The Western Union Company (NYSE: WU): A 15% Yield on a Shrinking Rail — and a Solana Settlement Bet
DividendChase LTD | Institutional Research
As of 30 September 2026
Western Union is no longer only a cash-to-cash agent network. It is a low-growth remittance franchise trying to become a regulated dollar pipe: USDPT on Solana, a Digital Asset Network into the existing payout grid, a Visa-linked Stablecard, and a cost program named Beyond Efficiency. The equity market has already voted. The stock trades near $6, the dividend is still $0.94 a year, and the yield prints ~15%. That is not a growth multiple. It is a payout-plus-execution price.
Q2 made the tension explicit. Revenue held around $1.01 billion (−1% YoY). Adjusted EPS was $0.31, down from $0.42 a year earlier and well below the ~$0.43 consensus print. Transactions in Consumer Money Transfer rose 3%; branded digital transactions rose 25%. Mix is moving to digital payout-to-account, which is good for volume and hard on take-rate and margin. Management cut the year. Buybacks are paused. The dividend was not.
What the company is, in 2026
Three engines:
| Segment | Q2 2026 snapshot | Role in the story |
|---|---|---|
| Consumer Money Transfer (retail + branded digital) | CMT revenue ~$866 million (−2%); 73.5 million transactions (+3%); cross-border principal $27.2 billion (+2%). Digital is ~32% of CMT revenue and ~43% of CMT transactions | Still the P&L. Americas retail and Middle East digital were the EPS hole |
| Consumer Services | GAAP +4%, adjusted +12% (bill pay, Travel Money, prepaid) | Diversification. Low-single-digit growth guided for the full year as eurochange annualizes |
| Digital assets / USDPT | Launched 4 May 2026. Not a disclosed material revenue line yet | The option the board is selling to investors |
CEO Devin McGranahan’s line from the spring still stands: it is no longer whether WU is in digital assets, it is how fast they scale. That is strategy. It is not Q3 EPS.
Future plans (the operating plan, not the keynote)
1. Stabilize CMT and buy corridor density.
The Intermex acquisition is the 2026 volume plug for U.S.–Latin America retail. Guidance assumed a 1 September 2026 close. As of late September, California approval was still outstanding. Cantor Fitzgerald cut its target to $6 (Underweight) on 29 September and removed $40–45 million plus $20–23 million of second-half Intermex contribution from its model. Until that deal closes, the 4–6% adjusted revenue guide is a pro forma that is slipping.
2. Push “Beyond Digital.”
Platform launches in Australia, Europe, and the U.S. before year-end 2026. Branded digital is the only CMT sleeve that is clearly growing. The constraint management keeps repeating: new-customer acquisition economics are still poor. Volume without contribution is not a turnaround.
3. Cut the cost base.
Beyond Efficiency: $50 million run-rate savings by year-end 2026, $200 million by year-end 2027 (including Intermex scale). That is how they defend $1.25–$1.35 of adjusted EPS after the Q1 guide of $1.75–$1.85 died. Second-half EPS is supposed to beat the first half on seasonality, agent wins, mix, and this program.
4. Hold the credit box, hold the dividend, stop the buyback.
Gross leverage was described around 3×, net 2×, with a target band of 2.5–3.0×. Buybacks paused to stay inside it. Dividend stance: maintained. Cash at Q2: $920 million; debt: $2.7 billion. YTD operating cash flow $214 million (+45%), helped by cash taxes — not by a sudden remittance boom.
5. Connect wallets to cash.
Management’s digital-asset KPI is tens of millions of digital wallets on the Digital Asset Network by end-2026, and 60+ markets for the Stablecard after a 37-market launch (WU + Rain). That is distribution, which WU actually owns. Issuance is someone else’s charter.
Blockchain positioning: settlement first, consumer second
WU already tried the Ripple corridor era. The 2025–26 design is different: a WU-branded dollar on a public chain, issued by a federally chartered crypto bank, settled into the agent grid.
| Piece | Design |
|---|---|
| USDPT | USD-backed payment stablecoin. Live 4 May 2026 on Solana. First corridors: Philippines and Bolivia |
| Issuer | Anchorage Digital Bank N.A. — OCC crypto bank, not WU’s own balance-sheet mint |
| Infra | Fireblocks (wallets, settlement, ops); Crossmint (enterprise wallets / fiat bridges); Rain (card); Solana Developer Platform with Mastercard/Worldpay cohort |
| Digital Asset Network (DAN) | Licensed exchanges and custodians → WU payout and liquidity. First four exchanges live and trading as of the Q2 call; 7+ partner activations targeted in 2026 |
| Treasury / agent settlement | The real P&L use case: 24/7 USDPT instead of SWIFT and idle nostro. Less trapped working capital across 200+ countries and hundreds of thousands of cash points (360k cited in partner copy; WU’s own grid is larger if you count every agent type) |
| Stable by Western Union | Consumer hold / send / spend of USDPT in 40+ countries through the WU app — dollar access without a U.S. bank |
| USDPT Stablecard | Visa-rail spend in 37 markets at launch, 60+ targeted by year-end 2026 |
This is the correct architecture for a remittance incumbent. Cash-out is the scarce asset. USDT and USDC already move cheaply; they do not, by themselves, pay out at a WU agent in Cochabamba on Sunday. DAN is an attempt to tax that last mile. Agent settlement is an attempt to cut WU’s own float. The consumer token and the card are distribution experiments on top.
What it is not:
- A replacement of CMT revenue in 2026.
- A GENIUS-licensed WU bank. Anchorage holds the charter; WU holds the brand and the agents.
- Immunity to USDC/USDT. If senders already hold those coins, WU is a cash-out utility, not a minting franchise.
- Solana equity. WU chose Solana for throughput and fees. Stripe can still route new merchants to Tempo; Circle can still settle Visa on several chains.
McGranahan on the Q2 call bundled four live items: USDPT launch, a treasury-bridge using USDPT, DAN, and the Stablecard. That is more product than most legacy MTOs have shipped. It is still a 2027–28 margin story if it works.
Income estimates
Company guide (updated 30 July 2026, Intermex at 1 Sept — now late):
| GAAP | Adjusted | |
|---|---|---|
| 2026 revenue growth | 3–5% | 4–6% (ex-FX / Argentina inflation rules) |
| 2026 EPS | Not reconciling | $1.25–$1.35 |
| Tax rate | 20–22% | 13–15% |
Q1 guide had been 6–9% adjusted revenue and $1.75–$1.85 adjusted EPS. The cut is the year.
Street (composite, late September; treat as a band, not a single house):
| 2026e | 2027e | |
|---|---|---|
| Revenue | ~$4.11 billion | ~$4.36 billion |
| Adj. / non-GAAP EPS | ~$1.29 (houses $1.23–$1.40; sits on the company range) | ~$1.52 (houses ~$1.27–$1.75) |
| Dividend / share | $0.94–$0.96 | ~$0.96–$0.97 |
| FCF | ~$0.37 billion | ~$0.50 billion |
Q3 2026 consensus EPS is ~$0.35; Q4 ~$0.37. That path only works if H2 really is better than H1 and Intermex shows up. Cantor’s 2027 EPS figure of $1.63 at 3.5× is how you get a $6 target. Other target lists still show midpoints in the $7–9 area with a $5–$8 live range after the latest cuts. The stock at ~$6 is already in the bearish camp.
Dividend math. Quarterly $0.235, annual $0.94, payable 30 September 2026 for the latest coupon. Trailing payout on stated earnings is high-70s%; on this year’s $1.25–$1.35 guide it is ~70–75%; on 2027 consensus it falls toward ~60%. Cash-flow payout is healthier (~40% on some TTM cuts). The yield is high because the price collapsed, not because the coupon was raised. Five-year DPS CAGR is ~1%. This is a flat dividend on a shrinking multiple.
Implications for investors
The bull case (the one management is selling).
Digital mix + Beyond Efficiency + Intermex + USDPT float savings produces mid-single-digit revenue and a path back toward $1.50+ of adjusted EPS in 2027. The agent network becomes the off-ramp for every exchange that joins DAN. You collect ~$0.94 while you wait. At 5–6× 2027 earnings the equity is a stub.
The base case.
Retail CMT keeps leaking to Wise, Remitly, WorldRemit/Sendwave wallets, and raw USDT. Digital grows transactions and not contribution. Intermex closes late and synergies slip into 2027. USDPT settles agents and does not move reported revenue much in 2026. EPS lands in the $1.25–$1.35 box. The dividend is covered on cash flow if nothing else breaks. The stock is a high-yield, low-duration instrument: you are paid to own a melting retail network with an option.
The bear case.
Intermex fails or is re-cut. Americas retail and a Middle East shock hit again. CAC stays ugly. A GENIUS-permitted USDC or bank coin turns WU into a cash-out commodity with no mint economics. Leverage forces a dividend trim if EPS stays near $1.20. At 15% yield the market is already pricing a cut; a cut would still re-rate the stub lower first.
How WU fits a DividendChase book
| Sleeve | Use |
|---|---|
| Income | Size for the $0.94 coupon, not for 15% forever. Cap the position. A 70%+ earnings payout plus paused buybacks is not a Dividend King setup |
| Payments / fintech | Do not pair WU with V/MA as the same theme. Those are networks. WU is a corridor operator |
| Solana / stablecoin | WU is a better cash-out proxy than SOL. It is a worse technology proxy than Circle (if listed) or COIN. USDPT success helps WU working capital first |
| Event path | Intermex close, Q3 print vs $0.35, DAN partner count, Stablecard market count, any change in the $0.235 |
Risks that are not in the stablecoin slide: U.S. immigration and corridor policy, Argentina inflation accounting, agent-commission inflation, FX translation, digital-asset regulatory perimeter (GENIUS listing rules, MiCA if they passport the consumer product), Anchorage as single-issuer operational risk, Solana outage risk on settlement days.
DividendChase stance
Western Union in September 2026 is a covered, high-payout equity on a structurally pressured remittance book, plus a credible but early Solana settlement stack (USDPT / DAN / Stablecard / agent treasury). Guidance credibility is impaired: EPS was cut ~$0.50 in one quarter and the Intermex date in the model has already slipped.
- Treat the dividend as the return, and assume it is flat, not growing.
- Do not underwrite $1.75 EPS again until Intermex is closed and H2 margins print.
- Underwrite USDPT as float and last-mile optionality, not as 2026 revenue.
- Prefer WU as a small income sleeve over a core payments holding. If you want blockchain payments beta, size SOL/BSOL or Circle/COIN separately so a WU miss does not masquerade as a Solana miss.
The agents are still the moat. The token is a better nostro. The stock is cheap because the moat is leaking and the token does not yet pay the dividend.
Intelligence for the Discerning Investor
DividendChase LTD
