Delphi Digital researcher’s ten predictions for 2025: DePIN’s market value will increase five times, and stablecoins will usher in multifaceted prosperity.

Author: Robbie Petersen

Compiled by: Shenchao TechFlow

Prediction #1: The front-end will dominate value capture

As the MEV supply chain matures, those participants with exclusive order flow will gain more value.

The reason is simple. Various participants located downstream of the order flow—such as DEXs, searchers, builders, and validators—will face more intense competition. The originator of the order flow (ie, the front-end) has a natural monopoly advantage in the MEV supply chain.

This means that the only role capable of improving yields without significantly losing market share is the front-end, especially those that master “less fee-sensitive” order flow (e.g. digital wallets).

Additionally, emerging technologies such as conditional liquidity (e.g. @DFlowProtocol) will further drive this trend forward.

Prediction #2: DePIN market cap will increase 5x by 2025

Market leaders in decentralized physical infrastructure networks (DePIN) such as @Helium and @Hivemapper are approaching breaking point in their network effects. And @dawninternet has become the most groundbreaking application of the year in the DePIN field with its significant technical improvements and encryption economic incentives.

Prediction #3: Crypto payment rails will have limited application in

agent trading

In the early days, transactions between humans and agents will still rely on traditional payment tracks. Stripe and PayPal will dominate the early stage of smart payment infrastructure through hold-for-others (FBO) account structures.

But only when the autonomy of the agent reaches a certain level, the high fee model of the traditional payment rail will expose its limitations. Due to the rise in demand for microtransactions and usage-based pricing, traditional payment rails (with ~3% fees) will be unsustainable.

However, this will not happen in 2025, as most transactions will still be interactions between humans and agents. (reference tweet)

Prediction #4: Stablecoins will cross the fintech adoption divide

The role of stablecoins will change from the "lubricant" of DeFi (decentralized finance) to a true medium of exchange.

This shift is driven by two main reasons why fintech companies adopt stablecoins: (1) to increase profitability and (2) to strategically control more payment chains.

As the widespread adoption of stablecoins becomes an inevitable choice for the survival of financial technology companies, the number of monthly active stablecoin addresses is expected to exceed 50 million.

Prediction #5: Visa launches stablecoin plan and proactively adjusts

profit structure

In order to cope with possible disruptive changes in the payment chain, Visa has laid out its stablecoin plan in advance. While that could cut into its card network's profits, the risk appears more manageable than being completely disrupted by the market. This logic also applies to other fintech companies and banks.

Prediction #6: “Yield Distribution” Stablecoin Market Share Will Grow 10x

“Yield distribution” stablecoins such as USDG @Paxos, “M” @m0foundation and AUSD @withAUSD change this by reallocating the revenue traditionally received by stablecoin issuers to those applications that provide liquidity to the network The economic model of stablecoins.

Although Tether will maintain its market dominance in 2025, the "yield distribution" stablecoin model is considered to be the way forward for the following reasons:

(1) The importance of distribution channels: Unlike previous revenue-based stablecoins that tried to directly attract end users, “revenue distribution” stablecoins target applications with distribution channels. This model achieves the simultaneous alignment of incentives for distributors and issuers for the first time.

(2) The power of network effects: By incentivizing the simultaneous integration of multiple applications, “revenue distribution” stablecoins can fully leverage the network effects of the entire distributor ecosystem.

In 2025, as distributors (especially fintech companies) and market makers cooperate, the market share of these stablecoins will increase significantly because they are able to create more direct benefits for distributors.

Prediction #7: The line between wallets and applications is increasingly

blurred

The wallet will gradually integrate application-like functions, such as deposit income (such as @fusewallet), credit accounts (such as @GearboxProtocol), native transaction functions, and a chatbot-like interface through which users can express their needs, controlled by AI agents and The background resolver performs the operation.

At the same time, the application will also try to maintain a direct relationship with the end user by hiding the existence of the wallet. For example, the mobile app launched by @JupiterExchange is an early example.

The biggest impetus for the vision of wallet centralization comes from exchanges like @coinbase, which view wallet products as the primary way for users to monetize on-chain. (reference tweet)

Prediction #8: Chain abstraction will be implemented from theory to

practice at the wallet level

Although the discussion of chain abstraction has previously focused on the chain and application layers, the optimal solution is to directly meet user needs. New technologies like @OneBalance_io’s Resource Locks, @NEARProtocol’s Chain Signatures, and @Safe’s SafeNet are driving a new paradigm of chain abstraction at the wallet level.

Prediction #9: General-purpose L2 will gradually lose relevance

The central trend of future blockchain activities can be boiled down to one question:

As an application, why should I choose to run on your chain?

For a few general-purpose chains with clear positioning (such as Solana and Base) and vertically integrated chains (such as HypeEVM and Unichain), the answer is clear.

However, for long-tail, general-purpose chains, the answer is less clear. By 2025, blockchain activity will become increasingly concentrated on those few chains that provide clear value to applications.

Prediction #10: The boundary between attention and value will gradually

disappear

As the most direct embodiment of the attention value theory, the value of AI agent tokens will continue to grow.