How “Senior Secured” Became Subordinated Scraps
A $15M “senior secured” loan to a class action law firm — how I lost $87,723, and what the offering documents said before I lost it
This is the most expensive lesson in my portfolio and the one I link back to most often, so I’ll put the takeaway up front: on a retail platform, the label on a deal is not the risk inside it. “Senior secured” sounded like the safest thing YieldStreet offered — top of the debt stack, secured by the case-portfolio collateral, backed by personal guarantees.
My second YieldStreet investment: $250,000 in, $162,277 back, a 35% loss and a -10.31% IRR against a 12.75% target. Four years in, the first lien became a junior note behind a new lender. The position is still open at 99 months, against a 36-month term.
The subordination is the obvious story here, and for years it was how I explained the loss. It was real. It was also the last thing that went wrong, not the first, so the rest of this runs in order: what page one left out, what the next eight years did to the loan, what is left of the position now, and why none of the protections held.
The Numbers
| Invested | $250,000 |
| Returned | $162,277 |
| Net Loss | ($87,723) |
| MOIC | 0.65x |
| IRR | (10.31%) |
| Target term | 36 months |
| Funding to last dollar | 77 months |
| Still open, as of Jul 2026 | 99 months |
| Target investor rate | 12.75% |
| Rate the borrower actually paid | 18.0% + excess LIBOR |
What Page One Left Out
In late March 2018, YieldStreet offered Law Firm Financing VII — a $15M senior secured term loan to a New York-based class action law firm, with a loan start date of March 27. The deal filled in minutes. I committed $250,000, and my money funded on April 10.
The tear sheet, which is page one of a ten-page memorandum:
- 12.75% target investor interest rate, estimated duration 36 months
- Senior Secured First Lien — the highest priority in the borrower’s capital structure
- Backed by the firm’s entire case inventory — 22 cases, mainly securities and antitrust class actions
- Loan-to-value of 4.3%–5.9%, against projected case revenues of $253.8M–$344.9M
- Experienced originator — Counsel Financial, $1.6B+ originated over 17 years, with defaults of $3M, or 0.4% of what it had advanced in the previous nine years
- Full recourse personal guarantees from three of the firm’s four principals, plus a three-month minimum interest guarantee
I have checked every line of that list against the source documents. Not one of them is a misrepresentation — though the page is careless: one bullet gives the loan-to-value as 4.3%–5.9% and the next gives it as 6.3%–8.4%, with nothing offered to reconcile them. The investment still lost 35% of its principal. In 2018 I read that page as a summary of the deal. It was a selection: every fact true, and every one chosen because it made the loan look safer.
Three things in the file cut against that page, each one widening the gap between the deal it described and the deal I actually bought. Together they explain the loss better than the subordination does.
1. The rate that mattered was the borrower’s 18%, not my 12.75%
The tear sheet puts the gross interest rate at LIBOR + 18.0%. A footnote on the same page defines it more narrowly: “the gross interest rate is 18.0% but if Libor is greater than .5% then the gross interest rate is 18.0% plus the excess Libor above .5%.” My 12.75% came out of that, alongside a 2.25% management fee for YieldStreet and a 3.0% servicing fee for Counsel Financial — three numbers that add to exactly 18.0%.
I spent a long time treating 12.75% as the deal’s risk signal. It wasn’t — it was the retail slice of the risk signal. The market price of this borrower’s credit was 18% compounding monthly, set by a specialist lender that had known the firm since 2002. Nobody pays that for working capital because they have good options.
The borrower’s cost did not even stay at 18%. Because the management fee was defined as “a fixed rate of 2.25% and whatever excess Libor rate is received,” the gross rate rose with LIBOR — and every basis point of that increase went to the platform, while my 12.75% stayed fixed. YieldStreet confirmed the size of it later, and to their credit did so voluntarily: in September 2022 they announced they were cutting their fee on this deal to 1.5%, noting that the contractual variable rate “in the current interest rate environment would exceed 4%.”
2. The money didn’t go to the law firm
Counsel Financial had a $64M line of credit outstanding to this borrower. It wanted to lend more but had hit its own internal limit on exposure to a single client. YieldStreet’s $15M term loan went to paying down that line. The memorandum lays it out in a table:
| Before | Amount | After | Amount |
|---|---|---|---|
| Counsel Financial line of credit | $64.0M | YieldStreet SPV senior loan | $15.0M |
| — | — | Counsel Financial, now junior | $49.0M |
| Total debt | $64.0M | Total debt | $64.0M |
Total debt before: $64M. Total debt after: $64M. The borrower received no new capital from this transaction. What changed is who held which slice — and the party that changed places was the one with the best information. Counsel Financial had been in this credit since 2002, had collected roughly $92M from the borrower over that period, and reviewed every single case rather than sampling. That lender took $15M of cash off the table and moved the remainder into a junior position behind retail investors who had a page-one tear sheet and a few minutes to decide.
The memorandum frames this as a benefit, and there is a fair version of that argument: having $49M sit behind me was real protection, and Counsel Financial genuinely accepted worse terms on what it kept. But it also cut its exposure to this borrower by 23% and was paid in cash for the slice it gave up. I read the subordination in 2018 and felt protected. I didn’t ask why the lender who knew this credit best wanted less of it and was happy for me to have more.
3. Twenty-two cases, one bet
“Backed by 22 cases” reads like diversification. Here is how the collateral value was actually distributed, from the memorandum’s own projection tables:
| Collateral group | Cases | Low projection | Share |
|---|---|---|---|
| Mutual fund advisory-fee cases (A–J) | 10 | $237.8M | 94% |
| Everything else (K–V), mostly antitrust | 12 | $16.1M | 6% |
Ten of the 22 cases carried 94% of the low-side collateral value, and all ten rested on the same legal theory. The memorandum states it plainly:
“[T]hese financial companies are alleged to have charged investment advisory fees for their own mutual funds at a rate significantly higher than they charged, for the same or similar sub-advisory services, to other mutual funds or managed accounts.”
The memo’s own illustration is a manager collecting 0.5% from its own branded fund and 0.25% from an unaffiliated one for the same work. That is one argument, brought ten times. If courts rejected it, they would not reject it in one case — they would reject it in all of them at once.
The memorandum makes its own case for that bet, and part of it is strong. YieldStreet called the theory “compelling and based on a sound theory of law,” had it reviewed by an outside expert, and noted that the firm had already survived multiple dispositive motions, with trial scheduled in several cases. Then, on page seven, comes the other half — in bold in the original:
“[T]he specific theory of law and approach to these mutual fund claims is yet to be proven and therefore the expected settlement amounts and resulting revenues, as estimated by the Borrower and Originator, are subject to a high degree of uncertainty. As such, YieldStreet has limited the Loan amount to $15M while tightly structuring its terms to account for the risk associated with these mutual fund claims.”
So here is the arithmetic worth doing. Strip the mutual fund cases out and the same $15M loan sits against $16.1M of remaining collateral: a loan-to-value of 93%, not 5.9%. The headline ratio was never a measure of how well secured I was. It was a measure of how confident the borrower and its existing lender were in a theory that had yet to be proven.
This is the part I can’t blame on anyone else. The tear sheet gave me a loan-to-value of 4.3%–5.9%, the body of the memo told me that range rested on an unproven theory, and I anchored on the tear sheet. Nobody hid it from me. It was just further from the number I wanted to believe than I bothered to read.
Averages on page one, losses on page six
The three findings above all come from reading the body against page one. This last one is on page one itself, in two lines: the borrower’s “Historic Three-Year Average Total Revenue” of $14.3M and its “Historic Three-Year Average EBITDA” of $8.5M. Both positive. Both averages. Page six prints the years they averaged:
| FY 2015 | FY 2016 | YTD 9/30/17 | Page-one average | |
|---|---|---|---|---|
| Total revenue | $2.2M | $12.1M | $28.6M | $14.3M |
| Interest expense | $2.5M | $20.5M | $12.0M | not shown |
| Expenses net of interest | $6.5M | $6.2M | $5.4M | not shown |
| Net income (loss) | ($6.8M) | ($14.7M) | $11.2M | not shown |
| EBITDA | ($3.2M) | $5.5M | $23.2M | $8.5M |
Two consecutive years of large losses, then one profitable nine-month stretch — which the memorandum itself attributes to a single antitrust settlement in 2017 that “accounts for most of the increase in 2017 revenue.” Both of the averages on page one rest on those nine months. Averaging revenue and EBITDA specifically, while leaving net income unaveraged, turns that record into a headline number that reads as steady profitability: a business healthy enough to carry $64M of debt without trouble. Net income — negative in both full years — is the line that would have said otherwise, and it is the one line nobody averaged.
The memorandum has a partial explanation, and it is fair as far as it goes. Contingency-fee revenue is lumpy by nature, because expenses accrue continuously while revenue only arrives when a case settles. The 2016 loss was worse than the underlying business because the borrower carried a $20.5M interest expense into that year on purpose, “to reduce what would have been an income generating taxable event.” Strip interest out of the expense table and the firm’s core costs were steady — $6.5M, $6.2M and $5.4M across the three periods, “consistent and in line with its 2018 budget.” None of that changes the number that matters: $64M of debt, compounding monthly at 18%, against a firm whose three-year average revenue was $14.3M. Steady running costs are no help when the debt is more than four times revenue and the interest compounds faster than the cases can settle.
The Number That Appeared Once
Fourteen months into a loan paying event-based distributions, I had received nothing. The June 2019 investor notice opens by conceding it: “no payments have been made to date as there have been no case settlements in the underlying collateral portfolio.”
That same notice contains the most useful paragraph YieldStreet ever wrote about this deal. They marked the case portfolio down and kept the marked-down total as the headline. Then, alongside it, they printed a second ratio that left the mutual fund cases out altogether — and called that one conservative:
| Source | Low projection all cases |
Headline LTV | What else it said |
|---|---|---|---|
| Mar 2018 memorandum |
$253.8M | 5.9% | 94% of that value sits in the ten mutual fund cases, whose legal theory is “yet to be proven” |
| Jun 2019 notice data as of Apr 30 |
$224.3M | 6.7% | Mutual fund cases marked down to $139.8M. Excluding them, collateral is $84.5M — a “conservative LTV of 17.8%” |
| Sep 2019 update data as of Aug 31 |
$228.2M | 6.6% | “To be conservative, we only focus on the low side case projections.” The 17.8% figure does not appear |
Leaving those cases out was the right way to be conservative about them; marking them down was not. Ten cases resting on one legal theory do not fail proportionally — they rise or fall together, on the same rulings. A partial markdown prices an outcome that could not happen. That is why the 17.8% is the figure that matters: three times the ratio on the tear sheet, and the only number in the entire file that reflects the deal’s real concentration. Even so, it arrives as comfort rather than warning — the notice calls it a ratio “assuring adequate coverage.”
The 17.8% needs one qualification, and it makes the number look worse, not better. The underlying collateral had changed: the same notice explains that “the updated projection includes five additional cases that the Borrower added to the case collateral.” In March 2018 the non-mutual-fund cases were worth $16.1M on the low side — a loan-to-value of 93%, not 17.8%. And newly added litigation joins at the back of the settlement queue, not the front. Coverage improved because the pool got bigger, not because anything had arrived that could pay before the loan came due.
Three months later, “conservative” is redefined silently:
- June: “to be conservative, we only focus on the low side revenue projections of the 16 non-mutual fund cases” — conservative means excluding the mutual fund cases. LTV: 17.8%.
- September: “To be conservative, we only focus on the low side case projections” — conservative now just means using the low side, which every prior figure already had. LTV: 6.6%.
The 17.8% is never mentioned again in any document I received over the following seven years. Both figures are real — arithmetically correct and disclosed. What changed between them was the label, not the number: one was informative, the other decorative, and the informative one had a single appearance.
Meanwhile the loan was getting bigger
Interest on this loan compounded monthly — a structure YieldStreet described as “relatively rare” on their platform. That detail is easy to skim past and it is the mechanical heart of what went wrong. At the 18% gross rate, a $15M balance compounding monthly grows to roughly $21M by the original March 2020 maturity and about $26M by the extended March 2021 date — and the real figure was higher, because LIBOR sat above the 0.5% threshold throughout and the excess was added to the rate. Even YieldStreet’s own reporting lost track of it: a July 2019 notice told investors that accrued interest had been counted twice, once on its own line and again inside “Principal Outstanding,” so the value shown for the investment had been reading high and was about to drop.
Now put that next to what the September 2019 update actually promised. Of the $228.2M in projected revenue, only $57.8M was expected to settle by the end of 2020 — and of that, only $25M came from non-mutual-fund cases. So on the conservative basis YieldStreet themselves had used three months earlier, the entire realistic collection pipeline through 2020 was roughly equal to what the loan alone would owe at maturity. Before the $49M junior line. Before the firm’s payroll.
That is the whole failure, and it was legible in September 2019 — 6 months before the original maturity, 18 before the extended one. Not a lien problem. An arithmetic problem. COVID gets the blame in a lot of the later updates, and court closures did delay resolutions. But the cases carrying the loan had already been written down 41% by June 2019, the firm had lost money in 2015 and 2016, and the loan was short of coverage on YieldStreet’s own conservative measure six months before the pandemic. COVID determined when the failure happened. It did not determine that there would be one.
The Restructuring Trap
The full timeline looks like this:
| Date | Event |
|---|---|
| Mar 2018 | YieldStreet raises $15M in minutes; I invest $250,000 |
| Apr 2018 | My $250,000 funds on Apr 10; interest accrues at 12.75%, compounding monthly |
| Jun 2019 | 14 months in, zero distributions. Collateral marked down; YieldStreet discloses a “conservative LTV of 17.8%” excluding the mutual fund cases |
| Jul 2019 | Founders disclose that accrued interest had been double-counted in “Principal Outstanding”; my reported position value is revised down |
| Sep 2019 | First distribution — $5,182, 17 months after funding. The 17.8% figure disappears from reporting |
| Dec 2019 | Second distribution — $21,386 (interest) |
| Mar 2020 | Original loan maturity — borrower fails to repay |
| 2020 | COVID-19 disrupts courts; case dockets backlog |
| Mar 2021 | Target investor maturity — still in workout |
| 2020–2022 | YieldStreet makes $4.2M in “protective advances” to keep firm operating — ranking ahead of investor principal |
| Oct 2021 | First money out of the borrower in a year — just under $440,000 — repays advances, not investors |
| Jun 2022 | RESTRUCTURING — ~50% principal returned; YS SPV becomes junior to new senior lender |
| Jul 2022 | Major distribution — $125,531 (partial principal return) |
| Sep 2022 | Small distribution — $3,912 (case settlement) |
| May 2023 | Small distribution — $3,991 (case settlement) |
| Oct 2023 | Investment re-designated to “Default” status |
| Q4 2023 | Borrower’s cash dwindles to <$750,000; considering bankruptcy |
| Q1 2024 | Borrower’s largest case dismissed; firm down to 5 attorneys |
| Q1 2024 | YieldStreet sells junior note to senior lender |
| Sep 2024 | Last distribution — $2,276. Promised for “mid-May”; arrived four months late |
| Jul 2026 | The $250,000 of case-settlement proceeds arrives — and is held in reserve, not distributed. Only remaining claim: the life insurance policies |
In June 2022, after more than two years of default and $4.2M in protective advances, YieldStreet closed a refinancing with a third-party lender:
- Investors received about $7.5M — roughly 50% of outstanding principal — net of management fees, protective advances, out-of-pocket expenses and fund expenses
- The SPV’s remaining position became a $12M subordinated three-year promissory note, entitled to 7.5% of the firm’s future collections
- The SPV was now junior to the new senior lender, and its relationship with Counsel Financial ended
- Interest on the new note no longer compounded
That last item is where the accounting catches up with reality. Work the monthly compounding forward at the gross rate and the contractual claim by mid-2022 was north of $30M on a $15M advance. What the SPV actually accepted was $7.5M in cash plus a $12M note — a combined claim of about $19.5M. So the trade cost more than $12M of accrued interest, and that write-off showed up in investor portfolios in August 2022 as an unexplained drop in balance. YieldStreet had to send a follow-up note explaining that the money hadn’t gone anywhere; the number had simply been wrong before. That is the second time the reported value of this position was revised downward after the fact.
It is tempting to read the restructuring itself as the failure: senior secured first lien became junior subordinated debt, and the protection that made the deal attractive was traded away to get half the principal out. But my position had been junior in practice long before June 2022. The $4.2M in advances ranked ahead of my principal, so when the first real money in a year arrived in October 2021 — just under $440,000 — it went to repay them, not me. I held the senior claim and stood behind the rescue financing that existed only because my senior claim was failing.
The seniority had been failing for two years by then, and for a different reason. The memorandum’s default waterfall is genuinely well drafted — while the loan performs, the junior line takes interest pari passu; the moment the loan defaults, the junior line is entitled to nothing and 100% of borrower revenue is redirected to the senior loan. After the March 2020 default none of it mattered, because a waterfall only allocates water.
By late 2023 the new senior balance had grown to $32.7M through further protective advances, and the SPV’s $12M junior note sat behind all of it, entitled to 7.5% of collections that were no longer arriving. YieldStreet said as much in February 2024: the junior position, together with the subordination to those advances, had “diminished the likelihood of any substantial recoveries.”
Selling the Junior Note
By early 2024, the law firm was in terminal decline:
- Down to just five attorneys, two paralegals, and two support staff
- Cash on hand below $750,000 — couldn’t make payroll
- Largest case dismissed by the court
- Two years of projected negative cash flows
- Actively engaging bankruptcy counsel
YieldStreet made a pragmatic decision: sell the junior note to the senior lender rather than fight for scraps in bankruptcy. The deal:
| Component | Value |
|---|---|
| Cash upfront | $350,000 |
| Accelerated case settlement payments | $250,000 |
| Life insurance rights — 50% of policy 1, 25% of policy 3 | TBD, 8 years to run |
What that meant for me: $162,277 back on $250,000 invested — a 64.9% recovery, and a 35% loss.
I want to be fair about this decision, because it is the one part of the story where I think YieldStreet acted well. Selling a junior note in a firm that was interviewing bankruptcy counsel, with its largest case dismissed and nothing ahead of it but negative cash flow, was the right call. A junior lien in a law-firm bankruptcy is a claim on nothing that costs money to assert. That was YieldStreet’s own reasoning, on the record before the deal closed: the February 2024 update said an exit would “provide some upfront cash recovery value while reducing the risk of having to pursue recovery in a potential protracted and messy bankruptcy for a junior lien position.” They took $350,000 in cash, $250,000 in contingent settlement rights, and a share of the life insurance, and stopped the bleeding. Had they fought, I would have gotten less. Competent workout management does not undo bad underwriting, but this was competent workout management.
The K-1 Confirms the Loss
My 2024 K-1 from YS CF LawFF VII LLC shows:
- Current year net income (loss): ($66,112)
- Beginning capital account: $121,301
- Withdrawals and distributions: ($2,276)
- Ending capital account: $52,913
The figures reconcile exactly: $121,301 less the $66,112 loss less the $2,276 distribution leaves $52,913. That write-off is my share of the SPV’s losses on the sale of the junior note.
The $52,913 still sitting there is worth understanding for what it is, because a capital account is a tax construct and not a valuation. It is not $52,913 of value waiting for me. It is the unrecovered basis I am still carrying against a partial claim on two life insurance policies, and the most likely path for that number is that it becomes another loss rather than a distribution. My real expectation for this position is zero; the K-1 just hasn’t caught up yet.
Where It Stands
This investment is not closed. Ninety-nine months after funding, the SPV is still open, against a 36-month term.
The latest update reports that the platform has now received the $250,000 of case-settlement proceeds that formed part of the 2024 sale price. And then this:
“The proceeds are being held in reserve to reserve for costs and to protect against any unforeseen expenses.”
So the money arrived and stopped. I don’t think that is improper — an SPV that has to stay open for years waiting on insurance policies carries real audit, tax and administrative costs, and those have to be funded from somewhere. But it is worth naming plainly, because it is the shape of the entire investment in miniature: a recovery event occurred, and the cash went to keeping the vehicle that holds my claim alive rather than to me. There have been no distributions since September 2024.
What is left is the life insurance, and only that: the July 2026 update says the platform “is not entitled to any additional proceeds other than the life insurance policies, if they are triggered.” There were three policies, one on each managing partner, and the sale swapped the SPV’s pro-rata share of all three for 50% of the first and 25% of the third — nothing of the second. They had eight years left to run in 2024, so roughly six now, with the senior lender paying the premiums and confirming they remain active. The only estimate anyone has published is $7.5M if every policy pays, and it dates from February 2024 — before the sale changed what the SPV was entitled to. It has not been restated since. Reporting has thinned to match: quarterly through 2023, as-needed after the sale, and now a promise of “at least once per annum.”
I’m not counting on it. Waiting for someone to die is not an investment thesis, and my capital account has been written down to reflect that. The remaining balance is an option on mortality that I did not intend to buy and cannot sell.
Why “Senior Secured” Didn’t Protect Me
1. Priority allocates cash flow; it doesn’t create it. This is the one that took me years to see, and it generalises past this deal. First lien, the UCC-1 filing, the waterfall that cut the junior lender out entirely — every protection in the structure described how money would be divided, and none of them caused money to exist. Against collateral made of unresolved lawsuits, first priority is a claim on a stream that may simply not arrive. Seniority is insurance against a competing creditor. It is not insurance against an asset that hasn’t happened yet.
2. Seniority is a negotiating position, not a fact. When rescue capital is the only alternative to bankruptcy, whoever supplies it sets the terms and the incumbent lender chooses between subordinating and getting nothing. YieldStreet took junior status because the alternative was foreclosing on pending litigation, which the January 2022 update candidly calls impractical given “the nature of the collateral.” A lien you cannot realistically enforce is a lien you will eventually trade.
3. The rescue money I was told would come from someone else came from the platform, and outranked me. The memorandum was explicit: “The Borrower’s operational expenses will be funded by the LOC and the Originator has agreed to continue funding as long as necessary unless YieldStreet SPV agrees otherwise.” Counsel Financial was contractually the one keeping the firm alive — subject to a carve-out that let the platform release it from that duty. In the event it funded roughly $1M of the $4.2M and YieldStreet funded the rest. So the commitment meant to insulate the senior loan from the borrower’s operating costs didn’t hold, and the money that replaced it was senior to mine.
4. Personal guarantees track the firm, not the person. Three principals guaranteed the loan. Their capacity to pay came from the same case portfolio securing it, so the guarantee and the collateral failed together. A guarantee only spreads risk if the guarantor’s wealth is independent of the borrower’s, and a law firm partner’s wealth is the firm.
5. The originator’s track record was measuring something else. Counsel Financial had defaults of 0.4% across the $760M it advanced after 2009. That statistic is real, and it was worth nothing to me, because it described revolving lines that Counsel Financial could monitor, resize and exit at will. I held one fixed-maturity term loan to one borrower with no ability to do any of those things. I borrowed a diversified lender’s default rate and applied it to a concentrated position.
The Date, Not the Lien
The seniority failure was real and it came at the end. What went wrong first was a date. Payments here were event-based, arriving as cases settled, which is the right design for litigation collateral. Two things were not event-based: the balance compounded every month, and the principal came due on March 27, 2020. Roughly $21M owed on a day certain, out of cases that answer to court calendars. When they had not resolved, the firm was not short of a number it could grow into — it owed money on a date its assets could not meet, and the protective advances, the restructuring and the subordination were all just the negotiation over who ate the gap. Priority is what you argue about afterwards. The date is what creates the afterwards, and I have since watched an entire funder die of the same mismatch.
The part that was entirely mine is the size. At $250,000 this was my largest single alternative position, and I sized it that way because the pitch was clean and the allocation was gone in minutes, which is FOMO dressed as conviction. The platform built the trap; I chose how much to put in it. That choice is what turned one bad deal into a portfolio-level event, dragging my overall YieldStreet IRR to -2.62% and erasing both of my winners there — the pre-settlement pool that paid 13.14% and the merchant cash advance portfolio that paid 12.82% — with room to spare.
Eight years later, when the same platform under a new name put another deal in front of me, I read all 27 pages of the note supplement before I read the marketing. That is what $87,723 bought.
Commentary and personal experience — not investment, legal, or tax advice. Investing carries risk, including total loss of capital. Always do your own due diligence.
