Author: Joey Shin @IOSG

Abstract

The encryption industry says every day that it lacks users, but this is not the case with data at all. The active users of consumer-grade encryption have long reached tens of millions, but they are just not on the radar of Silicon Valley and New York. These people in Manila, Lagos, Buenos Aires, and Hanoi use Coins.ph (18 million users), MiniPay (4.2 million weekly active users), and Lemon Cash (No. 1 in Argentina’s application list) every day, but the English media has almost no coverage at all. On the other hand, the activity volume of those protocols discussed by Western VCs every day is not enough for one hour of the Tron shadow liquidation network.

Seven core conclusions: The user problem of encryption is essentially a geographical issue; Tron is the most important consumer-level public chain, but no one talks about it in NYC and SF; on-chain e-commerce basically does not exist; the largest prediction market is centralized; revenue and user numbers tend to go in the opposite direction; the sustainable DEX war is over; there are indeed consumer-level encryption companies that really make money - they just don't look like DeFi.

Payment and new banks:Users have already existed, but they are not in the sight of VCs

General knowledge: Crypto needs to enter the mainstream and bring the next billion users, and wallet UX is the bottleneck.

Data shows: The next billion users are already here, and the biggest bottleneck is not customer acquisition, but monetization.

Look at the existing scale first. Telegram Wallet claims to have 150 million registered users (unverified – low confidence), but let’s put that number aside. Just looking at the verified data, the user base is already amazing: Coins.ph has 18 million confirmed users in the Philippines, mainly based on Tron's USDT track; MiniPay, as Opera's mobile stablecoin wallet on Celo, has 14 million registered users and 423 weekly active USDT users as of March 2026 million, with monthly transaction volume of $153 million, and on-chain activity increasing by 506% year-on-year (high confidence – from Tether/Opera/Celo joint disclosure). Chipper Cash covers 7 million users in 9 African countries and has recently achieved positive cash flow. With 5.4 million downloads, Lemon Cash ranks as the No. 1 finance app in both Argentina and Peru, with MAU quadrupling since 2021. Paga processes N17 trillion in annual transaction volume in Nigeria, but the crypto-related share is unknown (medium confidence).

The only payment company currently running through both scale and revenue is RedotPay: 6 million users, annualized revenue of $158 million, annualized transaction volume of $10 billion, and a valuation that has increased 16 times since the seed round (high confidence - The Block, CoinDesk, company disclosure). RedotPay's model is a crypto-to-fiat card processor for the Asia-Pacific region, with commissions based on transactions and zero chargeback risk - essentially a crypto-native Visa issuer-acquirer. It is the clearest case yet that consumer-grade encryption can generate real, recurring, non-incentive-driven revenue at scale.

Another bright spot on the revenue side is Exodus. According to its SEC 8-K filing, revenue in 2025 will be $121.6 million (high confidence). It is one of the few publicly listed and audited crypto consumer companies in the US stock market. Its revenue comes from exchange and staking fees on 1.5 million MAU, and the stock is listed on the NYSE American board under the symbol EXOD.

Ether.fi's Cash product is the most noteworthy native DeFi entrant: it was profitable in the first year, with more than 70,000 cards issued. Cash currently contributes about 50% of total revenue, with monthly revenue of US$2.8 million (high confidence - TokenTerminal daily verification). It proves that a DeFi protocol is capable of making real consumer-grade products—but 200,000 total users is still a niche.

The problem of customer acquisition in emerging markets has been solved, but the problem of monetization has not been. The gap between MiniPay’s 4.2 million weekly active users and its undisclosed (and presumably extremely low) revenue may be the crypto industry’s biggest unanswered question — and its biggest opportunity.

Marginal improvement vs. non-incremental value: fine-tuning screening criteria

A common rebuttal to consumer-grade crypto investments is that crypto must provide non-incrementalvalue relative to fiat solutions in order to hedge against integration costs. The data shows that the formulation of this test itself is wrong. Compare two of the clearest data points in the payments category. MiniPay’s advantages over traditional mobile money products like M-Pesa are marginal at best in the hands of users—slightly cheaper transfers, slightly wider U.S. dollar exposure, slightly wider cross-border coverage. It has 4.2 million weekly active users and essentially zero revenue. RedotPay's advantages over the traditional Visa issuer-acquirer are also marginal in terms of consumer experience - swiping a card, buying a hot dog - but the underlying mechanism is structurally different: zero chargeback risk, instant cross-border settlement, and no agency bank dependence. RedotPay generates $158 million in annualized revenue from 6 million users.

Both products have run smoothly and both have PMF. The difference is that RedotPay's "marginal but structural" advantages can compound into pricing power, while MiniPay's "marginal and superficial" advantages cannot. Zero Chargebacks isn't a feature that users will notice, but it's about a 1.5% gross margin difference that card issuers capture on every transaction, permanently. Slightly cheaper transfers are something that users only notice once and don’t assign value to once they get used to it.

It follows that the correct screening question is not "Is this non-incremental?" but "Does this marginal improvement map to the structural characteristics of the unit economy?"If the answer is yes - chargeback risk, settlement timeliness, correspondent banking, capital efficiency, custody costs - then a product that has almost no change in user experience can still compound into a big business. If the answer is no, then this product does not have investment value even if it has tens of millions of users. Consumer encryption exists in both categories, and conflating them has cost the category an entire generation of capital.

E-commerce

General knowledge: Encrypted payment is gradually being adopted by e-commerce, it is only a matter of time.

Data shows: No on-chain e-commerce agreement on DeFiLlama has a daily agreement revenue exceeding US$10,000. Not "very few", but literally zero.

This chapter is not about the battle between early competitors, but the absence of competitors. After auditing DeFiLlama, all protocols tracked by TokenTerminal, and all company public disclosures, we found only one player worth mentioning: Travala, a centralized travel booking platform with revenue of $7.17 million in February 2026 (medium confidence - self-reported, no independent verification). Travala is not a protocol, it is a travel agency that accepts cryptocurrencies.

UQUID claims to have 220 million users and 50 million monthly visits (the number 220 million actually represents users of partner platforms - Binance, etc. - rather than UQUID's own users). The headline numbers are misleading, but its product catalog is indeed quite large — 175 million physical items, 546,000 digital items — and Tron’s share of its trading volume doubled to 39% in the first half of 2025, with 54% of transactions denominated in USDT-TRC20. But there is no public revenue data, and user numbers don’t stand up to scrutiny.

Gift card and voucher service Bitrefill has monthly revenue of about $1 million (low confidence - Growjo estimate, historically imprecise). Apart from this, there are no other on-chain e-commerce protocols worthy of attention.

What really exists is a shadow e-commerce economy running on the Tron USDT track - but it is P2P and completely informal. Coins.ph handles OFW remittances and funds flow into retail consumption. Nigeria’s P2P ecosystem channels $59 billion in crypto trading volume annually via OTC desks and USD savings accounts (via Chainalysis), acting as an alternative to the broken banking system. In Argentina, SUBE bus recharge is done through Tron USDT and cash OTC channels. Vietnamese freelancers receive wages in TRC-20 USDT, which are then redeemed through the local P2P network.

This is real economic activity - but it is not e-commerce infrastructure. No protocol really captures any part of this. The entire crypto-native e-commerce stack—product selection, checkout, custody, performance tracking, dispute resolution, points—is almost all blank.

How much of these needs will remain after compliance?

Before declaring this area the biggest product gap in crypto, a harder question must be answered: How much of the existing demand is structural, and how much is regulatory arbitrage? The honest judgment is that most of it is regulatory arbitrage. The mainstream use cases on the Tron-USDT e-commerce track today fall into three categories: demand for U.S. dollar exposure from users in areas subject to capital controls (Argentina, Venezuela, Nigeria) - these users cannot legally hold U.S. dollars through traditional channels; avoidance of VAT, sales tax, import duties, especially on digital goods and gift cards - it is difficult for tax authorities to verify the buyer's identity; and cross-border freelance and gig payments that bypass bank controls - mainly in Vietnam, Iran and parts of Africa. UQUID’s catalog is heavily skewed toward gift cards, airtime top-ups, and digital goods—categories that exist precisely because they convert opaque crypto balances into spendable fiat equivalents with little to no identity friction.

This is critical to the investment thesis, as the survival rates of regulatory arbitrage demands under compliance vary greatly. Domestic VAT and tax evasion needs are zeroed out the moment KYC is mandated at the merchant level - these users are not paying for a better checkout experience, but for "the column without a tax number". Once the tax number is required, the value disappears immediately. The need for exchange control circumvention is more durable because the underlying problems (Argentina's capital controls, Nigeria's naira controls, Venezuela's bolivar) are structural and long-standing. But the platforms that serve these needs cannot legally operate in the corridors they need. They can get bigger, but they can’t register, get priced financing, or partner with local fintech issuance banks—these partnerships are what give them a moat.

The chances of surviving compliance are narrow but real. Traditional tracks are slow or expensive for cross-border merchant settlement - Latin America ↔ Asia, Africa ↔ Anywhere, freelance payment collection - can run smoothly under any regulatory framework, Because the underlying value proposition is "stablecoins are structurally cheaper than SWIFT", not "stablecoins help you bypass the rules". B2B settlements between SMEs in different jurisdictions also fall into this category. The same is true for merchant settlement of cross-border digital services.

It follows that the term "US$5 trillion global e-commerce" is the wrong framework for this opportunity. The real investable acreage is closer to the $200 billion to $400 billion cross-border B2B and freelance payments market - with value propositions that transition from gray areas to legitimate markets. Domestic crypto checkouts for Western consumers—what most “crypto payments” narratives envision—is not the opportunity and never has been. The protocols that win this category will look more like “Wise for stablecoins” than “Shopify for crypto.” The key question for investors is whether a team is building for a market that will survive or one that will disappear.

Speculation: The battle for sustainability is long over

General knowledge: Decentralized sustainability is a competitive market, with dYdX, GMX, etc. competing for share with Hyperliquid.

The data shows: Hyperliquid has won. GMX and dYdX are not competitors, but protocols in terminal decline.

Hyperliquid currently controls over 70% of open interest on all on-chain perpetual venues,with $105 billion in monthly notional trading volume and $58.8 million in fees in March alone - over $640 million annualized (high confidence - TokenTerminal, DeFiLlama, Dune). Its fees increased 56% sequentially in the most recent reporting cycle. It has executed over $800 million in HYPE buybacks and is one of the few protocols where token value capture is more than just talk.

Compare veteran players. GMX’s daily income is 5,000 US dollars, and its daily activity is about 500. dYdX’s daily income is US$10,000 to US$13,000, with 1,300 daily active users, and its handling fees fell by 84% year-on-year. This isn't a struggling contender -- it's an agreement that the runway has ended mathematically, not strategically.

edgeX's data is noteworthy: 30-day post-verification fee of $14.7 million, fee retention rate of 73%, running on StarkEx ZK-rollup. There was an aggregation error in our previous dataset that originally showed $2.5 million - after correction, edgeX firmly ranks as the second most on-chain sustainable venue by revenue (high confidence - verified daily by TokenTerminal). Whether edgeX can maintain growth or whether it will follow the old path of GMX/dYdX is the only unanswered question in this category.

Hyperliquid deserves to be looked at analytically because it doesn't win by better trading UX - the differences at the order execution level from GMX or dYdX are real, but only marginal. It wins in liquidity depth, currency listing speed and brand. Once perpetual liquidity is concentrated in one venue, the network effect is almost unshakable: traders go to the place with the narrowest spread, the place with the narrowest spread is where the largest trading volume is, and the trading volume returns to where the trader is. The sustainable DEX category has passed the winner-take-all stage. The capital deployed against Hyperliquid in this category is equivalent to burning money.

Prediction markets: This is the story of category selection, not decentralization

Another speculative category worth examining is the prediction market. The mainstream narrative is that Polymarket has verified the path of on-chain prediction markets. The data tells a different story—and the lesson of this story has nothing to do with decentralization at all.

• Kalshi is an off-chain/CEX-like. The contrast itself is where the insight lies.

According to Bloomberg reports (high confidence), as of March 2026, Kalshi has annualized revenue of $1.5 billion and a valuation of $22 billion. More than $10 billion in transaction volume was processed in February 2026 alone, with transaction volume growing 12x in six months. Sports betting contributes 89% of its revenue. On-chain alternative Polymarket has monthly revenue of $4.7 million to $5.9 million and MAU of 688,000. Kalshi’s monthly revenue is approximately 25 times that of Polymarket.

The lazy explanation is that Polymarket has a UX problem. From most product dimensions, Polymarket is the better built one - the order book is cleaner, settlement is faster, and the trader experience is even more mature than Kalshi. UX cannot afford a 25 times income difference for this reason. The so-called Polymarket "hasn't started charging yet" defense actually makes the comparison worse, not better: If Polymarket were to lose 25 to 1 without any fees, the gap in underlying revenue potential would only be wider than it appears.

The real explanation is category selection, distribution channels and jurisdictional positioning - these three things have nothing to do with decentralization.

Kalshi chose sports. Sports is a high-frequency, popular, structured recurring category: There are betting opportunities every day of the week, every day, the rules are generally understood, and the audience will update itself with the new season. Polymarket positions itself in political and event markets - these are decentralized, election cycle dependent, and structurally low frequency. Users who come to Polymarket for the 2024 election have no reason to come back in March 2026. Those who come to Kalshi for the NFL will have a reason to come back every Sunday. Regular participation in compound interest is liquidity, liquidity compound interest is the spread, and price difference compound interest is more users. Polymarket is on the wrong side of the flywheel.

The second factor is distribution. Kalshi has built a B2B2C model, integrating order books with brokerage platforms, financial technology applications and partners, rather than relying on direct customer acquisition. Polymarket only does DTC, and every active trader bears the full marketing costs. Crucially, Kalshi operates legally within the United States under CFTC regulation, while Polymarket — following a 2022 settlement with the same agency — is fully geo-blocked to U.S. users. The largest English prediction market audience is structurally unable to be reached by on-chain products. Kalshi didn't just win on execution;it owned a market that Polymarket was legally barred from entering.

The implications for evaluating prediction market projects are concrete. The correct due diligence questions are: (1) What is the frequency of regular participation in the selected category; (2) Does the project have a B2B2C distribution path, or does it rely on direct customer acquisition; (3) What is the regulatory posture in the largest accessible market. The degree of decentralization is basically irrelevant to the results. Polymarket lost 25 to 1 because it picked the wrong category, picked the wrong distribution model, picked the wrong jurisdiction—roughly in that order of importance.

Inferences from this chapter

There are two key points in the speculative sector: (1) If a category has already emerged as a winner, it has really emerged as a winner, and capital should not be invested there; (2) The mechanism for the winner to win is not decentralization, UX or token economic models - sustainability relies on liquidity concentration, and prediction markets rely on category selection and distribution. Both conclusions point to the DeFi mullet proposition: the most defensible consumer positioning is to wrap a compliant front-end around the encrypted native back-end. Ether.fi Cash is the cleanest example at the moment. CrediFi and next-generation payments-adjacent products are part of the same model.

Stablecoin infrastructure:Tron is the most important consumer-level public chain, but no one talks about it

General knowledge: Ethereum L2 and Solana are the main consumer-grade public chains, and Tron is an established network mainly used for cheap transfers.

Data shows: Tron has over $600 billion in monthly stablecoin trading volume - comparable to Visa - has 14.3 million MAU, 72.8 million USDT holders, and a stablecoin velocity ratio of 0.2–0.3x - proving that its activity is payment rather than speculation. It has a whole set of unlabeled protocol shadow economies that have absolutely zero coverage in the Western media.

The numbers are shocking. The supply of USDT-TRC20 on Tron is $86.4 billion. Monthly transfer volume between $600 billion and $1.35 trillion (lower bound high confidence - TronScan, TokenTerminal; upper bound includes recurring volume). The single-day transfer volume on March 29, 2026 reached $44.9 billion. The network handles over 2 million transactions per day, covering 13.8 million MAUs, with an estimated 80%+ of transactions being less than $1000, and 60%–70% of them being less than $100. This is a retail payment network, not a settlement layer dominated by whales.

Speed indicators are key analytical signals. Tron’s USDT velocity is 0.2–0.3x, meaning that on average every dollar of USDT on Tron only turns over approximately every 3 to 5 months. Compared to speculative public chains, the speed can be more than 10 times - quickly cycle between DeFi protocols, leveraged positions and Launchpad. Tron's steady, slow speed is characteristic of payment rails: money comes in, is used for a real-world transaction, and then sits in the wallet waiting for the next bill or transfer. The top ten USDT holders on Tron control only 8.7% of the supply – indicating widespread retail distribution and decentralization.

Then there is the shadow economy. Our audit of TronScan identified several unlabeled protocols that generate significant revenue but have absolutely no documentation in English:

CatFee daily fee is $82,000. No one in the Western crypto media knows what it is. TRONSAVE earns $863,000 a month and has unidentified owners. These protocols run within the shadow economies of Vietnamese P2P networks, Nigerian OTC desks, Philippine remittance corridors, and Latin American cash corridors. We estimate that billions of dollars flow daily through these unlabeled clearinghouses—dynamic addresses, collection settlements, and freelance payments infrastructure that, in effect, act as banking systems for people locked out of traditional finance.

Celo is the fastest growing public chain in its category, driven entirely by the integration of MiniPay and Tether. The number of unique users increased 506% year-on-year, the total number of wallets was 12.6 million, and the transaction volume in December 2025 was $153 million (high confidence). But it’s still a fraction of the size of Tron.

Ethereum remains an institutional settlement track - higher fees limit retail use. Solana’s stablecoin activity is dominated by transactions and Launchpad traffic (pump.fun, Jupiter, Meteora), not payments. BNB Chain carries a monthly stablecoin trading volume of US$60 billion, mainly CEX settlement. TON is the wild card – Telegram’s wallet integration has brought huge signups, but the depth of participation remains unclear.

Comprehensive: Regulatory Arbitrage Life Cycle and DeFi Mullet

Every successful consumer encryption category in this census has experienced the same arc. Started from regulatory arbitrage; accumulated capital and users in the gray area; withstood - or failed to withstand - a compliance incident; the part that crossed over became a legal financial infrastructure. Today's protocols and companies generating real revenue are at different stages of this life cycle, and where they are determines the risk and reward curve of the investment.

Phase 1 - Gray Zone Initiation. The emergence of a protocol or service to solve the problem of refusal or inability of traditional finance to serve is almost always due to certain regulatory constraints. The user base is small, highly technical, and tolerates legal ambiguity. Profit margins are extremely high because regulatory risks are priced into the commission. There is no upper limit to tail risk. Today’s examples: unlabeled shadow clearinghouses on Tron (CatFee, TRONSAVE), the Nigerian P2P USDT countertop, and the early days of pump.fun, NFTs, and even the early days of Hyperliquid.

Phase 2 - Users and capital accumulation. PMF becomes unquestionable. Transaction volume grew, and users began to come from outside the core technology circle. Western media are starting to take notice, but regulation has yet to take action. Tron's USDT economy is at this stage today - 14.3 million MAU, with a monthly transaction volume of more than $600 billion. Pump.fun in 2024, Polymarket in the 2024 election cycle, and Hyperliquid now are also at this stage.

Phase 3 - Compliance Transition. A pushback event—litigation, enforcement action, settlement, or proactive regulatory communication—drives a project's options to legitimize, fragment, or die. This is the stage with the highest variance and the most analytical value from an investment perspective. Polymarket's 2022 settlement with the CFTC, pump.fun's $500 million lawsuit, and any future enforcement actions against offshore sustainable venues are all here. Most projects do not complete this phase.

Phase 4 - Legal Economy. The part that passes through becomes durable, auditable, and bankable. Returns are compressed because businesses are now valued by fintechs rather than moonshots. Kalshi (CFTC regulated, $22 billion valuation), Exodus (NYSE American, SEC filing), Circle (S-1 disclosure), and RedotPay (funded at fintech comparable multiples) are all located here.

After unfolding the arc in this way, the issue of investment timing becomes more specific. Stage 1 has the greatest upside potential, but it is basically impossible for institutional capital to place bets - the underlying business will be reset to zero as soon as a law enforcement order is issued, and underwriting cannot actually be done. Stage 4 is fully priced; the multiples are fintech multiples and the asymmetry has disappeared. Stage 2 has historically been the stage with the best VC returns in this sector, but only if there is a credible path through Stage 3. The due diligence issue in Phase 2 is no longer "did the product run smoothly?" - Phase 2 clearly ran smoothly. The question is whether the business model can survive compliance.

Tron's shadow protocols cannot pass this level because their raison d'être iscircumvention itself. Once Vietnam implements KYC for Tron USDT flows, CatFee's daily handling fee of US$82,000 immediately disappears - what users pay is not utility, but "no identity". There is no legitimate business model underneath. This is the fundamental difference between a "protocol with a PMF" and a "protocol with only regulatory arbitrage fit". Both generate income, but only one is investable.

DeFi mullet proposition is directly derived from this framework. Products like Ether.fi Cash and the next generation of Latin American fintechs win because they wrap a compliant front-end around a crypto-native back-end. The user cannot see or care what the chain is. What the regulator saw was an ordinary financial technology company. The protocol captures the economics of the “cheapest track.” None of these projects are currently issuing coins - this in itself is a signal: value capture occurs at the equity level rather than the token level, and the institutional investors who win in this cycle will be those who hold equity positions rather than token shares.

The three structural opportunities that emerge repeatedly throughout the entire text of this briefing are also launched from this synthesis: the monetization infrastructure in emerging markets (users are already present, but the income has not yet been received); the e-commerce track of cross-border B2B and freelance payment (the part that can survive in the e-commerce gap); and the Tron-adjacent protocol ecosystem that is still not covered and is in stage 2 of the life cycle. All three are most suitable for entry into the DeFi mullet model;All three reward category selection rather than decentralized purity; all three are undervalued today because Western capital is still looking at the wrong dashboard.

Data Quality Appendix

All data in this report is accompanied by one of three confidence ratings:

• High—Multiple independent sources, verifiable on-chain, or regulatory filings (such as Exodus SEC 8-K, TokenTerminal daily verification, Tether/Opera joint disclosure)

• Medium—a single credible source, or the company's self-reported and partially independently verified (such as Travala's self-reported revenue, Coins.ph Latka estimates)

• Low—press releases, unverified claims, or Growjo-level estimates (e.g. Telegram 150 million signups, UQUID 220 million users, Bitget 90 million users)