GM. This is Milk Road PRO, making the contrarian case for the financial dinosaur the market left for dead.
Today's edition features our latest PRO report digging into Western Union: why a stock down 70% might quietly own the missing piece of crypto's stablecoin stack. You'll get the opening below, with the rest on the site.
Here’s what we’ve got for you today:
- 🌍 Why WU's 360,000 agent network is exactly what stablecoins need (and nobody else has it).
- 🛡️ The moat that makes this nearly impossible to replicate (and why crypto-native players keep failing).
- 💰 Whether the 10% dividend and contrarian setup are enough to buy.
Pharos is an institution-grade layer 1 (L1) designed specifically to bridge real-world assets onchain. Join the Pharos community (X & Discord) here.
Prices as of 2:00 p.m. ET. Powered by Coingecko.
THE MARKET’S PICTURE OF WU
Western Union (WU) is down 70% from its 2021 high.
Forward P/E of 5x. A 10% dividend yield priced as if the cut is coming.
Most investors see a dying remittance company getting eaten by fintechs and crypto. I think they're missing the second business growing alongside it.
My thesis: WU's agent network, compliance stack, and currency licenses are the rails stablecoin issuers need to reach real users. The market is currently fixated on WU's shrinking legacy business, overlooking the high-growth opportunity being built atop the existing infrastructure.
Wall Street's view of Western Union is simple: a dying legacy remittance giant losing share to digital players.
Revenue has been in structural decline for years.
Q1 2026 revenue came in at $983M, with the core business still under pressure.
Adjusted operating margin compressed to around 13% from 18-19% in prior periods. Adjusted EPS dropped to $0.25, missing estimates by 35-40%.
And it has two parts:
1. Fintechs: Wise, Remitly, and PayPal offer cheaper, faster, app-based transfers and have eroded WU's pricing power in developed corridors. Digital players don't carry the overhead of a physical network. They route bank-to-bank or wallet-to-wallet at lower fees.
2. Crypto: Stablecoins settle instantly, 24/7, at near-zero cost. The argument: WU's cash-heavy, agent-based model is obsolete in a world shifting to digital-native money movement.
The stock has paid the price.
WU is down roughly 70% from its 2021 highs and two-thirds since 2020, trading near $9.20 with a forward P/E of 5x and a 10% dividend yield, priced as if it's about to be cut.
Consensus is Hold/Reduce. Average 12-month price targets sit at $8.80-$9.60, implying flat-to-downside.
This is the consensus view. I think it's missing something.
So what? The bear case is priced in. Perhaps overpriced.
It treats WU as a static legacy player rather than a company quietly repositioning its unmatched physical infrastructure for the very technology (stablecoins) that everyone assumes will destroy it.
That contrarian gap is the entire thesis.
THE ASSET NOBODY CAN REPLICATE
The asset that matters here is the cash payout network.
WU has roughly 360,000 active agent locations across 200+ countries and 130 currencies.
About 90% of those locations are outside the U.S., concentrated in exactly the corridors where banking access is weakest, and remittance demand is highest.
Branded Digital (the app and website) handles 42% of transactions on top of that, so the company already runs a hybrid model: a sender can pay digitally, and the receiver can walk to a storefront for cash.
Wise, Remitly, and PayPal don't have this.
They route bank-to-bank or wallet-to-wallet, which is fine in developed corridors and useless in the places that drive global remittance volume.
Stablecoin issuers have the same gap in reverse.
USDC settles in seconds, but the recipient in rural Mexico or the Philippines still needs to convert tokens into pesos. Somebody has to handle that last mile.
WU already does.
The harder question is whether anyone catches up.
They won’t, but not for the reason most readers assume.
The agent count isn't the moat. Anyone with enough capital can sign up agents.
The moat is everything that has to exist around the agents for the network to actually function.
Three things make replication difficult:
1. Licensing.
Money transmitter licenses in nearly every U.S. state, plus the equivalent regulatory permissions in 200+ countries.
Each one is a separate process with separate timelines, capital requirements, and ongoing compliance obligations.
WU has been accumulating these for over a century.
A new entrant starting today is looking at a 5-7 year build to reach equivalent global coverage, assuming nothing goes wrong with any single regulator.
The GENIUS Act and parallel state-level rules raise this bar further by adding stablecoin-specific compliance requirements on top of the existing money transmission framework.
2. The compliance and FX stack underneath the licenses.
AML, KYC, sanctions screening, and FX conversion infrastructure that satisfies local regulators in every jurisdiction.
WU runs this across 130 currencies. The technology can be bought. The integrations with local banks, regulators, and payment systems in each country cannot. Those relationships are the part that takes time.
3. The network effect among agents.
Agents stay because customers come. Customers come because agents are there.
A new entrant signing up agents in a corridor where WU already has dense coverage faces a chicken-and-egg problem: the new agents need transaction volume to justify staying in the network, but customers won't switch until coverage matches what they're used to.
This is why crypto-native players have spent the past five years trying to build emerging-market off-ramp networks and have not done so at scale.
The moat doesn't help in app-based remittance in developed corridors. Wise, Remitly, and PayPal will keep eating that segment because the off-ramp problem there is solved by domestic banking. WU has lost share in those corridors and will keep losing it.
The moat is specifically about the cash off-ramp in places where banking infrastructure doesn't exist. That is exactly where the stablecoin opportunity sits.
To understand why the stock is down 70%, we must examine what the bear case sees and what it misses.
A LAYER 1 BUILT SPECIFICALLY FOR RWAs
This is one of the biggest narratives in crypto right now:
“Real World Asset (RWA) tokenization is one of blockchain’s biggest use cases.”
But here’s the catch:
Very few blockchains are actually built for RWAs.
One of the few is Pharos.
Unlike general-purpose chains, Pharos is an institution-grade layer 1 (L1) designed specifically to bridge real-world assets onchain.
Here’s what stands out:
- Purpose-built for RWAs
- High performance execution for institutions
- Infrastructure that connects onchain and offchain systems
This is not just another L1. It’s built specifically for RWAs.
Join the Pharos community (X & Discord) here.
CRACKS IN THE BEAR CASE
The bear case treats WU as one dying business. There are actually two businesses on the same network, and only one of them is dying.
The legacy retail business (walk into a storefront, hand over cash) is shrinking. That's where the revenue decline, the margin compression, and the U.S.-Mexico pricing pressure all show up.
Wall Street is right about that part. It's not coming back.
The digital business is growing fast on the same infrastructure.
Customers send through the WU app or website, and the receiver picks up at one of WU's 360,000 agent locations or directly into a bank account or wallet.
The Q1 2026 numbers tell the split clearly:
- Digital transactions grew 21% year-over-year and now account for 42% of remittance transactions.
- Travel money and bill pay revenue grew 24%, with adjusted growth of 33%.
- Cross-border principal volume grew 5% to $27.0B in the quarter.
- Total remittance revenue still declined 3%, and operating margin compressed to 13% on discrete items.
The headline is bad. The composition is what matters.
Behind the 2026 numbers is the pending Intermex acquisition. WU announced in August 2025 that it's buying International Money Express for ~$500M, with the deal expected to close in Q2 2026.
Intermex is a U.S.-to-Latin America/Caribbean retail remittance specialist with strong agent presence in the U.S.-Mexico corridor. It is the same corridor where WU has been losing share to fintechs.
The acquisition is part of management's 2026 growth math.
The legacy product is being run for cash flow while the digital and travel-money businesses grow at double digits on top of the same network.
Then there's the forward signal.
Management is guiding 2026 to 5-8% GAAP revenue growth and adjusted EPS of $1.75-1.85, the first growth year after multiple years of decline.
Consensus targets sit at $8.80-9.60 against a $9.20 stock price.
Either guidance is wrong, or the stock is.
The stock's valuation is driven by its declining legacy segment, completely failing to factor in the double-digit growth businesses operating on the same network.
The next question is what the digital business actually grows into. The market assumes a smaller, app-based remittance company.
I think it's something larger.
THE STABLECOIN OPPORTUNITY NOBODY SEES
So WU owns a global cash payout network nobody else can replicate.
Does that asset get more valuable or less valuable when money moves onchain?
The bear case assumes less.
The whole point of stablecoins, the argument goes, is that you don't need agents, storefronts, or 360,000 cash-out points.
Settlement happens onchain. The network is legacy infrastructure waiting to be obsoleted.
I think the network gets more valuable, not less, and the rest of the payments industry is already voting with its checkbook:
- Stripe paid $1.1B for Bridge in 2024 to own a stablecoin payments infrastructure.
- PayPal shipped PYUSD.
- Visa is integrating USDC for cross-border settlement.
- MoneyGram has been running a USDC off-ramp on Stellar since 2022.
None of these companies is betting that physical infrastructure goes away in a stablecoin world. They're betting it gets repriced.
The reason is that stablecoins haven't closed that gap.
Onchain settlement only matters if both ends of the payment can touch real-world money, and while the U.S. side is solved, the receiving side in the corridors that drive global remittance volume is not.
WU's network is the receiving-side infrastructure at scale. Stablecoins create more demand for it, not less.
The question for the rest of this report is how WU captures that value in dollars.
There are four ways:
1. Off-ramp fees from third-party stablecoins.
WU's Digital Asset Network connects external wallets to its 360,000 active agent locations.
A user holding USDC, USDT, or PYUSD cashes out into local fiat through WU's infrastructure. The wallet provider hands off, WU collects a fee on the conversion.
This is issuer-agnostic. It works whichever stablecoin wins.
Sizing it requires the right TAM.
The relevant market is not "all stablecoin payment volume" (most of which is B2B settlement, exchange flows, and treasury operations that never need a cash-out).
The relevant market is global cross-border consumer remittances: roughly $700-900B per year, per the World Bank, of which WU already handles a 5-7% share through its existing network.
Today, near-zero percent of those flows move on stablecoin rails.
Some estimates suggest 1-3% currently, growing fast.
A reasonable 3-5 year scenario is 10-20% of global remittances shifting to stablecoin rails as wallets, fintechs, and crypto-native players push adoption in emerging-market corridors.
Call that $70-180B of annualized stablecoin remittance volume by then.
WU is uniquely positioned to capture the off-ramp share because it owns the hardest piece of the stack: the cash-out infrastructure.
If WU captures 10-15% of stablecoin off-ramp volume at a 1-3% take rate (lower than WU's full remittance take rate of 5-6% because off-ramp-only services don't include the sender-side relationship), that's roughly $70-810M of annualized revenue at maturity.
2. Working capital release from USDPT.
WU's material says USDPT (their stablecoin coming in H1 this year) will be used "to facilitate high-speed payments between Western Union and its agents."
That's a settlement instrument, not a consumer product. The economics live there.
Today, WU has to prefund local-currency cash across its agent network so that receivers in Manila or Lagos can collect a transfer the moment they walk in.
With 360,000 agents in 130 currencies, that adds up.
WU reported $3.45B of settlement assets on its balance sheet at year-end 2025. Against $107.4B of annual cross-border principal, that's roughly 12 days of money flow trapped in the settlement system at any given moment.
USDPT compresses the cycle.
Instead of prefunding pesos in the Philippines, WU can settle with agents in USDPT in seconds, and the agent converts to local currency at the point of sale.
WU has been compressing this number for years through real-time payment integrations, agent network optimization, and treasury improvements.
What stablecoins unlock are the corridors where traditional rails don't reach: long-tail currencies, cross-border flows without correspondent banking, and 24/7 settlement in markets with the weakest local banking infrastructure.
That's where WU has the most network value, and it's exactly where traditional compression hits a ceiling.
Even with that ceiling, the unlock is significant.
Sizing it: if USDPT compresses the cycle by two days on half the eligible float, that's about $300M of one-time working capital release.
Compression won't be uniform (major pairs already settle fast, long-tail currencies offer the biggest savings), state regulators have to recognize stablecoin reserves as permissible investments for the full release to happen, and this is a one-time event, not a recurring revenue line.
3. Reserve income on USDPT float.
Under the GENIUS Act, stablecoin reserves have to be held in cash and short-dated Treasuries. Three-month T-bills currently yield around 3.7%. The gross yield on USDPT reserves is real money for whoever owns the issuance.
The sizing is smaller than some might expect.
WU's existing settlement assets are already earning yield, primarily through state and municipal debt securities held to comply with state money transmitter laws.
The float is mostly invested.
The incremental reserve income comes from the cash-equivalent slice of the balance sheet, which is roughly $500-700M of low-yielding cash that flows through USDPT reserves instead of operating accounts.
On that basis, the math: 3.7% gross yield, less roughly 25% for taxes and operating costs, gives a net spread of around 2.5 percentage points. On $500-700M of converted float, that's $12-18M of incremental after-tax income at maturity.
Distribution economics narrows it further.
If USDPT is held externally (in wallets, on exchanges, at partner platforms), WU has to share reserve income with distributors the same way Circle shares USDC yield with Coinbase.
WU is in the unusual position of being both issuer and distributor on its own network, which is what protects most of these economics. But to the extent USDPT lives outside WU's network, the share goes down.
Small but recurring, and only available because WU owns the issuance.
4. Internal economics WU currently pays others.
Today, WU pays FX spreads to banks, wire fees to correspondent partners, and carries the time-cost of capital tied up in 130-currency settlement.
The company statements bundle this into the cost of services without breaking it out, so I can't size it from the outside.
The savings show up as margin expansion on existing volume every year.
Owning USDPT is what turns WU from a stablecoin off-ramp partner into a stablecoin company that happens to own the largest off-ramp network in the world.
WHAT COULD BREAK THE (BULL) THESIS
Every contrarian thesis depends on assumptions that could be wrong.
Three are worth taking seriously, plus one piece of macro risk outside management's control:
1. State-by-state stablecoin reserve recognition stalls.
The working capital release and reserve income depend on state money transmitter laws recognizing stablecoin balances as permissible investments.
The federal GENIUS Act framework is the easier piece. State-by-state recognition is the harder slog, and each state has its own timeline. If recognition lags, the thesis still works at reduced magnitude, but the timeline pushes out by years rather than months.
What the bull case needs: at least the major remittance-corridor states (California, Texas, New York, Florida) recognizing stablecoin reserves within 18-24 months.
The rest of this report is for Milk Road PRO members only. And trust me, this is where the real alpha starts.
WHAT’S LEFT INSIDE? 👀
- Why stablecoins could make WU’s “obsolete” network more valuable than ever.
- The hidden moat crypto-native competitors still can’t replicate.
- The exact setup that could turn a “dead” dividend stock into a stablecoin infrastructure play.
Join Milk Road PRO and read the full report plus every other PRO deep dive (a single buck gets you seven days of access)!
BITE-SIZED COOKIES FOR THE ROAD 🍪
The moment is here. Midnight, the fourth generation blockchain designed for real world adoption, is officially live.*
Galaxy's Alex Thorn: Equities moving onchain creates a "ripe environment for AI." AI needs two things to act on financial data: the data itself, and the rails that move it.
Nice! Google Cloud just agreed to accept stablecoin payments from AI agents via a new service from the Solana Foundation.
The Great Re-Industrialization: Hormuz is reopening? And oil is sliding.
Higher earnings rates. Lower borrowing costs. Up to 0.5% cashback. New U.S. clients get these benefits on Nexo for 30 days.**
*this is sponsored content. **this is partner content.
Join the Pharos community (X & Discord) here.
MILKY MEMES 🤣
ROADIE REVIEW OF THE DAY 🥛
VITALIK PIC OF THE DAY
