In a Bankless podcast released on Sept. 7, 2026, Austin Barack, founder and managing partner of Relayer Capital and a former partner at Coin Fund, sat down with Bankless co-host David to discuss how he is framing the current crypto market and why he sees upside in VVV, ETHFI, Pump, Hype and ETH.
The episode ran for about 67 minutes. A disclosure attached to the discussion said Barack is the founder and managing partner of Relayer Capital, and that Relayer’s liquid token fund holds several of the assets mentioned in the show, including VVV, Pump, Hype, ETHFI and ETH. It also said he was an ETHFI Series A investor and led that round in February 2024. The target prices, valuation multiples and revenue forecasts discussed in the episode were presented as his personal views based on his own models, not as investment advice.
Relayer’s current framework: where growth meets value
Barack said crypto investing changes from cycle to cycle. A playbook that worked in 2017 may stop working in 2021 or 2024. One theme, in his view, keeps returning: the overlap between growth and value.
His argument was straightforward. Investors do not come to crypto looking for a company growing 10% a year at a 4x valuation. They come for assets with fast growth and room for re-rating. Because crypto capital is cyclical, valuations swing between periods of extreme overpricing and deep discounting, creating moments when both growth and value can show up at the same time.
Barack said Relayer works across both liquid markets and venture investing, but this year it has put 95% of its focus on liquid tokens, with the remaining 5% reserved for seed-stage projects. The two areas he is watching most closely are crypto tied to AI and tokenization built around 24/7 trading. He grouped Venice, Pump, Hyperliquid and Ether into the overlap of those themes.
On the old layer-1 and layer-2 trade, his view was blunt: the infrastructure era is mostly over, and capital this year has been moving toward applications.
How Barack gets to a $43.89 target for VVV
David brought up a recent post from Barack arguing that Venice’s VVV token looked deeply undervalued at a $1 billion fully diluted valuation and that his model implied a fair value of $43.9. Barack then walked through the logic behind that figure.
He said he does not build these estimates by instinct. His background is in corporate development and FP&A, and he works backward from revenue, gross margin and reinvestment needs. In his description, Venice is building a private, uncensorable AI gateway. Today it charges in two ways: subscriptions and credit purchases. The subscription tiers are priced at $18, $68 and $200, and users that go beyond their limits have to buy additional credits.
Barack said Venice raised a round in late June or early July at a $1 billion valuation using a combined equity-and-token structure. He acknowledged that markets often react nervously to those setups, but argued that Venice handled it in a clean way. The consumer AI business runs off-chain first, with users paying by credit card, while the token captures on-chain burns and tokenized compute.
He said Venice is currently sending almost all free cash flow, meaning what remains after reinvestment, back to the token in the short term. Two burn mechanisms are already in place:
- A VVV burn tied to each new subscription, scaled by tier.
- A 5% burn on credit purchases.
Over time, he expects more mechanisms to be added, including subscription renewals.
David pressed him on whether credit purchases were simply another form of subscription demand or a separate product. Barack described that piece as a smaller, "2-point" new product meant mainly to encourage heavier use. He drew a sharper distinction with Minds, an AI app store that has not launched yet. In that model, power users could package prompts and model combinations into products for sale to other users, Venice would take a cut, and token burns would follow. Barack called Minds the biggest variable in his 2027 burn model.
The figures he cited were specific:
- As of August, Venice was running at $107 million in annualized revenue.
- Annualized burns were $8.3 million.
- By 2027, he assumes revenue reaches $336 million.
- He assumes burns reach $70 million.
- Of that amount, Minds contributes $29 million, or 40% of total burns.
Under those assumptions, he takes $70 million in burns, applies a 50x multiple, gets to a $3.5 billion token valuation, and then divides that by the expected token supply at the end of 2027 to reach $43.89 per token. He said he would score the optimism of that model at 6 out of 10.
Barack added that he assumes the burn rate on credit purchases rises from 5% today to 10% by 2027. In his telling, Venice has already moved along a path from discretionary burns to new-subscription burns to credit-purchase burns, adding each mechanism as the model gets tested.
On the current market price, he said VVV was around $16, versus roughly $12 when he updated the model a few days earlier. He also said that 10x revenue growth in eight months was not something he had expected, and that the $29 million Minds assumption was impossible for him to model eight months ago. For that reason, he said the framework should be read as reasonably optimistic rather than fully stretched.
Barack also described his own history with the project. He said he had followed VVV since its listing in early 2025, first through an airdrop linked to the Virtuals and AIXBT wave. He later lost the thread for a while, then re-engaged in early 2026 after reading Eric’s long post on the token model and DEM, or tokenized compute, while riding in a taxi. He said he finished the thread in 40 minutes and opened a position for the fund right after.
Why he thinks ETHFI can at least double
Etherfi was another major focus of the conversation. Barack said it was the first venture investment made by the fund and that he led its Series A in February 2024. At that stage, the business was still centered on liquid staking and restaking. What convinced him, he said, was not only the team’s execution but its sense of where products were heading.
His point was that the team understood which products were customer acquisition tools and which ones were likely to become commoditized. Liquid staking and restaking, in his view, belong in the latter category.
That is why he believes the market has not caught up with what Etherfi has become. Barack said 65% of the company’s revenue now comes from what he called a "new bank" business, meaning card usage and borrowing interest, while only 35% still comes from yield and staking. To him, that represents a full reversal in the business mix.
He used market behavior around Ethereum’s inflation reduction plan as an example. On the day that proposal came out, Etherfi fell by more than 10%, and Lido also dropped. Barack said that move showed the market was still pricing Etherfi through a liquid staking lens even though, in his view, Etherfi and Lido are no longer comparable businesses.
When asked for a comparable, he named Nubank. He described Nubank as an $80 billion company with 139 million users built around friendlier consumer products. Etherfi, in his framing, is a global on-chain Nubank powered by stablecoins.
He then took that one step further. As Ethereum itself evolves from a new kind of bank into a new kind of brokerage, with tokenized stocks and real-world assets connected in the same system, Etherfi can package that stack for users and sell it with little incremental cost.
On operating performance, Barack said Etherfi is currently generating $3 million to $4 million a day in revenue, up 10x year over year. He stressed that the newer business lines are still ramping. He also said the main revenue engine for modern digital banking and brokerage models is borrowing interest. Nubank gets 60% to 70% of revenue from borrowing interest, while Etherfi gets only 4% from that category today, leaving what he described as a large runway.
He added that Etherfi has an 80/20 revenue split with Aave v4, with lending attached directly to an Aave instance, which in his view lets the team stay lean.
For valuation, Barack referenced a more conservative model from Blockworks that projects $21 million in buybacks and burns over the next 12 months. He said he leans more optimistic and sees at least $30 million. Using a 30x multiple, which he called reasonable for a new-brokerage trade, he said the token could move above $1. With ETHFI still around $0.5, that would imply at least a doubling.
He also said Blockworks cut its growth assumptions in half, while he thinks growth should be marked higher, not lower. Another part of the setup, in his view, is supply. Most of the token supply is already circulating, so there is less concern about new unlock pressure. Structurally, he said, it looks more like a mature stock with sustained buyback support.
Pump and the case for multiple expansion
The conversation then turned to Pump. David noted that the token trades at only a 5x buyback multiple, compared with 30x to 40x for Hyperliquid and Lighter, and asked whether that gap made sense.
Barack said it does not. He called Pump cheap.
He described the business as the biggest casino-style trade in crypto and compared it with Las Vegas Sands, MGM, DraftKings, FanDuel, prediction markets, zero-day options and speculative activity on Robinhood. In his telling, these are all versions of the same underlying business. Users do not always engage with negative-expected-value products to make money; often they do it for variance. That kind of business, he said, can last for decades.
He backed that up with operating data, saying Pump’s revenue over the past 90 days was up another 80% quarter over quarter and that the business had already shown durability for more than two years. He argued that the market is still in the process of recognizing that durability.
His valuation view follows from that. He said a 10x buyback multiple is more reasonable than 5x. Even with no growth, a pure re-rating from 5x to 10x would imply a doubling. If growth continues, he said, there is more room above that. He acknowledged that Pump had already risen 3x over the past six weeks to two months, but added that he does not think the move is finished.
Barack also raised a token-versus-equity risk. Pump originally committed 100% of revenue to buybacks, but that was never guaranteed forever. This year, the model shifted to 50% of revenue for buybacks over a 12-month commitment, with the other 50% reserved for reinvestment. For a multibillion-dollar operation holding a large amount of tokens and trying to build a durable business, he said the renewal decision next year remains a live variable.
He said the market is justified in discounting that risk. At the same time, he argued that for a top-tier team, the odds of simply walking away from the token are low. In valuation terms, he said that risk can be expressed through lower buyback multiples: 6x to 10x in a conservative case and 10x to 14x in a more optimistic one.
Why Hype stands out in his framework
Barack said Hype may be one of the clearest examples of the crypto model being fully expressed, excluding stablecoins, BTC and Zcash. He highlighted three features in particular: 24/7 settlement, every asset living on-chain, and an attempt to move the whole financial system onto blockchain rails.
One newer development he emphasized was price discovery for pre-IPO assets. He said names such as SpaceX, Cerebras and Unitary, none of which had formally gone public, were already finding market-clearing prices on Hype. He suggested that investment banks pricing future IPOs may eventually look to Hype screens as a reference point.
Financially, he said the latest surge has been driven by market volume on HIP-3, including commodities, equities and indexes, though he also made clear that these lines are still in investment mode and are not being monetized much yet. The real cash cow, in his telling, remains perpetuals tied to crypto tokens.
He gave a simple comparison to show the shift. About a week earlier, daily fees were still running around the low millions. More recently, on one day, they hit $5 million. His takeaway was that as capital flows back into the core crypto market, Hype is one of the clearest beneficiaries because it sits directly on top of crypto capital rotation and token volatility.
This cycle, in his view, belongs to applications and money
Barack’s broad cycle call was that crypto has already moved beyond the period when infrastructure captured nearly all the economics. He said that at one point infrastructure accounted for more than 95% of industry revenue, while now applications take roughly two-thirds and infrastructure about one-third. He expects that shift to continue, and said applications could eventually capture more than 90% of the revenue pool.
That is why, in his framework, the most durable token categories are likely to be applications and money-like assets. He explicitly mentioned BTC and Zcash in that context.
He said Bitcoin is not going away, and that Zcash-type assets now have a role again because they serve a different user base and carry the original purpose that crypto had a decade ago. He added that he has recently seen structural inflows from OG Bitcoin holders moving back into that trade.
On ETH, Barack said the asset sits in a very interesting position and could regain relevance as money, a lens he admitted he had not taken seriously for a long time. The reasons he listed included quantum computing risk for Bitcoin, concentration risk around Strategy’s holdings and other structural variables. Taken together, he said, those factors make ETH’s optionality as money stronger than at any point before.
He added that Ethereum has long relied on business activity to support its value proposition, and that the return of a money narrative changes the valuation discussion around ETH.
His added comments on Solana, Base and the Robinhood chain
David also asked about Solana, Base and the Robinhood chain. Barack said Solana is currently the leading chain for on-chain spot activity and underpins Pump’s broader business. Even so, he called its current position awkward: usage is extremely high, but the chain itself has not captured much revenue, in part because less value is being extracted through MEV. He said Solana remains one of the most important bets in crypto adoption, but the central question is how it converts usage into revenue.
His comments on Base and the Robinhood chain were shorter. He said both are heavily used public chains, and that the main thing to watch is how much value the application layer can pull out of that activity.
How he thinks 2026 will be remembered
At the end of the episode, David asked how the 2026 cycle might look in hindsight. Barack’s answer was that 2026 will be remembered as the year when the application trade and the money trade both started to pay off, and as the year when 0-to-1 products that connect crypto with the real world began to enter the market in force.
Looking back, he said, assets such as Venice, Hyperliquid, Pump, Etherfi, BTC and Zcash may end up being remembered as some of the defining entry points of this cycle.
The Chinese article was compiled and translated by TechFlow. The original podcast source was Bankless, with Austin Barack as guest and David as host.