Date: 2026-02-22 Quarter: Q3 FY25 (Sep-25) Market cap: ~$2.2B | EV/TTM Rev: ~2.5x | Revenue growth: 74% YoY First Bear analysis — no prior company-notes exist
I had no prior position in ARX and no written views going into Q3. This is my first systematic look. I'll apply my framework fresh.
What I expected (prior priors, framework-derived):
What the numbers show:
The numbers are better than I would have assumed going in. Updated.
| Q1'24 | Q2'24 | Q3'24 | Q4'24 | Q1'25 | Q2'25 | Q3'25 | |
|---|---|---|---|---|---|---|---|
| Mar-24 | Jun-24 | Sep-24 | Dec-24 | Mar-25 | Jun-25 | Sep-25 | |
| Revenue ($m) | 128.1 | 130.1 | 153.7 | 190.7 | 178.0 | 219.1 | 267.4 |
| YoY % | — | — | — | — | 38.9% | 68.4% | 74.0% |
| QoQ % | — | 1.6% | 18.1% | 24.1% | -6.7% | 23.1% | 22.0% |
| EBITDA ($m) [Adj] | 27.5 | 13.0 | 26.1 | 46.4 | 42.8 | 63.5 | 105.0 |
| EBITDA Margin [Adj] | 21.5% | 10.0% | 17.0% | 24.3% | 24.0% | 29.0% | 39.3% |
| EWP ($m) | 551 | 727 | 796 | 833 | 874 | 911 | 1,043 |
| Members | 170 | 186 | 204 | 217 | 232 | 248 | 265 |
| GLR % | 52.1% | 54.7% | 51.8% | — | 53.3% | 50.5% | 50.1% |
| NRR % | — | — | — | 146% | — | — | 135% |
| Take Rate % | — | — | — | 7.1% | — | — | 8.0% |
| 3P-DWP Mix % | — | — | — | — | — | 27% | 32% |
Note on Q3 EBITDA: $105m reported includes $39m one-time investment gain. Underlying = $66.3m (24.7% margin). I use the underlying figure for trend assessment. The 39.3% headline margin is not representative.
Quarterly revenue seasonality: Q1 typically soft (Q1'25 dipped -6.7% QoQ vs Q4). Management attributes this to reinsurance reset timing. Not alarming — Q2 and Q3 bounce back strongly.
The ARX thesis is a platform transition: shift from owning underwriting risk (balance-sheet-heavy, capital-intensive) to being an exchange where third-party insurers write the risk. ARX earns a fee on exchange volume.
Theory predicts: EWP grows as members and third-party capacity grows → ARX's fee revenue grows faster than balance-sheet commitments → capital efficiency improves → margins expand in the exchange segment even as the mix shift creates near-term noise.
Numbers check:
The numbers match the theory. That doesn't mean the theory is correct — it means it hasn't been falsified yet. Seven quarters of public data is thin. I need more.
FY26 guidance: EWP $5B+, 3P-DWP $2.1B, Adj EBITDA $269m.
Atlas describes this as "margin compression from 25% to 5.4%." I want to be precise here because sloppy framing matters.
The 5.4% is EBITDA as a percentage of EWP — not revenue. As more EWP flows through third-party insurers, a smaller percentage of EWP converts to ARX revenue (ARX keeps only the fee, not the full premium). So the denominator (EWP) grows faster than the numerator (EBITDA).
In revenue terms: Q3 underlying EBITDA run rate = $66.3m × 4 ≈ $265m annualized. FY26 EBITDA guide = $269m. That's roughly flat in absolute terms — on what will be significantly higher revenue. So yes, EBITDA margin as % of revenue is also compressing in FY26.
Here is my key question: is this the dip before a steeper climb, or a permanent reset?
Management's argument: the third-party model is asset-light. Once the mix shift is complete, ARX earns fee income on $5B+ of EWP with minimal capital at risk. The exchange economics should compound as volume grows. $269m EBITDA in FY26 could be $400m in FY28 if the model scales as designed.
My honest assessment: I don't know yet. The $269m FY26 EBITDA guide is real and large. But I want to see Q1'26 and Q2'26 numbers before committing to the "dip-then-ramp" narrative. The margin story requires trust in management's model, and I've had seven quarters to evaluate them, not twenty-seven.
Revenue growth (74% YoY, accelerating): A+. Accelerating growth in the third quarter of a company's public life is uncommon. The prior year comparables were not particularly easy. This is the most important signal.
NRR 135% (down from 146%): Still elite. 135% means existing members are growing meaningfully on the platform. But the directional decline matters. I want to understand whether this is mix-related (lower-NRR third-party members growing as share) or organic slowdown. Management did not give a clear answer. This is on my watch list.
GLR 50.1% (improving): Good. Sub-50% is the target. Trend is right.
Member growth +30% YoY: Solid. Network effects require continued new member onboarding. The pipeline is guided to remain active in FY26.
Valuation at ~2.5x EV/TTM Rev: At 74% growth, this is among the cheapest setups I've seen in high-growth companies. The market is pricing in the margin compression and doubting the FY26 ramp. If management delivers $269m EBITDA in FY26, the stock is cheap at any reasonable multiple. If they miss, the story breaks.
Jeff Radke (CEO) is measured and data-oriented on calls. The FY26 guidance is specific — EWP $5B+, 3P-DWP $2.1B, EBITDA $269m. Specific guidance is better than vague. He's not hiding behind ranges.
The IPO profits interest charge (-$1.37B GAAP loss) was non-cash, non-recurring, and fully disclosed. Management's explanation was clear. I'm not penalizing for this.
Promise tracking — this is their first specific FY26 guidance. I'll track it rigorously.
Q4 guidance specifics:
The Q4 EBITDA step-down from underlying Q3 to guide is something to watch. They're attributing it to the mix shift taking effect. Fair enough — but Q1'26 will be the tell.
FY26 EBITDA delivery. $269m is the promise. Q1'26 guide will reveal the ramp shape. If it front-loads or is flat-loaded, thesis intact. If it's aggressively back-half weighted with no visibility, I'll reduce conviction.
NRR trajectory. Does it stabilize at 130-135% or continue declining toward 120%? Below 120% changes the member-expansion story.
Member adds per quarter. 15-17 per quarter is the recent pace. Deceleration here would concern me.
Take rate. 7.1% → 8.0% is meaningful. Can it hold or grow as third-party mix grows? Or does competition force it lower?
GLR below 50%. Sub-50% sustained would validate the underwriting quality of the exchange members.
Using Bear's standard approach — P/S and market cap.
| Metric | Value |
|---|---|
| Market cap | ~$2.2B |
| TTM Revenue | ~$855m (Q4'24+Q1-Q3'25) |
| EV/TTM Rev | ~2.5x |
| FY26E EBITDA (mgmt guide) | $269m |
| EV/FY26E EBITDA | ~8x |
| Revenue YoY growth | 74% |
At 2.5x revenue and 74% growth, every comparable in my universe is priced higher. The catch: this is insurance, not SaaS. Insurance platforms trade at lower multiples because investors discount the complexity, cyclicality, and underwriting risk. But ARX is deliberately moving away from underwriting risk. If the market re-rates it toward a marketplace/exchange model, the multiple could expand significantly.
The embedded EV/EBITDA story is the most interesting: $269m FY26E EBITDA at $2.2B market cap = ~8x forward EBITDA. For a platform growing EWP 30%+ annually, that is cheap.
I would not pay 20x EV/EBITDA for ARX at this stage — not enough history. But under 10x on a $269m guide with 74% revenue growth is a setup that demands attention.
ARX is executing a deliberate platform transition in specialty insurance — replacing balance-sheet risk with exchange-fee income. Revenue is accelerating (74% YoY), network is growing (265 members, near-zero churn), loss ratios are improving, and the stock trades at 2.5x revenue after a 66% drawdown. The FY26 EBITDA guide ($269m at ~8x EV) is the market's test: can management deliver the ramp after the near-term margin compression?
Thesis status: Watching. Numbers match the theory after three quarters of public acceleration. But seven quarters of public data is not enough to build large conviction. I'd want a starter position — 2-3% — to participate if the thesis proves out, with room to add meaningfully after Q1'26 and Q2'26 results.
This is not a buy with full conviction. I don't invest based on hope, and the FY26 margin ramp requires trust in a management team I've watched for less than two years. The valuation is forgiving enough that I don't need to rush.
Action: Watchlist with potential small starter (2-3% max at this stage). Revisit after Q4'25 earnings and FY26 guidance confirmation.
Bear