As 2017 draws to a close (this will be the last Market Insight of the year so Happy Holidays everyone!) it is fair to say that the past 12 months weren’t especially kind to the bears. Global equities, despite looking expensive on standard valuation metrics, generated double digit gains in developed markets (EAFE Index +19%) while emerging market equities recorded even stronger gains (EEM Index +27%). Adding to the misery for the bears, the sell-off in bonds witnessed in the final weeks of 2016 failed to evolve into the multi-decade trend reversal many investors anticipated it would.
Thankfully, we were on the right side of these market moves (our currency calls were less successful – the JPY and GBP both displayed resilience we did not anticipate given the profound change to BoJ monetary policy operations in the case of the former, and extremely messy Brexit negotiations in the latter). However, our stellar call of the year was undoubtedly, Bitcoin. It’s not often an analyst can attach their name to favourable market move which in percentage point terms has three zeros in front of the decimal place.
After contemplating this year’s market moves what is clear to us is that many investors and financial market commentators do not fully understand speculative asset price bubbles or their associated components. After all, market prices seemed blissfully unaware of the plethora of articles in the financial media over the past 12 months containing bubble warnings for stocks, bonds and especially (and more recently) Bitcoin.
Former Fed Chairman Greenspan famously argued that it was difficult to identify asset price bubbles ex ante, which was part of his justification for pursuing a “clean versus lean” approach to monetary policy. However, this is far from obvious when one considers a standard definition of a speculative asset price bubble (courtesy of Investopedia).
“A speculative bubble is usually caused by exaggerated expectations of future growth, price appreciation, or other events that could cause an increase in asset values. This drives trading volumes higher, and as more investors rally around the heightened expectation, buyers outnumber sellers, pushing prices beyond what an objective analysis of intrinsic value would suggest.”
Based on this definition there are three key ingredients for a speculative asset price bubble:
Of the three, strong price momentum is the most visible and, as a result, it tends to be the one that most financial market participants focus on when deciding whether an asset price is in a speculative bubble or not.
However, as Bitcoin has repeatedly demonstrated, this is a flawed approach. In a Market Insight published earlier in the year, we included a series of charts plotting the evolution of Bitcoin’s price since 2012. It showed that over this period there were three occasions when its price rose exponentially (incorporating the latest move there are now four occasions). In each of the three previous occasions, there were fairly significant corrections but the underlying trend remained firmly upwards – for our view on this latest move keep reading).
Using the Bitcoin example shown in the exhibit above, the salutary lesson for investors is that exponential price dynamics are an incredibly inexact tool by which to assess whether there is an asset price bubble and, more importantly, whether the bubble on the cusp of bursting.
It has to be considered along with the other ingredients, that is to say, sentiment and fundamental valuations. Of these two additional elements, we are well-positioned to assess the former using our crowd-sourced sentiment indicators. Indeed, it was instrumental in keeping us bullish earlier in the year. As we noted in the aforementioned Market Insight (referenced in footnote 6),
“when we look at crowd-sourced sentiment towards Bitcoin it has been rising recently (positive sentiment momentum) but is far from extreme when compared with previous sentiment peaks. The absence of a strongly positive sentiment reading is important as it strengthens our conviction that Bitcoin is not in bubble-territory.”
Our crowd-sourced sentiment indicators were also critical in keeping us bullish towards equities, especially in the US. Equities may well have been overvalued, and hence ticked one of the boxes in the bubble list, but crowd sentiment was generally subdued – see exhibit below.
Exhibit 2: Crowd Sentiment vs. Price – US Equities
Simply put, there was not a sense of, to borrow Greenspan’s phrase, “irrational exuberance”. This was a perception not readily apparent to investors who use market prices, such as the VIX implied volatility index, as a proxy for sentiment. Indeed, consider CNN’s composite price-based US equity sentiment indicator, which as shown in the exhibit below has been in “Greed” territory for much of the year – (more fake news perhaps!).
Exhibit 3: CNN’s Fear And Greed Index
The key takeaway from these two examples is that to conclude that an asset price is in a bubble, one whose imminent bursting will create exploitable shorting opportunities, requires all three ingredients to be present.
Looking at the latest sentiment readings for global equities – see exhibit below – what is apparent is that the double digit gains witnessed over the past year has generated a more positive hue to our sentiment heatmap. However, across the major indices sentiment is far from extreme, suggesting that the bull market still has legs. (NB: Swiss, Indian and Hong Kong stocks have historically elevated crowd sentiment readings indicating a less constructive assessment).
Exhibit 4: Global Heatmap – Equity Indices
What about Bitcoin, the financial asset (a label many would quibble about as proof of the fact that Bitcoin really is digital gold and like its naturally occurring equivalent is financial marmite – you either love or hate) where bubble speculation is at its greatest?
Some 60 days and USD 10,000 higher than when we last published on Bitcoin signs of over exuberance are much clearer. As shown in the exhibit below, crowd sourced sentiment towards Bitcoin has risen to its highest levels seen since the cryptocurrency began to receive mentions beyond the geek world.
Exhibit 5: Crowd Sentiment vs. Price – Bitcoin
On this basis, and unlike earlier in the year, there is considerable frothiness of optimism on the part of the crowd which has, unquestionably in our view, been a key factor driving the price of Bitcoin higher. In that sense, RBS Chairman Howard Davies, who deployed Greenspan’s phrase “irrational exuberance” when describing recent price moves in Bitcoin is fully justified. Given this, we expect there to be a fairly significant downward correction as, or more probably when, crowd sentiment momentum starts to fade. Looking at previous episodes, a pullback of around 30% would not be an unreasonable expectation – an eye watering magnitude for those on the wrong side of it. Certainly, now is clearly not the time to be jumping on the Bitcoin bandwagon.
Exhibit 6: Crowd Sentiment vs. Price Drawdowns – Bitcoin
Does this also mean Bitcoin is unquestionably in a bubble? As mentioned we have two of the three ingredients for a bubble – excessive optimism (exuberance) and strong positive price momentum. The third ingredient – fundamental overvaluation – is much a much trickier proposition to assess.
For those who think that Bitcoin is nothing more than a hi-tech Ponzi scheme that relies upon the greater fool theory (presumably given its recent price dynamics there are a lot of them around) any value above zero is fundamentally overvalued. On this basis, Bitcoin is the shorting opportunity of a lifetime. That said, with price momentum so strongly positive, timing is everything. (This is a truism in investment, but the dangers of getting it wrong in Bitcoin’s case would be extremely toxic to one’s financial health.)
We find such arguments less convincing. In the Market Insight referenced in footnote 6 above we outlined our approach to generating a fundamental valuation for Bitcoin. We deployed a technique Tetlock and Gardner in their book Superforecasting called Fermi-izing, which breaks down a complex question into its component parts. This approach, they found, often generates superior predictions, especially when information is either unknown or missing. The same technique underpins the Drake equation that seeks to provide a framework for encapsulating all of the relevant information to calculate the number of intelligent civilizations that existed in the galaxy (a bitcoin valuation seems positively tame compared to that).
In the case of bitcoin the equation looks like this:
BIT_fv = (S . 1/N . R . C) / B
BIT_fv = Fundamental value of bitcoin
S = Stock of global fiat money
N = Total number of cryptocurrencies
R = Ratio of bitcoin’s market share to average cryptocurrency market share
C = Ratio of cryptocurrencies/fiat money
B = Total supply of Bitcoin
The outstanding stock of fiat money and the finite number of Bitcoins are both known numbers and serve to provide a crucial valuation anchor. The other three variables upon which the valuation relies need to be estimated and they are:
– how much money will the public hold in virtual form?
– how many cryptocurrencies will there be?
– what will Bitcoin’s market share be?
Assuming 10% of money will, eventually, be held in virtual form and that there would be 10 virtual currencies each having an equal share, we estimated the fundamental value for Bitcoin at USD 29,000. However, using the latest estimate for the stock of global fiat money (USD 83.6tr) and also taking into account that a recent study by Chainalysis that between 2-4 million Bitcoins have effectively been lost, the fundamental estimate for Bitcoin using the same assumptions rises to USD 46,000. Either way this is substantially higher than the current market price of Bitcoin – and hence suggests that short-term froth aside Bitcoin is not a bubble.
Obviously this approach relies on some fairly heroic assumptions, but how plausible are they?
The ratio we are most confident about is that there will be no more than around 10 globally traded cryptocurrencies. Given there are currently close to 1,000 in existence this seems like a big call. However, when one looks at trading activity of the 180 fiat currencies in the world today, what stands out is the very uneven distribution.
The latest data on OTC annual turnover from the BiS triannual FX survey (2016) show four currencies – USD, EUR, JPY and GBP – dominate (see exhibit below). Collectively they occur in more than 150% of all transactions (the data reports whether the currency was included in either leg of the transaction and hence the aggregates sum to 200% not 100%).
Exhibit 7: FX OTC Volumes – 2016
The comparison we are about to make may appear at first glance to be apples to oranges – one is based on turnover and the other is market capitalization – but as both reflect the impact of network effects we consider it to be legitimate. Indeed, looking at the distribution of market capitalization for cryptocurrencies we see a similar profile with the top ten cryptocurrencies accounting for almost 90% of the total – see exhibit below. Of these, Bitcoin is by far the largest at approximately 60%.
Exhibit 8: Cryptocurrencies By Market Capitalization (USD bn)
Even if hundreds, or thousands, of cryptocurrencies are in existence, network effects will ensure that only a handful dominate and, as evidenced by its current high share of market capitalization, Bitcoin has a strong first mover advantage.
Given that, it would appear that, if anything, there is upside risk to our assumption of Bitcoin having a 10% market share of all cryptocurrencies – possibly by orders of magnitude. With every percentage point increase in Bitcoin’s share of market capitalization (using the old economists ceteris paribus trick) adding USD 4,600 to its fundamental value it is not hard to understand some of the seemingly more extreme price predictions. (A 50% market capitalization would imply a fundamental value north of USD 200,000).
Regards the final assumption, the ratio of virtual currencies to fiat currencies held by the public, this is the one we are least confident in guessing. The total market capitalization of all cryptocurrencies stands at 0.48%. Our 10% ratio feels about right but even if we assume that we are too bullish and cut it in half, to 5%, it would still imply a fundamental value for Bitcoin of around USD 23,000 assuming all of the other assumptions are unchanged. (FYI: The break-even cryptocurrency/fiat currency ratio based on today’s Bitcoin price would be 3.5%).
To reiterate, the point of this exercise is not to generate precise valuation estimates for Bitcoin – that is practically impossible. It is to provide a practical framework for thinking about how to value Bitcoin, or any other cryptocurrency for that matter. Plugging in assumptions that appear reasonable (at least to us) generates numbers that exceed current prices, demonstrating that it is far from certain that, even with frothy investor sentiment and recent exponential price gains, Bitcoin is in a bubble.
That said, if we are not in a bubble yet, we are sure there will be one in the end. As Didier Sornette noted in his working paper referenced in footnote 5 above a bubble starts with a new opportunity or expectation, which could be a ground breaking new technology or access to a new market. Bitcoin, and the other cryptocurrencies, qualify on both counts and in that sense are ideally suited to bubble price dynamics -we just might not be at that point yet due to the fog relating to how to fundamentally value them.
Finally, as discussed, short-term we see Bitcoin’s price dynamics being driven by a sentiment battle. Longer-term the battle will be between investors in the cryptocurrency world and governments loath to give up their monetary sovereignty. Media headlines already make clear that policymakers are becoming concerned about their usage and in a recent CNBC interview Jamie Dimon (hardly a fan given his previous comments on Bitcoin) said,
“No government will ever support a virtual currency that goes around borders and doesn’t have the same controls. It’s not going to happen.”
We don’t always share Dimon’s views on Bitcoin, but on this aspect we think he is spot on. There is a regulatory risk element to consider. However, banning Bitcoin or other cryptocurrencies is far from straight forward given the borderless nature of the internet. To be successful it would likely require co-ordinated action at a global level and such things take time. Moreover, the greater the market capitalization of cryptocurrencies (including Bitcoin) and the wider ownership becomes, the more politically difficult it will be to deal with the losses generated by banning their usage.
In the end perhaps this is why Bitcoin is so important. Not because it could be the latest in a long line of speculative asset price bubbles, but because it represents the bulwark between centralized, hierarchically structured governments and decentralized nonhierarchical networks. We judge it to be the opening salvo in a much more important battle… something deep for you all to ponder over during the holiday period.
 The other element was that even if a bubble could be identified ex ante, it was unclear that central banks could calibrate monetary policy in such a way as to avoid triggering economic outcomes it was seeking to avoid – See: https://www.kansascityfed.org/publicat/sympos/2002/pdf/S02Greenspan.pdf
 Didier Sornette at the Financial Crisis Observatory has carried out some interesting research looking at the price dynamics in asset price bubbles – see: https://arxiv.org/ftp/arxiv/papers/1404/1404.2140.pdf
 We had a slight wobble to this bullish assessment in late January when our US equity Fear sentiment indicator ticked up – see: https://www.amareos.com/financialresearch/the-fear-factor/
 Several new approaches have been adopted by which to generate a fair value estimate for Bitcoin, which thankfully going beyond its black market usage which was a feature of early valuation models. One of the more interesting methods, was to use Metcalfe’s famous law, that the power of the network (any network) increases in proportion to the square of the number of users (or nodes). Although this method generates a relative measure of value for Bitcoin (compared with its own history), it does not really give us a good absolute measure.
 Named after the Italian America physicist Enrico Fermi who pioneered it.
 Consider this poor chap who accidently threw out his hard drive containing 7,500 Bitcoin private keys and is now considering digging up the local authority landfill site to find the drive worth an estimated USD85 million! – see: http://www.independent.co.uk/life-style/gadgets-and-tech/news/bitcoin-value-james-howells-newport-landfill-hard-drive-campbell-simpson-laszlo-hanyecz-a8091371.html