Crypto Derivatives: On Misleading Measurements

Abacus-1
19 October 2019

Noelle Acheson is a veteran of company analysis and CoinDesk’s Director of Research. The opinions expressed in this article are the author’s own.

The following article originally appeared in Institutional Crypto by CoinDesk, a weekly newsletter focused on institutional investment in crypto assets. Sign up for free here. For a primer on crypto derivatives, download our free report.


Data is useful. It enables us to simplify complex concepts into easy-to-visualize numbers, especially when we can apply shapes and colors and transform them into charts that tell a story.

Such as this one, comparing the size of the crypto derivatives market to that of the spot market:

 

The story is that crypto derivatives are booming, which points to increasing market sophistication and liquidity.

Yet even data with the best intentions can be confusing and misleading. Derivative volumes are almost always expressed in notional terms – in the chart above, we are not comparing like with like.

Notional volume represents the market value of the underlying asset to which the derivative contract gives exposure. It does not indicate how much was paid up front for the contract; it shows how much of an asset the derivative theoretically represents.

This is one of the main advantages of trading crypto derivatives vs the underlying asset: you can get exposure to a much greater amount that what you put in.

Spot market volumes, however, show how much was actually paid for the underlying assets. Leverage and credit in spot purchases are offered by a handful of venues, but it is not yet an established feature (few exchanges have the requisite balance sheets).

So, when comparing spot volumes to notional derivative volumes, we are comparing theoretical exposure to actual exposure. You’re starting to see the problem?

Future tense

But what’s the big deal? Doesn’t theoretical exposure represent actual exposure?

No, it doesn’t.

First, most crypto futures in the market today are cash-settled. They involve a promise to pay a stipulated price on a specified date, but no actual crypto assets are involved in the transaction. The exposure is financial, not “real,” and comparing these instruments to actual transactions in an asset is misleading.

Second, even with physically delivered contracts, most traders do not hang on to their positions until maturity. It is relatively easy for options holders to either sell their contract or let it expire without exercising, and even physical futures holders are likely to offset their positions before expiry to lock in gains or stem losses.

Third, notional volumes include a lot of double counting. When a futures trader decides to close her position, she will buy or sell an offsetting contract. Her position now nets to zero, but the notional includes the underlying exposure from her two contracts.

Fourth, comparing derivatives volumes to spot volumes is comparing the future to the present. Derivatives are bets on the future; the state of the spot market is a statement about present value. Comparing different time frames is meaningless. Of course there is much more future than present.

And fifth, notional volume does not give a reliable measure for overall risk exposure. It is an accounting construct that lumps together derivatives with a wide range of maturities; short-term has arguably much less risk than longer-term.

Furthermore, the statistic often includes various types of derivatives, with different exposure characteristics. A futures contract implies the obligation to buy bitcoin at a later date; the exposure is in the future. Options, on the other hand, give the holder the right to buy, but not the obligation; the actual exposure is in the up-front payment.

Options open

So, what is the solution?

Unfortunately, there isn’t an obvious one in sight.

The “notional” debate is not a problem specific to crypto markets. Former CFTC Chairman Chris Giancarlo has often spoken about the dangers of relying on notional volumes to form policy, and the CFTC has started looking at alternative calculations.

The task is mammoth, though. In fragmented markets, collating information gathered with uniform standards is tough. This is compounded by the varying margin rules across an asset, and even within an exchange. Throw in the growing use of credit on top of leverage (where the exchange lends you the money for the initial margin), and the actual exposure gets buried even deeper.

What’s more, as credit seeps into the spot markets, the situation will get even more confusing. Some exchanges offer investors the chance to buy bitcoin with a loan, a practice that is likely to grow – in spite of the business risk – since it is an attractive feature for users. Whether this counts as actual exposure or leveraged exposure depends on the rules of the exchange, as well as on your philosophical interpretation of what debt actually is.

While this would be beneficial to trading volumes (who doesn’t want more upside exposure for the same outlay?), it will obfuscate even further the actual state of the markets. Regulators will struggle to understand where risk might be accumulating, and the lack of insight could lead to poor policy decisions.

This is ironic, for an asset that promises enhanced transparency compared to traditional alternatives.

Foreseeable future

The situation highlights the need for more granular information sharing, and for reporting standards.

More detailed and useful data will not only enable regulators to get comfortable with the risk in the crypto markets; it will also help market infrastructure businesses with their strategy and product decisions. It could even provide a more useful barometer of sentiment, which would inform investment strategies and lead to a more efficient market.

But even more importantly, the confusion reminds us that we need to question the data we are using, and ask what it is trying to tell us.

Often the story is more complex than it seems, and – especially in such a young market as crypto – almost always more interesting.

Abacus image via Shutterstock