The Upside-Down Math of Growth Financing

Grab a beer and settle in. Today, we're going to tackle a bit of "upside-down" math that for some reason seems to be tripping up even the pros in finance.🤷‍♂️

It’s a classic head-scratcher: how can a loan with a 16% return be cheaper for a company than one with a 9.5% interest rate?

It sounds impossible, right? But as we'll see, when it comes to financing a company's growth, the headline interest rate is not the full story. The timeline and the way the money is paid back are what truly matter. This post was inspired by the many X posts from "experts."

For those who don’t know, Daniel is one of the more vocal Lemonade bears. Once a hedge fund manager… now an ex–hedge fund manager. After years of lagging the S&P 500 with ~9% average returns, the fund eventually shut down after heavy client losses. 😢 I don’t wish that on anyone, but when I looked him up it didn’t come as a massive surprise given the level of due diligence on display.🤦

My genuine hope and goal is not to throw shade at him or any others - and the purpose of this is to explain very simply why they should spend more time really looking at $LMND's capital structures and underlying unit economics vs spreading FUD on X.

Which brings us back to the reason for the post - Lemonades Synthetic Agents - you can see his view below. I'll try and explain as simply as I can what it is and compare it to an alternative financing setup that is also common in insurance, a surplus note.

For years I could never quite understand why he threw so much shade at $LMND. He would harp on about concepts that are frankly basic for a trained hedge fund analyst, and even when shown to be empirically wrong, he would double down loudly.

I often wondered whether he was shorting $LMND or simply had a personal issue with the leadership. Since he now runs a family office (which isn’t required to disclose assets the way larger funds do), I had never seen any clear disclosures from him on X or anywhere else about his positions. Then I came across this video.... and suddenly a lot of his posts made sense to me!

I will not comment directly, I will let you come to your own conclusions as to why he may be throwing shade while holding significant allocations in two competing InsurTechs. For the record, I am not saying anything negative about Root or Hippo. I actually hold some medium term calls in Root and I am evaluating whether to start a position in Hippo.

Google Trends Data - 31st Aug 2025 - (click for live)

Anyway, enough about him. Let us turn to what really matters: his posts and comments - which in my opinion are off the mark, misinformed, and at this point feel more like spreading FUD than analysis.

Because Daniel is such a vocal bear of $LMND and, in his own words, one of the largest shareholders of Hippo, I thought Hippo made the perfect comparative model. They chose a surplus note to raise external capital. So let us break down what both structures are, how they work, and the strengths and weaknesses of each.

Hippo Q2 2025 Shareholder Letter

The Two Financing Tools

Imagine a company wants $50 million to find new customers. Here are two ways they could get it:

1. The Surplus Note

Think of this like a special kind of interest-only mortgage.

• The company borrows $50 million.

• It pays only the interest every six months for years (in Hippo's case, seven years).

• The original $50 million principal isn't paid down during this time.

• Crucially, the company can't decide to pay it back early. It's locked in for a long time.

This type of loan is useful for insurance companies because it helps them meet regulatory capital requirements, but for our purposes, just think of it as a long, rigid loan.

Hippo pays a headline price of 9.5%

2. Synthetic Agents:

Think of this like a "Record Label Advance" - This is a much more modern and flexible approach.

• An investor gives the company $50 million to spend on marketing for a specific group (or "cohort") of new customers.

• The investor is then repaid from a small slice of the revenue generated only by that specific group of customers.

• Once the investor hits their agreed-upon return (say, 16%), the payments stop completely.

• The company then keeps 100% of all future profits from those customers. If that group of customers doesn't perform well, the investor's returns suffer, not the company's entire balance sheet.

  
  

Lemonade Pay a headline rate of 16%

On the surface, it's easy to come to the conclusion that 16% is more than 9% - but like most things - you can't make financial decisions using surface level observations, so let's dig it !

Why a Bigger Rate Can Cost Less 💰

So, why can the 16% "record deal" be cheaper than the 9.5% "mortgage"? It boils down to three simple reasons:

1. Duration: The 16% loan is paid back quickly (usually 18-24 months) and then it’s over. The 9.5% loan keeps charging interest year after year after year. A short-term cost, even at a higher rate, is often much less painful than a lower rate that lingers for the better part of a decade.

   
   

2. Recourse: The "record deal" is only paid back from the album's sales (the specific customer cohort). If the album flops, the record label takes the hit. The "mortgage," however, must be paid back no matter what, using money from anywhere in the company. It's a much bigger burden.

   
   

3. Cash Flow Matching: Payments for the "record deal" come directly from the money that investment generated. With the "mortgage," the interest bill comes due every six months, whether the company is having a good or bad half.

To demonstrate a single Q2 2025 view, the model below loads Hippo’s surplus note and Lemonade’s Synthetic Agents. Hippo runs on a seven year horizon, which is the earliest maturity redemption. Lemonade uses a 26 month loan style proxy at a 16% investor IRR, ADR set to 100% inside the proxy, and 100% of growth spend fronted so initial funding shows as $50m for like for like. There is little to change unless you want to stress test. Nothing here is financial advice, I’m just an Ironic Ape.

(Skip the next two paragraphs if you don't care how I arrived at the numbers in the tool... it's not the easiest of reads... but if you do like the detail, feel free to read on... it did after all, take nearly 14 hours today. 4am right now! so forgive me if you spot any errors!)

Full transparency: I reviewed six quarters, identified growth spend each quarter, and estimated the portion financed by General Catalyst. Because CAC and premium differ by line and the mix shifts, I apportioned growth spend by line using observed LOB growth QoQ. That yields an estimated blended CAC of $187 and a 26 month repayment profile for the Q2 2025 cohort (up from previous years where it was closer to 18 months when I last did this).

For clarity, the 26 month profile is a loan style proxy to estimate dollar cost of capital. Inside this proxy ADR is 100% and cohorts are assumed to repay in full, which isolates financing cost. CAC uses gross adds derived from IFP and ADR, so attrition is already in the add count even though ADR is not applied inside the proxy. In reality SA remits from cohort premiums and stops when the investor reaches a 16% IRR, so timing can differ. The proxy gives a clean Q2 2025 comparison against Hippo in my opinion. The short version is... you shouldn't need to touch anything, some quarters it may show 18 months to pay back, others 26+. This correlates to Q2 2025.

If you want to know more about the actual capital structures and background of the S.A, programme you can hear about it here from General Catalyst themselves:

So Let's Do the Math Together: A Side-by-Side Showdown

To compare apples to apples, let us assume both financing methods are used to fund $50 million of marketing growth spend, which is expected to generate $150 million in gross profit over the customers’ lifetime. Both sides have the same customer retention rate and the same LTV to CAC ratio (I cover LTV:CAC in more detail in my other articles and tools).

Anyways - back to it - here’s what our scenario shows after seven years:

That’s not a typo. By using the "more expensive" 16% financing, the company ends up with over $27.3 million more in net value....that's more than 42% higher....from the same $50 million investment. The short duration and fixed value makes all the difference.

The Break-Even Point

Just to hammer this home, how low would the surplus note interest rate need to be to match the total cost of the 16% Synthetic Agents program over 7 years?

The answer is shockingly low: around 1.89%. This is the power of "upside-down" math. A short-term, high-rate financing structure can be vastly more efficient than a long-term, low-rate one..... hardly "CCC-priced financing".

Synthetic Agents are short‑duration, self‑amortising, cohort‑level and non‑recourse. The investor is repaid from a single cohort’s premiums and stops getting paid when a capped IRR is reached. There is no long‑dated, unsecured corporate claim that a rating agency would label CCC. Think revenue‑share financing or asset‑backed style cash flows, not a 7 to 10 year junk bond!

So, Why Would Anyone Use a Surplus Note?

This doesn't mean surplus notes are "bad." They are a different tool for a different job. For a regulated insurance company, they are great for boosting their "surplus capital," which allows them to grow and provides a safety cushion. It’s a strategic tool for managing regulations and ratings, not just for funding marketing.

Think of it this way: if you need to fix a leaky pipe, you grab a wrench, that is the surplus note. If you want to build a deck, you need a saw and drill, that is Synthetic Agents. It is all about picking the right tool for the job, surplus notes could be used for growth spend or maintaining capital adequacy, Synthetic Agent capital cannot be counted as regulated surplus. (Surplus is the buffer insurers must hold back, not spend, so they can still pay claims even in a major catastrophic event. If you are investing in insurance companies you need to understand capital adequacy and how it impacts both growth and unrestricted cash - but that is a whole different blog).

A Quick Retail Investor Checklist

When you see a company raising money, ask yourself these simple questions:

• How long will the financing costs run for?

• Can the company pay it back early if they want to?

• Who takes the loss if the investment does not work out - the company or the lender?

• Does this change the company’s reported metrics, or only the timing of its cash flows?

• Will it dilute you as a shareholder? (Meaning: will your slice of the pie shrink, making each share you own worth less)

An Ironic Apes Bottom line 

Don’t let a headline rate fool you! The shape and duration of cash flows determine the real cost of money. A 9.5% loan that drags on for seven plus years is often far more expensive than a 16% fixed program that burns off in 18 - 26 months.

If the objective is efficient growth, a modern cohort based structure almost always delivers more value to shareholders for the same dollar of spend!

ohhh…and one last thing... if you are going to throw shade on a CEO for using Adjusted metrics you sure as shit better check the companies your "a major shareholder in", because Hippo leads with Adjusted Net Income as its key metric. That’s not analysis, it’s hypocrisy. ✌️

Disclosure and disclaimer: IronicApe currently holds a long position in Lemonade, Inc. (NYSE: LMND). Our positions may change at any time without notice. Nothing in this material is investment, legal, accounting, or tax advice, and it should not be relied upon for any investment decision. The information is provided for educational and informational purposes only, is believed to be accurate at the time of writing, but is supplied on an as-is basis and may contain errors or omissions. Any forward-looking statements are based on assumptions that may change. Investing involves risk, including the risk of total loss. Do your own research, consider your personal circumstances, and consult a qualified adviser before acting.