By Railnet Team —
Issuing a real fund on a public blockchain was long considered taboo in traditional finance. In this episode of Railnet Talks, Laszlo Szabo sits down with Theo Golden, Head of Digital Assets at Baillie Gifford, the $260B+ AUM asset manager that decided to do it anyway. They talk about the conviction it took to make that call, why blockchain is more than better record keeping, and what it will take for the rest of the industry to follow.
Discover the third episode of Railnet Talks featuring Theo Golden, Head of Digital Assets at Baillie Gifford
"The whole purpose of tokenization is that these assets can move. That only works if those transactions are legally true."
Baillie Gifford recently announced the Baillie Gifford Enhanced Yield Fund – ticker $BAGEY – the first fully natively issued UK fund, with Ethereum and Solana as the legal register for ownership. It's a notable step for an institution in a category where most managers have stayed on the sidelines of public blockchains.
Baillie Gifford is an independent, privately owned partnership with nearly 120 years of history and a track record of high-conviction, long-horizon investing. Its first move in crypto was a recommendation to buy Ethereum, made years before most institutions treated the category seriously. Its structure as an unlimited liability partnership shapes how it approaches this work: because the partners carry personal financial liability, a regulatory-first posture is built into the culture rather than adopted for messaging.
In this episode of Railnet Talks, host Laszlo Szabo sits down with Theo Golden, Head of Digital Assets at Baillie Gifford and the architect of $BAGEY, joined by Darshan Vaidya, Head of Asset Management at Kiln. They discuss the difference between a wrapped fund and a natively issued one, what happens to legal finality if a wallet is hacked, and why Theo sees vaults as the next fund structure. Below are the highlights – for the full conversation, watch the episode on YouTube.
From opera to bonds to digital assets -------------------------------------
You have an unorthodox background – opera, then Bloomberg, now digital assets. What's the common thread?
Theo Golden: What really fascinated me about music was systems. I spent a long time doing musical analysis, building mental models of how operas and symphonies can be thought of. Music in many ways is just maths and systems – how A correlates to B and how you get to C. That exercise of mapping sounds weird to port across into finance, but it felt incredibly natural. I went to the merger market, then Bloomberg and Business Insider, covering FX and investments – really it was about mapping the industry and talking to the people making markets move. Then I realised I wanted to get into the trenches myself. I joined Baillie Gifford as a junior investor, did everything from catastrophe bonds to FX to corporate credit, and found my home on credit. I just love bonds.
His first investment recommendation at the firm was to buy Ethereum. He'd been buying Bitcoin personally since he was fifteen.
Theo Golden: Something changes when you go from managing your own money to having a fiduciary duty to others. What becomes an investment thesis becomes an investment thesis evaluation – a much more rigorous process. When I made that recommendation for Ethereum with my multi-asset colleagues, the case was much more thought-through than when I was fifteen buying Bitcoin.
You can't value Ethereum with a DCF, so we had to start from first principles and build something new. The rigour that comes from fiduciary duty takes you to a whole new level of depth, and with that depth often comes conviction – though you have to be careful not to talk yourself into a case.
The fund is not a wrapper -------------------------
The clearest technical thread of the episode was the difference between a wrapped fund and a natively issued one, and Darshan took Theo through it directly.
What does it actually mean for the blockchain to be the legal register, rather than an accounting ledger?
Theo Golden: Most tokenized funds are wrappers. You have an offshore SPV that buys units of a traditional fund, holds them, and per unit issues a token. That's worked to an extent – it's given parts of the market exposure to real-world assets. But the SPV isn't institutional grade. You're adding an intermediary, which adds cost, and often that intermediary isn't an investment-grade counterparty.
If you take a AAA money market fund and issue it out of an SPV in an offshore jurisdiction through a counterparty that isn't investment grade, you've turned a AAA asset into a triple-C asset. The Sharpe ratio of that doesn't look very compelling.
So the first step was to strip out the SPV. $BAGEY is a UK UCITS – an OEIC – and instead of issuing shares, it issues tokens, with the minimum investment brought down to $100. It's a dematerialisation into another investment form.
Theo Golden: But once you've done that, you're still just wrapping. You've made a token of an off-chain books-and-records system. You've made the asset 24/7 on the distribution side, but not underneath. The only solution was to make the blockchain the legal source of truth – for who owned what, when, and how. So we worked with the FCA and proposed that Ethereum and Solana be the legal source of truth for holders of those tokens.
"The whole purpose of tokenization is that these assets can move around. If Laszlo and I are both clients, I can send him a unit of $BAGEY and he can send it back – regardless of what day it is. That only works if those transactions are legally true."
Off-chain, that transfer wouldn't be legally true until the next day's reconciliation. Make the chain the legal register, and the transaction is legally final the second it happens.
What happens when you get hacked --------------------------------
Laszlo pressed on the obvious risk: if the onchain record is the source of truth and your wallet is drained, doesn't the record now say the thief is the owner?
Theo Golden: Just because the onchain book is the legal source of truth doesn't mean we can't help you in the case of a hack. $BAGEY is a whitelisted product – only people onboarded with Baillie Gifford can hold it, so if a hacked asset were moved to a wallet that isn't whitelisted, that wouldn't be possible. That's the operational mitigation. Then there's the issuer mitigation: it's a centrally issued asset, not Bitcoin.
If something goes wrong, you can pick up the phone. You have an FCA-regulated institution with nearly 120 years of experience to say, please help me, what do I do – and we have you on the register as the owner. And because it's a UK-regulated fund, in the near-zero scenario of that not helping, you have legal recourse through the courts.
"There are three lines of defence there. Take a step back from being fully technologically focused and think about it as an asset which lives onchain – there are several mitigations to help."
The real prize is distribution and personalization --------------------------------------------------
The back-office efficiency story – one shared record instead of a custodian, transfer agent, and depository each keeping their own – is the part Theo finds easiest to underwrite. The bigger opportunity is on the other side of the coin.
Beyond back-office efficiency, you've argued the real prize is a shift in distribution and product design. What does that look like concretely?
Theo Golden: A shared register lets us work with other players on a more scalable platform. Rather than onboarding many different distribution networks, we can offer them a direct relationship with us. There's a good stat: a massive proportion of people under 25 in the UK only own crypto and cash. Managers like us have a bit of a fiduciary duty to provide products in the place where they're actually managing their financial lives, rather than forcing them through existing distribution channels. Tokenization is a neat way of doing that, because we're on the rails they're already using.
From there he went past distribution, to personalization.
Theo Golden: Today most people get a bucketed approach. You're a six out of seven on risk, you're young, you can bear a drawdown, so you get allocated this style of portfolio. That works because it's scalable, but it's not personalized. The future I see for tokenization is delivering hyper-personalized portfolios in-wallet – directly optimising for your preferences and your risk budget. In a world of AI, the monitoring process could be automated too: watching how someone saves month to month, and if they have a large capital call coming up, readjusting their risk budget for it.
He was measured on AI more broadly, pointing listeners to his colleague Tom Slater's note that AI "isn't coming for your job, it's coming for your mind."
Theo Golden: Human judgment stays incredibly valuable, particularly where Baillie Gifford lives – we're not tracking an index, we're trying to add value by asking the "what if" question everyone else assumes is impossible. AI is very complicated pattern recognition. Large parts of the process – data extraction, accumulation, rebalancing strategies that are rules-based – are ripe for it.
But agents aren't free. Just because something can be automated doesn't mean it should be, or that it'll be cheaper. We shouldn't forget the value of human capital.
Vaults, and a bold prediction -----------------------------
What's your boldest prediction for tokenized asset management?
Theo Golden: What ETFs did to mutual funds, vaults will do to ETFs. An ETF is a very elegant way of creating secondary liquidity – it's built to be traded. But a vault is built to do that too, in a way where it feels like an open-ended fund but can be traded like an ETF. It's the best of both worlds.
If you took an ETF and a mutual fund and made them have a baby, and it got all the best genes – that's roughly what a vault is going to look like. If a vault comes to be seen as a regulated collective investment scheme, there's scope for that to become transformational. In the next five years, maybe even three, I think that happens. The dependency is the underlying assets being high quality enough to be investable – which is why native issuance matters.
On an Aladdin-style operational layer for the onchain world, Theo was direct about the sequencing.
Theo Golden: Aladdin simplified and standardised a very process-driven part of pre- and post-trade. We need something like that across tokenized assets, tokenized securities, and crypto assets. But what needs to be defined first is the process – and that hasn't been clarified yet. Crypto hedge funds, vault curators – they all do very different things. We haven't even standardised what tokenized assets look like, native versus wrapped. That'll come as more fiduciary capital comes onchain, because managing your own assets is one thing, but justifying it to clients in a regulated capacity is the pressure that forces you to embed a process that's both scalable and defendable. Then a solution can be built for it.
Who's next ----------
We close every episode by asking the guest who should be in the seat next. Theo pointed to non-dollar stablecoins as the unlock that moves stablecoins from a distribution play to capital-markets infrastructure – you need FX at scale for that, and it doesn't exist yet. His nomination: Andrew from Agant, who recently received good news in the UK and is working on issuing a sterling stablecoin.
This is the third episode of Railnet Talks, a podcast series spotlighting the asset managers, builders, and partners shaping onchain finance through the Railnet ecosystem.
Watch the full conversation with Theo Golden on YouTube.