The initial coin offering, or ICO, is catching on as a new way for blockchain-based startups to raise capital — without having to go through the tougher vetting required by traditional lenders or having the track record needed for IPOs. As a new type of crowdfunding, ICOs let companies raise money faster and with less red tape so they can get started on building their businesses more quickly. But it also comes with caveats, as poorly designed ICOs can fail to take off. So how can firms ensure they will be maximizing their ICOs’ chances of success?

New Wharton research shows that getting the initial coin offering design structure right is a key ingredient for success, according to the paper, “Inventory, Speculators, and Initial Coin Offerings,” by Wharton professors Gerry Tsoukalas and Serguei Netessine as well as doctoral candidate Jingxing (Rowena) Gan.

It is the first paper to look at initial coin offering design for companies developing physical products, under demand uncertainty and in the presence of strategic investors. However, they found that even the best design has flaws. “ICOs have the advantage of being a low-risk means of financing for firms,” they said, “but this comes at a cost of lower production quantity, lower profit and limited flexibility in terms of product margin.”

So why do an initial coin offering at all? Startups that cannot get access to traditional financing could still get capital through ICOs if enough people back their business idea. And investors — or more accurately, speculators — can come from anywhere and face fewer restrictions. Also, unlike crowdfunding, startups don’t have to go through a platform like Kickstarter where if they don’t meet their fundraising goal they get no money at all. With an initial coin offering, the company can keep any amount raised. ICOs are largely unregulated as well, so firms have to deal with fewer bureaucracies.

But it makes ICOs riskier investments because of their high failure rate. According to the researchers, nearly half of all ICOs in 2017 and 2018 “failed to raise any money at all” and 76% did not even meet their minimum funding goal. Moreover, only 44% of projects remained active on social media five months after the initial coin offering. As for scams, 271 out of 1,450 ICO cases were “susceptible to plagiarism or fraud,” the researchers said, citing an investigation by The Wall Street Journal into offerings aimed at an English-speaking audience from 2014 until May 2018. “The profit-seeking yet ill-informed investors can become easy prey and have claimed losses up to $273 million,” they wrote.

Nearly half of all ICOs in 2017 and 2018 “failed to raise any money at all” and 76% did not even meet their minimum funding goal.

“Fundraising through an ICO is not regulated, which opens possibilities for all kinds of inappropriate behavior by fundraisers. Indeed, recent data demonstrates that in a majority of cases the company ceases activity without delivering the product or service,” Tsoukalas said. “Our research shows that, despite lack of regulation, moral hazard issues in ICOs can be mitigated when the initial coin offering itself is properly designed. In particular, we show how firms can determine the optimal number and type of tokens to issue, how to price them, and how to manage product inventory, when facing uncertain future demand.”

Despite its design shortcomings, ICOs remain alluring. Since taking off in 2017, the initial coin offering market has topped $22 billion in 2018, challenging the traditional ways of raising capital, the authors said. In the second quarter of 2018 alone, ICOs raised $9 billion — equal to 56% of the U.S. IPO market, or 39% of U.S. venture capital. With ICOs gaining in popularity, the authors set out to design the ideal structure from an operations standpoint and answer the following questions: What type and how many tokens should be issued? How should they be priced? How do these choices affect inventory decisions and odds of success? And how are they different from other types of financing?

How ICOs Work

The typical initial coin offering process begins with a startup publishing a white paper explaining the business idea it wants to fund with the proceeds from the offering. It will provide the number and price of digital tokens it plans to sell, the sales period, the sales cap, and other salient details. The startup may or may not have a prototype product to show potential buyers, the authors said.

Then the startup issues “platform-specific [digital] tokens” for sale to speculators, the researchers said. These tokens can be exchanged for future products or services (in the case of utility tokens) or alternatively they give holders the right to share in future profits (in the case of equity tokens). Buyers pay for the tokens using fiat money — or more likely, cryptocurrencies — and they may hold the tokens or trade them like a stock in the secondary market. There are more than 1,000 different types of digital tokens in circulation, according to CNBC.

For example, Honeypod, which makes an internet hub that protects user privacy, said in its white paper that it planned to sell 40 million tokens out of the 200 million it created, for a price of 5 cents apiece to the public. It expects to use the funds to make 50,000 devices in 12 months. The offering closed in April. Another one is Sirin Labs, which raised $150 million from a 2017 ICO to build a cryptocurrency-friendly smartphone. Backers who bought Sirin tokens could exchange them for its products, or trade them. But while Sirin was able to make its smartphone, demand “fell well short of expectations,” the authors wrote.

In the Wharton research paper, the authors only considered ICOs that offered products, not services, for their tokens. They analyzed both utility and equity tokens and identified three ICO participants: the company, speculators and customers who buy the product after it is made. They also looked at three stages — the ICO fundraising, production and market phases. In the first phase, the firm sets the number of tokens, sales cap, and token price, and conducts the first round of token sales (ICO).

Next, the firm makes decisions about production despite not knowing consumer demand. Finally, the product launches and the company gets a sense of actual market demand. The firm sells its remaining tokens jointly with other token owners (speculators) to the customers through a secondary market. The customers then purchase the firm’s products using the tokens.

The startup must be able to price the product at more than double the cost to make it, or else the venture will fail.

Their goal was to find the optimal initial coin offering design, price, token cap and production quantity to maximize sales and profits. “Are ICOs a viable means for product-based firms to raise capital? What drives their failure or success?” asked Gan. “What is the theoretically optimal way to design ICOs in a largely unregulated market environment?” Their key finding: “Despite rampant moral hazard such as risk of cash diversion and failure of the company to manufacture, both product-based utility and equity ICOs can be successful under the right conditions.” However, a crucial component is the existence of a secondary market where holders can trade their tokens.

The Successful ICO

What does a successful initial coin offering design look like? For utility tokens, the startup must be able to price the product at more than double the cost to make it, or else the venture will fail, the authors said. So if the cost of making a widget is $1, then the retail price should be at least $2.01. If the startup cannot price it at $2.01 or higher because it doesn’t think people will pay that amount, or competitors are cheaper, then the ICO will not be successful.

The authors compared this finding to that of a firm launching a product it finances itself, without having an accurate sense of market demand. In this case, the ‘self-financed’ firm just has to make sure it prices the product above the cost for the business to succeed; it doesn’t have to charge more than double the production cost as with ICOs. As such, “ICOs may be best suited for products with relatively high willingness-to-pay,” they said.

There are other drawbacks to utility-token ICOs compared to self-financed firms: They tend to produce fewer products and generate less profit. Their results show that the ICO startup produces up to 40% less than the self-financed firm and earns up to 50% less in profits. However, they said, “these gaps shrink when the market is bigger, more stable” or consumers are willing to pay more for the product. But ICOs have a big perk: Even if there is low demand for a product, losses are primarily borne by the investors. For self-financed firms, if they overestimate demand, they will have to bear the risk of losses themselves.

ICO firms should reserve more than half of the tokens for the third, or market, stage.

Another finding is that ICO firms should reserve more than half of the tokens for the third, or market, stage. If it sold more than 50% of the tokens in the fundraising stage, the startup will have given away more revenue-sharing rights than was needed. It also will end up with idle funds not used for production. “As a result, the firm produces less and collects less money,” they said. “The decrease in money raised has a bigger effect on the firms’ final wealth, which results in sub-optimal profit.”

The situation is a little different for equity tokens: There is less probability that the startup will run away with the money it raised instead of making the product. “As long as the firm does not sell out all the tokens during the ICO … it always produces some product if it raises money,” the paper said. That’s because token holders act more like shareholders who share in the future profit.

As such, “the firm’s incentives are better aligned with speculators’, making the firm less likely to divert cash from funds raised, to its own pocket.” However, for an equity token offering to be successful, the product’s price also must be more than twice the cost of production — similar to utility tokens. Importantly, this result holds even in the absence of additional regulations from which equity token holders often benefit.

Practical Implications

Their findings have several practical applications for ICOs. “With a capped ICO, the firm needs to set a wise ICO sales cap, which determines the fraction of all tokens that are sold during the ICO,” Netessine said. “The cap should be high enough to keep the investors optimistic about the token’s future value. At the same time, the cap should not be too high — the firm needs to save a substantial fraction of the tokens for secondary market trading to retain its revenue and profit-sharing rights.”

Netessine added that “products with higher price-cost ratio have better chances of being successfully financed through ICOs. High product margins effectively act as a deterrent to moral hazard,” such as fraud, cash diversion, and the like. “For firms planning to produce a physical product, ICOs with equity tokens perform better, or closer to first-best [optimal] outcomes, than those with utility tokens, all other things being equal,” he continued.

Startups that issue equity tokens do come closer to achieving first-best than utility-token firms. “Most ICOs, especially early on, issued utility tokens directly tied to the firm’s future products. This can be thought of as a form of revenue sharing: Selling more products increases the value of tokens, independent of the cost of the product,” Tsoukalas said. “Equity tokens, on the other hand, offer profit-sharing, akin to how traditional IPOs work. That is, the firm and investors share not just revenues, but also the costs. We show that profit-sharing in the context of ICOs better aligns incentives between the firm and its investors, and hence equity token issuance can lead to better outcomes.”

“We find that issuing equity tokens incentivizes the firm to produce more products.”

Overall, the authors found that equity-token firms have the potential to perform almost as well as a self-financed firm. “We find that issuing equity tokens incentivizes the firm to produce more products,” all other things being equal, they wrote. Moreover, good market conditions encourage the equity-token firm to push up production, unlike the case with utility-token startups. “This suggests that the first-best [performance of a self-financed firm] is almost achievable with equity tokens.”

The authors did find other interesting results. “The fact that equity tokens outperform utility tokens is intuitive, but nonetheless went counter to the fact that in the early days, most ICOs were using utility–based tokens,” Gan said. “Interestingly, recently there has been an increasing shift towards equity tokens in the industry, which is consistent with our results. Another interesting, and perhaps subtler, result, is that investor over-optimism in the ICO can make moral hazard issues worse. In particular, given the unregulated environment, when excess funds are raised, the firm has greater incentive to divert cash, and in some cases, even abandon production altogether.”

Looking ahead, the authors said there are opportunities to extend their research on ICOs to answer the following questions: To what extent will ICOs disrupt entrepreneurial financing? Are they sustainable for the long-term? And under what conditions do they replace or complement more traditional financing sources such as crowdfunding or venture capital? Future areas of exploration include token resale and inflation control; looking at the use of the funds in areas such as marketing and HR; diving into factors affecting the customer’s willingness to pay and demand, including network effects; how an initial coin offering informs future demand, and others.