Fire before Growth: The Likely Fate of Ethereum Killers — Placeholder

In the coming quarters, a high density of “Ethereum Killers” (EKs) plan to launch their mainnets, and in so doing, release their assets to the public crypto markets. The transition of price discovery from the private to public markets will be an important one to watch and understand, especially considering many EKs carry billion-dollar anticipated launch values.

To follow, I’ll reason through my expectations for asset prices and the prices of services offered by EKs. While maximalists may dismiss this as hopeless investigation of shitcoins, the behavior of newly launched EKs will impact the development of the sideways market we’re currently in, as well as the bull market to come.

To begin, I expect most of the EKs to face extreme downward pressure from their attempted listing prices. Two recent bellwethers for this pattern have been Algorand’s ALGO and Hashgraph’s HBAR (click the links to see price action). I write this with zero schadenfreude as the ramifications impact Placeholder and many of our entrepreneur and investor friends in the industry.

From simple intuition, most EKs set expectations too high in exuberant private markets, and are now attempting to maintain those valuations in a public market that’s looking for blood. In bull markets things can trade up from listing based on wishful thinking, whereas in bear markets the disposition is to trade down, especially if fundamentals are questionable. Fred Wilson recently wrote about how the public equity markets are providing the private equity markets with a reckoning, and I think we can expect a similar rektoning in crypto.

How did private market valuations get so high? Many warned us of the pattern, but Albert Wenger put it well when he wrote in May 2017 that in bubbles “everything is evaluated only in relation to other parts of the bubble and not the world at large.” In 2017, and sadly to a large extent still today, the most common means of evaluating cryptonetworks in relation is through their network values, similar to comparing companies’ market caps [1]. At the start of 2018, Ethereum was sporting a network value north of 130B USD, and un-launched but publicly traded EKs like Cardano and EOS were trading at 10-20B+ USD valuations (see below for January 2018 screenshot from CMC).

image

In that kind of environment, a 6B USD private-market valuation for a credible EK can be reasoned as having 20x upside if it gets to where ETH is, but offered at a 40-70% discount to similar-yet-to-launch-EKs that have (illegally) liquid assets. Fast forward to today and a 6B USD valuation feels like the nosebleeds, exposed to trading down to EOS’ current value (55% down to 2.7B USD), if not to Cardano (83% down to 980M USD), if not lower.

Going a level deeper than the psychology of exuberance and insular-network-value-indexation, EKs may be further weighed down by:

  1. continued wishful thinking for similar valuations as Ethereum due to “superior technology,” without understanding what’s supported Ethereum’s network capitalization to date;
  2. the market’s potential intolerance of inflation, if EKs are unable to develop robust fee markets amongst their users;
  3. the emerging game-theoretic nightmare of EKs deploying relatively undifferentiated systems in a short period of time, spreading developers and investors thin.

Before getting into the details of these three factors, I’ll fast-forward to a rosier conclusion: all of the above should create a glut of smart contract processing capacity, collapsing the cost to innovate on permissionless networks and attracting another bumper-crop of entrepreneurs. For the last few years crypto has felt stuck in a relatively high-cost innovation environment, and I expect the EKs, along with Ethereum 2.0, to help us enter a lower-cost epoch.

It’s not too dissimilar to what happened with the IT industry in the 90s into 2000s. The industry over-deployed capital and overbuilt capacities, leading to asset collapses and supply-gluts in certain subsectors of IT (e.g., the internet backbone), which then allowed entrepreneurs to experiment at lower cost. The experiments that stuck then received further injections of capital and labor, not to mention a whole host of copy-cats taking similar models to new markets. In fair systems, lower costs on the infrastructure side also get passed down to the consumer, increasing demand. Investors will win or lose depending on their timing, but entrepreneurs and consumers consistently come out victorious.

“Superior Technology” Alone Doesn’t Justify Network Capitalization

A common logic I’ve seen in almost every EK goes something like: our technology is superior to Ethereum’s, thus we deserve a similar, if not superior, network valuation. This logic may come from thinking of software as a means to extract centrally controlled profits (i.e., technology monetized through equity). In the equities world, superior technology differentiates a companies’ offerings and allows it to preserve margins, driving that company’s profitability and therefore market capitalization. But in the open-source world of crypto, technology is hardly defensible, and a protocol that attempts to maximally extract margins from its suppliers and consumers is a protocol that’s bound to lose the battle for its share of the market.

If a good enough network-technology gets sufficient third-party investment in supplementary tools and forms of distribution, then those network effects can outweigh the potential benefits of switching to a better-technology alternative that lacks such third-party investment. This is not to say that Ethereum’s headstart can’t be overcome, but rather that third-party investment and the value that results takes time. To unwrap this idea more, we’ll investigate the behaviors that we can expect of the early holders of EKs, and compare that with Ethereum’s capitalization pathway over the last 4+ years.

Let’s first agree that anyone who acquires an asset for profit purposes wants to sell it above the cost of acquisition. Otherwise, why acquire an asset in the first place? The cost of acquisition then serves as a psychological floor, and participants only start to consider selling beneath that floor when hope in the asset’s ability to recover is waning [2]. For an investor, the purchase price represents the cost of acquisition, whereas for a miner that cost is represented by amortized capital expenditures and ongoing operational expenses.

When an EK launches, at point of launch the only visible cost to the market is the amount of capital the founding team raised. Since EK investors often hold the majority of circulating tokens, the investors dictate the order books. And the progression of EK public launches to-date would suggest investors are in a race to the bottom, as they hope to exit above cost. Without lockups, we can expect cascades of selling as each cohort of investors tries to get out above the price of the round(s) they participated in [3].

How is Ethereum different? Time and proof-of-work (PoW). Ethereum’s been building out a robust set of stakeholders since its mainnet launch in the summer of 2015, making its order books more heterogeneous than the investor dominated markets of the EKs. Importantly, miner costs provide economic foundations to ETH’s value, just as they do for BTC. Miners should only sell beneath cost if they’re in distress or have lost faith in the future prospects of the asset. Hence, miner costs naturally create a price-floor for ASK prices of miners’ sell-orders [4].

Early on, the costs of ETH miners were low as the network wasn’t that competitive, and ETH was majority held by investors that paid $0.31/ETH. As a result, the asset had a choppy first half year of operation. That was okay, as crypto was much less in the spotlight then than now, and expectations around Ethereum started off relatively low. For example, on October 21, 2015 Ethereum was trading at $0.44, a 33M USD network value.

As Ethereum became more popular and expectations rose, it became more competitive to mine, increasing the cost to earn each unit of ETH. Rising costs required miners to hike the ASK-prices of their sell-orders, and required more selling to cover those costs. Forced selling injected daily liquidity into ETH’s markets, encouraging organic price discovery [5]. Now ETH is developing a monetary-premium, and so the flywheel continues.

A healthy double-digit billion-dollar network capitalization takes time to develop. The highest quality EKs will establish a baseline realized cap as their private assets turn public, and then will work up from there with adoption and sound cryptoeconomics. But that doesn’t mean the next year won’t be ugly, especially as the private market valuations of the EKs are starting off much higher than ETH did in 2015. That said, the ones that survive the turbulence will bond through the process, achieving a better balance between investors and everyone else. And that’s a good thing, as labor and shared experience earns more loyalty than capital does.

Everything I said above applies neatly to a PoW world, where it appears the markets are attempting to value the native assets as commodities. In contrast, nearly all of the EKs are pursuing proof-of-stake (PoS) out the gate, which I expect the market to consider as capitalized by a capital asset. As capital assets, value will be driven by how profitable it is to be a supply-sider in the network (the asset is the access token necessary to be a supply-sider).

I don’t know if being part of the PoS-transition is better or worse for the EKs. On one hand, PoS is newer than PoW, and a skeptical market tends to discount new, unproven things more than the older, proven things. On the other hand, the value of capital assets should be less anchored to the costs of production than commodities / PoW-assets are, driven instead by profitability (i.e., revenue - costs) of the supply-siders. If some EKs are able to provide high-value services to the world while requiring low-costs from their supply-siders, then these networks could be highly profitable for supply-siders, making the access token a coveted asset. Through this avenue, EKs may not need to “work their way up” cost curves the way PoW-networks have.

In order for the latter, more optimistic interpretation to be a credible pathway, there needs to be a competitive market of demand for the services that EKs provide. Long term, demand-side participants will be the ones who pay enough for the supply-side to have healthy margins (though, due to open & global competition and diverse stakeholders, those margins are still likely to be much lower than the maximally extractive margins we see in equity-capitalized and shareholder-governed services). As I’ll discuss in the section to follow, the high-throughput characteristics of these networks may make it hard for a competitive demand-side market to develop.

Intolerance of Inflation and Potential Downward Price Spirals

Within any network, if there is an excess of supply of that network’s service -- blockspace, smart-contract processing capacity, storage -- then there is little incentive for users to pay much to access that service (i.e., weak fee market). Talk to any staking-service provider and they’ll tell you that inflation revenues dwarf transaction fees in the high-throughput networks to date; conversations with such staking providers actually led me to some of the conclusions of this piece.

Inflation has been so commonly used as a supply-side subsidy to bootstrap towards strong fee markets that we’ve forgotten to question it as a strategy. But in my opinion, markets should only tolerate inflation if there’s a credible expectation that a fee market will develop to subsume it (see endnote [6] for an explanation regarding networks like EOS that don’t charge explicit transaction fees).

If the market loses faith in a network’s ability to ever charge fees of its users in some way, then the market may also stop tolerating inflation, abusing the network’s native asset and crippling the security of a PoS-network. The aggressively marketed throughput of EKs’ makes them predisposed to this pattern, as the advertised excess of supply creates headwinds for a robust fee market to develop.

If the point of these high-throughput networks is to enable applications that require Facebook-scale transaction volumes, then the weak fee market problem will be solved once such applications are deployed and in use. Fees can stay low because the supply-side collects them at scale, and PoS has low operating costs. So it all boils down to "killer apps" and the developers that build them, which brings us to our next section.

A Dense Launch Schedule Spreads Developers and Investors Thin

This factor is self-explanatory, but the way it’s playing out is a game-theoretic nightmare for the EKs. In 2017, the success of Ethereum combined with many technologists complaining about its shortcomings led dozens of teams to raise significant amounts of capital based on the promise of being the next Ethereum.

Those teams have all been building for the last two+ years, and while a handful have launched, the majority have yet to. Teams are feeling pressure as many are behind their roadmaps (an outcome of raising too much money and losing focus), or they want to grab market share before their competitors, and thus there’s a collective rush to launch.

Most of the engineers I meet don’t like to speculate about the systems they’re building upon. They want to build, and know that the infrastructure they’re building on is going to work reliably. If developers feel overwhelmed by the amount of choice and adopt a wait-and-see strategy, then I expect investors to do the same. And this means there will be even fewer BIDs to soak up the incoming supply of Ethereum Killer ASKs.

In Conclusion

Originally I had written my own conclusion, but while this piece was under review Brad Burnham summarized what he took away from it, which I preferred to my own summary:

  1. The 2017 crypto bubble capitalized a lot of worthwhile experiments even if it was irrationally exuberant
  2. In networks where multiple stakeholders invest time and money, technology does not always win
  3. Independent of the number of stakeholders, the diversity of ETH's stakeholders is a defensible advantage
  4. ETH built liquidity slowly and mining created a cost floor
  5. We understand the characteristics of PoW-created commodities, but are at an earlier stage in understanding PoS-created capital assets
  6. High throughput networks will create weak fee markets unless high throughput applications emerge to soak up capacity
  7. It could get worse before it gets better, as many EKs come to market simultaneously, and supply outstrips demand in the near term

While I don’t expect the EK shakeout to be fun, the public markets checking the private markets is by design, and in the name of optimal allocation of capital and labor. Sometimes a fire must run through the forest in order for saplings to grow again.

Endnotes:

[1] While network value is a useful tool, it has many nuances and should not be used in isolation.

[2] Different acquirers vary in their time-frames for tolerated recovery period, as can easily be seen in the delta of behavior between a trader and a VC.

[3] Long term investors, such as Placeholder, USV, and a16z, are accustomed to J-curves, and if involved can be expected to hold through the initial plummet, with the expectation that in 5-years time they will be the beneficiary of the slope of enlightenment. Unfortunately, in the hot-money rush of 2017, such investors were in the minority.

[4] Such thinking fits with the common refrain that the price floor of commodities should be at marginal cost, as beneath marginal cost is where suppliers start shutting down, decreasing supply, and therefore re-balancing supply-and-demand. Any PoW-based asset I think of as being a commodity, roughly abiding by this rule. If raised in the Bitcoin community, though, marginal cost being a price floor is often contended due to the difficulty adjustment theoretically allowing for recursive downward adjustments if miners continually go offline. Practically, however, in the two big bear markets I have lived through, bitcoin bottomed in the range of prices where chatter surfaced that many miners were operating at cost (~200 USD in 2015, 3000 USD in 2018/19), and the same goes for Ethereum miners in the recent bear market.

[5] As a crypto market participant over the last 5-years, anecdotally the networks that allow significant amounts of their asset to be earned develop the healthiest markets in the space. Earning implies a cost structure, which then forces selling of the asset to cover costs. Forced selling is vital, akin to putting chum in the water to catch the interest of fish, sharks and in crypto, whales. As the scale of the market grows, this activity draws in more participants, instruments, and standardization, all of which improve price discovery. Without forced selling as a kickstart mechanism, a cryptoasset’s market tends to remain relatively thin, inactive, and brittle (XRP is an anomaly here due to the amount Ripple invested in improving its markets).

[6] In the case where discrete transaction fees aren’t paid but instead capacity is secured by staking, then as long as there’s increasing demand to use the network (and capacity is constant), then the value required to be staked should still rise. The rising staking value should counteract the effects of supply-inflation, and one would expect the asset to rise in value (capital gains) if the growth in demand (PQ) outstrips asset supply growth (M), at constant velocity. In this sense, a fee market develops in the amount of stake required, and thus the market tolerates and absorbs the inflation.

Thank you to the Placeholder team, Fred, Albert and Nick at USV, Kristen Stone, Yassine Elmandjra and Staked for providing information and feedback that led to the formation of this piece.