When 60% of people are losing money, Bitcoin will rise

(Image from CoinTelegraph)

Key points:

• After a bubble occurs, the market can only recover when prices fall to the point where only about 40% of Bitcoin holders have "unrealized gains", and only then will holders feel the greatest pain and surrender.

• The current distribution shows that holdings are becoming less concentrated and more expensive over time.

The following is a translation of the original text:

Financial asset bubbles and crashes have occurred frequently in economic history, largely due to deep-rooted cognitive biases in human psychology. In Bitcoin’s ten-year history, at least three complete bubble and crash cycles have been observed.

By estimating the distribution of investors’ cost basis, we examine the state of investor psychology at six key moments in Bitcoin’s history, including its most recent state.

Estimated cost basis distribution

Market capitalization values ​​each token at its current price, whereas “realized capitalization” values ​​each token at the time of its last on-chain transfer.

We extend this concept further by assuming that an on-chain transfer represents a transaction between a willing buyer and a willing seller, such that the price at the time of the transfer represents the buyer’s cost basis.

Therefore, “realized capitalization” can be thought of as the total cost basis of all token holders.

Of course, this assumption does not always hold true.

Because, in many cases, transfers do not occur as transactions, such as due to wallet maintenance, transfers between cold and hot wallets, or other reasons (withdrawing money from an exchange to your own wallet).

Moreover, most transactions are conducted on centralized exchanges. Users’ transactions on these exchanges are bookkeeping transactions. They do not need to be registered on the chain, and no on-chain transfers will occur.

For all of the above reasons, we remind readers that viewing “realized capitalization” as the total cost basis for token holders is an incomplete estimate of the true value of the token.

However, this assumption is still valuable. By estimating investors' cost basis, we observed their two emotions during asset bubble & collapse cycles - joy (unrealized gains) and pain (unrealized losses), which in turn affect their investment decisions.

(Note from Odaily Planet Daily: "Unrealized gains" refers to investors who have made profits on their books but have not sold their assets and continue to hold their coins; "unrealized losses" refers to investors who have made losses on their books but have not sold their assets and continue to hold their coins.)

Six key moments in Bitcoin's history

We examine the state of investor psychology by taking a snapshot of the estimated cost basis distribution during the last market cycle and the current cycle.

For each market cycle, three points are selected: the peak of the bubble, the low point of the cycle after the bubble completely bursts, and the mid-cycle adjustment. The following figure marks these six points:

The following table briefly describes each moment, and the details are described in detail later in the text:

1. Peak of the previous cycle (2013.11.29)

We first examine the estimated cost basis distribution on November 29, 2013, the peak of the previous cycle, when prices increased from approximately $100 to $1,100 in two months.

Below we present a visualization that contains a snapshot of the estimated cost basis per Bitcoin.

For example, nearly 5.5 million Bitcoins have a cost basis between $0 and $50 (some of which are believed to be lost forever), and nearly 400,000 Bitcoins were purchased at the peak of the bubble with a cost basis between $1,100 and $1,150.

At the time of the snapshot below, Bitcoin’s price hit an all-time high and 100% of Bitcoin holders had “unrealized gains”.

Despite the explosive growth in Bitcoin prices during this phase, Bitcoin holdings were still largely concentrated in the hands of early holders. At the time, there were a total of 12 million Bitcoins, and approximately 6.4 million Bitcoins had a cost basis between $0 and $100.

As the price of Bitcoin broke through the all-time high of $200, 100% of holders gained. Investor sentiment was very positive and they did not sell their tokens. In the price range of $200 to $700, there were few net transfers and holders continued to hold their coins, which also resulted in a limited supply of Bitcoin.

However, when the price broke through $800, some holders began to sell their tokens because they had received a 100% return in less than two months and wanted to keep their paper gains. In the price range of $800 to $1,150, more early holders began to sell their dormant Bitcoins, and Bitcoin was transferred from miners to later investors. A large amount of selling pressure also caused the price to gradually fall, and eventually the bubble of this cycle began to burst.

2. Low point of the previous cycle (2015.1.14)

After reaching its peak on November 29, 2013, Bitcoin fell 84% in a little over a year. Below we show the estimated cost basis distribution when Bitcoin price was as low as $176 (January 14, 2015). At that time, only 43% of Bitcoin holders had unrealized gains and 57% had unrealized losses.

In a little over a year, the shape of the estimated cost basis distribution has changed significantly. In contrast to the distribution at the peak of the bubble, holdings above $800 are few and early holders have seen a significant reduction in holdings.

Here we introduce another visualization method, which shows the change in the cost distribution between two points in time, rather than a snapshot of the cost distribution at a single point in time. For example, 1.5 million bitcoins were sold between November 29, 2013 and January 14, 2015, with the original price range between $0 and $50.

This change in distribution indicates complete psychological capitulation by investors, which is necessary for prices to bottom out.

As prices fall, both investors who bought at the peak of the bubble (above $800) and the large number of early adopters (below $150) reach their most painful level and their strongest beliefs begin to waver, at which point they will sell large amounts of existing assets to investors who bought during the decline.

We also observed that out of the 13.7 million bitcoins at the time, a total of 5.7 million bitcoins moved from one price range to another, and the cost of holding coins rose and fell.

3. Adjustments in the previous cycle (August 24, 2015)

Bitcoin price corrections during cycles are generally overlooked, but we decided to examine them anyway, especially the correction during the last cycle, when prices fell sharply from their highs, which is most similar to the current situation of Bitcoin: Bitcoin prices fell from about $13,000 in the summer to a low of $6,500 in late November this year.

In the previous cycle (November 2013 to January 2015), after hitting a bottom of $176, Bitcoin price recovered to around $310 in the following months; the price then corrected 35% in the next two months to $200. Here we show a snapshot of the estimated cost basis distribution on August 24, 2015, when the price had corrected 35%.

We also show the distribution changes after the cycle low and mid-cycle correction. Over the 8-month period, we observe a significant change in the distribution. We see some early holders with costs below $200 selling in small amounts, which are much smaller than the sales at the cycle low. This also shows that early holders have basically capitulated, and even after a 35% drop, the market's trading volume has not changed much.

Importantly, we see essentially zero selling pressure from investors who bought at the peak of the bubble (above $800), suggesting that this group of investors has also capitulated.

Instead, most of the selling pressure we are seeing is coming from investors who bought on the dip and those who bought on the rally from the bottom, and who have yet to reach their maximum pain level before prices fall significantly.

A study of investor psychology shows that prices cannot truly bottom out until all investors reach their maximum pain point and give up completely.

After the mid-cycle adjustment was completed, as the investor psychology reached the level of "most investors who wanted to sell have already sold", the price of Bitcoin had no room to fall any further and began to rebound.

4. Peak of this cycle (December 17, 2017)

Within a year, the Bitcoin price surpassed its previous all-time high ($1,100) and peaked at nearly $20,000 on December 17, 2017.

Below, we again show a snapshot of the estimated cost basis distribution, but this time in larger intervals ($500 instead of $500).

At the time, approximately 7.4 million of the 16.75 million Bitcoins were held by early holders at an estimated cost between $0 and $100, leaving 98% of Bitcoin holders with unrealized gains.

The distribution of peaks in this cycle is similar to that of the previous cycle, but with important differences. We can observe a large number of transfers of Bitcoins at prices between $1,000 and $7,000, while the previous cycle's transfers at intermediate prices were much lower. This also led to a slowdown in the pace of price increases over a period of about 8 months, with several major corrections in between.

On the other hand, we see almost no trading movement in the $8,000 to $16,000 price range for Bitcoin. This is not due to a lack of buying interest, but rather because holders have no strong incentive to sell.

At the same time, the price of Bitcoin has continued to rise at a very rapid pace, rising by $8,000 in just two weeks. As the price surpassed $16,000, we saw an increase in transfers as high prices once again incentivized early holders to sell their long-held Bitcoins.

5. Low point of this cycle (December 15, 2018)

The speed and magnitude of the decline in this market cycle was very similar to the previous one – both experienced a decline of about 84% and took about a year to fully deflate the bubble.

At the low point of this cycle, 39% of Bitcoin holders had unrealized gains, which is comparable in size to the 43% unrealized gains in the previous cycle.

This also suggests that after a bubble, only when prices fall to the point where only about 40% of Bitcoin holders have unrealized gains will holders feel the greatest pain and capitulate.

The snapshot of the distribution of the estimated cost basis and the changes in the distribution are also very similar to the previous cycle, which provides support for the assertion that the bubble and crash of Bitcoin (and other financial assets) are driven by deep-rooted cognitive biases that also cause the cycle to repeat itself.

Here we show the change in distribution between the bubble peak and the trough of the current cycle. Similar to the previous cycle, we see that investors who bought near the bubble high (cost basis above $16,000) and early holders (cost basis below $1,000) have capitulated and sold. Investors who bought near the bubble peak and early holders were the first to capitulate.

During the last drop between $6,000 and $3,000, investors started buying as the price dropped, which also meant that the last of the early holders began to capitulate.

We also observe that there are few holders with a cost basis above $12,000, so such investors are unlikely to be a source of selling pressure in the future.

In total, 5.4 million bitcoins (currently 17.4 million) have seen changes in their holding costs in 2018. We should also consider the transfer factor between Coinbase wallets - Coinbase moved its assets to another set of "cold wallets" in early December 2018, with a total of nearly 900,000 bitcoins.

6. Mid-cycle correction (2019.11.25)

The current distribution shows that holdings are becoming less concentrated and more expensive over time.

At the peak of the bubble, early holders held 7.4 million Bitcoins with a cost basis between $0 and $1,000, and today those holdings have dwindled to 5 million Bitcoins.

Without considering early holders, the current cost basis of Bitcoin is normally distributed around $8,000, with a clear peak around $3,500, which is consistent with the low point of this cycle. Consistent with the low point of this cycle, the holding of cost basis above $12,000 is at a very low level, indicating that this group of holders has completely capitulated.

Below we show the change in allocation from June 26, 2019 (peak of the cycle retracement at $12,863) to November 25, 2019 (cycle retracement price dropped to $7,139).

The peak-to-trough decline during this time has been close to 50%. This is a very large decline for a bull market, which has caused many observers to express concern about the current market.

Analysis of the sources of selling pressure shows that investors’ concerns are justified. Unlike the correction in the previous cycle (August 2015), the selling pressure in this cycle correction is very strong, mainly coming from the following types of investors:

• Investors who bought at the peak of the correction and whose cost basis is around $12,000-$12,500 have been the big sellers;

• Investors who bought the dip also began to sell in large quantities, their costs were between $7,500 and $8,000;

• Investors who successfully bottom-fished also sold in large quantities, with their costs ranging from $3,000 to $6,000.

Compared with previous distribution changes, the current distribution change is most similar to the cycle of price bottoming. The optimistic interpretation is that most investors have surrendered and the selling pressure will decrease in the future.

But in fact, those who bought the bottom around $3,000-6,000 are still profitable, and there will be a lot of selling pressure in the future. The capitulation and selling that was originally thought to be completed when the price bottomed in December 2018 may actually take more time or the price may fall further.

The number of dormant Bitcoins increases regularly

The number of Bitcoins that have been dormant for a year (not moved for a year) has been declining after reaching an all-time high in May this year.

As of November 31, the number of bitcoins dormant for a year was 3,174,760; in comparison, on May 18, 2019, the number was 4,500,526, a month-on-month decrease of 41%.

The chart below shows the amount of Bitcoin that has not moved in X years, where X can range from 1 month to 5 years.

From the above chart, we can also find that the number of stocks with a dormant period of more than five years and more than two years has been on the rise in the past year; in addition, the number of stocks with a dormant period of more than three months and more than six months has also been gradually increasing, indicating that some investors are buying on dips and gradually building positions.