I am just coming out of a week of Zoom meetings at Consensus 2020 aka Consensus Distributed and it really helped me clear up my ideas about tokenisation and fractionalisation. So first of all, shoutout to everyone who took the time to network with me – you were awesome.

Some terminology

To get some definitions fixed up-front, here some terminology before we dive into the core part of this discussion.

Tokenisation. In this context tokenisation simply refers to financial instruments (we are only looking at financial instruments here) whose ownership is linked to the “ownership” of a token that lives on some blockchain. Ownership of a token of course means that either you control the private key associated with the address that owns the token, or someone else (aka a custodial wallet in crypto speak) controls this private key on your behalf.

Fractionalisation. Fractionalisation is a bit of a fuzzy term in this respect – it refers to the fractional ownership of certain assets, but of course almost all financial instruments refer to a situation where an asset is fractionally and jointly owned by more than one person. So in this context here I want to define “fractionalisation” in the sense that the pieces are relatively small, so that it is possible to own say less than $100 worth of a particular asset. Fractionalisation does not nessecarily need to happen on a blockchain – eg Robinhood’s fractional shares operate in a traditional centralised context.

Asset backed tokens. When I talk about asset backed tokens then I mean tokens that are backed by relatively simple, cash flow producing assets. Again, the definition is somewhat fuzzy because technically a company is also a cash flow producing asset, but in this context I refer to things like real estate, vessels, income-producing infrastructure such as solar power plants, and also more abstract items like future royalty payments say of a song or a movie. I would however exclude royalty payments relating to works-not-yet-produced, because in my view they are, like companies, on the wrong side of “simple” in this context.

Liquidity. A term that is often thrown around in the tokenisation context is “liquidty” and it can mean either of two things. The first one is what I like to refer to as settlement liquidity, which indicates how easy it is to actually execute a transaction once you have agreed to it with another party. The second one is what I’d call trading liquidity or discovery liquidity and this refers to how easy it is to find another party that wants to take the opposite side of the transaction one wants to do.

Tokenisation vs issuance of traditional instruments

The big question that everyone operating in the tokenisation space is trying to understand is what advantages tokenisation offers vis-a-vis a more traditional way of issuing and registering financial instruments. It should be noted though that the “more traditional way” also includes modern variations thereof – for example it is not only important to understand the advantages of issuing an investment in a tokenised format vis-a-vis registering it with Clearstream and trading it on the London Stock Exchange. One should also compare it to more nimble and purpose built platforms like Crowdcube or Seedlegals.

Also, not all tokenisations are equal. In the limit, a token that is registered on a private chain might not look all too different from a financial instrument that is registered on one of those centralised, purpose-built platforms I was just referring to. And with the plethora of not interoperable chains we currently have we might run into similar problems that we run into the traditional world: if we want to exchange two assets that live on different chains then this is not necessarily much easier from a security and user experience point of view than operating on multiple centralised systems.

The key differences between traditional instruments and tokenised instruments are the following:

Custody. The right to transfer – and therefore the ownership – of a token is tied to the possession of one or multiple private keys in a manner immutably enforced by the system. This is very different from a centralised system where that system always can make the transfer, but it internally choses to only perform it if and only if it has authenticated the user with the proper credentials.

In a traditional system there must be a chain of trust between the system that authenticates the user and all subsequent system that handle the user’s instructions. On a blockchain that is different: every system involved can independently check the credentials associated with a transaction (or any other user action) so there is no need for such chain of trust.

Where this becomes important is for example for cross-border transactions: if an asset is registered in country X and the investor is located in country Y then either the investor needs to open a custody account in country X, or there must be a trusted custody link between his custodian in country Y and the depositary of the investment in country X. The complexity of such an interlinked system becomes mind-boggling the more the number of countries increases, which typically limits cross border investments, or at least makes some uneconomical for smaller transaction sizes.

Operations. Traditional financial systems are put together out of a few, loosely coupled systems for the different functions that need to be performed. The depositary and custody functions we already mentioned above, and the issues related to linking them. Payments go through different rails, typically through the regular banking system, and again cross border payments can be a major source of complexity, especially when correspondance banking is involved. Last but not least there is a need for communication, both broadcast communication from the issuer to the investors, and bilateral authenticated communication between issuer and investor, for example for voting at the annual general meeting.

It takes a significant effort to tie all those diffent systems together, and ultimately this is what a lot of the traditional financial system spends its time doing. For example, when a trade has been agreed on an exchange, then this information is submitted to the settlement system. The two parties send the security and the money to the settlement system, using custody and payment rails. The settlement system checks that it received what it meant to receive, and then sends the cash and security on its way. What gets really complicated – and expensive – is exception handling, say if the securities or moneys don’t arrive, or in wrong quantities, or if securities or moneys arrive that can’t be matched to a specific trade.

In a blockchain based system all this operational complexity goes away because everything – including payments and voting at an AGM – can be operated on-chain and exceptions simply can not occur. This brings with it a major decrease in transaction and servicing costs, especially cross border where custody and payment chains get long, slow and complex. This allows for issuance – in particular global issuance – of a minimum size that would not previously have been possible.


Regulations are a thorny topic with respect to tokenisation, for a number of reasons, some of them good, some of them bad. To start with the good ones, regulations tend to be there for a reason, for example to protect investors, to ensure that the market structure remains sound even in times of crisis, and last but not least to avoid illicit activity being enabled by the financial system. For example in the ICO craze the investor protection angle fell somewhat short, and many investors got scammed. A crackdown here is arguably a good thing. Likewise, the permissionless and distributed nature of blockchain and DLT makes it often harder to control payments and to enforce KYC, AML and associated regulations. Again, this is something that needs to be addressed. On the other hand we have seen massive breaches in this respect in the traditional financial system, so maybe overzealus regulation is not warranted here.

However, there is regulation that is fundamentally misguided and bad in this situation, typically because it is not technology agnostic and it assumes that the system works in a specific way, and blockchain and DLT-based system do operate differenly. One prime example here is CSDR, the European Union’s Central Securities Depositary Regulation, which is very hard to bring in accordance with how blockchain and DLT based depositary systems work, and which severly limits the possibility of some instruments registered on-chain.


In my view in order for tokenisation and fractionalisation to work, there is a hierarchy of problems to solved. From the most to the least complex areas those are

  1. Investors
  2. Regulations
  3. Technology

We have already dealt with regulations and technology above, so now let’s move on to the most complex area of all, investors.

Investors come in two major classes: professional investors and retail investors. As the name implies, the former invest in a professional capacity, ie it is their job doing it, typically for someone else, and the latter invest their own wealth. Those categories are not homogenous of course: Bill Gates and Jeff Bezos are retail investors (albeit not their respective family offices) as much as Johnny Retail who saves a few bucks a week from his paycheck. Similarly professional investors come in all shapes and forms, from a small family office where someone spends a few days a month looking after the money, over some billionaire’s family office, to the largest investment managers and insurance companies in the world.

A lot of those professional investors are pooled investment vehicles (eg investment funds) that aggregate the investments of a large number of retail investors. Important reasons for this aggregation are:

  1. Many investments opportunities are not sufficiently granular (or “fractionalised”) for small retail investors to include them into their portfolio.

  2. The operational costs associated with managing very small investments (eg paying dividends, or inviting everyone to the AGM) are too high, which often is the reason for (1) because of how those costs scale with the minimum investment size.

  3. The cognitive load on small investors is too high – the amount of analysis and due diligence that need to be performed does not really scale down with the size of the investment, and retail investor do neither have the time nor the mental capacity to deal with this for too large a number of investments.

  4. Liquidity – in the sense of trading liquidity or discovery liquidity – gets the worse the smaller the investment is. It comes as a surprise for many people not involved in the markets how bad liquidity generally is outside the major equities and the on-the-run government bonds. A token representing a fractional interest in some shopping mall in a small town in Idahoe has literally no chance of any discovery liquidity in the current system.

Tokenisation – and tokenisation technology in particular – can address the first two issues above, in that it allows to create tokens that represent a very small interest in a single asset – the aforementioned shopping mall in Idahoe for example – in a cost-efficient manner. However, tokenisation in its current form does not address neither the points (3) nor (4), and we have to do some more work before fractionalised retail distribution becomes feasible. In the meantime, if we want to distribute tokens in meaningful amounts then professional investors must be a very important part of the picture.

Vision for a future market structure

To summarise the conclusions of the previous section, in order to make fractionalised investments feasible for retail investors in a large scale – and thereby disrupt the market of pooled investment vehicles who’ll see much smaller inflows once investors start investing directly – we need to address two related issues. Firstly we must reduce the cognitive load associated with every single investment a retail investor makes, and secondly we must deal with the (lack of) liquidity that is associated with the fractionalisation of investments. I do not know of course what form this future market structure will take, but my vision is that it will involve three distinct players

  • anchor investors
  • robo advisors
  • robo market makers

I will go through those three kinds of players in turn.

Anchor investors. An anchor investor is a big investor anchoring the investment, typically holding between 20-80% of it. The presence of this anchor investor – who typically is a reputable professional investor – reassures other investors that this is a good investment and that the price paid is fair. This in turn means that those other investors do not have to do much due diligence themselves. They still should ask themselves whether they like shopping malls generally as investments. However, they can be reasonably reassured that, as far as shopping malls go, this is a reaonable investment. So the presence of anchor investors deals with the cognitive load problem because it moves the level of analysis required from detailed to strategic.

Anchor investors would also typically underwrite a deal, ie they would ensure that the deal can be placed even in the presence of unpredictable interest from retail investors: if retail demand is high they invest a bit less, and vice versa. They also sometimes make markets in their investments, so if some retail investors want to sell their position – say because they need a new car, or have medical bills to meet – the anchor investor might buy them out, and if other retail investors want to get in they might be able to get some tokens from the anchor investor.

Anchor investors of course don’t do all of this out of the goodness of their hearts, but simply because they can make money out of it: they’ll get an anchor investor fee for performing the analysis, an underwriting fee for underwriting, and they’ll be able to earn the spread whenever a retail investor wants to get into or out of an investment.

Robo advisors. Anchor investors are a well known concept in capital markets. Robo advisors – automated algorithmic investment advisors – on the other hand are a very recent development. They first and foremost address the issue of the cognitive load the investors face. First they need to “understand” the goals, circumstances and possibly investment views of the investors. They then search the market for potential deals, and ranks those both in terms value (possibly by simply checking for an “approved” anchor investor) and fit into the investor’s portfolio. They then either may make the investment decisions autonomously, or they may leave the final choice to their investor, like a Tinder, but for investments where the investor swipes left or right on the opportunities presented.

Robo advisors not only address the issue of cognitive load, but they also somewhat address the issue of liquidity, to the extent that it is associated with discovery: if some robo advisor knows that their investor wants to buy some shopping malls and there is a shopping mall coming to the market, the robo advisor will have a look at it and, if it corresponds to the overall investment profile, will either purchase it or propose to its investor to purchase it. One issue in this context however is valuation, especially to the extent that it relates to asymmetric information, as it is a priori not clear whether the seller wants to sell because they want to buy a new car, or because they’ve just learned that the anchor tenant is about to go bust.

Robo market makers. The final piece in the puzzle is robo market makers, ie trading bots that are willing to show a bid an any token offered to them, and offer their inventory to the market. In order to fulfill their function properly they need to carry a substantial inventory, which exposes them to financing cost as well as price risk.

Financing is easy to arrange as those positions are positive carry, ie the investment yield on the tokens is higher than the cost of financing them. Moreover it is pretty straight forward to arrange asset-backed financing, ie the lender has full recourse against the tokens they financed in case the robot market maker goes bust. Valuation is a bit of an issue in this respect, but at least with tokens there is a real time view on the volume of each and any token held in the portfolio of the market maker.

The token price risk also seems manageable. Provided the portfolio is large enough it will behave more or less like a related index, so the dealer can put on macro hedges to stabilise the overall value of the portfolio – and any residual basis volatility means that the dealer will have to charge a bigger bit offer spread. Alternatively the dealer can sell exposure to its portfolio on a wholesale basis, or they can construct their own synthetic index. This index can then be sold either to (the robos of) retail investors who want to create a more diversified portfolio in one go, or to institutional investors who want to take wholesale risk.

However, like for secondaries with robo advisors, asymmetric information is an issue here. If people with local knowledge are trying to capitalise on their superior information in the market then a priori market makers without local knowledge are at a disadvantage. It seems to me however that ultimately this is a problem that can be solved with (a) strict position limits, (b) AI monitoring of flows, and (c) appropriate documentation with clawback provisions.

What is the fun in that?

A key value proposition of the individualist, robo-advisor assisted investment strategy vis-a-vis traditional pooled investment vehicles is the fun that can be had. It is much more satisfactory to be in control of ones own destiny, and choosing ones investments oneself with assistance of a robo advisor is part of this. A particularly good investment always creates bragging rights, whilst a bad investment can be easily swept under the carpet, so net/net the satisfaction level goes up.

It is also a very interesting way for projects and companies to raise capital. For one thing, they can offer perks. For example if the investment is in a single feature film, then investors can be invited to the premiere, or a pre-screening, or they can get some real world or digital memorabilia. As other examples, investors in shopping centers might get discount on the parking, or be invited to special events, and investors in small (or large) companies might be offered discounts on products or be the first in line to buy it when a new exciting product is sold.

This in turn creates a much more interesting investment environment. Cool companies or projects can reduce their financing cost. Similarly, selling tokens to end investors helps customer acquisition – Monzo (the UK challenger bank) is a good example here. They offered “investor” cards to their early investors who all became Monzo customers right away, just to get their hands on those cards, and on the associated bragging rights.

This works on a national scale, but it probably works even better on a local scale. For example the local bakery might finance their new oven via token issuance that is mostly taken up by their customers who in turn might be more inclined to buy their bread there instead of elsewhere, especially if they get the royal treatment just because they are token holders.


In conclusion I find the potential how tokenisation and fractionalisation can and will disrupt traditional market structure fascinating, with the biggest impact in my view on (a) pooled investment vehicles aka investment funds that will lose volume to investors who run their own portfolio with the help of robo advisors, and (b) local or tribal projects or companies where the investment token becomes part of their product offering and helps with customer acquisition and retention.

Liquidity will be an issue, and on top of robo advisors there is a need for market makers who inject a minimum amount of liquidity into the market so that people who do need money for specific purposes – say for buying a car – can get it by liquidating their investments without being ripped off.

None of this however can happen without the presence of anchor investors, ie professional investors who underwrite the offering and also take up a big part of it for themselves, and who are ultimately responsible for fairly pricing the deal. At the beginning, anchor investors will have to take up a very large proportion of the deals offered so that, even with limited retail appetite, there is sufficient diversity in the market for it to become worthwhile developing robo advisors and robo market makers.

The problem that we as an industry that wants to develop tokenisation and fractionalisation are facing is that to disrupt this market we need the cooperation of the very participants that will be disrupted. This in my view is one of the reasons why the tokenisation has been off to a slow start. However, it will happen eventually – the tide will turn, and traditional investors will either go with the flow or will be swept away by it.

image credit: Sharon McCutcheon on Pexels

    podcast link