Slide Deck

⚡ Kimi’s Take

Kimi’s own subjective opinion — the author’s independent view, provided as general information only and not investment advice. The analytical body (§1–§15) below is evidence-first and carries no position; this block is the single place an opinion and a valuation zone appear.

Verdict: HOLD / do not chase — accumulate only on a material derating. Enova is the best-executed company in a structurally hard business, and the execution is real: a distressed-priced OnDeck acquisition that became 70% of the book, nine straight quarters of SMB net charge-offs in a 4.4–4.8% band, consumer credit range-bound at 14–16% while funding costs fall, ~24% ROE, and a decade of disciplined buybacks that cut the share count 31%. But none of that is the question at $232. The question is the price: the stock has doubled in twelve months to the richest valuation in its own history — ~96th composite percentile of its own multi-year range, ~4.1x book and ~5.5x tangible book against a peer anchor (OneMain) at ~7–8x earnings, and ~14–15x the FY26 guidance it must still deliver. The 10-K itself concedes the price embeds completion of the Grasshopper Bank charter. This is a momentum one-way street (1-yr Sharpe ~2.5) that is +45% above its 200-day EMA and gap-prone around every earnings print — a great business at a frothy price, in a business with zero customer captivity.

Zone: a defensible accumulation zone is roughly $130–170 (~8–10x FY26 guided adjusted EPS of ~$16, ~2.5–3.2x Q1’26 book) — where the stock traded as recently as Q1 2026. Above $190–225 (~12–14x) the buyer is underwriting all three of: clean Grasshopper approval, ≥25% EPS growth delivered, and no credit-cycle turn — simultaneously. At $232 the risk/reward is skewed the wrong way for new money; for holders, the fundamental momentum is too strong to short and the insider tape (one-directional selling, including the chairman at $231 this month) argues against adding.

Rationale and framing: the framing is quality-compounder-at-a-price, late-momentum. The twelve-month re-rating is predominantly idiosyncratic — factor data show financials, fintech, size and beta were all headwinds while the stock doubled — i.e., the market is specifically repricing Enova from “high-beta subprime lender” to “bank-charter-in-waiting compounder.” That story is genuinely good. It is also now fully subscribed: Grasshopper approval (base case, but with organized NCRC/CRL/NCLC opposition and a visible consent-order history), deposit-funded cost-of-capital transformation, and a continuation of the best credit tape in company history are all required to hold the multiple, while the bear case needs only one of: an SMB delinquency uptick (already 6.2%→7.1% QoQ in Q1’26), a deal delay past H2 2026, or a return of the subprime-contagion tape that knocked 13.5% off the stock in six weeks last fall.

Conviction: medium. Flip bullish: Grasshopper approval on clean terms plus proof deposit funding actually reprices the stack (the first structural moat ingredient in company history) with SMB NCOs holding ~4.6% at the July 23 print. Flip bearish: OCC grants opponents a public hearing or delays the deal, or SMB 30+ delinquency breaks above its four-year range — either punctures the charter-compounder narrative the multiple now depends on.

---

📈 Stock Price Action — Five-Year Event Map

Over the past five years Enova’s stock has made a full round trip from a rate-shock trough to a record high: from ~$30 at the start of the window (Jul-2021) it fell to a five-year closing low of $26.63 on 2022-06-16, then compounded to an all-time closing high of $242.18 on 2026-07-01 — roughly 9x off the low — and closed the latest session at $232.52 (2026-07-17). The stock sits -4.0% off its high, at the 93rd percentile of its 52-week range ($100.56 low on 2025-08-01 / $242.18 high on 2026-07-01), and +45% above its 200-day EMA (~$160) — deep into the upper end of its own cycle, with the largest gains concentrated in the last twelve months (+102% 1-yr). (Price figures: AZI Trading daily adjusted OHLC CSV, pulled 2026-07-19.)

#	Period	Approx. move	Price (~from → to)	Primary driver(s)	Fact / Interp
1	Jul-2021 → Jun-2022	-36% from Jan-22 peak; 5-yr low	$46.59 (Jan-22 hi) → $26.63 (2022-06-16)	2022 rate shock / non-prime consumer-credit fears (Interp); Q1-22 print followed by -8.5% on 2022-05-05 amid market-wide selloff (Fact: move/8-K date; Interp: cause)
2	Jun-2022 → Feb-2023	+100%	$26.63 → $53.30 (2023-02-03)	Q3-22 earnings (8-K 2022-10-27; +11.4% 2022-10-28) and Q4-22 earnings (8-K 2023-02-01; +14.5% 2023-02-02) — strongest single-day move of the 60 months (Fact: moves/dates; Interp: earnings as driver)
3	Mar-2023 → Oct-2023	-16% SVB window; -30% in 10 sessions into Q3 print	$49.83 (2023-03-08) → $36.14 (2023-10-30 wk lo)	SVB regional-bank scare Mar-2023 (Interp); Q1-23 print 8-K → -12.6% next day; Q3-23 print 8-K → -18.2% on 2023-10-25, largest down day of the period, even as a new $300M buyback was authorized the same day (Fact: moves/8-K contents; Interp: credit-cost concerns dominating the buyback news)
4	Nov-2023 → Jan-2024	+33% in 2 months	$41.49 (2023-11-01) → $55.36 (2023-12-29)	Nov–Dec-2023 rates rally lifting high-beta financials (Interp); Q4-23 print (8-K 2024-01-30) → -7.4% on 2024-01-31 (Fact: move/8-K; Interp: print vs. elevated expectations)
5	Feb-2024 → Nov-2024	+81% across the year	~$54 → $105.54 (2024-11-27)	Steady earnings-driven grind with an -11% early-Aug-2024 interruption (global unwind selloff, Interp); +10.6% on 2024-11-06, the post-election session, extending to +18% for November (Fact: moves; Interp: election/deregulation trade in financials)
6	Dec-2024 → Apr-2025	-17% into Apr-7 trough, sharp V	$116.49 (Feb-25 hi) → $85.91 (2025-04-07 lo) → ~$99	Apr-2025 tariff shock: -11.1% on 2025-04-03 after “Liberation Day” tariffs, +11.7% on 2025-04-09 on the 90-day pause (Fact: moves; Interp: macro tape, not fundamentals); Q1-25 print (8-K 2025-04-29) → -7.8% next day
7	May-2025 → Oct-2025	+39% to Sep hi, then -13.5% scare, then recovery	$100.56 (2025-08-01, 52-wk lo) → $127.32 (Sep-25 hi) → $104.01 (2025-10-10) → ~$121	Subprime-credit contagion scare: Tricolor Holdings Chapter 7 (2025-09-10) and PrimaLend Chapter 11 (2025-10-22) weighed on all non-prime lenders (Interp — ENVA has no disclosed Tricolor/PrimaLend exposure); Q3-25 print (8-K 2025-10-23) → +9.4% on 2025-10-24 (Fact: move/8-K; Interp: print re-asserting idiosyncratic credit story)
8	Nov-2025 → Jul-2026	+93% to ATH, now -4% off	~$121 → $242.18 (2026-07-01) → $232.52 (2026-07-17)	$400M buyback authorization (8-K 2025-11-12); Grasshopper Bancorp merger agreement ~$369M for a national bank charter (8-K 2025-12-11; +11.7% that day); strong Q4-25 print (8-K 2026-01-27); a -7.7% day on 2026-02-03 during the Feb-2026 AI-led market selloff (Interp); Q1-26 beat with raised 2026 guidance (2026-04-23) and a +24% June run into the 2026-07-01 record (Fact: moves/8-Ks; Interp: charter-deal + earnings momentum as drivers)

Cycle narrative. The 2022 drawdown was macro, not company-specific: ENVA fell with every rate-sensitive, credit-exposed small-cap as the Fed hiked, bottoming in June 2022 when consumer-credit fear peaked; the recovery that followed was built almost entirely on earnings prints — each of the four largest single-day gains of the 60-month window is either an earnings reaction or a macro reversal day (FACT as to price, INTERPRETATION as to cause). The October 2023 episode is the clearest illustration that headline capital returns do not protect the stock when credit is questioned: the board authorized a $300M repurchase in the same 8-K as the Q3-23 print (Item 8.01, filed 2023-10-24), and the stock still lost 18.2% the next day and ~30% over ten sessions. From November 2023 onward the tape shifted to a persistent uptrend punctuated by macro shocks that were each retraced — the Aug-2024 unwind, the Apr-2025 tariff crash, and the Sep–Oct-2025 subprime-lender failure scare (Tricolor Chapter 7 filed 2025-09-10 amid fraud allegations; PrimaLend Chapter 11 filed 2025-10-22) — with Enova-specific fundamentals reasserting at each earnings date. The final leg to the 2026-07-01 record is, in this read, the market repricing Enova from “high-beta non-prime lender” toward “bank-charter-in-waiting compounder”: the ~$369M Grasshopper Bancorp merger (agreement dated 2025-12-10, 8-K filed 2025-12-11; regulatory applications filed 2026-01-16; expected close H2-2026) coincided with a +11.7% single-day move and preceded a near-doubling over seven months (INTERPRETATION; the price moves themselves are FACT).

---

1. Executive Summary

Enova International is the largest scaled online lender to non-prime consumers and small businesses in the United States — a Chicago-based, 100%-online lender (CashNetUSA, NetCredit, OnDeck, Headway Capital) that has funded $65B+ to 13M+ customers over 21 years. FY2025 revenue was $3,151.7M (+18.6%), net income $308.4M, diluted EPS $11.52 (adjusted $12.96), ROE ~24% on my computation, ROA 5.3% — the best returns in its peer set by a wide margin.

The SMB pivot is the story. Small-business lending (OnDeck/Headway) is now ~70% of the $5.3B portfolio, up from 52% at YE2021, growing receivables +38.6% y/y in Q1 2026 at 48–49% yields with only ~4.6% net charge-offs and ~69–71% net revenue margins. The legacy consumer book (90–147% yields, 14–16% NCOs) is ex-growth to low-growth. Enova is becoming a small-business lender with a consumer legacy — a materially better business mix than its “online subprime” label suggests.

The moat is real but narrow. There is zero customer captivity (marketing is 19.7% of revenue — the price of non-captivity), no network effects, and no structural cost advantage: 8.2% wholesale funding is a disadvantage against deposit-funded or IG-rated rivals. The genuine edge is economies of scale plus 21 years of learned underwriting data inside a regulatory maze — an execution advantage that shows up in best-in-class ROA (5.3% vs OneMain ~3%, everyone else ~0–2%) and in casualties around it (CURO bankrupt 2024; Oportun, Regional, World Acceptance sub-scale or barely profitable).

Quality of earnings carries one structural flag: Enova elected fair-value accounting for its entire loan book in 2020 — no CECL allowance exists; credit cost flows through a Level 3 mark-to-model “change in fair value” line, and the book carries a $774M premium over principal (~58% of equity). Marks have been stable and corroborated by realized losses for two years, but growth mechanically flatters the P&L and there is no loss-assumption sensitivity disclosed.

The swing factor is Grasshopper Bank. The pending ~$369M acquisition (announced 2025-12-10, OCC/Fed decision pending, close guided H2 2026) would make Enova a bank holding company with ~$3B of deposits against an 8.8% corporate funding cost — the first structural moat ingredient in company history — while inviting a political fight (NCRC/CRL/NCLC opposition citing Enova’s two CFPB consent orders). The 10-K concedes the stock price embeds completion.

Capital allocation is a genuine strength with two blemishes: ~$1.0B of buybacks since 2019 (share count −31% since 2020), the outstanding OnDeck deal, and falling securitization coupons — against long-term incentives with zero performance conditions and uniformly one-directional insider selling (~$99M in 24 months, one open-market buy).

The price is the problem. At $232.52 the stock trades at the richest multiples in its own history (~96th composite percentile): ~4.1x book, ~5.5x tangible book, ~18.9x TTM GAAP EPS, ~14–15x FY26 guided adjusted EPS — versus a direct-comp anchor (OneMain) at ~7–8x earnings and Enova’s own 2019–2024 history of 5–9x. The market is underwriting deal completion, ≥25% EPS growth, and no credit turn, all at once.

---

2. Business Overview

What Enova is. Enova is a technology-and-analytics consumer and small-business lender, 100% online, with no branches (contrast OneMain’s ~1,300), 1,836 employees, headquartered in Chicago. Founded as CashNetUSA in 2003–04, it was spun off from Cash America and IPO’d in November 2014. It operates a single reportable segment managed as two portfolios — consumer and small business (SMB) — plus a small “other” bucket containing the Pangea remittance unit. Through YE2025 it had processed ~69.3M customer transactions and accumulated >95TB of behavioral data; in FY2025 it extended ~$7.8B of credit across ~4.3M transactions (FY2025 10-K, filed 2026-02-20).

The product grid and its unit economics (FY2025 10-K, Business section):

Product	Ticket	Term	Avg annualized yield FY2025	Footprint
Consumer installment	$300–$10,000	3–60 mo (avg 39 mo)	~90%	37 US states + Brazil (39 states by Q1’26)
Consumer line of credit	limits $100–$7,000	revolving	~147%	31 states (+2 legacy)
SMB installment	$5,000–$400,000	6–24 mo (avg 15 mo)	~48%	49 states + DC
SMB line of credit	$5,000–$200,000	12–24 mo	~49%	49 states + DC

The consumer customer earns ~$42,000 on average with FICO scores mostly 500–680 (Enova says it generally does not use FICO in underwriting — its own models use 1,000+ variables across 100+ algorithms). The SMB customer has median annual sales of ~$585K and ~11.4 years of operating history. The economic engine in one line: triple- or double-digit asset yields, minus mid-teens consumer credit losses (or ~4.6% SMB losses), minus ~8.2% wholesale funding cost, minus ~20%-of-revenue customer acquisition, equals ~5% ROA — exceptional for this industry (INTERPRETATION from disclosed figures).

The SMB pivot. The defining fact of the business today is the mix shift. Combined receivables (principal + accrued):

Date	Consumer ($M)	SMB ($M)	Total ($M)	SMB %
Q4 2021	941.4	1,016.6	1,958.0	52%
Q4 2022	1,056.2	1,797.3	2,853.4	63%
Q4 2023	1,263.0	2,050.4	3,313.4	62%
Q4 2024	1,506.8	2,483.5	3,990.3	62%
Q4 2025	1,596.1	3,328.5	4,924.6	68%
Q1 2026	1,578.9	3,696.7	5,275.6	70%

SMB receivables grew +34.0% in FY2025 and +38.6% y/y in Q1 2026; consumer grew +5.9% and +7.6%. Q1’26 originations of $2.3B were +33% y/y, with SMB at a record $1.7B (+42%) and consumer at $559M (+10%). Revenue mix FY2025: consumer 55.5%, SMB 43.2%, other 1.3%. The SMB book is lower-yielding (48–49% vs 90–147%) but far cleaner: NCOs ~4.6% vs 14–16%, net revenue margin ~69–71% vs ~51%, and shorter duration (15-month average term). Management claims ROA is “pretty similar across the two portfolios” (Q1’26 call) — a HYPOTHESIS; the fair-value marks (consumer carried at 122% of principal vs SMB at 112%) imply the consumer book still carries more embedded profit per dollar of principal.

How it actually makes money — and the treadmill. Marketing is the single largest operating expense: $621.1M in FY2025 = 19.7% of revenue, rising to 22% in Q1’26. This is a paid-acquisition business: search, direct mail, lead generators, and broker/“commissionable originations” (a large SMB driver). Cash is a commodity and switching costs are zero; revenue is quasi-recurring only in that returning customers re-borrow — and returning customers do default less (10-K MD&A). The SMB book turns in ~15 months average term, so revenue requires continuous origination. Nothing about the revenue stream is contracted or recurring (INTERPRETATION).

Origination structures. Four channels: (1) direct balance-sheet lending under 37–39 state licenses (CashNetUSA, NetCredit) and 49-state SMB licenses (OnDeck, Headway); (2) Bank Programs — five partner banks originate, Enova markets and services with an option (not obligation) to purchase, no performance guarantee — 31.6% of FY2025 originations, the rent-a-bank structure that carries true-lender legal risk (§3); (3) a legacy Texas CSO/CAB program, a $22.3M-guaranteed remnant of the old payday model, immaterial; (4) Pangea money transfer (~1% of revenue, acquired March 2021 for $32.9M). Brazil (Simplic, organic since 2014) runs through a third-party lender and is 2.4% of revenue — and still cumulatively loss-making after 11 years, with $73.6M of gross NOL carryforwards at YE2025 (up from $39.0M at YE2024) under a full valuation allowance. Treat Brazil’s “$49B TAM” claim as marketing; the business is a zero-valued option (INTERPRETATION).

Funding. Total debt $4,832.5M at 3/31/26, 69% securitization-based (10+ warehouse/ABS shelves), weighted-average cost 8.18% in Q1’26, down from 9.31% in 2024. No recourse debt matures before December 2028. Liquidity ~$1.1B. Funding is the structural weakness the Grasshopper deal attacks (§7, §8).

Verdict: a genuinely good, genuinely simple business to understand — lend at 48–147%, lose 5–16%, fund at 8%, spend 20% of revenue finding borrowers — whose mix is improving as SMB (cleaner, shorter, faster-growing) displaces consumer. The model’s economics are proven across a full cycle including COVID. What it is not is a franchise with captive customers: every revenue dollar is re-earned in the acquisition market every quarter.

---

3. Industry Dynamics

Structure and size. Enova competes in two adjacent markets. US non-prime consumer lending: the company cites an $85B opportunity, against a demand backdrop where 37% of US adults cannot cover a $400 expense with cash (Fed SHED, cited in the 10-K). US SMB lending: a claimed $313B market, with sub-$250K loans representing 78% of SMB loan applications (2025 Federal Reserve Banks small-business study) and banks having vacated the segment. Brazil adds a claimed $49B consumer market that this analysis discounts to zero value (§2). These headline TAMs are marketing-grade (FACT: company estimates); the serviceable market is materially smaller — roughly 20 states plus DC cap small-loan APRs at or below ~36%, which excludes high-APR consumer product, and SMB addressability is limited to firms too small or young for bank credit but with real cash flow. Enova’s $5.3B book is a mid-single-digit share of its serviceable market: growth is a share-gain story, not a market-growth story (INTERPRETATION).

Competitive map. Consumer non-prime (FY2025 figures, ROIC.ai; market caps at 2025-12-31):

Company	Model	Revenue	Net income	Mkt cap	Read
ENVA	online, consumer + SMB	$3,152M	$308M	$3,957M	ROE ~24% (computed), ROA 5.3%
OneMain (OMF)	branch + online personal loans	$4,970M	$783M	$7,917M	scaled incumbent, IG funding, ~$24B receivables
Upstart (UPST)	AI marketplace	$1,024M	$54M	$4,199M	fee model, thin GAAP profits
LendingClub (LC)	marketplace-turned-bank	$999M	$136M	$2,171M	deposit-funded, prime/near-prime tilt
Oportun (OPRT)	branch + online, Hispanic-focused	$406M	$25M	$246M	lost $79M FY24, $180M FY23
Regional Mgmt (RM)	branch installment	$561M	$44M	$365M	~8% net margin
World Acceptance (WRLD)	branch installment	$536M	$35M	$629M	shrinking branch model
CURO	storefront + online payday	—	bankrupt	—	Chapter 11, March 2024

Plus OppFi (closest pure consumer-model comp; Q4’25 revenue $159.2M, +17.3%), private Avant (unverified scale), and sub-scale venture-backed names. On the SMB side: AmEx (Kabbage/Business Blueprint), Block/Square Loans, PayPal Working Capital, BlueVine, Fundbox, Funding Circle, and legacy MCA shops. The most dangerous SMB competitors are the embedded-finance players — Block, PayPal, AmEx — whose captive merchant bases and payments data give them structurally cheaper customer acquisition than Enova’s broker-and-advertising model (INTERPRETATION).

Profit pool. The consumer non-prime profit pool concentrates in exactly two names: OneMain (scale + IG funding + branch-led collections) and Enova (online cost structure + analytics). Everyone else is sub-scale (OPRT cumulative losses FY23–24), shrinking (WRLD/RM), unproven through-cycle (UPST), or dead (CURO, Chapter 11 March 2024 under a debt-laden LBO structure and rate-cap pressure). In SMB, the pool is contested between OnDeck/Enova and the embedded-finance giants, with fragmented legacy MCA shrinking as states impose commercial-financing disclosure laws (INTERPRETATION).

Capital cycle (Marathon lens). 2020–21 saw VC/SPAC capital flood fintech lending — Upstart, Oportun (SPAC), OppFi (SPAC), MoneyLion — and the returns were competed away; the 2022–23 funding shock then destroyed the sub-scale (UPST −90%+ from peak, CURO bankrupt). 2024–26 is the discipline phase: private-credit forward-flow and ABS demand returned selectively for proven performers. The evidence Enova sits on the right side of that selectivity: in Q1’26 it upsized four warehouse facilities by $377M at unchanged terms and spreads while other subprime funding markets widened (Q1’26 call), and its new ABS coupons have fallen from 6.84% (2024-1) to 5.65% (2025-2). The caution signal at the margin: early-2026 private-credit retail redemption stress (a Stone Ridge interval fund honoring only ~11% of redemption requests; Morgan Stanley forecasting higher private-credit defaults — press reports 2026-03) — marginal funding for subprime paper tightening at the edges historically helps incumbents with diverse funding and hurts origination-for-sale models. Verdict: capital is not flooding into non-prime balance-sheet lending today; returns are not being competed away — but embedded finance is a structural source of new low-CAC capacity in SMB (INTERPRETATION).

Regulatory landscape. Three layers, in ascending order of threat:

1. Federal — currently a tailwind. The CFPB under Acting Director Vought is largely defunded (staff cut >80% per contemporaneous press reports; operations suspended; nominee lapsed per American Banker 2026-01). The Small Dollar Rule’s payment provisions took effect 2025-03-30 but the CFPB announced it will not prioritize enforcement and plans to narrow the rule. Section 1071 SMB data reporting is delayed to 2028 with scope-narrowing proposed. Enova’s own record: a 2019 consent order ($3.2M, unauthorized debits) and a November 2023 second order ($15M) as a repeat offender — terminated in full by the CFPB effective 2025-08-29/09-02, with alleged non-compliance waived. The largest regulatory overhang on the equity was quietly removed (FACT, consumerfinance.gov). The tail: pending federal bills in both directions, including a “Predatory Lending Elimination Act” (36% MLA-style extension). A federal 36% cap would terminate the consumer business as configured — low probability under this Congress, the single largest long-horizon tail.
2. State — the binding constraint. ~20 states + DC cap small-loan APRs at/below ~36% (NCLC surveys). Illinois’ PLPA (2021) is the poster child: 36% all-in MAPR with an anti-evasion clause aimed at bank partnerships; a Fed/academic study found post-PLPA subprime loans fell 38% and licensed lenders halved — in Enova’s home state. Enova consumer-lends in only 37–39 states; the excluded set maps closely to the cap universe (ASSUMPTION — filings don’t enumerate). The cap ratchet is slow but one-directional; Virginia’s 2025 bill threatening the bank-partnership model shows the frontier.
3. Legal-doctrine — regime risk. 31.6% of originations run through five bank partners under valid-when-made rules (OCC/FDIC 2020, upheld 2022); the CashCall “predominate economic interest” true-lender test remains live doctrine. An adverse ruling or hostile OCC could impair a third of the origination engine — a regime risk, not a probability-weighted earnings risk. The Grasshopper charter is partially a hedge against exactly this.

Verdict: a structurally mediocre industry with unusually good perches. Demand is structural (liquidity-stressed consumers, bank-abandoned micro-SMBs), but returns are statute-capped, the product is a commodity with zero switching costs, credit cost dominates the P&L and is only partially controllable, and funding is wholesale and confidence-sensitive. The industry favors scaled specialists with proprietary data, diversified funding, and regulatory adaptability — precisely Enova and OneMain — and grinds down everyone else. Enova’s perches (online cost structure, SMB unit economics, funding access) are among the best available in a hard neighborhood.

---

4. Competitive Position

Applying the Greenwald taxonomy strictly, candidate moat by candidate moat:

Demand-side advantage / customer captivity — NONE, and it bears saying plainly. Switching costs are approximately zero: a borrower can apply to ten lenders in ten minutes, cash is fungible, and there is no contract, ecosystem, or lock-in of any kind. The proof is in the P&L: marketing is 19.7% of revenue (FY2025) and rising (22% in Q1’26) — a business with captive customers does not spend a fifth of revenue re-acquiring them. High repeat-borrower rates reflect habitual liquidity stress in the customer base, not captivity. Brand (high Trustpilot/BBB ratings) reduces acquisition cost at the margin but is replicable spending, not a moat.

Network effects — NONE. More borrowers do not make the product better for other borrowers. Enova’s data advantage (21 years, 69M transactions, 95TB, 100+ algorithms, the OnDeck Score) is a learning-curve/scale effect — it improves Enova’s loss models, not each customer’s experience of other customers. Any “data network effect” framing is speculative and is not used here.

Supply-side / cost advantage — WEAK, and currently negative vs the best rivals. Weighted-average funding cost is 8.2% (Q1’26) — wholesale money, structurally more expensive than LendingClub’s deposits or OneMain’s investment-grade funding. There is a real operating-cost edge over branch models (no ~1,300-branch network; operations & technology at 8.2% of revenue) and a funding edge over sub-scale online rivals who lost market access in 2022–23 — but versus the strongest competitor, Enova is the higher-cost producer of funds. This is precisely what the Grasshopper acquisition targets (§7).

Economies of scale — REAL but narrow, and the core of the actual advantage. Fixed-cost leverage shows up in three places: (a) underwriting learning curves — loss-given-curve knowledge across 69M transactions that a new entrant must buy with real losses (the 10-K itself notes new portfolios incur “early losses associated with unseasoned loan portfolios”); (b) regulatory infrastructure — licenses and product variants across 37–39 consumer states, 49-state SMB, five bank programs, Brazil, with a claimed 3–6 months concept-to-launch; the ability to re-engineer products state-by-state is what storefront operators like CURO could not do when caps hit; © capital markets access — 10+ securitization shelves and a bank revolver syndicate that kept lending at ~8% through the 2022–23 funding winter while rivals were cut off (and upsized $377M of warehouse capacity at unchanged spreads in Q1’26 while other subprime funding markets widened).

The tests. Greenwald’s two empirical checks: (1) Share stability — consumer receivables +5.9% FY25, roughly holding share in a low-teens-growth origination market; SMB +34% in a market growing high-single-digit — clear share gain where they lean in. No evidence of share loss anywhere. (2) Returns — ROA 5.3% and computed ROE ~24% (FY2025) versus OneMain ~3% ROA, Regional ~2%, Oportun/LendingClub ~1%, Upstart ~breakeven, CURO dead. Best-in-group returns persisting across years is the signature of some advantage; the question is width and durability.

The honest moat sentence. Enova has scale economies plus learned underwriting data inside a regulatory maze — a process/execution advantage, durable only while underwriting discipline holds and no federal rate cap arrives. Zero captivity, zero network effects, statute-set pricing power. In Greenwald terms: weak-to-moderate economies-of-scale advantage without captivity; in Morningstar vocabulary: no moat to narrow. The first erosion signals to watch are marketing as % of revenue (rising: 19% Q1’25 → 22% Q1’26) and the SMB NCO spread versus any competitor that reaches scale (the embedded-finance threat).

Verdict: weak/narrow advantage — the best operator in a crowded, commodity market with weak structural differentiation. Execution, not franchise, is what separates Enova from the graveyard (CURO) and the barely-alive (OPRT, RM, WRLD). Execution advantages are real and monetizable — but they must be re-proven every quarter, which is exactly why the market historically charged only 5–9x earnings for them, and why the Grasshopper charter (deposit funding + national preemption) would be the first structural advantage in company history if approved.

---

5. Growth History and Forward Opportunities

History. Revenue grew from $1,083.7M (2020) to $3,151.7M (2025) — a 23.8% CAGR — and from $1,174.8M (2019), a 17.9% six-year CAGR. Net income over the same 2020–25 span went $377.8M (COVID credit-release flattered) → $256.3M → $207.4M → $175.1M → $209.4M → $308.4M: earnings only regained and surpassed the 2020 level in 2025, on nearly triple the revenue — margin-normalization math worth bearing in mind when extrapolating growth. Diluted EPS went $11.70 (2020, credit-release year) → $5.49 (2023 trough) → $11.52 (2025) → $3.46 in Q1’26 alone (+28.5% y/y).

Originations tell the cleaner story: FY2025 ~$7.8B extended (consumer $2,274.7M + SMB $5,475.1M); Q1’26 $2.3B (+33% y/y), with SMB originations growing ≥20% for eight-plus consecutive quarters. Consumer originations re-accelerated to +10% in Q1’26 after management deliberately slowed line-of-credit growth in spring 2025 when credit wobbled — a tightening episode that self-corrected within two quarters (management commentary, corroborated by NCO/delinquency data).

Organic vs acquired. The SMB engine is acquired at the root — OnDeck (October 2020, ~$113–122M at distressed pricing) — and organic thereafter (SMB receivables have roughly tripled since 2022). Consumer growth is organic. Brazil is organic and loss-making. Pangea was a small 2021 acquisition (~$32.9M), immaterial.

Forward vectors, ranked by evidence:

1. US SMB (the engine). Record $1.7B originations in Q1’26 (+42%); a claimed $313B market banks have abandoned below $250K; post-2020 business-formation wave. Durability question: can 48–49% yields and ~4.6% NCOs survive either (a) seasoning of a book growing ~40%, or (b) embedded-finance competition (Block, PayPal, AmEx) scaling down-market? Enova’s answer is the OnDeck Score data moat and the broker channel; the skeptic’s answer is that embedded players have cheaper CAC and better data on their own merchants.
2. Consumer re-acceleration. State count 37 → 39; returning-customer mix improving credit; +10% Q1’26 originations. Structural ceiling: the 36%-cap state map keeps shrinking the addressable footprint; the business is managed to a ~15% NCO equilibrium, not maximized for growth.
3. Grasshopper (the step-change, pending). Deposit funding + national preemption + BaaS/SBA optionality. Management claims >15% adjusted-EPS accretion year one and >25% with full synergies — unvalidated HYPOTHESES; the deal has not closed and faces organized opposition (§7).
4. Bank Programs growth. Already 31.6% of originations; asset-light-ish growth with no performance guarantee — but carries true-lender regime risk (§3).
5. Brazil (Simplic). Option value only: 2.4% of revenue, $73.6M NOLs fully reserved, record Brazilian household delinquency and a 15% Selic. Zero-valued in this analysis.
6. Pangea. ~1% of revenue. Immaterial.

Guidance (management HYPOTHESES). FY2026 raised at Q1’26: originations ~+20% (from ~+15%), revenue growth similar to originations, adjusted EPS ≥ +25% (from ≥20%). Q2’26 guided: revenue +15–20% y/y, net revenue margin 55–60%, marketing ~20% of revenue, adjusted EPS +20–25% y/y. Guidance excludes any Grasshopper contribution.

Verdict: high-quality growth — for a lender. It is predominantly organic post-OnDeck, funded (not promised), profitable at the unit level (SMB net revenue margin ~70%), and gaining share against bankrupt and sub-scale competition. The two discounts: growth consumes balance-sheet capacity roughly 1:1 with receivables at constant ~3.4x leverage (distributable earnings ≪ reported earnings during fast growth), and the marginal growth dollar is being bought with rising marketing intensity (22% of revenue in Q1’26) into a book whose recent vintages have not yet seasoned.

---

6. Financial Quality

The accounting election that frames everything: fair value. Effective January 1, 2020, Enova elected the fair-value option (ASU 2019-05) for its entire loan portfolio — CECL was never adopted and no allowance for loan losses exists. Credit cost flows through the “change in fair value” line (a deduction from revenue: −$1,321.4M in FY2025), and loans are carried at Level 3 model values: 122.9% of principal (consumer) and 112.2% (SMB) at Q1’26 — a $774M premium of fair value over principal, ≈58% of total equity. Three consequences:

• Growth mechanically flatters earnings. New originations enter near par and are marked up over following quarters if assumptions hold; inside the ΔFV line, marks offset realized NCOs (Q1’26 consumer: ΔFV charge $215.5M vs NCOs of $227.6M — marks were net positive offsets). A growth slowdown or assumption shock reverses this mechanism. This is the mark-to-model earnings-quality trap in full; the mitigant is two years of stable marks (consolidated ~115% premium) corroborated quarterly by realized losses.
• The balance-sheet reserve is a model, not a number. Disclosed sensitivity covers only the discount rate (±100bp = ∓0.54%/+0.56% of loan FV, ≈$30M, ~2.2% of equity). Loss-assumption sensitivity (e.g., the P&L hit from a +10% charge-off assumption) is not disclosed — a genuine transparency gap and a candidate management question.
• Peer comparability breaks. Allowance-coverage benchmarks (OMF ~11–12%, SYF ~10%) do not map to Enova; the pre-FVO analog is the 12/31/19 allowance of $176.9M = 14.3% of gross receivables, on a then shorter-duration, consumer-only book.

P&L trend (FACT, reconciled ROIC-to-EDGAR; $M):

Year	Revenue	ΔFV / credit cost	Net revenue	Marketing	Income from ops	Interest, net	Net income	Dil. EPS	Adj. EPS
2021	1,207.9	183.7	1,024.2	271.2	413.1	76.5	256.3	6.79	n/a
2022	1,736.1	618.5	1,117.6	382.6	384.0	115.9	207.4	6.19	n/a
2023	2,117.6	887.7	1,229.9	414.5	422.1	194.8	175.1	5.49	n/a
2024	2,657.8	1,128.4	1,529.4	523.6	584.8	290.4	209.4	7.43	9.15
2025	3,151.7	1,321.4	1,830.2	621.1	739.4	339.3	308.4	11.52	12.96
Q1’26	875.1	—	529.0	189.4	—	94.0	91.1	3.46	3.87

Margins and efficiency. Net revenue margin has expanded from 57–58% to 60.4% (Q1’26) as SMB mix rises. Total opex fell from 42.3% of revenue (2022) to 34.4% (2025) — genuine operating leverage — though Q1’26 showed negative jaws (marketing +36% y/y on +17% revenue) on SMB commissionable originations and online ad costs. Adjusted EBITDA $820.9M FY2025 (+25%), 26.0% of revenue in Q1’26.

Credit. Consumer NCOs have been range-bound at 12.7–17.3% since 2022 with normal seasonality — Q1’26’s 14.3% is the best Q1 since 2022 — with 30+ delinquency stable at 7–8%. SMB NCOs stepped up from COVID-era ~1–3% to a 4.4–4.8% plateau and have held it for nine straight quarters. The one yellow flag: SMB 30+ delinquency ticked up 6.2% → 7.1% QoQ in Q1’26 (still down y/y from 7.4%, and within its four-year range) — on a book growing +38.6%, recent vintages are unseasoned; this is the metric to watch at the July 23 print. Versus the mid-2026 cross-industry read (prior analysis of Synchrony, Ally, and Credit Acceptance filings): near-prime is improving (SYF NCO 5.42% and falling; ALLY five straight quarters of improvement) while the deep-subprime tail is at record stress (subprime auto 60+ DQ ~6.9%, a 32-year high; FirstCash’s AFF book deteriorating across every metric). Enova’s book currently behaves like the improving tier — which so far cuts against the bear case.

Funding and balance sheet. Debt $4,832.5M at 3/31/26: funding (securitization) debt $3,346.6M at 7.08% weighted average across warehouses (6.41–9.17%) and term ABS (5.65–8.31%); corporate debt $1,510.0M at 8.82% ($825M revolver at SOFR+3.25% due 2029; $400M 11.25% notes due 2028; $500M 9.125% notes due 2029). Cost of funds has fallen from 9.31% (2024) to 8.18% (Q1’26); new ABS coupons are down to 5.65% from 6.84%. No recourse maturity before December 2028; four warehouses hit revolve-end within 15 months — routine, and all four were upsized in March 2026 at unchanged spreads. Leverage has risen every year: debt/equity 1.10x (2020) → 3.39x (2025); equity itself declined in 2024 despite $209M of net income because buybacks outpaced earnings. Liquidity ~$1.1B ($421M cash incl. restricted + $654M available facility capacity).

Returns. Computed ROE (NI / average equity): 18.2% (2022) → 14.4% (2023) → 17.2% (2024) → 24.3% (2025); ROA 5.3%. The decomposition is honest: ~5.3% ROA × ~4.6x assets/equity. A >20% ROE is leverage-amplified and depends on holding consumer NCOs ≤16% and SMB ≤5% while funding at ~8.2%. Durability drivers in order: (1) loss-rate stability, (2) SMB mix shift, (3) funding-cost drift. It is vulnerable to a consumer-credit cycle, warehouse repricing, and rate-cap shocks to 90–147% yields. (Note: ROIC.ai’s 16.7% ROE does not tie to the statements; the computed figures here reconcile to EDGAR exactly.)

Per-share quality. Share count 32.97M (YE2019) → 24.72M (YE2025), −25% (−31% from the 2020 OnDeck-issuance peak). Tangible book value per share $35.13 → $42.64 (2024→2025, +21%). The GAAP→adjusted EPS gap (+12.5% FY25, +23% FY24) is structural, not one-time: ~80% of it is the stock-comp add-back (~$1.24/diluted share FY25, 10.7% of NI) plus intangible amortization. Treat “adjusted EPS” as roughly 10–12% generous. Q1’26’s 19.3% tax rate flattered EPS (excess SBC tax benefits on share appreciation); normalize ~23%.

QoE traps checked and resolved. No pre-charge-off delinquent-loan sales (the loan-sale NCO-suppression trap does not apply — NCOs are clean). Pangea was acquired ($32.9M, 2021), not divested. OnDeck goodwill $279.3M unimpaired since 2020. The only material one-offs in five years: the Linear JV write-down (−$16.5M, 2024) and Grasshopper deal costs ($6.6M FY25 + $2.7M Q1’26). Marketing is 100% expensed; no capitalized acquisition-cost asset. Texas CSO guarantees immaterial ($17.9M off-balance-sheet). Lender “free cash flow” is not meaningful the industrial way (2025 CFO of $1.82B adds back the non-cash ΔFV charge and excludes loan-book growth); the correct lens is PPNR-style earnings power: net revenue $1,830M − opex $1,084M = ~$746M, less $339M interest = $401M pretax.

Verdict: economics improve with scale — genuinely, but with a mark-to-model asterisk. Operating leverage is real (opex/revenue −8pp since 2022), returns are best-in-class, credit is stable-to-improving, and funding costs are falling. The asterisk: the entire earnings stream passes through Level 3 marks that growth flatters and whose loss assumptions are not stress-disclosed. The marks today are believable — two years of stability corroborated by realized losses — but they should not be taken on faith through a recession without the sensitivity disclosure Enova declines to give.

---

7. Capital Allocation

Buybacks — the headline virtue, with a price-sensitivity caveat. Enova has repurchased ~$1.0B of stock since 2019 ($33.8M in 2019 rising to a $289.3M peak in 2024; $214.6M in 2025), cutting the share count 31% from the post-OnDeck peak (35.8M → 24.7M). The 2023–24 purchases at ~$49–65/share (≈1–1.5x book) were outstandingly accretive — EPS went $5.49 → $11.52 while shares fell 15% in two years. A new $400M authorization (announced 2025-11-12, through 2027-06-30, replacing the $300M August 2024 program) had $367.8M remaining at 3/31/26. The caveats: the pace has decelerated sharply — $38M (Q3’25) → $35M (Q4’25) → $15.6M (Q1’26, at ~$152.75 average) — as management preserves flexibility for the Grasshopper cash leg; and at current prices (~4x book) repurchases are EPS-accretive but BVPS-dilutive (the Q1’26 10-Q itself notes buybacks partially offset BVPS growth). No dividend; none anticipated.

M&A scorecard — one of the better records in fintech. OnDeck (closed October 2020; ~$90M announced / ~$122M final, 0.092 shares + $0.12 per ONDK share) bought a distressed SMB lender near its COVID trough for ~1x book; SMB is now ~70% of a $5.3B portfolio with its own securitization shelf. Arguably one of the best fintech acquisitions of the decade (INTERPRETATION). Blemishes are small: the Linear JV (OnDeck’s ODX platform) took a $16.5M write-down in 2024; Pangea (2021, $32.9M) is retained but immaterial; Brazil has consumed eleven years of losses.

The Grasshopper fork. The pending ~$369M (fixed 50% cash / 50% stock per the December 18, 2025 amendment) acquisition of Grasshopper Bancorp — a digital national bank with $1.4B assets and ~$3.0B deposits — is simultaneously the boldest and riskiest allocation in company history: roughly 28% of YE2025 equity plus ~4–5% stock-leg dilution (ASSUMPTION pending final share count), deployed against an 8.82% corporate funding cost that deposits could transform. Management claims >15% adjusted-EPS accretion in year one and >25% with full synergies ($125–220M adjusted net income) — HYPOTHESES. The capital-allocation question is explicit: $369M for the charter versus $400M of authorized buybacks at 4x book — the company is choosing the charter. If deposits reprice even a third of the funding stack by 300–400bp, the deal pays for itself in funding savings (INTERPRETATION); if approval fails or is conditioned into irrelevance, the 10-K’s own risk language concedes the price embeds completion.

Debt management — quietly excellent. Liability management has been consistently intelligent: $400M of 11.25% notes issued at the 2023 rate peak to retire 8.50% paper (painful but necessary), $500M at 9.125% in August 2024, the revolver upsized to $825M at SOFR+3.25% (maturity 2029), and securitization coupons ridden down from 6.84% (2024-1) to 5.65% (2025-2, with the A-class at 4.84% AA(sf)). Funding cost has fallen ~113bp from the 2024 peak while the book grew ~35%.

Compensation and incentives — the weak link. Short-term incentives are one-third each Revenue (GAAP), Adjusted EBITDA, Adjusted EPS (2025 payout: 123.2% of target on 103%/110%/113% achievement) — the EPS metric at least aligns with per-share outcomes. But long-term incentives are 50% time-based RSUs and 50% quarterly-granted stock options with zero performance conditions — no performance shares, no ROE/TSR/credit hurdles. Options granted quarterly at market function as appreciation rights. Fisher’s 2025 total comp was $11.70M; his Executive Chairman package (~$8–9M/yr: $825k base, 130% STI target, 520% LTI target) is rich for a non-CEO role and a watch item. Say-on-pay was 95.5% in 2026 — shareholders do not object while the stock doubles.

Insider behavior — uniformly one-directional. Across 233 parsed transactions over 24 months: exactly one open-market purchase (a director, 1,700 shares at $115.95 in September 2025) against ~$99.3M of aggregate sales. Fisher sold $54.5M (monthly 10b5-1 sales plus large discretionary blocks, including ~$231.50 on July 14–15, 2026 — days before this report); Cunningham sold $18.9M via routine exercise-and-sell; two directors liquidated their entire direct stakes in February 2026 at ~$172; one director sold $2.7M in May–June and then resigned from the Audit Committee on July 10. Nobody bought during the February–March 2026 dip into the ~$120s. Interpretation: consistent with an equity-heavy comp model monetizing a stock that 5x’d off its lows — but the complete absence of discretionary buying, and the chairman’s discretionary selling into all-time highs ahead of the Grasshopper close, sits uncomfortably next to the corporate buyback narrative (INTERPRETATION).

Governance notes. Single-class stock, annual elections, no dual-class, anti-hedging/pledging policy, clawback in place. Orderly succession: Cunningham (CFO since 2016) to CEO effective 2026-01-01, Fisher to Executive Chairman (≥2-year commitment), Cornelis (ex-JPM securitized products; Enova Treasurer since 2017) to CFO. The 2023 CFPB consent order was terminated in full in August 2025. Two Audit Committee directors resigned simultaneously on 2026-07-10 (“planned,” stated non-disagreement) weeks before the bank-holding conversion — a monitoring footnote, no more.

Verdict: management has allocated capital intelligently — among the best in the sector — with two honest deductions. The OnDeck deal at the trough, the 2023–24 buyback prices, and the liability management are evidence of genuine skill. Deductions: performance-free long-term comp and an insider tape that points only one way. The Grasshopper bet will define the next chapter of this scorecard.

---

8. Changes and Headwinds — Last Two Years

The last 24 months have been the most eventful in Enova’s history as a public company. In chronological order (all dated from the 8-K corpus and validated press releases):

• 2024-08-12: $500M of 9.125% senior notes due 2029 issued; new $300M buyback authorization. The funding-cost peak was already passing.
• 2025-07-24: CEO transition announced — David Fisher (CEO since 2013) to Executive Chairman, CFO Steve Cunningham to CEO, Treasurer Scott Cornelis to CFO, effective 2026-01-01. An orderly, continuity-preserving insider succession; it also means the architect of the OnDeck deal remains in the building while the operator of the securitization machine now runs the company.
• 2025-08-28 / 09-02: Revolver upsized to $825M at SOFR+3.25%, maturity extended to 2029.
• 2025-08-29 / 09-02: CFPB terminated the November 2023 consent order in full and waived alleged non-compliance. The seven-year repeat-offender injunction overhang — the single largest regulatory discount on the equity — was removed without fanfare.
• 2025-09-10 / 10-22: Sector scare — Tricolor Holdings (subprime auto) filed Chapter 7 amid fraud allegations; PrimaLend followed. ENVA fell 13.5% in six weeks on contagion fear despite no disclosed exposure. The episode demonstrates the stock’s sensitivity to subprime-credit tape regardless of company specifics.
• 2025-11-12: New $400M buyback authorization through June 2027.
• 2025-12-10/11: Grasshopper Bancorp merger agreement — ~$369M (fixed 50/50 cash/stock after the 12/18 amendment) for a digital national bank ($1.4B assets, ~$3.0B deposits). Enova becomes a bank holding company on close; CashNetUSA will operate outside the bank subsidiary; Brazil sits outside as a nonbank affiliate. Stock +11.7% on the announcement.
• 2026-01-16: OCC and Federal Reserve applications filed (including BHC/FHC elections). Grasshopper holders approved the deal 2026-02-02.
• 2026-02-23: NCRC and the Woodstock Institute filed formal opposition urging denial, followed by Center for Responsible Lending hearing requests and an NCLC-led coalition letter to Congress (2026-05-07) framing the deal as creating “the first national bank dedicated to directly making predatory loans.” Their ammunition is real: 100%+ APRs, two CFPB consent orders, complaint data. Approval under the current administration is the base case; conditions (CRA/product undertakings), delay past H2 2026, or a granted public hearing are the risk paths.
• 2026-01-27: Q4’25 print — revenue +15%, adjusted EPS $3.46 (+33%).
• 2026-04-23: Q1’26 beat with FY26 guidance raised (originations ~+20%, adjusted EPS ≥+25%); consumer credit best Q1 since 2022; four warehouses upsized $377M at unchanged spreads.
• 2026-07-10: Two Audit Committee directors (Goodyear, McGowan) resigned simultaneously — “planned,” stated non-disagreement; board cut to 10. A footnote, but an odd one weeks before a bank conversion that will multiply audit complexity.
• Headwinds to name plainly: marketing cost inflation (19% → 22% of revenue y/y in Q1); SMB delinquency uptick (6.2% → 7.1% QoQ); buyback deceleration; the pending close’s binary-ish risk; and a macro tape (private-credit redemption stress, subprime-lender failures) that can reprice the stock 10%+ regardless of fundamentals.

Verdict: the changes strengthen the thesis — materially — but they are also already in the price. The consent-order termination, the charter bid, the funding-cost decline, and the raised guidance are all genuine improvements. They are also precisely the drivers of a stock that has doubled in twelve months to its richest-ever multiple. Strengthening fundamentals and weakening margin of safety are not contradictory here; they are the same event.

---

9. Risk Analysis

#	Risk	Likelihood	Impact	Evidence basis
1	Consumer/SMB credit cycle turn — NCOs revert above the 14–16% consumer band or SMB breaks its 4.4–4.8% plateau as unseasoned +39%-growth vintages season	Medium	High	SMB 30+ DQ already 6.2%→7.1% QoQ (Q1’26 10-Q); 2021–22 vintage lessons across the sector (from CACC/ALLY filing analysis); deep-subprime tail at 32-yr-high stress
2	Grasshopper denied, delayed, or conditioned — OCC/Fed action past H2’26, public hearing granted, or CRA/product conditions that gut the economics	Low-Medium	High	NCRC/Woodstock/CRL/NCLC opposition with real ammunition (two consent orders); 10-K concedes price embeds completion; ~28% of equity committed
3	Fair-value mark risk — Level 3 premium ($774M, ~58% of equity) proves optimistic in a downturn; growth-flattered marks reverse	Medium	High	FVO accounting (10-K); no loss-assumption sensitivity disclosed; marks offset NCOs in growth quarters (Q1’26 detail)
4	Federal 36% rate cap — terminates the consumer business as configured	Low (this Congress)	Severe	Pending bills (S.3721/S.3793); management itself notes 17–18 years of failures; the tail is fat, not probable
5	State rate-cap ratchet — more states adopt 36% caps, shrinking the consumer footprint	Medium-High (slow)	Medium	~20 states + DC already capped; IL PLPA cut subprime lending 38%; VA 2025 bank-partnership bill
6	True-lender / Madden regime shift — bank-program loans (31.6% of originations) challenged	Low (current doctrine)	High	CashCall doctrine live; hostile OCC/DOJ or adverse circuit ruling is a regime event, not an earnings risk
7	Funding market disruption — warehouse repricing/closure; ABS windows shut	Low-Medium	High	Wholesale-funded (8.2%); 4 warehouses roll within 15 months; mitigants: 10+ shelves, no recourse maturity before Dec 2028, $1.1B liquidity, Q1’26 upsizes at unchanged spreads
8	Embedded-finance SMB competition — Block/PayPal/AmEx scale down-market with cheaper CAC	Medium (multi-year)	Medium-High	Structural CAC advantage of captive merchant bases; Enova’s marketing already rising (22% of revenue)
9	Key-person / execution — new CEO’s first year; Fisher transition; dual Audit Committee departures	Low	Medium	Orderly succession (8-K 2025-07-24); board changes 2026-07-10 stated non-disagreement
10	Macro tape / beta — subprime-contagion or risk-off repricing independent of fundamentals	High (recurring)	Medium	β ~1.36; ~39% vol; eight ±10% days in 60 months; Tricolor/PrimaLend episode cost -13.5% with zero company exposure
11	Brazil / Pangea drag — continued losses, FX, impairment	Medium	Low	$73.6M NOLs fully reserved; 2.4% of revenue; already zero-valued in this analysis
12	Governance/comp — performance-free LTI entrenches; chairman monetization continues	Medium	Low-Medium	DEF 14A 2026; $99.3M 24-month insider selling vs one buy

Aggregate read (INTERPRETATION): the distribution is asymmetric at the current price — not because the business is fragile (it is the strongest it has ever been) but because three of the four high-impact risks (credit turn, deal failure, mark reversal) strike directly at the two pillars of the current multiple: the charter-compounder narrative and the cleanest credit tape in company history. A recession is the scenario that activates risks 1, 3, and 7 simultaneously; the last real stress test (2022) took the stock down ~45% peak-to-trough while the business remained solidly profitable — the equity’s beta to its own sector’s fear is the practical risk to size.

---

10. Valuation Discussion

All multiples recomputed from filings at $232.52 (close 2026-07-17); shares 24.92M at 3/31/26 → market cap ~$5.80B. TTM GAAP diluted EPS $12.29; TTM adjusted EPS $13.85; TTM revenue $3,281M; BVPS $56.25 (3/31/26); TBVPS ~$42.6–44.9. EV/EBITDA is not meaningful for a lender (interest is COGS); the appropriate frame is PPNR-less-NCO.

Current multiples (FACT):

Metric	Value	Basis
P/E — TTM GAAP	18.9x	$232.52 / $12.29
P/E — TTM adjusted	16.8x	$232.52 / $13.85
P/E — FY26 guide-implied	~14.3x	guide ≥ ~$16.20 adj EPS
P/B	4.13x	BVPS $56.25
P/TBV	5.2–5.5x	TBVPS $42.6–44.9
P/S — TTM	1.77x
P/PPNR (FY25)	~7.8x	PPNR $746M → ~12.8% pre-tax earnings yield
Buyback capacity	6.3% of market cap	$367.8M remaining through 2027-06-30

Against its own history — the single loudest datum. AZI’s own-history percentile ranks (≈10-year window, 2026-07-17): composite 96.5th percentile; P/B 99.6th; P/S 99.6th; P/E 90.3rd. The P/B and P/S reads are the reliable ones (the P/E series is distorted by the 2019 discontinued-ops year and the 2020 COVID mark-recovery year). Corroboration from year-end multiples: P/B at year-end 2016–2025 ran 0.93–2.18 — the current 4.13x is roughly double the highest year-end print of the prior decade; year-end P/E 2021–2025 ran 5.8–12.8 — even the FY26-guide multiple (~14.3x) exceeds all of them. Part of the P/B rise is mechanical (buybacks shrank book: 2024 equity fell despite $209M of net income), but the dominant driver is price (+102% in twelve months). Percentile context is never a target — it says the re-rating has already been paid for, not that it is wrong.

Against peers (unofficial live prices 2026-07-17/18, screening-grade): OneMain — the direct comp — trades at 9.0x TTM earnings and ~2.0–2.6x book with ~24% ROE and a 7% dividend yield. Enova trades at 2x the P/E and ~1.6–2x the P/B of OneMain at essentially the same ROE; the premium is a growth premium (+33% originations vs OneMain’s low-single-digit) plus the Grasshopper option. Elsewhere: RM ~1.0x book, OPRT 0.7x, LC 1.5x, SLM 1.0x; the only names at Enova-like P/B are CACC (4.4x — extreme niche moat, 28% ROE) and FirstCash (5.3x — scarcity asset). The bank-charter lens is sobering for the re-rating narrative: Synchrony earns 20–26% ROTCE with deposit funding and gets 7.6x earnings and 1.6x book. Deposit funding buys a lower cost of equity, not a higher multiple — if Grasshopper makes Enova more bank-like, the cross-section says its multiple should compress toward SYF/OMF, not expand. Downside anchor: the Santander Consumer take-private (2022) went at ~1.0–1.1x book / ~6–7x earnings — on Enova’s current book that is ~$56–62/share; on FY26 GAAP-ish EPS, ~$85–105/share.

Justified-multiple identity. A standard justified-multiple identity: justified P/B = (ROE − g)/(COE − g), COE 10–12% for consumer-credit cyclicals. At Enova’s actual 24.3% ROE and g of 2–4%, justified P/B is 2.5–2.8x; on ~31% ROTCE, justified P/TBV ~3.5–3.8x. The market pays 4.13x / 5.2–5.5x. Inverting: to defend the current price you need either ~36% ROE sustained (no US consumer lender sustains that through a cycle — CACC’s ~26–28% is best-in-class) or ~24% ROE with ~5.4% perpetual growth at a 10% COE — i.e., you must believe the regulatory tail never lands, funding costs fall durably (Grasshopper), and credit never mean-reverts. The price is underwriting persistence, not level.

Embedded expectations — what must be true for $232 (INTERPRETATION):

1. Grasshopper approved and closed H2 2026 without crippling conditions — the 10-K itself concedes the price embeds completion.
2. ≥25% adjusted-EPS growth delivered in FY26 and double-digit beyond — which requires the +39%-growth SMB book to season at ~4.6% NCOs. At a historically normal 8x earnings, $232 requires ~$29 of EPS — 1.8x the FY26 guide, i.e., FY2028 delivery with zero multiple contraction; at the decade-top 12x it still requires ~$19.40, roughly FY2027 delivered flawlessly.
3. No credit recession in the non-prime tier — the record deep-subprime stress one tier below stays below Enova’s customer.
4. No federal 36% cap and a static state cap map.
5. Funding cost keeps falling and the buyback resumes post-close.

What the market is pricing correctly: the franchise is genuinely the best in the cohort (ROA 5.3%, ~24% ROE, nine-quarter SMB NCO plateau, funding access improving while peers’ markets widened); the capital cycle favors incumbents; the federal posture is benign. What it is pricing incorrectly or not at all: (a) any discount for credit cyclicality — the tier-divergence question is assumed resolved in Enova’s favor before the evidence exists; (b) deal-break risk — near-zero discount despite the 10-K’s own concession; © quality-of-earnings haircuts — the $774M Level-3 FV premium (~58% of equity) and SBC-free “adjusted” EPS taken at face; (d) the bank-lens evidence that deposit funding compresses rather than expands lender multiples; (e) insider distribution into the highs.

Scenario analysis (12–24 months, ranges not targets, ASSUMPTION-driven):

Scenario	Key assumptions	EPS basis	Multiple	Implied value range
Base — guide delivered, deal closes	FY26 adj EPS $16.2–17.0; Grasshopper closes H2’26 with manageable conditions; consumer NCO ≤16%, SMB ≤5.5%; partial funding benefit FY27	FY27E ~$18–20	10–13x	$180–260
Bear — tier divergence resolves against ENVA	SMB DQ uptick broadens; SMB NCO reverts 6–7%; FV premium compresses through the ΔFV line; de-rate to historical	FY27E ~$10–13	7–9x	$70–117
Bull — charter + funding transformation + compounding	Clean approval ≤Q4’26; ≥25% EPS growth FY26 and FY27; deposit migration begins; “proven compounder” multiple holds	FY27E ~$21–23	13–16x	$273–368
Deal-break (overlaps bear)	Denial, or hearing/delay past 2026; deal premium exits; standalone EPS intact but trust discount	FY26E ~$15.5–16.5	7–9x	$109–149

The asymmetry is the point: the base case is centered on the current price ($180–260 vs $232.52) — the market has already paid for the base case. The bull case requires everything right and a multiple at the top of any lender cross-section. The bear and deal-break ranges sit 35–70% below. Deal-break and credit-bear are additive, not mutually exclusive.

Verdict: at $232 the market is underwriting a flawless base case plus a free option on the bull case, at the richest own-history multiples on record. The underlying earnings power is real and growing ~20–25%; the price has simply pulled two-plus years of that growth forward. Embedded expectations leave no room for the ordinary accidents of consumer credit.

---

11. Variant Perception

Consensus belief (read from the multiples and the tape): Enova is a quality compounder being re-rated, not a subprime lender at a cyclical peak. Evidence: composite own-history percentile 96.5th; +102% in twelve months with a 1-yr Sharpe ~2.5; ~14x FY26 guidance versus a 5–9x historical home; a buy-side narrative of SMB pivot + Grasshopper funding transformation + buyback capacity. The factor data add a useful nuance: the re-rating is idiosyncratic — financials, fintech, size and beta factors were all headwinds over the past year while the stock doubled, meaning the market is specifically and deliberately repricing this company, not lazily beta-lifting the group. Consensus is paying attention. Consensus may still be wrong.

Strongest bull case: this is OneMain-in-2017 — a proven, self-funding, high-ROE lender early in a favorable mix shift (SMB at 70% of book and ~70% net revenue margin), with a pending structural funding upgrade and $368M of buyback capacity equal to 6.3% of the cap table. If ≥25% EPS growth compounds through FY27, today’s ~14x guide is cheap on FY27 numbers even at a flat multiple — and the 5–9x history was the mispricing, not today. Grasshopper approval converts the weakest structural feature (8.2% wholesale funding) into a potential moat ingredient while retiring the true-lender vulnerability over a third of the origination engine.

Strongest bear case: 4.1x book and 18.9x TTM GAAP earnings for a non-prime lender at the 96.5th percentile of its own history, priced for perfection exactly when: deep-subprime credit is at 32-year delinquency highs one tier below its customer; the growth engine is the least-seasoned book the company has ever carried (SMB +39%, nine quarters into a plateau that has never been tested by a downturn); earnings embed a $774M Level-3 premium that growth mechanically inflates; insiders are distributing ($99M sold in 24 months, one open-market buy, the chairman selling at $231 this month); and ~$350M of the thesis rests with regulators whom organized advocacy is lobbying to block. The justified-multiple identity requires ~36% ROE or ~5.4% perpetual growth to defend the price. And the bank-lens precedent (SYF) says successful deposit funding compresses the multiple.

The 3–5 assumptions that matter most:

1. SMB credit holds (NCO ~4.6% ±0.5%, 30+ DQ ≤7.5%) as the +39% book seasons. Falsifies bull: Q2’26 print (July 23) shows SMB NCO >5.5% or DQ >8%. Falsifies bear: 2–3 more quarters on the plateau.
2. Grasshopper approved H2’26 without product/CRA conditions that gut the economics. Falsifies bull: OCC grants a public hearing, or no approval by Q4’26. Falsifies bear: a clean approval letter.
3. The non-prime consumer tier does not crack (consumer NCO ≤16%). Falsifies bull: consumer NCO >17% / DQ >9% while SYF/ALLY keep improving — the divergence resolving against Enova.
4. Funding cost keeps falling. Falsifies bull: warehouse repricing wider, or a concession-priced ABS print during private-credit stress.
5. No 36% federal cap / no acceleration of the state-cap ratchet. Falsifies bear: legislative calendar stays dead through 2027.

The variant perception (INTERPRETATION): the disagreement with consensus is not about the business — this analysis arguably holds Enova’s execution and competitive position in higher regard than the average holder — it is about what is being paid for it. Consensus prices the charter as done, the credit cycle as exempt, and the re-rating as permanent. The factor-positioning read says the move is idiosyncratic and momentum-driven, which cuts both ways: real buyers, real conviction — and a tape that will gap violently (eight ±10% days in 60 months) the first time an assumption breaks. The July 23 print is the first falsification checkpoint.

---

12. Fact vs. Interpretation

Claim	Status
FY2025 revenue $3,151.7M, NI $308.4M, diluted EPS $11.52; Q1’26 revenue $875.1M, NI $91.1M	FACT (10-K/10-Q, EDGAR-reconciled)
SMB = ~70% of the $5.3B portfolio; SMB receivables +38.6% y/y; originations +42% Q1’26	FACT (10-Q)
Entire loan book at fair value since 1/1/2020; no CECL allowance; $774M FV premium (~58% of equity)	FACT (10-K)
Consumer NCO range-bound 12.7–17.3% since 2022; SMB on a 9-quarter 4.4–4.8% plateau; SMB 30+ DQ 6.2%→7.1% QoQ Q1’26	FACT (10-K/10-Q tables)
Computed ROE 24.3% FY25 (ROIC.ai’s 16.7% does not tie); ROA 5.3%	FACT (computed from statements)
~$1.0B buybacks 2019–2025; share count −31% since 2020; $367.8M remaining on $400M authorization	FACT (cash-flow statements, 10-Q)
Grasshopper ~$369M, 50/50 cash/stock, pending OCC/Fed; NCRC/Woodstock/CRL/NCLC opposition; close guided H2 2026	FACT (8-K, S-4, validated PRs)
Own-history valuation percentiles: composite 96.5th, P/B & P/S 99.6th	FACT (AZI, as computed; context not target)
Insider tape: one open-market buy vs ~$99.3M sold in 24 months	FACT (133 EDGAR ownership filings parsed)
CFPB 2023 consent order terminated in full 2025-08/09	FACT (CFPB enforcement page)
Moat = narrow scale/execution advantage with zero captivity	INTERPRETATION
Fair-value growth-flattery and mark risk are the central QoE issue	INTERPRETATION (mechanics FACT, materiality judgment)
Brazil is a zero-valued option; management ROA-parity and accretion claims are unproven	INTERPRETATION
Market is underwriting deal completion + ≥25% growth + no credit turn at $232	INTERPRETATION
Insider selling is “taking chips off” rather than routine diversification	INTERPRETATION (mostly discretionary blocks; FACT that only Fisher’s stream is partly 10b5-1)
Mid-2026 credit tier-divergence will resolve against Enova’s tier	OPEN QUESTION — Q2’26 print is first test

---

13. Open Questions

1. Q2 2026 print (July 23, 2026 — after this report’s date): does SMB NCO hold ~4.6% and SMB 30+ DQ stay ≤7.5% as the +39% book seasons? Did buyback pace re-accelerate?
2. Grasshopper: will the OCC grant opponents’ public-hearing request? Final consideration ($350M cap vs $369M press), stock-leg dilution (~4–5%, unverified), deposit costs and pro-forma BHC capital ratios — not yet quantifiable from filings.
3. FV loss-assumption sensitivity: what does a +10%/+25% charge-off assumption do to the $774M premium and to earnings? Not disclosed; candidate management question.
4. Marketing efficiency: is Q1’26’s 22%-of-revenue marketing a mix effect (SMB commissionable originations) or CAC inflation? Determines whether growth’s marginal cost is rising.
5. The excluded-state map: precise overlap of Enova’s 11–13 non-consumer states with the 36%-cap universe (filings don’t enumerate).
6. Avant’s scale/profitability (private): the largest unverifiable consumer competitor.
7. 2019 segment-level NCO/delinquency rates (FY2019 10-K predates the mirrored corpus) for a true pre-COVID credit benchmark.
8. Whether May–July 2026 insider sales were 10b5-1-planned (Fisher’s July sales read as discretionary blocks; plan status unverified).
9. The two Audit Committee resignations (2026-07-10): genuinely planned, or friction ahead of BHC conversion? Stated non-disagreement is unverifiable from outside.

---

14. What Must Be True

Bull case — what must be true, and the falsification test:

• SMB credit holds the 4.4–4.8% NCO plateau (DQ ≤7.5%) while the book seasons. Test: Q2’26–Q4’26 prints — SMB NCO >5.5% or DQ >8% breaks it.
• Grasshopper is approved by Q4 2026 on terms that preserve the funding-cost thesis. Test: an OCC public-hearing grant or silence past year-end breaks it; a clean approval letter confirms it.
• FY26 adjusted EPS ≥ ~$16.20 (guide) with FY27 growth ≥ mid-teens. Test: a guidance cut or a print below $3.60 adjusted in Q2 breaks the trajectory.
• Funding cost keeps falling (≤8.0% by YE26) and the buyback resumes post-close. Test: warehouse repricing wider or a concession ABS print breaks it.
• No federal 36% cap through 2027. Test: committee movement on S.3721/S.3793-type bills.

If all five hold: FY27 EPS ~$18–23, and even a de-rating to 10–13x supports $180–300 — the price is roughly fair-to-cheap on FY27 delivery.

Bear case — what must be true, and the falsification test:

• The mid-2026 credit tier-divergence resolves against Enova — record deep-subprime stress migrates up-tier. Test: consumer NCO >17% or SMB plateau breaking while SYF/ALLY keep improving confirms it; 2–3 more clean quarters kill it.
• The unseasoned SMB book is the weak point: +39% growth originated adverse selection that seasons into 6–7% NCOs. Test: vintage-level disclosure at the next two prints.
• The FV premium proves pro-cyclical: marks compress through the ΔFV line as growth slows, turning the accounting tailwind into a headwind. Test: FV/principal ratios (122.9%/112.2%) falling >150bp in a quarter.
• Grasshopper fails or is conditioned into irrelevance, removing the funding-transformation leg while ~$369M and management bandwidth are spent. Test: OCC/Fed docket.
• The multiple mean-reverts to the 5–9x historical home for execution-advantage lenders. Test: any of the above arriving — the multiple does not de-rate on its own, it de-rates on evidence.

If the bear path activates: FY27 EPS ~$10–13 at 7–9x = $70–117, i.e., 50–70% downside from $232 — the same round trip the stock has made twice in five years.

---

15. Source Appendix

See Appendix B — Source Appendix in the combined report (ENVA_source_appendix.md), which lists every primary source relied upon: the FY2025 10-K (filed 2026-02-20), Q1 2026 10-Q (filed 2026-04-23), the 60-month 8-K/DEF 14A/S-4 corpus, 133 EDGAR ownership filings, the Q3’25–Q1’26 earnings-call transcripts, CFPB/OCC/regulatory docket items, validated press releases, and the third-party data feeds (ROIC.ai, AZI Trading, FactorsToday) with their reconciliation status.

The analytical body above (§1–§15) carries no recommendation and no price target; the only opinion in this document is the labeled Kimi’s Take block at the top, which is the author’s own view. This report is independent analysis and general information, not investment advice.

APPENDIX A — Standard Diligence Questionnaire

Enova International, Inc. (NYSE: ENVA) | Report date: 2026-07-19 | Data as of Q1 2026 (10-Q filed 2026-04-23) Companion diligence appendix to the ENVA research report above. Every answer is grounded in the underlying filings and the sources listed in Appendix B; claims are labeled Fact (from filings/data), Interpretation (analyst judgment on facts), or Assumption (unverified premise) where the distinction matters. This appendix carries no recommendation and no price target.

---

1. General

What thoughtful questions have other investors asked about this company?

The sell-side Q&A record across the last three earnings calls (Q3’25 2025-10-23, Q4’25 2026-01-27, Q1’26 2026-04-23 — full transcripts in output/ENVA/transcripts/) shows analysts pressing on exactly the fault lines this report identifies:

• Capital return constraints vs. the Grasshopper cash leg. Scharf (Citizens), Q3’25 and Q4’25: pressed on covenant-limited buyback capacity (~$80M into Q4’25, ~$106M into 2026), the possibility of a dividend, and post-close buyback continuity. Fisher’s answer: “everything is on the table… a dividend at the right time, although that’s… usually a better tool when the stock is more fully valued.” Cunningham’s capital hierarchy: organic growth first, buybacks second, M&A “down the list,” with tangible capital maintained around ~17–18%. (Fact: commentary on calls.)
• Marketing efficiency vs. guidance. Scharf (Q3’25) and Caintic (BTIG, Q1’26): marketing ran below guide through 2025 (~18% vs ~20% guided) then overshot to 22% of revenue in Q1’26 vs an “upper teens” guide — is the model structurally more profitable, or is CAC inflating as growth is pushed? Management’s answer: demand-driven and unit-economics gated. This remains an open question in the memo (§13.4): mix effect (SMB commissionable originations) vs. genuine CAC inflation. (Fact: the overshoot; Interpretation: what it means.)
• Credit deterioration pockets. Caintic (Q3’25) and Orenbuch (TD Cowen, Q1’26, on April reads): any pockets of weakness? Fisher flatly denied any; Cunningham said early Q2’26 reads show “consumer charge-offs towards the lower end of the range.” (Fact: exchange; management’s credit claims are hypotheses — the falsification test is the July 23 print.)
• Rate caps and regulation. Caintic and Hecht (Q4’25): Cunningham noted federal 36%-cap bills have been introduced “17, 18 years in a row… probability very, very low,” called January 2026 credit-card rate-cap talk “probably… a positive for us” (card-declined borrowers migrate to Enova), and described the state tape as “relatively quiet.” (Fact: commentary.)
• Funding-market stress. Caintic (Q1’26, on private-credit nerves): Cornelis answered with the four warehouse upsizes (+$377M) executed at existing terms with spreads held firm — no widening “like you’ve seen maybe in some of the other funding markets.” (Fact.)
• SMB competition and share. Joseph (Stephens, Q1’26) and Caintic (Q3’25): Cunningham said the competitor set is unchanged and banks are getting more conservative in SMB; Fisher added that prime lenders who dip into near-prime “pull back very, very quickly… they’re not good at lending above 36%.” (Fact: commentary; the embedded-finance threat — Block, PayPal, AmEx — was notably not the frame analysts used, and remains the report’s own forward-looking concern.)
• Deal timing. Caintic (Q1’26) mis-stated a “second quarter close” for Grasshopper; Cunningham corrected him — “second half of 2026 is still our expectation.” (Fact.)

The memo’s own open questions (§13) extend the list: FV loss-assumption sensitivity (what does a +10%/+25% charge-off assumption do to the $774M fair-value premium — not disclosed); the excluded-state map vs. the 36%-cap universe; whether May–July 2026 insider sales were 10b5-1-planned; and whether the two simultaneous Audit Committee resignations (2026-07-10) were genuinely planned.

---

2. Cyclicality & Earnings Nature

Are earnings at a cyclical high or low?

Interpretation: earnings are at a cyclical high in level, though credit metrics themselves are at mid-cycle, not peak-stress, levels. FY2025 net income of $308.4M is an all-time high on an all-time-high revenue base ($3,151.7M, +18.6%), and Q1’26 net income of $91.1M (+24.9% y/y) extends it. Computed ROE of 24.3% (FY2025) is the best clean-year figure of the past decade (2020’s ~58% was a COVID fair-value-recovery distortion). Two caveats keep this from being a simple “peak earnings” call: (Fact) consumer NCOs of 14.3% (Q1’26) sit inside the 12.7–17.3% range that has held since 2022 — credit is at its normal steady state, not a trough; and (Fact) the 2020 COVID troughs (consumer NCO 4.6%, SMB 0.7% in Q2’21) show how much room there is for losses to improve in a genuinely benign environment — meaning today’s losses are not cyclically depressed. The cyclicality risk is the mirror image: the last true stress test (2022) took the stock down ~45% peak-to-trough while the business remained solidly profitable (Fact: price; Interpretation: the equity’s beta to sector fear exceeds the business’s beta to credit).

Are earnings driven by the external environment or the company’s own actions?

Both, with an unusually clear separation (Interpretation from the record). Company-driven: the OnDeck acquisition (2020) and the SMB pivot it enabled (now ~70% of the book), nine straight quarters of SMB NCOs in a 4.4–4.8% band, operating leverage (opex/revenue 42.3% in 2022 → 34.4% in 2025), liability management that cut funding cost from 9.31% (2024) to 8.18% (Q1’26), and ~$1.0B of buybacks. External: the 2022 rate shock drove the drawdown; the CFPB’s termination of the 2023 consent order (2025-08/09, Fact) removed the largest regulatory overhang without any company action; and the mid-2026 credit tape — near-prime improving (SYF/ALLY prints) while deep-subprime sits at 32-year delinquency highs (Fact, cross-read of Synchrony and Ally filings) — is an environment Enova benefits from but does not control.

How stable are revenues?

Fact: revenue has grown every year 2019→2025 ($1,174.8M → $3,151.7M, 17.9% CAGR; +18.6% in FY2025, +17.4% y/y in Q1’26). Interpretation: the growth is real but the revenue stream is not recurring — nothing is contracted. The SMB book turns in ~15 months average term, consumer LOC is revolving but re-underwritten in practice, and marketing at ~20% of revenue is the price of continuously re-acquiring demand. Repeat borrowers (who default less, per the 10-K) are the only quasi-recurring element. Revenue is stable so long as origination capacity and funding hold; it is not stable the way a subscription or servicing book is.

What is the outlook for products/services?

Fact/Interpretation: the product grid is deliberately shifting mix — SMB installment/LOC (48–49% yields, ~4.6% NCOs, ~69–71% net revenue margin) growing receivables +38.6% y/y, versus consumer installment/LOC (90–147% yields, 14–16% NCOs) growing +7.6%. Management’s FY2026 guidance (raised at Q1’26; management hypotheses): originations ~+20%, revenue growth similar, adjusted EPS ≥+25%. The consumer book is managed to a ~15% NCO equilibrium, not maximized for growth; the SMB book is the growth engine and the open seasoning question.

How big will this market be — growing, shrinking, domestic or international?

Fact (company estimates, marketing-grade): $85B US non-prime consumer opportunity, $313B US SMB loan market (sub-$250K loans = 78% of SMB applications, 2025 Federal Reserve Banks study), $49B Brazil. Interpretation: the serviceable market is far smaller — roughly 20 states plus DC cap small-loan APRs at/below ~36%, excluding the high-APR consumer product, and the state-cap ratchet is slow but one-directional (Illinois PLPA cut subprime lending 38% post-2021). Enova’s $5.3B book is a mid-single-digit share of its serviceable market: growth is a share-gain story inside a slowly shrinking consumer-addressable map, against a genuinely growing SMB demand backdrop (bank retrenchment sub-$250K, post-2020 business formation). International is immaterial: Brazil (Simplic) is 2.4% of revenue and cumulatively loss-making after 11 years ($73.6M of fully-reserved NOLs at YE2025) — a zero-valued option in this report’s work (Interpretation).

---

3. Business Quality & Competitive Moat

Is the industry getting more or less competitive?

Interpretation: less competitive at the balance-sheet-lending core, more competitive at the embedded-finance margin. The 2020–21 VC/SPAC flood destroyed sub-scale returns and the 2022–23 funding shock killed the weak (CURO Chapter 11, March 2024; Oportun’s cumulative FY23–24 losses; Elevate’s divestitures; UPST −90%+ from peak). 2024–26 is a discipline phase where funding returned selectively for proven performers — Enova upsized four warehouses by $377M at unchanged spreads in Q1’26 while other subprime funding markets widened (Fact, Q1’26 call). The countervailing force: embedded-finance players (Block/Square Loans, PayPal Working Capital, AmEx Business Blueprint) bring captive merchant bases and payments data — structurally cheaper CAC — into SMB lending, and early-2026 private-credit retail redemption stress (Stone Ridge honoring ~11% of redemption requests) signals marginal funding tightening at the edges (Fact: press reports 2026-03; Interpretation: this historically helps funded incumbents and hurts origination-for-sale models).

How profitable is the business (ROIC, ROE)?

Fact (computed from EDGAR-reconciled statements): FY2025 ROE 24.3% (net income / average equity), ROA 5.3%, honest decomposition ~5.3% ROA × ~4.6x assets/equity. ROIC in the industrial sense is not meaningful for a lender (debt is inventory, not capital structure); the sector analog is ROA and ROTCE (~31% estimated). Peer comparison (Fact, ROIC.ai FY2025): OneMain ~3% ROA, Regional ~2%, Oportun/LendingClub ~1%, Upstart ~breakeven — Enova is best-in-group by a wide margin. Caveat (Fact): ROIC.ai’s own “ROE” (16.7%) and “BVPS” do not tie to the filings; the 24.3% figure is this report’s computation and reconciles to EDGAR exactly.

How profitable is the industry — how many competitors, what barriers to entry?

Fact/Interpretation: the consumer non-prime profit pool concentrates in exactly two names — OneMain (scale + IG funding + branch collections) and Enova (online cost structure + analytics). Everyone else is sub-scale (OPRT), shrinking (WRLD/RM), unproven through-cycle (UPST), or dead (CURO). Barriers to entry are real but moderate: state-by-state licensing across 37–39 consumer states and 49-state SMB, 21 years of loss-curve data across ~69M transactions (the 10-K itself notes new portfolios incur “early losses associated with unseasoned loan portfolios”), and capital-markets access (10+ securitization shelves) that kept Enova funded at ~8% through the 2022–23 winter while rivals were cut off. But pricing power is set by statute, not by scarcity — entry barriers protect returns only up to the rate caps.

Can the business be easily understood?

Yes (Interpretation) — unusually so for its return profile: lend at 48–147% yields, lose 5–16% to credit, fund at ~8.2% wholesale, spend ~20% of revenue on customer acquisition, earn ~5% on assets. The one genuinely hard-to-understand layer is accounting, not economics: the entire loan book is Level 3 fair value, and the $774M premium over principal (~58% of equity) rests on management’s own models with no disclosed loss-assumption sensitivity (Fact).

Can it be undermined by foreign low-cost labor?

No (Interpretation). The product is domestic US credit, the labor content is analytics and servicing (1,836 employees, mostly US), and the regulatory perimeter is state-by-state US law. The foreign exposure is the other direction: Enova’s own Brazilian venture is the one losing money.

Do brands matter?

Marginally (Interpretation). CashNetUSA/NetCredit/OnDeck carry high Trustpilot/BBB ratings that reduce acquisition cost at the margin, but brand here is replicable spending, not captivity — the borrower is rate- and speed-shopping, and marketing at 19.7% of revenue (22% in Q1’26) is the quantified proof that brand does not retain anyone.

What is the nature of competition?

Interpretation: competition happens on acquisition spending and speed, not on product differentiation — cash is a commodity. In consumer, the fight is for the same liquidity-stressed borrower across online, branch, BNPL/EWA-adjacent substitutes; in SMB, it is increasingly against embedded-finance distribution (Block, PayPal, AmEx) with structurally lower CAC. The competitive variable that actually separates winners is underwriting accuracy per dollar of marketing — an execution contest re-run every quarter.

What are customers’ switching costs?

Approximately zero (Fact-adjacent Interpretation, and the report says so plainly): a borrower can apply to ten lenders in ten minutes; there is no contract, ecosystem, or lock-in. High repeat-borrower rates reflect habitual liquidity stress in the customer base, not captivity.

---

4. Financial Condition & Balance Sheet

Are there assets not fully recognized on the balance sheet?

Interpretation: nothing of the classic hidden-asset kind (no appreciated real estate, no understated intangibles of consequence). If anything the balance sheet runs the opposite risk: loans are carried at fair value above principal — a $774M premium (~58% of equity) whose existence depends on Level 3 models (Fact). The closest thing to an unrecognized asset is the franchise value of the data/regulatory infrastructure (21 years, 69M transactions, 37–39 state licenses), which is real but unbooked. Brazil’s $73.6M of NOL carryforwards sit under a full valuation allowance — an asset the balance sheet conservatively values at zero (Fact).

Off-balance-sheet liabilities?

Fact: small and disclosed. The legacy Texas CSO/CAB guarantee (obligation to purchase defaulted third-party loans) is $17.9M (P+I) at 3/31/26 — under 0.4% of the book, immaterial. Bank Programs (31.6% of FY2025 originations) carry an option, not an obligation, to purchase loans, with no performance guarantee — but the true-lender doctrine is the real contingent exposure: if Enova were deemed the true lender on bank-program loans, the consequences (usury, enforceability, licensing) would be regime-level, not balance-sheet-level (Fact: structure per 10-K; Interpretation: materiality).

How conservative is the accounting?

Mixed, and this is the report’s central quality-of-earnings finding (Fact + Interpretation). Aggressive-leaning: the 2020 fair-value election means no CECL allowance exists; credit cost flows through a Level 3 mark-to-model “change in fair value” line; growth mechanically flatters earnings (new loans enter near par and are marked up over following quarters — Q1’26 consumer marks net-offset NCOs, $215.5M ΔFV vs $227.6M realized); disclosed sensitivity covers only the discount rate (±100bp = ~$30M, ~2.2% of equity), with no loss-assumption sensitivity — a genuine transparency gap. “Adjusted EPS” structurally excludes stock comp (~$1.24/share, ~80% of the FY25 GAAP→adjusted gap with intangible amortization). Conservative-leaning: all marketing is expensed as incurred (no DAC asset); origination fees/costs are not deferrable under FVO; marks have been stable (~115% consolidated premium) for two years and are corroborated quarterly by realized losses; and Enova does not sell delinquent loans pre-charge-off (charge-off at 65 days consumer / 90 days SMB, then sell — so NCOs are not suppressed by the loan-sale trap found at peers) (Fact).

How CapEx-hungry is the business?

Not CapEx-hungry in the industrial sense — the question doesn’t map directly (Interpretation). Enova is a balance-sheet lender: its “capital expenditure” is loan-book growth, which consumes funding capacity roughly 1:1 with receivables at constant ~3.4x leverage. The correct sector analog to maintenance-vs-growth CapEx is the split between earnings that can be distributed and earnings retained to fund receivables growth: FY2025 PPNR-style earnings power was ~$746M (net revenue $1,830M − opex $1,084M), less $339M interest = $401M pretax — but distributable earnings are materially below reported earnings while receivables grow 23–27% (Fact: figures; Interpretation: the framing). Industrial CapEx itself is trivial (no branches; operations & technology 8.2% of revenue, expensed).

---

5. Capital Allocation & Management

How much FCF does the business generate, and how does management use it? What is the philosophy?

“Free cash flow” does not map to a lender (Interpretation, stated explicitly in the report): FY2025 operating cash flow of $1.82B adds back the non-cash ΔFV charge and excludes loan-book growth (−$2,445.8M in investing), so it is not FCF the industrial way. The correct analog is PPNR/earnings power: ~$746M FY2025 pre-credit-cost operating earnings, ~$401M pretax after interest (Fact). Uses of that earnings power, in management’s stated hierarchy (organic growth first, buybacks second, M&A “down the list” — Cunningham, Q4’25 call): ~$1.0B of buybacks 2019–2025 (share count −31% from the 2020 peak), a pending ~$369M acquisition, and continuous reinvestment in receivables. No dividend; none anticipated (Fact).

Significant acquisitions recently?

Fact: two of consequence in the record. OnDeck (closed October 2020, $113.2M consideration — $104.3M stock + $7.2M cash + $1.6M awards), bought near its COVID trough at ~1x book; SMB is now ~70% of the $5.3B portfolio with its own securitization shelf — Interpretation: arguably one of the best fintech acquisitions of the decade. Grasshopper Bancorp (pending; agreement 2025-12-10, amended 2025-12-18 to fixed 50% cash / 50% stock; ~$369M per press vs ~$350M cap per 8-K — open item): ~28% of YE2025 equity for a digital national bank with ~$3B deposits — the boldest and riskiest allocation in company history, deployed against an 8.82% corporate funding cost (Fact). Small blemishes: the Linear JV write-down (−$16.5M, 2024); Pangea (2021, $32.9M) immaterial; Brazil’s eleven years of losses.

Is the company buying back shares?

Fact: yes, but decelerating. $214.6M in 2025; a new $400M authorization (2025-11-12, through 2027-06-30) with $367.8M remaining at 3/31/26 — 6.3% of market cap. Pace: $38M (Q3’25) → $35M (Q4’25) → $15.6M (Q1’26 at ~$152.75 average) as management preserves flexibility for the Grasshopper cash leg. Interpretation: the 2023–24 purchases at ~$49–65 (~1–1.5x book) were outstandingly accretive; at ~4x book, repurchases are EPS-accretive but BVPS-dilutive — the Q1’26 10-Q itself notes buybacks partially offset BVPS growth.

Is the company issuing large amounts of new shares to insiders?

Fact: no large issuance — but comp is heavily equity-based. SBC was $33.1M in FY2025 (~10.7% of net income); long-term incentives are 50% time-based RSUs and 50% quarterly-granted at-market stock options with zero performance conditions; the diluted/basic share gap is ~6%. Net share count still falls because buybacks outpace issuance. The Grasshopper stock leg will issue ~1.1–1.3M new shares (~4–5% dilution — Assumption pending final 424B3 verification).

What is the compensation policy of directors/management?

Fact (DEF 14A filed 2026-04-02): short-term incentives are one-third each Revenue (GAAP), Adjusted EBITDA, Adjusted EPS, with a 50–200% payout range; 2025 achievement 103%/110%/113% → 123.2% payout. Long-term incentives carry no performance shares, no ROE/TSR/credit hurdles. Fisher’s 2025 total comp was $11.70M; Cunningham $6.83M; Fisher’s Executive Chairman package (~$8–9M/yr: $825k base, 130% STI, 520% LTI targets) is rich for a non-CEO role. Say-on-pay: 86.6% (2025), 95.5% (2026). Interpretation: the EPS metric in STI at least aligns with per-share outcomes; the performance-free LTI is a genuine alignment weakness.

What are the motivations of management?

Interpretation from the observable record: an equity-heavy comp model monetizing a stock that 5x’d off its lows. Fact: across 233 parsed transactions over 24 months (133 EDGAR ownership filings), there was exactly one open-market purchase (a director, 1,700 shares at $115.95, September 2025) against ~$99.3M of sales — Fisher $54.5M (monthly 10b5-1 plus large discretionary blocks, including ~$231.50 on July 14–15, 2026, days before this report), Cunningham $18.9M (routine exercise-and-sell), two directors liquidating entire direct stakes in February 2026 at ~$172, and one director selling $2.7M in May–June before resigning from the Audit Committee July 10. Nobody bought during the February–March 2026 dip into the ~$120s. Interpretation: the complete absence of discretionary buying, and the chairman’s discretionary selling into all-time highs ahead of the Grasshopper close, sits uncomfortably next to the corporate buyback narrative — it reads as taking chips off the table, though Fisher’s stream is partly plan-driven (Fact) and the 10b5-1 status of the May–July sales is unverified (open question).

---

6. Valuation & Market Data

Is the stock an ADR, MLP, or K-1 issuer?

Fact: none of the above. Enova is a US domestic C-corp (NYSE: ENVA), single class of common stock, no dual-class, directors elected annually, preferred authorized (25M shares) but unissued. US investors receive a standard 1099; there is no partnership, trust, or ADR structure. Post-Grasshopper close, Enova would become a bank/financial holding company — a regulatory status change, not a security-structure change.

What is the dividend policy?

Fact: no dividend is paid and none is anticipated (10-K Item 5). Capital return is exclusively repurchases. Fisher (Q3’25 call) left the door open — “a dividend at the right time, although that’s… usually a better tool when the stock is more fully valued” (management commentary).

How profitable is the business?

Fact: FY2025 net margin 9.8% ($308.4M / $3,151.7M), ROA 5.3%, computed ROE 24.3%, adjusted EBITDA $820.9M (26.0% of revenue Q1’26). TTM GAAP diluted EPS $12.29; TTM adjusted EPS $13.85 (SBC-inclusive haircut ~$1.2–1.3/share lower). PPNR-style pretax earnings power ~$401M (FY2025). These figures are recomputed from filings; third-party feeds (ROIC.ai) tie on revenue/net income but not on ROE/BVPS (Fact, reconciliation documented).

Is net income diverging from cash from operations?

Fact + Interpretation: for a lender this divergence is structural, not a red flag — a standard lender convention is to judge growing balance-sheet lenders on PPNR, not NI-vs-OCF. FY2025 CFO of $1.82B vastly exceeds net income because the $1.32B ΔFV charge is non-cash and loan growth sits in investing. The more meaningful checks all pass: the ΔFV charge tracks realized NCOs closely quarter-by-quarter (Q1’26 consumer: $215.5M vs $227.6M); marks are stable at ~115% of principal for two years; and no pre-charge-off loan sales suppress the NCO series. The residual divergence to watch is not NI-vs-cash but NI-vs-distributable earnings: at 23–27% receivables growth and ~3.4x leverage, a meaningful share of reported earnings must be retained to fund the book (Interpretation).

---

7. Risks & Downside

What factors would cause the stock to decline?

From the memo’s risk register (all Fact-grounded; likelihood/impact judgments are Interpretation): (1) a credit-cycle turn — consumer NCOs breaking the 14–16% band or SMB breaking its 4.4–4.8% plateau as the +39%-growth book seasons (SMB 30+ delinquency already ticked 6.2%→7.1% QoQ in Q1’26); (2) Grasshopper denied, delayed past H2 2026, granted a public hearing, or conditioned into irrelevance — the 10-K itself concedes the stock price embeds completion; (3) fair-value mark reversal — the $774M Level 3 premium (~58% of equity) compressing through the ΔFV line, the mechanism that flattered earnings in growth quarters running backward; (4) the macro tape itself — beta ~1.36, ~39% vol, seven ±10% days in 60 months, and demonstrated contagion sensitivity (the Tricolor/PrimaLend failures cost the stock 13.5% in six weeks with zero company exposure); (5) state rate-cap ratchet and the federal 36% tail; (6) funding-market disruption against an 8.2% wholesale stack with four warehouses hitting revolve-end within 15 months; (7) embedded-finance SMB competition compressing CAC economics; (8) governance/comp drift. The 2022 precedent is the practical calibration: the stock fell ~45% peak-to-trough in a scenario where the business stayed solidly profitable (Fact).

What is the risk of a catastrophic loss?

Interpretation: low for the enterprise, real for the equity. Catastrophic business loss requires the simultaneous failure of diversified funding (10+ shelves, no recourse maturity before December 2028, $1.1B liquidity — Fact), a credit event far beyond any in its 21-year history including COVID, and regulatory destruction of the yield model. The genuinely catastrophic single-scenario paths are regulatory: a federal 36% cap would terminate the consumer business as configured (low probability under this Congress — the single largest long-horizon tail), and an adverse true-lender regime shift could impair the 31.6% of originations run through bank partners (regime risk, not an earnings risk). For the equity, a 50–70% drawdown is a demonstrated, repeatable outcome of this business model’s beta (2008-era ancestor experience aside, the stock has made ~45–50% round trips twice in five years — Fact: price history; Interpretation: severity reading).

What is the chance of a total loss?

Interpretation: remote but not zero, and the honest answer requires separating the company from the stock. A total loss of the enterprise would require funding-market closure plus credit collapse plus regulatory assault simultaneously — against a franchise that stayed profitable through COVID, the 2022 rate shock, and the 2025 subprime-contagion scare, and whose nearest historical casualty (CURO) died of a debt-laden LBO structure Enova does not have (Fact: Enova’s leverage is 3.4x debt/equity against loan collateral, not an LBO stack). The equity can, however, lose most of its value without the company failing — at 4.1x book and the 96.5th percentile of its own valuation history, the margin of safety is thin precisely because the business is priced as if the three pillars (charter approval, credit stability, ≥25% EPS growth) cannot all crack at once (Interpretation, no recommendation implied).

---

8. Recent News & Events

Has the business environment changed recently?

Fact: yes, materially, in both directions. Favorable: the CFPB terminated Enova’s 2023 consent order in full (effective 2025-08-29/09-02) and the federal posture is the most benign in years (Small Dollar Rule enforcement deprioritized; Section 1071 delayed to 2028); funding costs fell ~113bp from the 2024 peak; Q1’26 delivered a beat with raised FY26 guidance. Adverse/watch: SMB 30+ delinquency ticked up QoQ; marketing intensity rose to 22% of revenue; the Tricolor (Chapter 7, 2025-09-10) and PrimaLend (Chapter 11, 2025-10-22) failures showed sector contagion can reprice the stock 13.5% regardless of fundamentals; and early-2026 private-credit retail redemption stress is a live funding-sentiment factor.

Significant acquisitions?

Fact: the pending Grasshopper Bancorp acquisition — ~$369M (50/50 cash/stock per the 2025-12-18 amendment) for a digital national bank ($1.4B assets, ~$3B deposits). Applications filed with the OCC and Federal Reserve 2026-01-16 (including BHC/FHC elections); Grasshopper holders approved 2026-02-02; organized opposition from NCRC/Woodstock Institute (2026-02-23), CRL hearing requests (Feb–Apr 2026), and an NCLC-led coalition letter to Congress (2026-05-07); close guided H2 2026, still pending as of this report. Management claims >15% year-one adjusted-EPS accretion and >25% with full synergies — unvalidated hypotheses; 2026 guidance excludes any deal contribution (Fact).

Any change in accounting policies?

Fact: no change in the window. The consequential election is older but frames everything reported today: the fair-value option for the entire loan book, effective January 1, 2020 (ASU 2019-05) — CECL was never adopted, no allowance exists, and credit cost flows through the Level 3 “change in fair value” line. Q1’26’s 19.3% effective tax rate (vs ~23% normalized) flattered the quarter via excess SBC tax benefits — a one-period item, not a policy change (Fact).

Recent changes — new markets, facilities, management?

Fact: (a) Management — orderly succession effective 2026-01-01: CFO Steve Cunningham (CFO since 2016) to CEO, CEO David Fisher (since 2013) to Executive Chairman (≥2-year commitment), Treasurer Scott Cornelis to CFO; (b) Board — two Audit Committee directors (Goodyear, McGowan) resigned simultaneously 2026-07-10, stated non-disagreement, board cut to 10, Maria Veltre appointed (8-K/A 2026-07-15 assigns her to the Compensation Committee) — a monitoring footnote weeks before a bank-holding conversion that multiplies audit complexity (Interpretation); © Footprint — consumer state count expanded 37 → 39 by Q1’26; management named California, Pennsylvania, and Ohio as post-charter expansion targets (Q4’25 call commentary); (d) Facilities — revolver upsized to $825M at SOFR+3.25% (2025-08-28/09-02), four warehouses upsized $377M at unchanged spreads (March 2026), new ABS coupons down to 5.65% (2025-2) from 6.84% (2024-1); (e) Capital — new $400M buyback authorization (2025-11-12). Q2 2026 results land after market close on July 23, 2026 — four days after this report’s date; the SMB delinquency, buyback pace, and Grasshopper language in that print are the immediate event risk (Fact).

---

Diligence companion to the ENVA research report. No recommendation, no price target. All facts traceable to the underlying filings and the sources listed in Appendix B; interpretations and assumptions are labeled as such.

APPENDIX B — Source Appendix

Enova International, Inc. (NYSE: ENVA) | Report date: 2026-07-19 | Companion source appendix to the ENVA research report Every source the report relies on, organized by class. Local paths are relative to the repository root; the full filings corpus is mirrored under output/ENVA/sources/ (442-row MANIFEST.csv with SEC URLs). All web sources accessed/validated 2026-07-19. This appendix carries no recommendation and no price target.

---

(a) Primary — SEC filings

Core statements

• FY2025 Form 10-K (filed 2026-02-20) — output/ENVA/sources/10-K/2026-02-20_enva-20251231.htm — https://www.sec.gov/Archives/edgar/data/1529864/000119312526060461/enva-20251231.htm — Business section (product grid, yields, customer profile, TAM claims), MD&A (credit/funding tables), FVO accounting policy, FV sensitivity, risk factors (Madden/true-lender/CSO/MLA/1071), Note 11 contingencies, the concession that the share price embeds Grasshopper completion.
• Q1 2026 Form 10-Q (filed 2026-04-23) — output/ENVA/sources/10-Q/2026-04-23_enva-20260331.htm — https://www.sec.gov/Archives/edgar/data/1529864/000119312526174021/enva-20260331.htm — Note 2 FVO/loan tables (FV/principal 122.9% consumer / 112.2% SMB), Note 3 debt table, MD&A credit tables (SMB 30+ DQ 6.2%→7.1%), issuer-purchase table ($367.8M buyback remaining), liquidity.
• Prior-year 10-Ks FY2021–FY2024 (filed 2022-02-28, 2023-02-24, 2024-02-23, 2025-02-18) — output/ENVA/sources/10-K/ — FVO election history (1/1/2020, ASU 2019-05), OnDeck purchase-price note, Pangea note, prior-year quarterly credit tables, COVID +500bp discount-rate disclosure.
• Q4/FY2025 earnings release EX-99.1 (2026-01-27, accession 0001193125-26-024499) — https://www.sec.gov/Archives/edgar/data/1529864/000119312526024499/enva-ex99_1.htm — adjusted EPS/EBITDA reconciliations (GAAP→adjusted bridge; SBC add-back quantification).
• SEC EDGAR companyfacts API, CIK 0001529864 — pulled 2026-07-19 via scripts/edgar.sh — ROIC-to-EDGAR reconciliation (revenue/NI/LT debt tie exactly 2020–2025) and 2012–2019 allowance-rollforward tags (pre-COVID 2019 anchors: NCOs $570.7M; allowance $176.9M = 14.3% of gross).

Key 8-Ks (local corpus output/ENVA/sources/8-K/ unless noted)

• 8-K 2024-08-12 — 2024-08-12_d736055d8k.htm — https://www.sec.gov/Archives/edgar/data/1529864/000119312524198919/d736055d8k.htm — $500M 9.125% senior notes due 2029 issuance + $300M buyback authorization.
• 8-K 2025-07-24 (Item 5.02) — 2025-07-24_enva-20250724.htm — https://www.sec.gov/Archives/edgar/data/1529864/000095017025098354/enva-20250724.htm — CEO transition: Fisher → Executive Chairman, Cunningham → CEO, Cornelis → CFO, effective 2026-01-01 (comp terms in 8-K/A 2026-01-02, sources/8-K-A/).
• 8-K 2025-09-02 (event 2025-08-28) — 2025-09-02_enva-20250828.htm — https://www.sec.gov/Archives/edgar/data/1529864/000095017025112478/enva-20250828.htm — revolver upsized to $825M, SOFR+3.25%, maturity extended to Aug 2029.
• 8-K 2025-11-12 (Item 8.01) — 2025-11-12_enva-20251112.htm — https://www.sec.gov/Archives/edgar/data/1529864/000119312525277671/enva-20251112.htm — new $400M repurchase authorization through 2027-06-30, replacing the Aug 2024 $300M program.
• 8-K 2025-12-11 (Item 1.01, event 2025-12-10) — 2025-12-11_enva-20251210.htm — https://www.sec.gov/Archives/edgar/data/1529864/000119312525315025/enva-20251210.htm — Grasshopper Bancorp merger agreement (~$350M aggregate cap per 8-K vs ~$369M press; stock +11.7% that day).
• 8-K 2025-12-18 — 2025-12-18_enva-20251218.htm — https://www.sec.gov/Archives/edgar/data/1529864/000119312525325071/enva-20251218.htm — Merger Amendment No. 1: fixed 50% cash / 50% stock ($9.69 cash or 0.07637 ENVA shares per Grasshopper share), elections eliminated.
• 8-K 2026-07-10 + 8-K/A 2026-07-15 (Item 5.02) — 2026-07-10_enva-20260710.htm / sources/8-K-A/2026-07-15_enva-20260710.htm — https://www.sec.gov/Archives/edgar/data/1529864/000119312526300931/enva-20260710.htm — Audit Committee directors Goodyear and McGowan resign (stated non-disagreement); board cut to 10; Maria Veltre elected, then appointed to Compensation Committee.
• Earnings 8-Ks 2021-10-28 → 2026-04-23 (17 quarterly Item 2.02 filings) and ~15 securitization facility 8-Ks (ODAS IV series, ODR/RAOD/NCR/HWCR/NCLOC amendments) — output/ENVA/sources/8-K/ — the 60-month price-event map and funding-execution timeline.

Proxy & deal registration

• DEF 14A (filed 2026-04-02) — output/ENVA/sources/DEF_14A/2026-04-02_enva-20260330.htm — https://www.sec.gov/Archives/edgar/data/1529864/000119312526138937/enva-20260330.htm — CD&A (STI metrics: ⅓ Revenue / ⅓ Adj EBITDA / ⅓ Adj EPS; LTI with zero performance conditions), Summary Comp Table (Fisher $11.70M), say-on-pay 95.5%, ownership table.
• Form S-4 (2025-12-19) + S-4/A (2025-12-29) — output/ENVA/sources/S-4/2025-12-19_d27021ds4.htm, sources/S-4-A/2025-12-29_d27021ds4a.htm — https://www.sec.gov/Archives/edgar/data/1529864/000119312525326859/d27021ds4.htm — Grasshopper deal registration and target financials.
• 424B3 proxy/prospectus (Dec 2025) — https://www.sec.gov/Archives/edgar/data/1529864/000119312525338295/d27021d424b3.htm — Grasshopper stockholder vote mechanics (vote held 2026-02-02); final consideration and stock-leg share count still to be verified here (open item).
• Form 425 filings (2025-12-11, 2025-12-18) — output/ENVA/sources/425/ — deal-day communications; DEFA14A 2026-04-02 — output/ENVA/sources/DEFA14A/ — supplemental proxy solicitation.
• 133 Form 3/4/5 ownership filings, 2024-07 → 2026-07 — fetched live from EDGAR ownership XML (accessions in output/ENVA/sources/MANIFEST.csv), parsed to output/ENVA/2026-07-19/_scratch/form4_txns.csv (233 transactions) — insider-trading tape: one open-market purchase (Rice, 1,700 sh @ $115.95, 2025-09-10) vs ~$99.3M aggregate sales (Fisher $54.5M; Cunningham $18.9M; Gray/Tebbe full liquidations; Goodyear pre-resignation sales).

---

(b) Earnings calls & company releases

Transcripts (via ROIC.ai; full copies saved locally)

• Q1 2026 earnings call (2026-04-23) — output/ENVA/transcripts/ENVA_Q1_2026_earnings_call_2026-04-23.txt — FY26 guidance raised (originations ~+20%, adj EPS ≥+25%); Q2’26 guide; warehouse upsizes at unchanged spreads; ROA-parity and mix-agnosticism claims; analyst Q&A (Scharf, Caintic, Orenbuch, Joseph).
• Q4 2025 earnings call (2026-01-27) — output/ENVA/transcripts/ENVA_Q4_2025_earnings_call_2026-01-27.txt — initial FY26 guidance; capital hierarchy and covenant capacity; Grasshopper accretion framing ($125–220M synergies, >25% adj-EPS accretion by yr-2); rate-cap commentary; post-charter state “hit list.”
• Q3 2025 earnings call (2025-10-23) — output/ENVA/transcripts/ENVA_Q3_2025_earnings_call_2025-10-23.txt — spring-2025 LOC credit blip and self-correction; “everything is on the table” capital-return commentary; subprime-auto non-correlation pushback.

Company press releases (PRNewswire, validated at source)

• Q2’26 earnings date announcement (2026-07-09) — https://www.prnewswire.com/news-releases/enova-announces-date-of-second-quarter-2026-financial-results-conference-call-302822240.html — establishes the July 23 post-report-date print risk.
• Grasshopper definitive agreement (2025-12-11) — https://www.prnewswire.com/news-releases/enova-announces-definitive-agreement-to-acquire-grasshopper-bank-302639059.html — ~$369M deal terms, >15% year-one accretion claim, H2 2026 close guidance.
• $400M share repurchase program (2025-11-12) — https://www.prnewswire.com/news-releases/enova-announces-new-400-million-share-repurchase-program-302613087.html — authorization terms through 2027-06-30.
• Q1 2026 results (2026-04-23) — https://www.prnewswire.com/news-releases/enova-reports-first-quarter-2026-results-302752097.html — beat + raised guidance figures.
• Q4/FY 2025 results (2026-01-27) — https://www.prnewswire.com/news-releases/enova-reports-fourth-quarter-and-full-year-2025-results-302671738.html — record FY25, initial FY26 guide.
• Board appointment/departures (2026-07-10) — https://www.prnewswire.com/news-releases/enova-announces-board-of-directors-appointment-and-planned-departures-302823026.html — “planned departures” framing vs 8-K non-disagreement language.
• Needham TMC conference fireside announcement (2026-04-30) — https://www.prnewswire.com/news-releases/enova-to-present-at-the-needham-technology-media–consumer-conference-302759389.html — Cunningham/Cornelis public appearance.
• OnDeck/Ocrolus SMB Cash Flow Trend Reports (2026-01-28, 2026-05-01) — https://www.prnewswire.com/news-releases/new-report-93-of-small-businesses-expect-growth-in-2026-302759435.html — SMB demand evidence (93–94% of SMBs expect growth; ~75% bypass banks; 46% of bank-first applicants denied) — vendor-survey grade, used as demand color only.
• OnDeck acquisition PRs (2020-07-28 announcement, 2020-10-13 close; ir.enova.com) — deal terms (0.092 shares + $0.12 per ONDK share) and synergy claims for the M&A scorecard.

---

© Regulatory & legal docket items

• CFPB enforcement page — Enova 2023 consent order + termination — https://www.consumerfinance.gov/enforcement/actions/enova-international-inc-2023/ — documents the 2023-11-15 $15M repeat-offender order and its termination in full effective 2025-08-29/09-02 with alleged non-compliance waived; the single largest regulatory-overhang removal in the story.
• NCRC / Woodstock Institute opposition letter + release (2026-02-23) — https://ncrc.org/ncrc-woodstock-institute-urge-occ-federal-reserve-to-deny-bank-merger-between-enova-and-grasshopper-bank/ — formal comment letter to OCC/Fed (application 2026-Combination-344662) urging denial of the Grasshopper merger; basis for the deal-risk path analysis.
• NCLC-led coalition letter to Congress (2026-05-07) — https://www.nclc.org/broad-coalition-urges-congress-to-oppose-unaffordable-100-apr-bank-loans/ — “first national bank dedicated to directly making predatory loans” framing; tracks pending federal bills (Whitehouse S.3721 / Reed S.3793 “Predatory Lending Elimination Act”; Moreno S.3889 industry-side).
• NCLC state rate-cap surveys — https://www.nclc.org/wp-content/uploads/2022/10/report-installment-loans.pdf — ~20 states + DC at/below 36% APR; the state-exclusion map behind the serviceable-market haircut.
• Illinois PLPA impact study (Elliehausen et al.) via Maine legislative testimony (2025-02) — https://legislature.maine.gov/testimony/resources/HCIFS20250212Miller133837847655643665.pdf — post-PLPA subprime loans −38%, licensed lenders halved; the empirical template for cap-driven market exit in Enova’s home state.
• Chapman & Cutler marketplace-lending survey (Nov 2024) — state commercial-financing disclosure law landscape for the SMB/MCA read.
• American Banker — “High-cost lender Enova’s plan to buy a bank sparks backlash” (2025-12-11) — https://www.americanbanker.com/news/high-cost-lender-enova-to-buy-digital-only-grasshopper-bank — deal-day coverage; CashNetUSA to operate outside the bank subsidiary.
• American Banker — CFPB posture / nominee lapse (2026-01-09) — https://www.americanbanker.com/news/with-cfpb-nominee-lapse-vought-continues-as-acting-director — acting-director status underpinning the benign-federal-posture read.
• Sheppard Mullin — CFPB payday-rule limited enforcement (2025-04-04) — https://www.sheppard.com/insights/blogs/cfpb-plans-limited-enforcement-of-payday-lending-rule — Small Dollar Rule payment provisions effective 2025-03-30 but enforcement deprioritized with a narrowing NPRM planned.
• PYMNTS — Virginia bank-partnership bill / AFC veto plea (2025-03-04) — the frontier of state-level true-lender pressure.
• Banco Central do Brasil data via Setcesp/Brasil247 (2026-01-29) — Selic 15%, 47.2% average bank lending rate, record household delinquency — Brazil macro for the Simplic zero-valuation call.
• CURO Chapter 11 docket (S.D. Tex. Case 24-90165, via FTI Consulting) — http://cfcanada.fticonsulting.com/CuroGroup/ — the sector-casualty anchor for the capital-cycle and profit-pool analysis.

---

(d) Third-party data feeds (with reconciliation caveats)

• ROIC.ai MCP tools (accessed 2026-07-19): get_income_statement / get_balance_sheet / get_cash_flow / get_profitability_ratios / get_enterprise_value / get_valuation_multiples / get_per_share_data / get_company_news / earnings-call transcripts (ENVA FY2016–FY2025 + 9 quarters; peers OMF, OPRT, RM, WRLD, UPST, LC FY2023–2025). Caveats documented in the log: revenue, net income, and LT debt tie to EDGAR companyfacts exactly (2020–2025), but ROIC’s “ROE” (16.7% FY25) and “BVPS” ($79.71/$83.99) do NOT tie to filings — the report uses computed ROE 24.3% and BVPS $56.25 instead; ROIC’s “COGS”/“operating income” are re-mapped constructs; ROIC’s latest-price feed ($235.88) disagreed with AZI/FactorsToday ($232.52) for the 2026-07-17 close — treated as stale, immaterial. Comp per-share data flagged ordinal-only where definitional (OPRT/UPST/LC negative BVPS; SLM TBVPS suspect).
• AZI Trading (pulled 2026-07-19): daily adjusted OHLC CSV (https://azitrading.com/controls/download-data.php?t=ENVA, 2,942 rows → output/ENVA/2026-07-19/_scratch/ENVA_prices.csv) — 60-month price/event map, returns, drawdowns, EMAs; and valuation_index percentiles dated 2026-07-17 (_scratch/azi_valuation_index.json) — own-history composite 96.5th percentile (P/B 99.6th, P/S 99.6th, P/E 90.3rd).
• FactorsToday API (https://www.factorstoday.com/api, pulled 2026-07-19; loadings dated 2026-07-17): stock-loadings, leaderboard, stock-info, stock-specific-vol, related-stocks, factor-returns JSONs in _scratch/ — factor fingerprint (Market 1.36, SmallSize 1.05, Financials 0.69; Quality/Growth/Momentum zeroed), idiosyncratic-vs-factor attribution of the +102% 1-yr move, 1-yr Sharpe 2.55, factor-similar peers (OMF 0.894 cosine).
• yfinance via scripts/fetch.py (pulled 2026-07-19) — live comp prices and stats for OMF OPRT RM WRLD CACC UPST LC SLM FCFS PRG SYF COF — flagged UNOFFICIAL, screening-grade; all headline ENVA multiples recomputed from filings before use.
• SEC EDGAR companyfacts (see §a) — the reconciliation anchor all third-party feeds were checked against.

---

(e) Trade / financial press

• TipRanks — OppFi Q4 2025 results (2026-03-11) — https://www.tipranks.com/news/company-announcements/oppfi-reports-record-q4-2025-revenue-and-profitability — closest pure consumer-model comp datapoints (Q4’25 revenue $159.2M, +17.3%; receivables $493.1M).
• BadCredit.org / WSJ — private-credit redemption squeeze (2026-03-25) — https://www.badcredit.org/news/private-credit-flight-squeezes-subprime-lending-capital/ — Stone Ridge interval fund honoring ~11% of redemption requests; Morgan Stanley private-credit default forecast; funding-sentiment factor for the capital-cycle read.
• Tricolor Holdings / PrimaLend bankruptcy coverage (web-verified 2026-07-19) — Chapter 7 (2025-09-10, fraud/double-pledged collateral) and Chapter 11 (2025-10-22) anchors for the Sep–Oct 2025 contagion episode (−13.5% ENVA with no disclosed exposure).
• PYMNTS — OnDeck acquisition coverage (2020-07-29) — contemporaneous deal-terms corroboration.
• PYMNTS — Virginia bank-partnership bill (2025-03-04) — see §©.
• AInvest (2026-01-27) — Grasshopper deal size/deposits corroboration and management accretion claims (treated as hypotheses).
• Seeking Alpha — Grasshopper synergies bull note (2026-06-25) — https://seekingalpha.com/article/4917710 — secondary corroboration of the synergy framing; used as sentiment evidence only.
• Zacks / defenseworld.net aggregator items (BTIG PT raise $144→$199, 2025-12-15; Zacks Rank #2, 2026-07-09; insider-sale notices Dec-2025–Feb-2026; Feb-2026 pullback note) — triaged as secondary/noise; beat/miss consensus figures only where validated (Q4’25 $3.46 vs ~$3.20; Q1’26 $3.87 vs $3.66).
• Market Research Future — US MCA market size (2025-09) — https://www.marketresearchfuture.com/reports/merchant-cash-advance-market-24003 — vendor-grade market sizing (~$35.8B), used with explicit caution flag.

---

No source was used that is not listed here. No recommendation, no price target.