Nick Cherney (Janus Henderson): “In a tokenized world, tokenization looks a lot like a transfer agency”

Rather than launching a standalone, in-house tokenized share class, Janus Henderson went through Centrifuge and integrated directly with DeFi protocols. Why prioritize protocol integration?

Our long-term vision is that fundamentally everything we do is going to move on-chain, and that extends across financial services. We see a future in which the basic architecture of the financial system operates on blockchain rails. But getting there requires a very complex shift in the operating system of entrenched legacy companies, ourselves included, our clients, and all of the infrastructure around them.

A lot of market participants, asset managers and banks alike, have started building walled-garden, proprietary ecosystems in view of that future. I think that misses the point. We are not going to jump from A to B; we have to bridge the gulf. There are huge businesses evolving in DeFi that are in some cases competitive with, new versions of, or totally disruptive to traditional finance. So from the very beginning our strategy has been to optimize our efforts around the large economy that has grown up natively on-chain, because those are the people who deeply understand the technology and get real benefits from it today. We integrate there, provide services, learn with them, and build the expertise and credibility to eventually bring traditional finance into that ecosystem. Traditional finance is where most of the money is, but the shift required for it to move onto this operating system is still quite high. So we pursue both paths: educating our traditional clients while building where the activity already is.

"The critical diligence is on the partner bringing us on-chain. The protocols using our funds are clients."

How do you vet the protocols that become your partners?

It is worth separating two things. We are a traditional asset manager investing in traditional, off-chain assets. We wrap those assets in a fund structure and tokenize the fund for delivery. The critical partnership, where we do deep diligence on every issue at play, is the one that brings us into the on-chain ecosystem: in this case, Centrifuge.

From there, the tokenized funds get placed all over the place, and that is a different dynamic. A large part of what we have done has been with the largest on-chain native stablecoin issuers: Sky with USDS, Ethena with USDe. They are very important partners and clients, and we are very comfortable with them and bullish on their prospects. But the technology isn't flowing the other way; we are delivering our product for them to use. So the level of diligence is different. Behind them there is a long tail of participants using the funds for various purposes, and our focus there is that they integrate appropriately with Centrifuge and use the funds appropriately.

Who is actually buying these tokenized funds today?

About 200 institutions have onboarded, though not all of them are active at any given point in time. It is a direct subscription process; you can't just buy the funds anywhere.

The key use case today is balance sheet and collateral management, and that is most critical for what I would call blockchain-native stablecoins. Tether and Circle are the dominant players, but, somewhat counterintuitively, I don't view them as blockchain-native, because all of the assets backing them sit off-chain. Sky and Ethena have both pioneered models that rely end to end on what happens on-chain. For them, it is critical to access other sources of yield in a digitally native format rather than moving money off-chain into a money market fund.

Beyond those two, I would put demand into two categories. First, balance sheet management: a lot of the large players in this ecosystem are really complex fintechs with large balance sheets that they manage on-chain, and helping them generate additional yield matters. Second, protocols building vaults, looping strategies and similar business models, where these assets enhance the product they deliver to their end clients.

How do you measure success?

Two metrics. The number of protocols we work with, because we want to expand the places where these solutions deliver value. And total assets. At the end of the day, we are an asset manager. We would like to manage more assets for more people.

"Guy said: what more can we do together?"

You recently announced a partnership with Ethena. What does it involve?

Like Sky, we see Ethena as one of the leading businesses in this space, delivering stablecoins whose utility and yield come from very unique on-chain sources. With Sky, we helped them roll out their first move to bring real-world assets on-chain, with a lot of partnership alongside Centrifuge. When Ethena was looking to bring real-world assets into its reserves for the first time, we had very similar conversations with Guy [Young, Ethena's founder]. To his credit, he asked: what more can we do together?

The result has several layers. Ethena uses tokenized JAAA as a key part of the reserves backing USDe. We invested in the protocol through a small venture fund we run, focused on very early-stage companies we see as critical to bringing traditional finance on-chain, and we talk with them regularly about how they are scaling. And there is a joint product development partnership: thinking about other assets we can bring on-chain for on-chain investors, and about how on-chain products such as sUSDe can be used within traditional finance, for example to manage cash reserves.

To be clear, it is complex for a traditional asset manager to integrate on-chain assets into traditional products. That is a long tail. The pace is increasing, but today it remains relatively hard.

"When clients pull billions in a day, they are proving the case."

USDe supply has contracted sharply during past market stress. Isn't that a risk for you?

I see that more as a feature than a bug. If you have a short-term liquidity and yield vehicle, users should be able to move in and out at scale, easily. The fact that this has been stress-tested is valuable.

I also wouldn't call USDe an algorithmic stablecoin. That label implies the fatal flaws of certain earlier designs. USDe is on-chain native and generates its yield through on-chain activity, but it is dollars backed by yield-generating positions, whether that is exposure to AAA CLOs or a carry trade. The real question is whether the strategy can handle large inflows and outflows, which comes down to how liquid the backing assets are.

We see the same thing in our own funds. JAAA is one of our largest strategies both on-chain and off-chain. The US-listed JAAA ETF has more than $32 billion in assets and is the largest AAA CLO ETF globally. We tell clients they should be able to move literally billions of dollars in a day. We regularly see block trades of $300 to $500 million executed in AAA CLOs. Obviously we prefer inflows, but outflows demonstrate the utility of the product. Ethena and Sky have both been battle-tested repeatedly. The inner workings are complex, but the net result for investors is simple: a stablecoin that can deliver meaningful yield and may have structural advantages over USDC and USDT.

Does that make the model superior to Circle's?

No. They are differentiated business models. Circle has built an incredible business; it is just a different one. Its reserves are managed off-chain, almost entirely in a single institutional money market fund managed by BlackRock. The threshold for us to invest in a new layer 1 is very high, because launching and scaling one is hard today. The only one we have invested in is Arc, Circle's new layer 1. USDC has become the default currency for transacting in DeFi, not necessarily for generating yield, and Circle is one of the most buttoned-up companies in crypto on regulatory compliance and institutional infrastructure. I think Arc has a fighting chance of being adopted by large institutions.

There is room for more than one. Tether has its place, USDC has its place, and Sky and Ethena will both keep scaling significantly. As long as they serve a differentiated purpose, we won't end up in a world of one or two stablecoins.

"No AAA CLO has ever defaulted. It is one of the highest Sharpe ratio asset classes in the world."

What separates JTRSY from JAAA?

JTRSY is a very traditional short-term cash product. It is literally just exposure to US T-bills, the single most conservative version of the liquidity vehicles we run for institutional clients. Insurers and banks, for capital and other reasons, keep significant assets in rolling T-bills. On-chain, stablecoin issuers and other businesses managing balance sheets need the same anchor at the conservative, lower-yield end of the spectrum. US government debt is no longer AAA-rated, but Treasuries remain the most liquid and widely held short-term asset in the world. It is a product you don't need to think about; it simply accrues yield.

JAAA is the next step out. CLOs have existed for 30 years, used mostly by banks and insurers. We pioneered JAAA as a US ETF to extend access to financial intermediaries, advisers and individual investors. It has grown to more than $32 billion, making it the largest AAA CLO ETF globally, and we have since extended the franchise into Europe. AAA CLOs combine three features. Credit: no AAA CLO has ever defaulted, including through the financial crisis. Duration: they are floating rate, so you take no meaningful interest rate risk. And yield: roughly 100 to 200 basis points over SOFR, probably averaging around 110 to 120, without taking credit or duration risk.

The risk is short-term price movement driven by supply and demand. AAA CLOs trade in a liquid secondary market, and in a liquidity crunch large institutions often sell them first because they are the most liquid part of their portfolios. Historically, the largest drawdown has been around 10%, typically recovered within weeks, and on most days you are talking about maybe a few basis points or less. So if you need certainty that you can liquidate today at a stable price, Treasuries are the base layer. AAA CLOs are what many add on next, for a large jump in yield with modest mark-to-market impact.

Those historical drawdowns also predate JAAA. We have brought more than $32 billion of new investors into the asset class, and we have not seen a drawdown like that since. Our thesis, not yet proven in a stressed environment, is that a broader buyer base dampens that risk: when a bank or insurer needs to sell, there is now a large pool of investors who would love nothing more than to buy AAA CLOs at a 5% discount.

"Tokenized JAAA is not a tracking vehicle. It is direct ownership of the fund."

Who can actually hold the token? Is there whitelisting, a minimum, a difference between European and US banks?

The key point is that we did not take an existing fund and bolt on a tokenized share class. Tokenized JAAA with Centrifuge is a separate, fully native fund. The only way to subscribe and redeem is in USDC, it is smart contract-driven, and it is integrated into the ecosystem. On the back end it is still a traditional fund: you own a proportional share of the assets, not a tracking instrument.

The fund is incorporated in the BVI and open to non-US professional investors, which includes European banks. Every institution onboards through KYC and AML, as for any fund subscription. Once onboarded, they hold their interest 100% natively on-chain, and wallets are tied to whitelisted investors, so the token cannot be passed to any wallet. Most of the holders are entities like Sky and Ethena, which are themselves eligible investors. If you buy USDS or USDe, part of the yield you receive comes from assets held in JAAA, but you are not invested in JAAA directly.

Separately, there are freely transferable versions available across a range of platforms, the deRWA tokens. They are one-to-one backed, issued by a separate entity and unaffiliated with the fund itself. Think of them like an ADR, or a GDR trading in London that represents an interest in a US-listed company.

"We own the asset management. Centrifuge handles almost everything after that."

Where does Centrifuge sit in the stack, and what do you keep in-house?

Today, we own the asset management. Tokenized JAAA is run by the same team that manages our flagship CLO products; JTRSY by our cash liquidity teams. Almost everything after that, integration, onboarding and offboarding, is handled by Centrifuge.

Choosing the technology partner was the critical decision. It would have been relatively easy to spin up our own tokenization effort; there are probably hundreds of companies offering fund tokenization. But representing an asset on-chain is not the same as a usable tokenized product that integrates into DeFi. We needed a partner that clears our diligence on technology risk and partnerships, and that is trusted and deeply integrated by builders across the ecosystem: stablecoins, Aave, Morpho, layer 1s. Our assessment was that Centrifuge is quite far ahead on most of those metrics. It was one of the first tokenization platforms, integrated into DeFi from the early days, and not in a walled-off way. It has been battle-tested.

That model will evolve. We are getting closer to a world where demand to consume tokenized products extends into traditional fund structures such as UCITS and '40 Act funds. In that world, tokenization looks a lot like a transfer agency function. We have a high degree of trust in Centrifuge as that transfer agency partner, while many functions that still operate off-chain would be filled by our traditional model and service providers. That is probably the more scalable long-term solution for bringing traditional clients on-chain.

So, for now, partnering beats building everything in-house?

In general, yes, with nuances. Our approach is to partner wherever we don't have a relative competitive advantage. We could have built our own digital transfer agent, and some competitors have. But I think they have constrained themselves, because what they built amounts to a fairly closed ecosystem: to do business with them, you have to come inside it.

Our vision is to let our asset management be consumed in any ecosystem. There are regulatory hurdles to that, and we are part of the Coalition for Tokenized Markets (CTM), which is pushing policymakers and regulators to free up some of those restrictions. But this is the core of what blockchain is meant to do: decentralize and reduce intermediaries. Using it to build a walled-in system is antithetical to the promise of the technology. So we own what is core IP for us, differentiated asset management, and use the best partners we can find to expand into the broader ecosystem.

"Eventually we won't talk about DeFi and TradFi. We'll just talk about finance."

What is the one thing European asset managers should start, or stop, doing now?

Start figuring out how to build blockchain infrastructure internally. The end goal is a place where we no longer talk about decentralized and traditional finance; it is just finance, running on blockchain rails. That is a complex problem spanning regulation and technology, and there is no single answer. But clients will only move if they get something they can't get elsewhere, and creating that benefit means bringing supply and demand, asset managers and investors, together.

The part we can own as asset managers is integrating blockchain into our own operations so that clients see the benefit. That is hard, which is why many build separate things instead. We haven't deeply integrated on our side yet either. It is a goal we are seriously working on, and we are investing in companies we think are key to that transition. Most asset managers don't even have broad-based infrastructure to buy assets on-chain if they wanted to. But whether we push to use the technology ourselves is the part we control. That is the crux of it.