Token Sales: Data & Returns

The exponential growth in Token Sales (ICO’s) has been staggering. Does it make sense for the average investor in digital currencies to pick individual tokens or invest in a portfolio of tokens rather than in Bitcoin or Ethereum? Early data says no.

If you’re interested in digital currencies, blockchain technology or venture finance, you have probably heard about Token Sales and/or ICO’s (Initial Coin Offerings). There are many excellent articles about the underlying business model, use cases of tokens and rise of the token sale as primary fundraising method. Every day sees a new Token Sale announcement and even mainstream business outlets have published articles about the amount of capital raised through this new method (Economist, Forbes, Fortune, HBR, FT).

We should look at the little data that is available to see what’s been happening and the financial returns for digital currency investors. My conclusion is that tokens are risky investments that, on average, do not provide a greater return than core digital currencies such as Ethereum.

Fertile Days, Limited Data

Once you go down the rabbit-hole of investigating the space of tokens, you quickly discover 3 things:

  1. There is a lack of terminology: Coins, App Coins, Tokens, Token Sales, Token Offerings, ICO’s, etc.
  2. Information is scattered: You have to check twitter, reddit, token rating websites, and digital currency discussion forums to stay up to date on the latest sales and final results.
  3. Need data? Good luck!: High quality and consistent data on token sales is extremely rare.

On the one hand, these struggles are a positive. As mentioned by Chris Dixon and Olaf Carlson-Wee in a great a16z podcast: when no one knows what to call something it means that we’re still in the “fertile, creative days” of innovation. On the other hand, it makes it really hard to comprehend this space and/or make investment decisions…

All data used in this analysis can be found in this Google sheet.

Close to $400M has been raised

The dataset that I’ve put together comprises 117 Token Sales (ICO’s) that have been announced since July 2014. Of these 117 Token Sales, 56 projects successfully raised a total amount of $378M of capital.

Figure 1: Total USD Raised Incl. The DAO

Given the outlier nature of The DAO token sale and its failure, a more correct statement is that $228M of capital has been raised through Token Sales. A staggering $134M has been raised in 2017 year-to-date and $101M in April 2017 alone. At the time of writing 18 Token Sales are still open for fundraising and this number increases every day.

Figure 2: Total USD Raised Excl. The DAO

Market Capitalization of Tokens > $1.2B

While it’s interesting to look at Token Sales in terms of in dollars raised, it’s equally important to calculate the value of the tokens after they have been issued.

Of the 56 successful Token Sales in the dataset, 34 tokens have started trading and are listed on Coinmarketcap. The market capitalization (the USD value of available supply of tokens) for these 34 tokens currently stands at $1.16B.

Figure 3: Market Capitalization of “Listed” Tokens

The 33 projects that these tokens represent raised $150M in their respective token sales. A back of the envelope calculation shows a 8x increase in dollar value, and our impulsive reaction might be to rush into every single token sale we can get into.

Tokens ARE Risky

The theoretical case for Token Sales is that it they are the optimal way to launch decentralized business models, ranging from core blockchain protocols to Decentralized Applications (dApps) on top of existing blockchains. Ethereum itself started as a Token Sale raising $18M worth of Bitcoin. In these decentralized business models tokens should be the core value unit that represent some combination of usage, work and/or ownership. As Nick Tomaino explained, we should think of tokens primarily as a product feature and not as a fundraising feature.

The majority of projects behind the tokens have not yet launched. They’re in various stages, ranging from beta launch to being stuck in endless ideation. Simultaneously, many projects are taking advantage or “exploit” the exponential growth and interest from unsophisticated investors interested in digital currencies. This ranges from projects that have no obvious network effects at all or are outright scams (Google “ico scam” for starters).

This indicates that Token Sales are a very young and very risky market. It will take time and more data to get a sense of how risky tokens are and what factors determine winners and outperformance. Sophisticated investors and VC’s in the cryptocurrency space look at the development teams, network potential and core technology as factors that could mitigate some of the risks. The average investor and many digital currency enthusiast will have very few of these skills. Nonetheless, there is some data that we can use to conduct a straightforward analysis and answer the following question:

Have Token Sales returned more money than Bitcoin and/or Ether?

Tokens Do Not Outperform Ethereum

Given the limited amount of data needed for financial return analysis, I was able to calculate the returns for 28 tokens since their sale and compared them with returns on Bitcoin and Ether during the same timeframe. The 28 tokens and the projects they represent were the only ones on which I could collect data from multiple sources to calculate the price at issuance, price now and market capitalization.

Figure 4: Absolute Returns and Hypothetical Portfolio Returns

If you were to invest $1 in each of the 28 Token Sales, held onto the tokens and sold them at the time of writing you would have made roughly 9x of the notional investment. This compares with “only” 2x if you had invested $1 in Bitcoin tokens at exactly the same times as the Token Sales. More importantly, a portfolio of Ethereum tokens would have yielded 11x, beating the portfolio of tokens.

Because of the limited data, outliers skew the results drastically. For example, Augur’s token sale took place in October 2015, when Ether was trading at $0.70. Since then, Augur REP-tokens have increased by 29x vs 134x for Ether. Taking Augur out, the results swing in favor of tokens, as per the figure below. When I remove the 2 largest return outliers for Tokens and Ether (i.e. remove 4 in total) the Ether and Token portfolios return roughly the same amount.

Figure 5: Multiples for various portfolio combinations

It’s not the goal of this analysis to determine which tokens are the best investments. We can slice and dice the dataset in many ways that tip the returns in favor of either Tokens, Ether or even Bitcoin. We must not forget that many of the Token Sales are an application layer on top of the Ethereum blockchain. By investing in tokens you are essentially exposing yourself to some (undetermined) amount of Ethereum exposure, and taking on additional risk. One could argue that the performance swings in my hypothetical portfolios are merely reflections of this additional risk. Until there is more time series data the jury is still out on whether this is the case.

Figure 6: Market Cap of Tokens vs Market Cap of Ethereum


The conclusion that I am confident to make is that the data shows that most digital currency enthusiasts are better off staying away from Token Sales. Unless you’re planning to actually use the tokens for the services that the underlying projects will offer, tokens as a speculative investment come with great risks.

Investors who are unable to do a full due diligence (development team, technical specifications, network model) on a new Token Sale are probably better off investing in core cryptocurrencies such as Ether. This will yield better results than rushing into every single Token Sale or even token portfolios. It’s common sense and is backed by some data now.

— — — — — — — — — — — — — — — — — — — — — — — — — — — — —


Any questions and or requests for the dataset can be sent to me directly ( Thanks to Mark Murphy, Valantis Vais, Paul Veradittakit and Steven McKie for commenting on earlier drafts of this article.

All data used in the analysis is publicly accessible on the websites or via the public API’s of Coinbase, Coinmarketcap, Smith & Crown, ICOrating & Tokenmarket

Most of the terminology that I’ve used were first coined by the following persons on a variety of websites and podcasts:
Chris Dixon & Olaf Carlson-Wee:
a16z podcast on tokens and polychain capital
Olaf Carlson-Wee & Ryan Zurrer:
Ether review podcast
Fred Ehrsam:
one, two, three, four
Fred Wilson:
Decentralized Startup
Naval Ravikant:
The Fifth Protocol