Why CME CF BRR and Coindesk XBX are far from "nearly identical"

Whilst CME CF BRR and Coindesk XBX prices may closely resemble each other, as indices, they are mathematically and characteristically far from "nearly identical". CF Benchmarks CEO Sui Chung explains why

There has been much comment on the recent rejection of the proposed Grayscale Spot Bitcoin ETF by the SEC. CF Benchmarks, like many market observers and participants is a strong believer that the time for a US spot Bitcoin ETF has come.

Furthermore, from its near decade long experience of managing the biggest fund in GBTC we believe Grayscale to be a fitting manager for such a spot Bitcoin ETF.

However, during the course of the analysis as to the rights and wrongs of the SEC's decision, an assertion has been made that we believe needs to be corrected. Namely that the Coindesk XBX and the CME CF BRR are "nearly identical". They are not.

The genesis of this assertion seems to stem from the comment letter submitted to the SEC during the NYSE ARCA 19b-4 process by Robert E. Whaley, Valere Blair Potter Professor of Management, and Director of the Financial Markets Research Center at the Owen Graduate School of Management - Vanderbilt University.

In this comment letter, Prof. Whaley conducts a rigorous statistical analysis of the performance of XBX and CME CF BRR. We take no issue with the manner of the analysis. However, we don't agree with the statement made in the comment letter that "the key issue is whether XBX performs differently than BRR".

In our view, assessing a benchmark index and comparing it to another index requires the analysis of multiple facets, not just performance alone. After all, past performance is no guarantee of future performance.

A Fully Transparent and Replicable Methodology

For CF Benchmarks, the foremost principle that needs to be adhered to in providing any benchmark is that it is fully transparent and replicable. Whilst there can be a degree of interpretation as to what "fully transparent and replicable" means, let us be very clear that when CF Benchmarks uses this term it means that the full methodology is fully publicly available, such that any market participant with access to the same underlying input data is able to replicate the index calculation.

The CME CF Bitcoin Reference Rate Methodology that has been publicly available since the index's inception in 2016 is a prime example. It details not only the calculation algorithm of the index, but also every single calculation step, parameter, time period, and contingency measure that is or can be employed, as well as when the methodology fails and cannot be implemented.

This is done in both descriptive language and mathematical representation, leaving zero room for misconception or doubt. Any changes to the methodology are also recorded under version control, so that the index history can be accurately replicated.

The Coindesk XBX does not adhere to these principles. Whilst a Coindesk XBX methodology guide is publicly available, at least one critical element of the methodology appears to have been omitted from guide.

The Coindesk XBX employs something called the Constituent Weighting Adjustment Algorithm (CWAA).

Whilst there is a written description of the outcomes CWAA seeks to drive, and the market phenomenon it is intended to mitigate against, there are no details of the algorithm itself, nor are its input parameters published in the methodology guide.

This means that market participants with access to the same underlying input data are not able to replicate XBX index calculation.

Adherence to the principle of full transparency in order to enable replicability is not merely some shibboleth or dogma. It has very real impacts on a benchmark, its users and market participants and investors in the financial products that reference the benchmark.

For the sake of unequivocal clarity, below, I will detail the two underlying pillars that form the basis of our adherence to the principle of transparency.

1. Accountability

We believe that providers of financial benchmarks that can impact the outcomes of financial trading and investments for thousands of market participants bear a very specific responsibility. This means the provider needs to be accountable for the production of the benchmark. This is a key tenet of the IOSCO Principles for Financial Benchmarks, and the BMR regulatory regime in operation in the UK and Europe. The only way a provider can be accountable to market participants is if market participants can replicate the calculation in full, and validate the provider's calculation and implementation of the stated methodology.

Because all CME CF Constituent Exchanges (Bitstamp, Coinbase, Gemini, itBit, Kraken, LMAX Digital) make their data publicly accessible through APIs for no fee (indeed, this is one of the conditions exchanges must meet to become a CME CF Constituent Exchange) then the BRR is not just "fully transparent and replicable" in name, but also in deed.

2. Replicability

For financial benchmarks to be referenced by exchange traded instruments such as futures contracts and ETFs, it is important that market participants can replicate the index fully. They can only do this if the full methodology is available. In the case of any bitcoin benchmark this means actually being able to buy and sell the underlying asset, for the benchmark value, every single day.

Again, this is more than mere principle alone, or posturing. If a benchmark cannot be fully replicated with a high degree of surety, then the instruments that reference the benchmark will suffer. In the case of futures contracts or other derivative instruments, this will manifest itself in wide spreads, as liquidity providers cannot be sure of their ability to hedge risk through buying and selling the underlying asset in the spot market.

In the case of ETFs, inability to fully replicate a benchmark will manifest itself in 'NAV drift', whereby the manager is not able to buy and sell the underlying asset at the same price that it strikes the fund NAV. ETFs are especially susceptible to this risk due to the ongoing creation and redemption process. Only by having a "fully transparent and replicable" benchmark methodology can these risks be adequately mitigated.

Let us re-iterate that we have the utmost respect for Prof. Whaley and his analysis. However, we believe that analysis of index performance is not enough when comparing indices. And we certainly do not think that the term "near identical" can be applied to the Coindesk XBX and CME CF BRR, or indeed when comparing any two indices by analysis of the performance alone.

We believe this misconception can lead to downstream effects for markets and financial products, and that such effects can have very real negative impacts for investors and market participants.

Find out more about the CME CF Bitcoin Reference Rate by visiting our BRR's index page.