CoinMarketCap Cut Korean Prices From Its Crypto Averages

CoinMarketCap excluded prices from three major South Korean exchanges from its cryptocurrency averages on January 8, 2018, abruptly changing the reference values watched by a market already operating without a consolidated global price.

The platform said it removed some Korean exchanges because their prices diverged sharply from the rest of the world and arbitrage opportunities were limited. Contemporaneous reports identified the affected venues as Bithumb, Coinone and Korbit. CoinMarketCap also said it was developing tools that could provide averages more relevant to different users.

The methodological case for treating inaccessible regional premiums as outliers was defensible. The implementation was disruptive because the calculation changed before CoinMarketCap publicly explained it, making part of the resulting decline look like a sudden market selloff.

A benchmark change that resembled a crash

CoinMarketCap’s preserved January 8 snapshot listed bitcoin at $15,170.08, down 8.01% over its stated 24-hour window. XRP was $2.4562, down 26.53%, while bitcoin cash was $2,421.47, down 13.20%. Those were CoinMarketCap aggregate observations after its methodology change, not synchronized closing prices from a single exchange.

XRP was especially affected because its Korean prices carried a large premium. Contemporaneous reporting from The Next Web documented CoinMarketCap’s displayed XRP price moving from $3.19 to $2.67 around 05:00 UTC, while its displayed market capitalization fell from more than $123 billion to approximately $103 billion.

That roughly $20 billion change did not mean $20 billion of cash had necessarily left XRP markets. CoinMarketCap calculated market capitalization by multiplying a reference price by reported circulating supply. Lowering the reference price mechanically lowered the displayed capitalization even if no corresponding quantity of XRP changed hands at the new average.

The preserved daily snapshot was captured on a different observation schedule and subsequently showed XRP at $2.4562. The intraday and snapshot figures should therefore not be combined into one continuous exchange candle.

Trading losses were real, but causation was uncertain

The episode was not purely a visual artifact. Reuters reported that bitcoin was down 7.1% at $14,980 on Bitstamp around midday in New York and had traded below $14,000, its lowest level in about a week. Reuters also reported that XRP had fallen more than 30% during the day.

Those exchange observations establish that assets were actually trading lower, not merely being displayed differently. They do not establish how much selling resulted from CoinMarketCap’s change. Cryptocurrency markets were fragmented, prices were already weakening, and South Korean financial authorities were inspecting six banks that provided virtual-currency account services.

Market participants quoted by Reuters attributed some confusion and selling to the unexplained data exclusion, but that remains contemporaneous interpretation rather than a controlled causal finding. No surviving record can separate orders prompted by CoinMarketCap from those driven by regulation, profit-taking, leverage or unrelated positioning.

Why January 8 mattered

By January 2018, CoinMarketCap had become an important informational layer for cryptocurrency traders. Its front page compressed activity from many exchanges, currencies and jurisdictions into apparently simple global prices and rankings.

January 8 demonstrated that those figures were constructed benchmarks rather than universal market facts. Venue selection, access restrictions, capital controls and local demand could materially affect an aggregate. Changing the included venues could also rewrite charts and rankings without an equivalent change occurring on every underlying order book.

The defensible conclusion is narrow: CoinMarketCap removed premium-priced Korean venues from its averages, its displayed values fell sharply, and real exchange prices declined during the same session. The records support the event and measurements, but not a precise estimate of how much market selling the methodology change caused.