Interview with frontier tech investor Zheng Di: SEC's "innovation exemption" opens a compliant bull market — which targets are the potential winners?

Interview and editing: Jae, Tapbit news

Over the past week, Bitcoin has broken strongly above $87,000, driving a broad rally across the crypto market, and market sentiment has clearly improved.

The moment the SEC formally implemented a five-year "innovation exemption" for tokenized equities on September 18 may well have been the trigger for a new crypto bull market. "On-chain US equities" quickly became the strongest main theme, and valuation logic across multiple tracks — from exchanges to public chains to DeFi protocols — has been rewritten.

On September 23, Tapbit news invited frontier tech investor Zheng Di for an interview, engaging in a dialogue around the topics the market cares about most: the essence of the policy, the beneficiary targets, and the investment logic. He believes the "innovation exemption" is a national-level move by the United States to push the financial system toward on-chain migration, and a "compliance bull market" driven by RWA on-chain migration is unfolding. The endgame of blockchain, meanwhile, is to embrace the coming machine economy era.

Below is the full edited transcript of the interview with Zheng Di.

Administrative exemption replaces legislation: a compliance pass valid for at least two years

The CLARITY Act's defeat in Congress did not stop the advancement of US crypto regulation. The SEC used a single five-year innovation exemption to carve out an administrative alternative path.

Zheng Di said the value proposition of the "innovation exemption" is to press the pause button on the "securities test" for stock tokenization. Within five years, qualified institutions can freely experiment with putting stocks on-chain without worrying about being enforced against by the SEC for "unregistered securities issuance." Although there are still restrictions on the types of underlying assets and transaction amounts, it substantively implements the part of the CLARITY Act that the industry most looked forward to.

It is not just the SEC — the CFTC is acting in tandem as well. With the two working together, a scaled-down version of the CLARITY Act can operate through administrative means.

Zheng Di emphasized that the certainty of the time window is an important premise. SEC Chair Paul Atkins' term runs until 2031, and midterm elections cannot replace the chair; at the earliest, it would only be possible to nominate a new chair after the November 2028 presidential election and the inauguration of a new president in January 2029. In other words, unless an extremely crypto-hostile Democratic president takes office, the industry still has at least two more years of a moderate regulatory cycle.

Zheng Di also pointed out that the innovation exemption does not fully realize the original intent of the CLARITY Act. The CLARITY Act's main demand was to transfer most regulatory authority over cryptocurrencies from the SEC to the CFTC, dividing regulatory boundaries through a "maturity test": most tokens would not need a securities test and would instead be regulated by the CFTC as commodities; the few tokens requiring a test would also be given a 3–4 year exemption period. Simply put, the ultimate goal of the CLARITY Act was that, regardless of who the future SEC chair is or their attitude, crypto assets could escape the securities regulatory framework and be regulated under commodity logic.

To this day, the SEC still holds enormous discretionary and enforcement power over securities-type tokens. Precisely for this reason, this five-year "get-out-of-jail card" is especially precious: it grants the US equities tokenization business a safe "trial-and-error period" without triggering regulatory enforcement at every turn. This is the most solid policy floor of this round's "compliance bull market."

Beneficiary targets: Coinbase has the inside track, Robinhood is gearing up, Circle wins by default

Once the policy gate opens, who can cross the finish line first?

Zheng Di believes there are three institutions in the first tier: Coinbase, Robinhood, and Circle, corresponding to the three major roles of exchange, internet broker, and stablecoin issuer.

Coinbase: the best-prepared top contender

Among all players, Coinbase's product form is the closest to SEC requirements. It simultaneously holds securities issuance registration service qualifications, AMM technical capabilities, and distribution channels. Its existing stock token products already closely align with the innovation exemption's requirements on terms such as shareholder rights and dividend mechanisms. Its Verified Pools have already integrated the Base chain and Uniswap V4, and added a KYC identity verification module. This means Coinbase is not starting from scratch — it only needs to fill in details such as AMM operating rules, asset rights delineation, and notification mechanisms to quickly roll out the business.

Of course, a product being close to compliance does not mean the platform is approved to operate; final rollout still requires implementing various regulatory requirements. But it is undoubtedly the one currently closest to the finish line.

Robinhood: the customer-experience king with a shortcoming

Robinhood's advantage lies on the user side: its customer base, product experience, and stock-trading user habits are all better than Coinbase's, and its PFOF (payment for order flow) business model is also more mature. However, its shortcoming is equally obvious: it has not yet launched a stock token product in the US that meets the requirements, and its existing European stock token business leans more toward CFD (contract for difference)-type linked products. The legal structure cannot be copied to the US, and it needs to rebuild a product system that meets the innovation exemption's requirements. In other words, Robinhood's front-end capabilities are stronger, but its back-end compliance overhaul workload is larger.

Circle: the invisible winner of the stablecoin track

The explosion of on-chain trading will inevitably bring growth in stablecoin demand, and Circle is the direct beneficiary. If stock tokens are only traded on exchanges, fiat settlement suffices; but if trading moves fully on-chain, especially toward AI Agent trading scenarios, stablecoins become the mainstream medium.

The stablecoin track's landscape is also gradually clarifying: Tether's strategic focus is shifting toward the gold token XAUT, USDT's scale growth continues to slow, and Circle is all-in on USDC and on-chain money market funds.

In on-chain trading scenarios, USDC itself has an advantage over USDT. USDT's base is in CEX trading pairs, while the growth of on-chain native trading will benefit USDC and Circle.

Comparing the three platforms horizontally, Coinbase's management execution capability is relatively weak. Over the past six months, a large number of executives have departed, indirectly confirming the management problems.

Compared with Robinhood, its entrepreneurial spirit is also somewhat lacking. Robinhood CEO Vlad Tenev can take a symbolic $1 annual salary for four years and bet a huge incentive package against the board — an entrepreneurial spirit Coinbase does not possess. Circle is similar, leaning more toward a defensive, steady type.

Even so, Coinbase is currently at a reasonable valuation level. Zheng Di offered a valuation anchor: Coinbase acquired Deribit last year for a mix of cash and stock, with a total consideration of about $2.8 billion, of which the stock portion was priced at $200 per share. That is to say, Deribit, as the counterparty, recognized Coinbase's value at $200. The current stock price still has room relative to this level, and the catch-up logic holds.

Beyond opportunities, risks also need to be watched. Zheng Di also flagged a structural flaw in Coinbase's business model: its biggest competitor is actually ETFs, not other exchanges. Coinbase's retail trading fee is as high as 1.3%, while Bitcoin ETF annual fees are only 0.2%. As ETFs for BTC, ETH, and SOL roll out one after another, retail investors will increasingly shift to ETFs, the share of trading fee revenue from mainstream coins will keep declining, and Coinbase will become increasingly dependent on altcoin trading revenue.

Even more worth watching is the fragility of Coinbase's business structure: taking Q1 last year as an example, of $2 billion in revenue, $300–400 million was block rewards, which basically all had to be returned to users and does not count as effective revenue; of the remaining $1.6 billion, $1 billion was trading fees and $300 million was the USDC revenue share.

Of the $1 billion in trading fees, $900 million came from retail investors, and only $100 million from market makers. Retail investors contribute only 10%–15% of trading volume, yet contribute 85%–90% of the fees.

This means the more ETFs there are, the less mainstream coin trading there is, and the more sharply retail fee revenue declines. On the surface it is "revenue diversification," but in reality it is declining mainstream coin revenue and forced reliance on altcoins — this is the main reason its stock price has kept falling over the past few years.

Its incremental space will depend on the acquired Deribit options business: the acquired Deribit has about $1.8 trillion in annual trading volume, charging 3–4 bps, roughly $400 million in annual revenue. At a 40% net margin, that is only $200 million. It currently contributes to the consolidated business but cannot yet support the core fundamentals.

Protocol-layer dividends: Uniswap is the strongest, Arbitrum next, Base has variables

Drilling down to the protocol layer, dividends will be transmitted layer by layer along the tech stack. Uniswap is the DeFi protocol with the highest degree of fit. The permissioned pools introduced in V4 specifically provide a technical solution for securities-type assets subject to transfer restrictions to enter AMMs, which is also why it can enter the SEC's innovation exemption candidate system. As the underlying infrastructure for on-chain trading, Uniswap's value capture logic is the clearest, and it is also the crypto target most easily accepted by traditional financial institutions.

Arbitrum's logic is tied to Robinhood: Robinhood Chain is deployed on Arbitrum, and 10% of its revenue is paid to the Arbitrum ecosystem. Of that, 2 points go to the developer program and 8 points go to the treasury. This gives ARB a clear earnings expectation, but compared with UNI, its position is further back, and value transmission is one layer more distant.

Base is also worth watching, but the situation is more complex. As Coinbase's Layer 2, Base once accounted for 70% of Ethereum Layer 2 network trading volume, with an operating margin exceeding 80% — an excellent business. Zheng Di pointed out that if Coinbase's tokenized equities business fully takes off, Base's imagination as the underlying settlement network is enormous, and Coinbase's catch-up logic will also be supported.

However, it carries two hidden risks:

First, the entanglement of interests with Optimism. Early on, the agreement stipulated that Base pay Optimism 2% of revenue and 15% of operating profit. Before the Blob upgrade, L1 data service fees accounted for 50%–70% of L2 revenue, and the 15% profit share was acceptable; but after the Blob upgrade, L1 toll fees dropped 90%, and the 15% profit share became a huge expense. Base might have to pay Optimism tens of millions of dollars a year, which is also why there are market rumors of Base leaving the Optimism system.

Second, regulatory risk. Base's sequencer is almost monopolized by Coinbase alone, which could be deemed by the SEC as an exchange rather than technical infrastructure, thus requiring an exchange license and stricter regulation.

The deeper reasons for guiding US equities on-chain

There is one question no discussion can avoid: US investors have extremely rich channels for buying US equities, and traditional brokers and internet brokers offer excellent experiences — why migrate on-chain? And additionally endure a series of problems such as wallet operations, tax handling, and insufficient liquidity?

Zheng Di's answer is: if it were merely moving human stock trading from brokers to wallets, on-chain has no advantage at all. The real incremental demand comes from scenarios that traditional finance cannot satisfy.

The first type of demand: the cure for the asset shortage of on-chain native capital

The biggest lesson of the last bear market was the lack of quality assets — everyone was speculating on air coins. Among native crypto assets, apart from a few targets like BTC and ETH, the vast majority have no real cash flow and are essentially a zero-sum game among in-circle capital. In a paper early last year, the SEC did the math: excluding mainstream coins with ETFs, market makers alone siphon $4.8 million per day from the market for the remaining crypto assets, not even counting exchange fees. Without external fresh capital and with only continuous internal draining, the collapse of the altcoin market was inevitable.

On-chain US equities will solve this pain point. For the first time, the crypto industry's native capital can allocate at scale to standard assets with earnings, dividends, and real value, rather than cutting each other in air coins.

The more practical significance is that on-chain trading eliminates the friction cost of fiat conversion and also circumvents the constraints of cross-border tax information reporting such as CRS (Common Reporting Standard). In the last DeFi Summer, there was a batch of funds on the Ethereum mainnet in the $100–300 billion range that would rather bear high gas fees than enter CEXs or go to Layer 2 networks. Clearly, they have a strong on-chain allocation demand, and US equity tokens are their natural new asset.

The second type of demand: US national strategy and offshore capital

From a higher vantage point, promoting the on-chain migration of financial assets is essentially a US national strategy. The US will have a large batch of frontier tech company IPOs: Anthropic's valuation is heading straight for $2 trillion, OpenAI is at least $1.5 trillion, and Blue Origin and others are all in the hundreds of billions to trillions of dollars. Such enormous financing demand cannot be absorbed by US domestic capital alone and requires global offshore hot money to participate. However, as capital controls in various countries tighten, capital inflows through traditional broker channels are becoming increasingly difficult.

On-chain US equities provide a new channel for global capital, both helping US assets complete financing and promoting the global penetration of dollar stablecoins. Just as the 1960s saw the shift from paper trading to electronic trading, this time it is the shift from electronic records to on-chain records — a new round of settlement revolution. If the US completes the on-chain transformation of its financial system first, it can seize the initiative in future global financial competition.

The third type of demand: the endgame of the machine economy

From first principles: the native users of blockchain are machines, AI Agents. Industry research institutions have conducted tests, letting different AI models choose transfer methods, and 70% chose stablecoins over fiat. Compared with traditional banking and brokerage systems designed for humans and requiring manual intervention, machines are naturally suited to on-chain settlement, programmable, automatically executing financial systems.

This also explains why blockchain, after nearly 20 years of development, has never achieved mass adoption. It is a new technology that appeared too early — it is the financial infrastructure for machine trading, and was never prepared for humans. This adoption cycle may be 5 to 10 years, but capital markets price it in advance, and the current market is reflecting this long-term vision.

Target analysis: who truly benefits, and who is riding the hype?

Under the policy tailwind, various targets have risen in turn. Stripping away the emotional filter, which ones have solid fundamental support, and which are merely hype-driven themes?

Breakout assets: trading infrastructure

Uniswap is the DeFi protocol with the highest policy fit. Its essence is an on-chain exchange, with clear fee revenue and a clear value capture mechanism that can be valued with a DCF model. If everything goes on-chain in the future and Uniswap becomes the base layer of global trading, its value ceiling will be extremely high. In addition, its attribute of being "understandable and calculable" allows it to break out of the crypto circle and be added to traditional investors' watchlists. Zheng Di estimates that if Uniswap had a corresponding DAT, a large number of traditional institutions would be expected to trade its stock.

Similarly, Hyperliquid was a target recognized by the traditional finance circle in the last round, with institutions entering to allocate, and it also follows the exchange logic. However, Hyperliquid still needs to resolve compliance issues.

One layer further down, Uniswap is Ethereum-native infrastructure. Although it is deployed across multiple chains, most of its trading volume is still in the Ethereum ecosystem. In the long term, this is also a major positive for Ethereum: the utility improvement brought by RWA on-chain migration will redefine Ethereum's value.

Some argue that the Blob upgrade has affected Ethereum mainnet's value capture, but from another angle, Layer 2 networks' operating margins can now exceed 80%, taking initial shape as future exchanges — requiring little manpower, only modular KYC/AML systems, with costs far lower than traditional exchanges.

Zheng Di gave an example: today's Ethereum is like Pinduoduo in its "ten-billion subsidy" phase — using low fees to subsidize the Layer 2 ecosystem and grow the ecosystem. The bull market logic is: grow the ecosystem, and there will always be a day to raise prices; cash flow is only a matter of time. This is also one of the drivers of Ethereum's rise from $1,800 to $2,700.

Looking back at history, Ethereum first rose through ICOs, then fell to $80 after the bubble burst, and once faced an existential crisis; later it reached a second peak through DeFi, with applications like Uniswap bringing massive gas consumption that re-supported Ethereum's value.

However, this round differs fundamentally from DeFi Summer: this round is RWA on-chain migration, backed by real trading volume and cash flow, with a more solid foundation.

Today, the industry is entering its third phase: blockchain itself no longer insists on "native crypto assets," but serves RWA on-chain migration as underlying technology, with enormous space. The current pricing is trading the future ceiling, not current revenue.

As for the exchange track, all CEXs will eventually compete head-on with traditional brokers. The assets traded on each platform will converge into stocks, but different platforms' strategies will clearly diverge. Take Binance, for example: its strategic focus is tilting toward primary-market projects, because the pricing power of US equity tokens lies with the NYSE and Nasdaq, and CEXs are only shadow markets — they cannot earn the money from pricing power, and profit margins are limited. Primary-market equity is different: hot unlisted companies have no continuous trading and can only be traded inefficiently off-market through financial advisors, while the market has a strong demand for continuous trading. Binance can acquire primary-market equity and use it as collateral to issue on-chain equity tokens or perpetual contracts, holding the pricing power itself.

Recently, Binance announced upgrading its funding account to a stock account, which is also a clear signal: accelerating toward RWA. Non-standard assets like real estate trusts and wine are hard to scale; the RWA track's best opportunities are standard assets, and the biggest standard assets are US Treasuries and US equities. US equities have high volatility and better match crypto capital's risk appetite, naturally becoming the best vehicle for RWA.

All CEXs still in operation will eventually become "offshore US equity brokers." And the new battlefield where pricing power can truly be seized is the continuous trading of primary-market equity.

As for other targets, Arbitrum's logic is slightly weaker but has clear earnings support. Robinhood Chain's revenue sharing is visible cash flow, but its market recognition is still mainly confined to the crypto circle, with a low degree of breakout; Near is similar — its logic mainly lies in cross-chain execution and user experience, further from value capture, and its trading group is more skewed toward native users, making it harder to break out.

Zheng Di concluded that beyond BTC and ETH, the crypto assets that have truly "broken out" and been recognized by traditional financial institutions are, for now, Hyperliquid, followed by Uniswap.

Potential track: prediction markets, the prototype of the next-generation exchange

Prediction markets are the most undervalued track in the entire crypto industry, and may become the next-generation exchange.

Zheng Di pointed out that the prediction market track basically has two and a half players: Kalshi and Polymarket are the top two, and Robinhood counts as half. Currently, Kalshi's monthly trading volume is about $300 billion, Polymarket's is up to $100 billion, and Robinhood's prediction market is only $40 billion — a significant gap.

Kalshi is almost all-in on sports, with sports accounting for 3/4 of trading volume, and fees starting at a minimum of 1%, with high-demand markets charging 2%–5% — several times that of ordinary exchanges. This year, it spared no expense on advertising to acquire users, achieving great success during the World Cup, with extremely rapid user growth. Kalshi has also obtained a perpetual contract license and begun launching related products, and the closed loop of traffic monetization is being completed.

Polymarket is relatively balanced, with sports, crypto, and political futures forming a three-way split. It has not been long since it returned to the US compliant market, only starting to charge fees in March, so its revenue side is much weaker than Kalshi's, with a valuation gap of double. Not long ago, it completed financing at a $200 billion valuation backed by the Trump family, while Kalshi's next round is heading for $400 billion.

The advantage of prediction markets lies in the potential for cross-selling: sports event predictions serve as a customer acquisition entry point. The user base of sports betting is far larger than that of stock trading, and the user base of stock trading is far larger than that of crypto trading. Acquiring users through high-frequency sports events and cross-selling stocks, options, cryptocurrencies, and perpetual contracts to them — this customer acquisition logic is hard for traditional CEXs to match.

The biggest pain point of traditional crypto exchanges is the difficulty of acquiring new users; they can only compete internally within the existing user base. Prediction markets, by contrast, continuously acquire new users from outside the circle and then convert them into trading users. Robinhood's prediction market has the same problem: it does not acquire users through sports, but only funnels its existing stock and options trading users into the prediction market — this is deep mining of old users, not incremental customer acquisition.

Zheng Di believes prediction markets are the most noteworthy new force in the crypto industry, and because they are unlisted and have not issued tokens, the market underestimates their explosive power. They will reshape the landscape of the entire trading market, and if traditional CEXs do not quickly follow up, they may be struck by a dimensionality-reduction blow.

Supplementary opportunity: privacy demand under compliance

As the main tracks become increasingly "white" and KYC/AML requirements become increasingly strict, a portion of capital with high privacy needs will inevitably seek an outlet.

In the future, when enterprises transfer stablecoins among themselves, they will need both compliant identity and transaction privacy, and zero-knowledge proofs will become a rigid need. "Bitcoin's public key is public, which is unfriendly to many institutions and high-net-worth users; privacy coins like ZEC will absorb this demand," Zheng Di added.

The undertone of the bull market: market-wide underweight + two market catalysts

Returning to the market itself, the strength of this round has exceeded many people's expectations. What is its underlying support, how far can it go, and what signals should be watched?

The anchor of bull and bear: the 200-day moving average and market-wide underweight

Technical analysis sometimes seems like mysticism, but when the entire market is watching the same indicator, it becomes self-fulfilling. Zheng Di pointed out that Bitcoin's 200-day moving average is the dividing line between bull and bear.

On August 21, Bitcoin effectively broke above the 200-day moving average (around $70,000), which was the signal for the official start of this bull market. This round's strength far exceeded expectations, with almost no meaningful pullback, leaving a large amount of sidelined capital behind.

Zheng Di believes the main reason behind this is the extreme market-wide underweight. Not only are outside institutions' positions low, but a large number of OGs within the crypto circle also have extremely low positions; many even sold coins to allocate to AI stocks, waiting to buy the dip in October. As a result, the market started two months early, causing widespread missed rallies.

This pattern of scarce counterparties determines that the probability of a deep correction is low. Every pullback has a large amount of sidelined capital waiting to get on board. The $82,000–85,000 range was the previous dense trapped zone, and after breaking through it has turned into strong support. As long as $82,000 is not broken, the market will continue its strength; as long as the 200-day moving average is not broken, the bull market structure will not change. And once a bull market starts, it lasts at least a year — this is a basic law.

Two key variables that determine the market

On the question of how long the bull market will last, Zheng Di offered two main catalysts:

The first is oil prices and US Treasury yields. All liquidity-driven assets are ultimately anchored to the 10-year Treasury yield. Around 5% is already the "pain zone" for US fiscal policy, and the government has sufficient incentive to suppress yields through diplomatic and energy means.

This round's collective breakout of risk assets is highly correlated with oil prices falling below $100. Oil prices determine the direction of CPI, which in turn determines the direction of yields. If oil prices are unstable, inflation rebounds, yields rise, and all risk assets will come under pressure.

The second is the progress of AI white-collarization. AI should be deflationary, but it has not yet produced a deflationary effect. OpenAI's internal timeline is March 2028, to achieve fully automated AI white-collar workers that can work continuously for a week with a low error rate, but the models are not yet strong enough — the success rate of working 16 hours continuously without human correction is only 50%, so they cannot replace white-collar workers. Meta originally planned to cut 60% of staff last year, but later found that Agents could not meet the requirements, so it hired people back. Large-scale white-collar layoffs have not happened, so inflation naturally cannot come down.

If highly autonomous AI Agents are achieved in 2028, large-scale white-collar layoffs will arrive, and inflationary pressure will drop significantly. The market may gradually price this in by late 2027 or early 2028, at which point yields will fall and capital markets will see a great boom — possibly the final main upward wave of this bull market.

Zheng Di expressed agreement with Fed Chair Kevin Warsh's view: AI is inflationary in the first 1–2 years, as infrastructure construction pulls employment, just like China's real estate development back then.

Once it truly begins to replace labor, it will become a strong deflationary force. Today, the market is on the eve of the tipping point, and the expectation of peaking yields combined with extremely low positions has jointly given rise to this bull market.

Epilogue: the wind has changed

From ICOs to DeFi Summer to the Meme wave, the crypto industry's past narratives have always revolved around "native assets" — creating new assets on-chain and completing self-circulation within the circle.

The SEC's five-year innovation exemption has formally pushed open another door. US equities, the largest and most liquid standard assets in the real world, are about to migrate on-chain at scale. The crypto industry's mission will also shift from "creating new assets" to "serving real assets."

In the short term, the innovation exemption has opened a compliance bull market, injecting new assets — US equities — into a crypto market that has experienced an asset shortage; in the long term, it is the first step in the global financial system's migration on-chain, building the payment and settlement foundation for the coming machine economy.

After September 18, the wind has quietly changed.