Trading Cryptocurrencies: The Role of the Resale Option and How to Exploit it.

“Theirs is nothing to support Bitcoin except the hope that you will sell it to someone for more than you paid for it” — J.Bogle, Vanguard

Is there a Bitcoin bubble? Are cryptocurrencies efficiently priced? How do I determine the fundamental value of a cryptocurrency?

These are the questions I often hear when talking about cryptocurrencies.

I come from an equities background, and what I teach students at university in my finance course is that you value a security by examining its net present value.

Consider a stock.

A stock is valued by summing its discounted dividend payouts into perpetuity. How do you work out its dividends in the future? You first need to estimate its net profit. Net profit is determined by looking at the stock’s underlying business. What is the business selling? What’s the growth rate? What are the operating costs? And so on. … You can soon see that something as simple as discounted cash flows can be transformed into a deep dive research on company fundamentals. If you’re good at this, you get rewarded with a six digit tax bill.

What about cryptocurrencies?

If you apply the traditional discounted cash flows (DCF) approach, you run into some trouble. Holders of cryptocurrencies are not entitled to any future cash flows. Most ICO tokens do not promise any dividends. With no future cash flows, how do you value a cryptocurrency? Or is it quite simply just worth zero?

Not a security.

Not quite. Because not all cryptocurrencies are securities. Most cryptocurrency coins, for instance, would consider themselves to be a currency rather than a security. You don’t value the US dollar with DCF now do you? The sole purpose of a coin is the act as money: a unit of account, a store of value and a medium of transfer. There has been a number of attempts to value cypto-coins, using the theory of money or levering upon network value — in particular Metcalfe’s law (see Peterson, 2018 and Van Vliet, 2018). Tokens, however, cannot be valued in such a manner, as they are tied to some sort of entrepreneurial venture, which makes them more like securities than currency.

There are some differences. Traditional securities gives holders some claim to company profits. Tokens do not give holders claims to bottom line profits, but gives them exposure to top line revenue. Tokens are designed to have a certain use case with the issuer’s business model.

Starbucks gift cards.

Think of tokens as being Starbucks gift cards. Except, now pretend Starbucks only accepts their gift cards as the method of payment and one gift card entitles the holder to one cup of coffee. And so on Monday mornings, as you would expect, the value of these gift cards might trade a little higher on the market…

Japanese speculators.

Much of this ICO frenzy reminds me of the Japanese secondary market for golf club memberships in the late 80s and early 90s, which were traded like stocks and bonds. The average price of Japanese golf-club membership reached circa $250,000 in the early 90s. Many were sold prior to the golf-club being even built. For example, Ibaraki Country Club sold more than 50,000 memberships between 1988 and 1991 ahead of its scheduled completion. Many of the buyers did not have any intentions of playing golf at the country club, but simply saw it as a speculative asset. Would it be correct to say buyers of the Filecoin ICO are no different?

A Thought Experiment.

Say I offered you a one dollar bill. But this dollar bill is special because it’s slightly different. It’s still legal tender, but it has a slightly different colour and texture to your normal standard bills. Perhaps it was printed incorrectly. Now, pretend you know you have this one bizarre uncle, who’s probably willing to buy it off you for five dollars due to its novelty effect. How much are you prepared to pay for it?

Let’s change it slightly. Say, instead of this one uncle. You now know there’s a whole group of bizarre people online who are probably willing to buy it off you for various prices ranging from one dollar to several hundred of dollars. How much are you willing to pay for it now?

The value of the dollar bill here, is no longer just its fundamental value, i.e. 1 dollar, but also incorporates what is called a “resale option value”. The value of the option of selling it to someone else down the track for a higher price. In a market with no short selling and a lot of dispersion in beliefs, this resale option value is especially high. In other words, if you think you know there’s going to be a lot of divergence of opinion on the price in the future, then perhaps today, you are willing to pay slightly more than what you think fundamental value is. Because you can always sell it to that one guy with the most optimistic opinion sometime in the future.

The resale option hypothesis was first explained in Schienkman and Xiong’s (2003) paper “Overconfidence and Speculative Bubbles” in the Journal of Political Economy. It has since been used to describe equity market mispricings (Cheng, Lung and Wang, 2009, Journal of Financial and Quantitative Analysis) and out-of-the-money Chinese warrants (Xiong and Yu, 2011, American Economic Review).

Preparing to pay a little more.

Are cryptocurrency holders willing to pay that little bit more in the hope that one day they can sell it for a higher price to some random person with an optimistic view?

I think so.

The more uncertain the information, the more divergent the opinion. The more divergent the opinion, the greater the resale option value. (This is why Bitcoin rallied hard in 2017, but government bonds not so much.)

In my research, I find strong empirical evidence in favour of the resale option hypothesis in cryptocurrency markets, i.e. resale option characteristics drive prices.

The Exploit.

If a cryptocurrency demonstrates strong resale option characteristics, what does that mean about its future returns?

Intuitively, coins with higher resale option values are trading at higher values, have larger bubbles and are likely to have lower future returns.

I test this out over 3.5 years across 89 cryptocurrency coins. Each month, I sort cryptocurrencies based on their resale option characteristics and went long coins with low resale option and short coins with high resale option. On average my long/short “crypto-neutral” portfolio made 13.2% per month.

I test this out again, for 131 cryptocurrencies over a shorter sample series.

Table from “Resale Option and Cryptocurrency Mispricing” working paper

I find the same effect.

My paper can be found at: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3253208