Bitcoin Value Indicator Report: December 1, 2018

|
About: Bitcoin USD (BTC-USD), Includes: COIN, GBTC, TWTR
by: Hans Hauge
Summary

The BVI gets complicated! For the first time the price is below the prediction of a single metric.

The BVI has been revamped and undergone some positive changes.

The Bitcoin Z-Signal has been in negative territory since November 20th.

I'm buying the dip.

Introduction

If this is your first time reading a Bitcoin Value Indicator report, you may want to get caught up with the initial post which explains the reasoning and methodology behind it.

The TLDR; version is that the Bitcoin (BTC-USD)(COIN)(OTCQX:GBTC) price grows along with the network. That means that when we look at the data in log scale, we can see a strong relationship between the total transactions, the unique addresses, and the hash power of the network.

At times, the price is bid up to levels beyond the natural growth rate of the system. These times tend to precede a Bitcoin bubble. At other times, the price of Bitcoin tends to be below the price that we would expect from looking at the other network stats. These are optimal times to purchase Bitcoin, which tend to precede a large price run-up.

Therefore, the Bitcoin Value Indicator is a tool to help you maximize your return on investment; should you decide to venture into the dangerous waters of Bitcoin. Never invest more than you are willing to lose. This is a new asset class and anything could happen.

Important Changes

Before we dig into this month's report, I want to mention a few things that have changed about the system itself.

  1. I have finally automated the chart creation, which up until now I had been doing manually in Excel. With this change, I have also switched over to a new charting framework, which I think enhances the visual appearance and readability of the report.
  2. Instead of looking at the Market Cap of the system, we are now looking directly at the price. The strength of the relationships between the input metrics and the price is on par with the relationship to the market cap. However, I don't personally like the term "market cap" or even the way it's usually derived, which is not the same for crypto as it is for stocks (many cryptos manipulate their "circulating supply" to bump themselves up in the market cap charts). Also, I think talking about the price is a more direct approach that will make using this tool even easier.
  3. Our data feed has improved. From 2009 to 2017 I have been only able to get full network stats in 48 hour increments. But, starting near the end of 2017 and going forward, we have samples every 24 hours. This will improve the quality of the output in the long term, but it does also give more "weight" to recent measurements because now we're getting 365 per year instead of 183.

Hash Rate of the Network

The hash rate of the network is a response to the increasing price of Bitcoin. However, unlike the price of Bitcoin; the hash rate tends to be more stable because getting into or out of the mining business is a big decision. Investment in a large mine requires CAPEX and time. Getting out is also a big decision not to be taken lightly.

What we've seen this year was that the Bitcoin price spiked back around the start of the year, but the hash rate continued to increase until late August. Now, the hash rate of the network has started to decline.

Some have pointed out that the price drives the hash rate, and have questioned the validity of using the hash rate to determine the price. To this I will make the following arguments:

  1. The hash power of the network did increase by several orders of magnitude between January 2009 and August 2010 (when the first price was recorded). If the price always drives the hash power, then how did this happen?
  2. While it's true that a higher Bitcoin price will attract more miners, an increase in the number of miners and the resulting hash power actually does benefit Bitcoin. The massive hash power that we see securing the Bitcoin network hardens it against attack and makes the system more secure. A more secure system is more attractive, more robust. This creates a positive feedback loop which I think leads to a higher price.

In Zen they ask, "does the fish move the water, or does the water move the fish?" I think the proper answer is "yes."

Let's look at the hash power regression in isolation first.

hash power regression Source: blockchain.com and author's new charts

Isn't that a nice looking chart? Feel free to comment below and let me know what you think of the new charting framework provided by plotly.js.

What you're seeing here is the price of Bitcoin compared with the predicted price based on the network hash rate. The Bitcoin Value Indicator looks at monthly averages, but here you're seeing more fine-grained detail.

As you can see, the hash rate prediction near the end of last year was sending a warning signal; but now it seems to be back in the negative Z-score range (safer time to buy, but not necessarily the bottom of the market).

Is declining hash rate an issue?

I have been asked if the declining hash rate of the network is an issue. Earlier I said that more miners make the network more robust. So, if we have less miners are we in danger?

This is a good question. In the long term, declining miners would be an issue. However, after bubbles in the past we have seen hash power decline before. The longest decline that has been observed was a six month decline between 2011 and 2012. So far, we've seen the hash rate declining for just over three months.

I expect that we'll see the hash power fall further for two reasons:

  • I don't think we're at the bottom of the market yet
  • There's a delayed effect from when the price moves and when the hash power responds.

The hash power will stabilize in the next few months in my opinion because there's a bunch of new equipment hitting the market that's more efficient, and I think the price will stabilize soon. These are just my opinions, we'll have to see. We should monitor this situation, but I think this correction is the natural reaction to a market that was overheated too much for too long.

Total Bitcoin Transactions (confirmed, on chain)

The total Bitcoin transaction count will continue to climb as long as people (or software platforms) are using the network. The rate at which this count grows is an indicator of the adoption of the network and the length of time that it's been alive. Like the hash power, this is a more stable metric; but it measures total lifetime activity of the network instead of hash rate of the miners. See below.

total transaction regression Source: blockchain.com and author's new charts

The total transactions prediction make a much bigger deal out of what happened in 2011 than what happened in 2013. But, like the hash rate, we are seeing that the price is finally below the predicted levels.

Remember that since the Bitcoin Value Indicator uses monthly averages, and the price only really started to drop in the last two weeks, on a monthly basis we will not see as large of a movement as we might expect.

Unique Addresses

The unique addresses in use represent one way of quantifying the number of users. This metric has its own strengths and weaknesses, as I wrote about here. Let's look at the chart.

Unique address regression Source: blockchain.com and author's new charts

The highest spike in price/predicted recorded by this metric was back in 2010. However, the widest mountain range is the one we're in now. Note that using this approach, the Bitcoin price is still too high.

Bitcoin Value Indicator Chart

Without further delay, let's look at the Bitcoin Value Indicator chart for December, 2018.

Bitcoin Value Indicator Chart Source: blockchain.com and author's new charts

For the first time since June of 2017, the average Bitcoin price for the month was low based on a single metric — the price predicted by the hash rate. However, the other two metrics are still predicting lower prices (total tx and unique addresses).

You may notice that in 2014, the price first crossed the hash power indicator before crossing the other lines. Will history repeat itself? Do you guys think this could end up being the longest bear market in Bitcoin's history?

So far the longest downturn in price was a period of 15 months, set after Mt. Gox collapsed. We're at 11 months right now... guess we'll have to wait and see.

Signal Breakdown

Indicator Name Unique Addresses Hash Power Total Transactions
Predicted Price $1,802 $5,554 $3,969
Actual / Predicted 3.02x 0.98x 1.37x
Metric Signal Overbought Oversold Overbought

Source: Bitcoin Value Indicator tables and blockchain.com

December 2018 Signal

It's complicated!

fry can Ah, complexity. The fly in the ointment of decision making. What are we to do when one input says the price is too low, and the other two say it's too high?

One option is to dollar cost average in as we enter the bottom of the market, but another way is to change your perspective. Consider the Bitcoin Z-Signal.

The Bitcoin Z-Signal

The newest member of my family of metrics is the Bitcoin Z-Signal, which I wrote about here, and here. I have been posting daily updates to Twitter (TWTR) which you can see if you follow me on that platform.

The Z-Signal takes the total transactions, unique addresses, and the daily transactions and rolls them up into a single chart (using multiple regression, instead of independent linear regressions in parallel). The idea here is to simplify the decision making process. If you don't want to deal with multiple predictions and the fact that they are rarely going to say the same thing, then the Z-Signal is for you.

Importantly, the Z-signal actually has a higher R^2 and lower standard error than all of its inputs. So, it may be the best way I know of to gauge the current price at a single glance.

Let's first look at the long chart.

Bitcoin Z-Signal Source: blockchain.com and author's new charts

As you can see, when we use all three input metrics we get a simplified perspective. Bottoms of previous bear markets have tended to be lower than where we are now, but at the same time, we are no longer in "dangerous bubble" territory.

If we look at where we are now, compared to the last two bubbles; this is how it breaks out.

Low Price High Price Max ROI Low Z-score High Z-score
2011-2013 $2.29 $1,151 502.6x -0.36 1.53
2015-2017 $243.99 $19,498.68 79.9x -0.41 1.21

As of today, we have a reading of -0.12. The lower this goes, the more attractive the buying opportunity (in theory).

The Z-score has been negative since November 20th. Let's zoom in a bit to the last two years and focus on them.

z signal z score only Source: blockchain.com and author's new charts

You can see the depth of the last winter still lingering on the two year chart. Are you ready to buy the dip? I am.

Conclusion

I've said before, and I'll say again; please use these metrics with caution. Nobody can predict the future. I happen to believe in the power of the Bitcoin network, and I think that we will see these trends continue unless they outright fail completely. This was the topic of my article on Bitcoin's Binary Future.

If you are looking for a way to time your Bitcoin entry and exit points, then I humbly suggest that the approaches I have shown here be a part of your research cannon. There were many high quality researchers who predicted the last bubble and were not surprised by the downturn. It would be wise to seek out these people and follow their work if you're going to be in this space.

Cheers,

Hans

Disclosure: I am/we are long BTC-USD. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.