio on an ongoing basis - to keep up with coupon payments and make sure the cash received in their bank account ties out to expectations. Other firms will use S&P’s WSO to handle all of the portfolio and loan-level data management via a combination of services and software. Some of his peers have even said to hell with the middle and back office, and outsourced the entire function to BNY Mellon. Ben will now sit tight and wait for the JPS to build out the trade in ClearPar so that he can begin processing the paperwork.
Monday, July 10, 5:16pm: After a busy day of trading Vertexify loans, Avery updates her trade blotter. At some firms this might exist as an internally built software tool or an extension of the order management system, but at many firms it’s just an excel spreadsheet that the traders and trade assistants update manually. After entering the trade details into her blotter, she emails it to Brad (loan agency, JPS).
Tuesday, July 11, 9:17am: Brad opens up his queue of pending loan settlements. The first trade is for an $8m sale of Vertexify TLB 2L. He pulls up LoanIQ, a Finanstra-owned syndicated lending software application that his bank uses to build out the loans and record the lenders of record. “Ugh.” JPS only owns $6m of the Vertexify TLB 2L. He looks back at his queue. There’s a later trade where JPS purchases $23m of the Vertexify TLB from Atlas Investment Group. He needs to process that first because there’s no shorting in the loan market so it’s impossible to sell something you don’t have. The sale to Yellowstone is pending an upstream purchase and can’t be processed until that clears. Put more simply, JPS can’t sell $8m of Vertexify to Yellowstone because it only owns $6m of Vertexify. Brad opens up an Excel spreadsheet that he uses to Pro Forma for these kind of mutli-stage trades. His loan books and records system shows him these trades as PDFs so he needs to manually calculate how they will work out in Excel.
Tuesday, July 11, 9:34am: As Brad is working through the buyside trade, he comes across an issue. Atlas holds its position across 15 different legal entities but the contact information is wrong for one of them. The Admin Details Form (ADF) for Atlas Grand Cayman 12 is outdated and the email address returns a bounce back. He emails his friend who works in loan operations at Atlas and asks him to update the ADF for AGC-12. Well, better luck tomorrow.
Wednesday, July 12, 10:23am: No response from Atlas. Brad calls his friend who finally agrees to update the ADF after lunch.
Thursday, July 13, 11:10am: Brad gets the Admin Details situated for Atlas and can finally move on to beginning settlement for the $8m Vertexify sale to Yellowstone. He pulls up ClearPar, a loan settlement platform built by a lawyer in 2000 that was sold to FIS in 2004, IHS Markit in 2010, and then rolled into S&P Global via their acquisition of IHS Markit in 2022. It doesn’t look pretty, but it’s used by 99% of the market and it’s where he spends most of his day. It’s a combination of document generation, e-signature, calendar management, and distribution portal. He starts to build out the trade so ClearPar can generate the required materials. For a loan trade to close, both counterparties must sign the Trade Confirmation, Assignment Agreement, and Funding Memo. While entering the information into ClearPar, he realizes Yellowstone doesn’t have Borrower Consent for the legal entity that wants to conduct the trade. In the loan market, the borrower gets to approve whether a new lender can hold its outstanding credits. In this case, Yellowstone doesn’t have that approval so Brad emails Derek (CFO, Vertexify) and Samantha (Capital Markets, Blackpebble).
Thursday, July 13, 11:15am: Both Derek and Samantha are actively monitoring the trading prices of the outstanding loans and quickly respond to Brad, grant borrower consent to Yellowstone, and sign the Docusign link. Brad can’t believe his luck. Oftentimes Samantha will approve via email but disregard the Docusign link. Technically the consent will be granted with or without her signature, so if she’s in a rush, she’ll just ignore the email. The eventual outcome is the same, but trade settlement can get delayed by an extra 2+ weeks if Samantha forgets to click on the signature link. Not this time. Brad finalizes his work in ClearPar which sends the Trade Confirmation over to Yellowstone and creates action items and timelines for the Assignment Agreement and Funding Memo.
July 14 through August 7: Over the course of the next few weeks, many hurdles arise that delay the trade confirmation, assignment agreement, funding memo, and ultimately, wire processing required to settle a loan. Forgoing our previous narrative flair, here are a few example hurdles that might arise:
KYC: JPS needs to run a Know-Your-Customer (KYC) check on Yellowstone. This needs to be done for each of the different legal entities, which can be numerous. The KYC process also touches a variety of different teams internally. The burden of KYC is on JPS, so as you can imagine, Yellowstone isn’t champing at the bit to submit all of their required paperwork, signatures, and tax documents. JPS has to continually chase to get these documents compiled in a timely fashion.
Gaming Delayed Compensation: A large bank might be working on settling 100s of trades at a given time. When a credit fund is buying a loan and the bank is delayed in settling, they start to accrue something called Delayed Compensation. Basically, the buyer starts to accrue interest on paper for every day past T+7. When they finally settle and need to pay cash for their purchase, this accrued interest is subtracted from their purchase price. When a credit fund’s counterparty is a large investment bank with an understaffed loan agency team, the fund can pick up a $100-200k+ discount just by dragging its feet on closing.
Upstream Issues: It’s impossible to short loans, so to sell a loan, you need to own the loan. We saw Brad dealt with this issue from the bank perspective. On the other side, in order for a credit fund to buy a loan, they need cash. If they expected a loan sale to settle in 2 weeks, but it takes 2 months, they might not have a sufficient cash buffer to close on new loan purchases.
Trading Blackouts: Agents will freeze trading for a variety of reasons - primarily attributable to operational complexity. For example, loan agencies will stop settlement activity a few days in advance of a coupon payment so they can make sure the wiring information is correct across the updated roster of payees. Similarly, the loan agencies will stop processing secondary trades during ongoing amendments to the credit agreement.
Europe: If you’re trading European loans, the same issues arise but the timelines are dragged out. Time zones and strict working windows magnify the impact of paperwork ping pong, and all signatures are signed physically with wet ink.
Typos: Most of these loan trades are conducted via Bloomberg Instant Message. In fact, only 1% of this $800bn market is traded via an electronic matching engine. Everything else is basically traded via BBG IM. It’s not uncommon for the buyer to believe he bought X while the seller believe she sold Y.
Human Capital Risk: Loan agency is not always a very fun job. Both the agency departments at large investment banks, and the outsourced loan agency providers like Alter Domus, are known for high human capital churn. If a loan agency analyst is checked out, or quits, a wrench can be thrown into any ongoing trade settlements.
Leverage Grid Triggers: Some of these loans will implement leverage grid pricing. In short, the interest rate on the loan will increase if the leverage ratio increases. This is outlined in an exhibit in the the credit agreement called a “leverage grid” or “pricing grid.” If the agent misses a change in the leverage grid, they might over or under charge the borrower interest for a given period and disburse the wrong amount of interest to the lenders in the syndicate. Depending on the magnitude of the error, and time it goes unnoticed, this can be quite messy to unwind.
Monday, August 7, 3:00pm: After a few weeks of endlessly chasing parties on paperwork, and a hectic Monday corralling wiring details, JP Salomon is ready to settle. They receive $8m in cash from Yellowstone and happily update the lenders of record in LoanIQ to reflect the new creditor. From now on, they’ll distribute Yellowstone’s pro rate interest payments and loop them into any important governance discussions. The situation at Vertexify seems to be improving and Raj and Nate are happy that this loan trade settled in such timely fashion. They still have a European distressed debt trade outstanding from 2 years prior, so one month feels pretty good by comparison.
Why are loans so painful to trade?
When I embarked on this post, I intended to make it comprehensive. 5,000 words later, I realized that will be impossible. We’ll save the exciting details around Settlement Date Coordination, and distressed vs. par papers for another day. In the meantime, I think it’s important to highlight a few reasons WHY this process has evolved into such a cluster.
Loans were not meant to be traded. Shakespeare’s Shylock never intended for Antonio’s TLB 1Lb to be traded on secondary markets. Loans are complex legal agreements and bespoke in nature. They’re definitionally non-fungible.
Loans are not TRACE-securities. Dealers do not need to report on trade prices or volumes to any central intermediary like they must in other markets. The flows in this market are opaque. Information, inventory, and volumes are hoarded by the administrative agents who stand to collect juicy spreads and regular $3,500 assignment fees.
Coordination is a nightmare. There are too many people involved in signing too many documents. Worst of all, a bunch of them simply cannot be bothered. The PM at a large CLO shop wants his loans to settle on time, but has no interest in dealing with the paper pushing. The middle and back-office professionals across this industry struggle to chase signatures for this, approvals for that, documents for X, Y, and Z. Their requests and pleas often land on deaf ears until the problems grow too big to ignore.
Loan administration and admin agencies are error prone. Agency is handled by a 3rd party, or at the very least, a separate division. Yet, admin agencies are responsible for updating cash flows in accordance with the credit agreement and subsequent amendments. Sometimes, they’re left out of the loop and updates fall between the cracks.
Is there an opportunity to streamline this space with better technology? Maybe. But I don’t think it would do much to slap a sexier UI on ClearPar, and a lot of this stuff is already in the cloud. The banks are working in consortiums to address some of these challenges but it’s too soon so see how those initiatives will play out. Versana and Octaura are two interesting examples to watch. The LSTA is also an important party here. They’ve been the driving factor for standardization in this industry and have done some cool stuff over the years like working with OpenLaw to build some blockchain-based smart contracts.
A real 10x product in this market would need to be equal parts operational, administrative, and technological. I’ve thought a lot about this and haven’t personally figured it out, but if you have ideas, hit me in the comments below or in my Twitter DMs. I might make a future post about some of the leading initiatives, potential solutions, and wildcard ideas for fixing loan trade settlement.
Finally, does this matter?
I’ll leave you with 1 negative and 1 positive example about why loan settlements matter.
Citi’s $1bn fat finger:
The linked NPR article details the story in all of its glory but the short version is that Citi accidentally wired a lot of money to some very feisty hedge funds on behalf of Revlon, a distressed debtor. Despite a “six eye approval process,” some poor loan operations guy in Delaware clicked the wrong button on Citi’s janky software system and sent $900m to some distressed credit hedge funds who refused to give it back.
Settlement speed as a source of…alpha?:
The following block quote is snipped from Patrick O’Shaughnessy’s Invest Like the Best Interview with Scott Goodwin, co-founder of Diameter Capital Partners. The trade is detailed below and in the linked podcast, but in short, Scott secured a 10% discount on a $500m portfolio of loans from a forced mutual fund seller because the bank knew Scott’s fund could settle, close, and wire the cash extremely quickly.
Scott: [00:18:03] Sure. Sure. I'll give you an example from COVID. That's maybe the most interesting example. The levered loan market is a market that is very opaque. 70% of the market is private issuers, which means there's no public stock you can file. You have to go on the interlink 1819 site to get the financials.
And levered loans don't settle like stocks or bonds. It's mind-boggling, but levered loan settlement process could take anywhere from a week to months. Hopefully, someday blockchain will fix that, but it hasn't yet.
Scott: [00:18:54] I'm sending them an IB and then I'm also talking to them. For each bank, sort of nuance the list a little bit, as are the traders on our team. And the head of loan trading at -- BofA calls me. He says, "Hey, at 7:00 a.m., I've got a mutual fund that's got a $1 billion outflow in loans." They're calling us because we are the fastest settlement process for loans.
"Well, okay. They own these names on your list. Can you buy $500 million by 8 a.m. because I want to make some progress?" I call Jon. We're like, "Let's not buy cyclical stuff. We don't know what's going to happen here." We're starting to buy a little bit of IG because we think the government is going to start buying IG, but this is junk-rated loans.
And we had had our analysts in software learn all the software loans in 2019 because we said, "Well, if there's a recession and there is a cyclical environment, the whole loan market is going to trade down because that's where a lot of the excesses are building up, but software will be the most defensive place. It's stickier."
So we bid that firm for $500 million of loans, of which $350 million was software loans. Let's say the average price on them the prior day was in the high 80s 1954. We bid around 80s, so down, say, 10% or 8 points, and they sold it to us.
And I think there are probably 2 firms in the world that could have responded to that call within 15 minutes. And we responded, I think, within 5 minutes. But they called us because, a, we had shared the list with them, and, b, they knew we had a track record of providing liquidity into these dislocations and responding fast.
So that speed of capital in that situation provided a lot of alpha.
I’m not a loan expert, but I find this market fascinating and have spent lots of time speaking to people in this industry. It’s an enormous market and critically important to our economy, but incredibly slow and inefficient when compared to similarly sized markets. It’s grown in unexpected and unwieldy ways, but presents opportunity for innovation.
Let me know what you think in the comments.
Great stuff Van. I’m going to think on this and let’s circle up for coffee again soon.
Great read, I feel like I'm quickly becoming a FIN expert! Love the characters!!