FIGR — Earnings Review (Saul)

Date: 2026-02-22 Quarter: Q3 FY25 (reported) + Q4 FY25 (preliminary) Task type: Earnings Review Atlas baseline: Read and used


Verdict

The numbers are, frankly, preposterous in the best possible way. Revenue accelerating from 12% to 31% to 55% to 91% YoY over four consecutive quarters. Marketplace volume growing at 131% — faster than revenue, which means the platform is building head of steam. EBITDA margins of 51-55% while growing at 91%. I have been investing for decades and I can count on one hand the number of companies that grow at 90%+ while generating 50%+ EBITDA margins.

But I am not ready to call this a conviction hold. Here is why.

Thesis status: Intriguing but Unproven. Hold at current ~5% weight. Watch Feb 26 call carefully.


What the Numbers Say

The revenue trajectory is what drew me to this company, and it has not disappointed:

Q1'25 Q2'25 Q3'25 Q4'25 (prelim)
Revenue ($m) 84.5 106.1 156.4 ~160
YoY % 12.3% 31.4% 54.8% 90.8%
QoQ % +0.8% +25.5% +47.4% +2.3%

Revenue growth accelerating at scale — from 84Mtoarunrateof 640M annualized. That is the single most important thing I look for. I have seen it in Axon, in AppLovin, in Samsara. When it shows up like this, you pay attention.

The leading indicator that makes me even more excited: Marketplace Volume is growing at 131% YoY — faster than revenue. That is how platform businesses work. Volume today is revenue tomorrow. Figure Connect went from literally zero to 26% of revenue in nine months. Ecosystem & Tech Fees went from $7.3M to $35.7M YoY — nearly 5x. That is the highest-quality revenue stream in the company, and it is exploding.

EBITDA margins: 44.9% → 20.2% → 55.4% → 51.0% (Q3'24, Q4'24, Q3'25, Q4'25). The Q4'24 dip was real — gains on loan sales collapsed from $57M to $24M in a single quarter. That is the lumpy part of this business. But Q3-Q4 FY25 at 51-55% is outstanding. The company went from (5.6%) adj EBITDA margin in FY23 to nearly 50% in FY25. That kind of margin trajectory is extraordinary.

Revenue Quality is Improving

This is what separates a maturing platform from a commodity lender:

Stream Q3'24 Q3'25 Growth
Ecosystem & Tech Fees $7.3M $35.7M +388%
Gains on Sale of Loans $57.4M $63.6M +11%
Origination Fees $18.9M $21.4M +13%
Interest Income $12.8M $17.9M +40%
Servicing Fees $6.5M $7.9M +22%

The transaction-based revenue (Gains on Sale of Loans) is growing slowly. The platform revenue (Ecosystem & Tech Fees) is growing at 388% YoY. That shift in mix is exactly what I want to see. It means the business is becoming stickier, more recurring, less dependent on transaction volume in any single quarter.


What Keeps Me Up at Night

Customer Concentration — This Is Serious

Top 2 customers = 76% of UPB purchased. Let me say that again: two customers account for three-quarters of the business. This is the single biggest risk in the entire thesis. I care deeply about companies being masters of their own fate. A company where two customers can walk away and devastate the economics is not master of its fate.

Now, I understand that large financial institutions naturally dominate early in a marketplace's lifecycle. The 246 → 200 active partner decline quarter-over-quarter needs explanation. Is it seasonal? Is it consolidation? Or is the platform less attractive to smaller participants than it appears? The Feb 26 call must address this.

No Transcript — I Cannot Hear Management

This drives me crazy. I read every earnings call transcript myself. Every time. I want to hear how management handles tough questions, what they emphasize, what they dodge. For FIGR I have no transcript. I have only the press release. For a company this early in its public life, that is a meaningful gap. I am flying partially blind.

SBC of $40M in Q4 — Is This a Pattern?

$40M of stock-based compensation in a single quarter — 64% of the full-year total. If this is one-time IPO grants, fine. If this is a recurring run rate, then GAAP earnings are going to disappoint meaningfully and EBITDA becomes misleading. The Feb 26 call must address SBC normalization.

Mike Cagney — I Cannot Ignore This

Cagney co-founded SoFi and built it into a real business. He also left under a cloud of sexual harassment allegations in 2017. He is now Executive Chairman of Figure. I have no idea what he does day-to-day or how much influence he exerts. But I have seen reputational risk erupt without warning. This is in the back of my mind.

No Guidance — Cannot Track Management Credibility

One of the most important things I do is hold management accountable to their stated goals. FIGR has provided zero forward guidance. I understand this is common for new public companies, but it means I have no track record to evaluate. Are they growing as they expected? Are they surprised by the rate of adoption? I don't know.


The Platform Story — Real or Hype?

This is the key question, and I think it is real. Let me explain why.

Figure built Provenance Blockchain as a settlement layer for financial assets. They tokenize HELOC loans on it. Lenders, servicers, and investors all transact on the same chain — eliminating the reconciliation friction that makes traditional capital markets expensive. The 75% market share in RWA tokenization is not marketing; it's the result of being first and being right.

Figure Connect — the secondary market for these tokenized loans — went from zero in early 2025 to $1.13B in volume by Q3. YLDS, their stablecoin, hit $376M in circulation and is growing 15% per month. These are not vanity metrics. YLDS earning yield on reserves while circulating as a settlement medium is a real business model.

DART (their AI credit model) has 91% adoption among marketplace participants. When 91% of your network is running on your proprietary credit model, you have deep switching costs. Nobody is ripping out Figure's rails and rebuilding elsewhere.

Is this comparable to a SaaS company with NRR of 120%? No. The revenue streams are lumpier, more transaction-dependent, more rate-sensitive. But the platform dynamics are real: network effects, high switching costs, expanding product surface. These are characteristics I recognize.


Revenue Durability Assessment

For a non-SaaS company I ask: is this growth secular or cyclical? Durable or fragile?

Arguments for durability:

Arguments against:

My assessment: Revenue durability is partially proven and improving as platform mix increases. Not yet at SaaS-level confidence, but better than a pure origination lender.


Prior Beliefs / Updated Beliefs

Topic Prior Belief Updated Belief
Revenue trajectory Unknown (first analysis) Accelerating sharply — best in portfolio on growth-adjusted basis
Platform reality Skeptical — blockchain lending sounds like marketing More convinced — Provenance, DART, Figure Connect are real rails
Customer concentration Not tracked Serious risk — 76% top-2. Must improve.
SBC Unknown Concerning — $40M in Q4. Need run-rate clarity.
Management No track record Cagney history is real. Tannenbaum untested publicly.
Valuation Appeared cheap Confirmed cheap — 7.5x TTM rev on 91% growth is exceptional

What I'm Watching at the Feb 26 Call

  1. SBC clarification. Is the $40M one-time or recurring? If recurring at $160M annualized, GAAP earnings are roughly zero.
  2. Partner count decline. 246 → 200 active partners. What happened? Seasonal? Quality-over-quantity? Or attrition?
  3. Customer concentration trend. Is the top-2 concentration improving? What is the roadmap to <50%?
  4. Figure Connect growth trajectory. What were Q4 participants? Is 33 growing?
  5. FY26 guidance. If management provides targets, that is a major catalyst. First opportunity to establish credibility.
  6. Rate environment commentary. Management's view on HELOC demand in 2026. Is origination volume holding?

The Bottom Line

I am holding FIGR at ~5%. I am not adding until the Feb 26 call clarifies the SBC question and the partner concentration trend. If the call is strong — SBC is one-time, partner decline is seasonal, Figure Connect is accelerating, and they give FY26 guidance — I would seriously consider moving this to 8-10%.

The growth numbers are exceptional. The platform story is real. The valuation is among the cheapest in the portfolio on a growth-adjusted basis. But customer concentration at 76% and an untested management team at a company with only two quarters of public history demand patience.

Tell me again that growth investing doesn't work when a company grows 91% YoY at 50%+ EBITDA margins and trades at 7.5x revenue. But I follow the money and the results — and the Feb 26 call is where I will learn whether this story has legs beyond the extraordinary numbers.


Action

Hold at ~5%. Watch Feb 26 call. Decision point for add or trim based on SBC, partner concentration, and guidance.

Best, Saul


Sources: Scout brief (2026-02-21), SEC EDGAR 10-Q Q3 FY25 (accession 0002064124-25-000032), Figure IR press releases (Q3 and Q4 preliminary), Atlas earnings review (2026-02-21)