恭喜發財,利是逗來 Photo by Kevin Bluer on Unsplash

Bitcoin and The Chinese new year

Ancient traditions and new tech

By Percy Venegas www.EconomyMonitor.com

The recent dip is part of the natural cycle of Bitcoin: every year, a few weeks before the Chinese New Year a sell-off is triggered in order for liquidity to make its way into the informal lending sector in Asia. This year big money ventured to cash out early due to the combined effect of two favorable conditions, a) the whales who traditionally control the market had cover from the negative news from China-Korea b) the new money from institutional investors has more instruments to bet against crypto, in the form of derivatives. Since bitcoin is still the main gateway for onramping/offramping the other coins and tokens, many of them were also affected by the downward price pressure.

The news is not in the news

It was amusing to read a recent piece by The Economist and see how they too succumbed to the hype of the news cycle,

Although it is true that “The crypto sun sets in the East”, it is not entirely because of the reasons circulated in the press. Part of the problem is that some of the explanations have a misleading appearance of sound logic, even if they are weakly based on facts. But also, mainstream media journalists are relying on “specialized” outlets in the hope that they fact-checked their analysis: in reality, many of the writers of crypto news sites are copywriters without training in crypto economics or digital assets finance —they themselves are repeating the opinions of someone else (usually a commentator, with no knowledge scientific methods), and their writing exists for the purpose of converting search engine traffic into ad sales.

In reality, some of the drivers of short-term price changes are rooted in human behavior and traditions, and in the long term are not changing much after all.

The Lunar New Year (Spring Festival)

For context on how a big deal the Spring Festival is, we just need to have a look at this interesting coverage from Quartz. Takeaways: a)to give money, you need to settle debts and have it well in advance (it is bad luck otherwise); b)the tradition is moving to the digital realm; c)we are talking about a celebration at scale — 2 billion souls.

The seasonality of Bitcoin

Now, let’s take a step back. The first thing that we need to understand is that despite the increasing demand for bitcoin, as depicted by rising transactions,

Left: Seasonal Decomposition of BTC transaction count signal. Right: Autoregression function of BTC transaction count signal.

The price of bitcoin is cyclical in nature.

Left: Seasonal Decomposition of BTC price signal. Right: Autoregression function of BTC price signal

In a nutshell, what the plots above show is that the trend of transactions started increasing long before the price of bitcoin did and that there’s a seasonal component to both. However, while transaction count is non-stationary (changes in time), the price of BTC is strongly seasonal.

The Lunar New Year and the Bitcoin cycle

Put bluntly, you can expect that 2–4 weeks before the Chinese new year there will be 2–3 noticeable sell-off events. Those can be very extreme (as it happened in 2015) or more subdued, resembling an “Adam and Eve pattern” (in technical analysis speak), as it happened in 2018 before the bear market of Jan 28th.

BTC sell-off events (in red) before The Chinese New Year.

A few weeks ago there was a rumor spreading, about institutional investors building a short position before the CBOE bitcoin futures contract expiration. Taken out context that is juicy material for conspiracy theorists — and some sensationalist press. But confronted with the facts, one could have realised that a) CME (where there is a bigger upfront commitment, 5 BTC as contract unit) is a better place for institutions looking to engage in that kind of market manipulation b) it makes sense that the bigger players in CBOE are predominantly short, they are also the most sophisticated investors who have actual hedging strategies c) besides speculation, the raison d’être of futures markets is to hedge risk. In light of this, it is more likely that MMs aligned for the dip in price that was already expected (because of the need to provide liquidity to mainstreet in Asia in time for the lunar new year) with the new facilities available (derivatives settlement schedule), rather than engineer an attack per se.

The behavioral economics perspective

But where is the evidence about the price-demand relationship? For a start, we need to understand that in China, exchange and OTC activity are countercyclical — that is, when the regulators are not focusing on one, the other flourishes.

BTC-CNY and BTC-CNY OTC exchange trading volumes (Paxful and LocalBitcoins). Source: National Commitee of Experts on Internet Financial Security Technology, People’s Republic of China

And we can use signals from the off-chain digital economy to measure the changes in demand. First, we need to understand the user journey: when not using a centralized exchange, a person transacting BTC in China would normally initiate a trade in a messaging app, or in an online platform. Online platforms themselves are cyclical: in early 2017 mainland-based Bitkan was dominant, in early 2018 Hong Kong-based CoinCola is. The offramp to cash typically will be through a payments service such as Alipay, in-app within the messaging application itself (e.g. WeChat), or via some cash-equivalent (vouchers, gift-cards or the like). At each point user activity is measurable.

User journey, China P2P BTC transactions.

Behavior similarity

Using a quantitative methodology to compare the correspondence of signals (e.g. canonical warping, a signal processing technique), one can define a distance metric that shows how similar is the behavior of the users of each service to a transactional signal, either transaction count or BTC price.

To visualize the level of similarity between the services under study and the price of BTC we use a statistical technique, multidimensional scaling, and the rendering is in the form of a network graph. Here the edges represent data points, and length is proportional to the distance metric: the similarity between price movement over time and the usage of WeChat, QQ, Bitkan, CoinCola, BitcoinWorld, and Alipay.

Behavior similarity network

The network reveals that the most representative service is the peer to peer OTC CoinCola. We can plot the time series to verify this relationship; there is a very close correspondence in January of 2018 — before the Spring Festival and when CoinCola had gained prominence following a more strict posture of the regulators against its mainland competitors.

Coincola use vs BTC price

What else the press might be missing

So, now that we have established that bitcoin prices move to accommodate human activity (i.e. Chinese New Year), a far more interesting question might be how the money from mining operations in Inner Mongolia may flow to the informal lending sector in conveniently close Liaoning (i.e small business needs access to credit to stack inventory, or to extend consumer credit, especially around the New Year).

The signs are there. A curious journalist simply needs to connect the dots.

Photo by Sang Huynh on Unsplash

新年快樂

一本萬利

Selamat Tahun Baru

Maligayang bagong Taon

새해 많이 받으세요

Chúc mừng năm mới

Tiền vô như nước