The miner-trader disparity
Bitcoin hashrate has been increasing at an alarming pace despite the spot price having been butchered year-to-date. Since January this year, Bitcoin miners and traders have lived in completely separate universes.
Why is this happening? Do miners simply have unwavering conviction to the future of Bitcoin? The truth is, despite the blood bath in secondary markets, the current price level hasn't been painful enough to trigger a mass exodus of miners yet. A simple scenario will illustrate why.
Suppose you are mining with the common Antminer S9i. Assuming you are amortizing the machine cost linearly for 12 months, and operating the machine at $4c per kwh, your all-in production cost should be roughly around $5,000 per BTC (source: whattomine.com). As long as the trading price doesn't tank below production cost for prolonged period of time, miners are better off keeping the power on.
However, the average production cost is likely going to change very soon. Competitors of Bitmain are racing to ship significantly more powerful machines.
Panda and Ebang just announced 12nm & 10nm BTC miner respectively. With these and the Avalon and ShenMa machines, Bitcoin hashrate is going to get very interesting in Q4 (and painful for Bitmain’s IPO)— Leo (@Leorzhang) August 27, 2018
Why are both demand and supply still so strong for Bitcoin ASIC miners? On the demand side, most miners entering the market pre-ordered the machines during the peak bull market last year. When new machines are announced, customers can seldomly get the shipments right away. Typically they have to wait for 4 months or even longer. This is because TSMC requires most chip designers to fund the tape-out with full upfront cost. Pre-ordering machines is basically buying futures contract. Once the machines arrive, the miners need to operate the machines regardless of how thin profit margin is to recoup the hardware expense. This is why demand for hashpower is never aligned perfectly with price movements.
On the supply side, as alluded in the previous post, Bitmain is losing its technological edge. Designing cryptocurrency ASICs is not nearly as complicated as designing chips for advanced computing. ASIC design is quickly getting commoditized. The machines coming from new competitors are much better in performance than Bitmain's latest and greatest, a water-cooled version of the S9. Now well-equipped from the pre-order fundings, the new chip designers are able to put down orders for large allocation of wafers at foundries. Bitcoin price is a general indicator of changes in mining capacity, but foundry wafer allocations give more timely information. Intuitively, in the next few quarters we will likely see demand for wafers from cryptocurrency ASIC designers decline.
The disconnect between hashpower and trading price is a known characteristic of cryptocurrency markets. This is due to the different pace of hardware cycle and information structure in capital markets. The two factors are deeply intertwined. Hashpower crystalizes the security of the blockchain, making the network more valuable; in turn, price increase attracts demand for hashpower, and pushing the competition among hardware vendors to be fiercer than ever.
For intelligent investors, how cycles evolve through dynamics of the two factors, is the key to understanding where the opportunities are.
"As on-chain governance systems emerge, staking is expected to play an increasingly important role in the process. Staking gives investors a seat at the governance table and the opportunity to have a profound impact on long-term network value. For example, in Tezos, protocol amendments are submitted to stakeholders for approval by quorum before being automatically deployed on the network. For stakeholders to exercise their governance rights and help shape protocol evolution, they will need to be actively staking on the network."
"There are several reasons why. The first is that layer 1 solutions require ongoing protocol change to happen at the base protocol layer, base layer protocol change requires governance, and it has still not been shown that, in the long term, highly 'activist' blockchain governance can continue without causing ongoing political uncertainty or collapsing into centralization."
Why blockchains don’t solve the voting problem. Part 1/833837— Matthew Green (@matthew_d_green) August 28, 2018
Large-scale voting requires a number of complicated properties. People need to be assured that their vote will be accurately recorded and counted. But votes also have to remain secret.
"This leaves AMD (and IBM, which sold its manufacturing capabilities to GloFo in 2014) without its preferred manufacturing partner. Accordingly, AMD has switched to TSMC for the "breadth" of its 7nm products. This arrangement will cover both CPUs and GPUs, with AMD previously having announced that Vega GPUs (due later this year) and future Epyc processors, codenamed Rome (due 2019), will be built by TSMC."
Market concentration means price control. Coupled w/ tech limitations, transistor cost over time flattens. Then the fabs will have little incentives to go to shrink the lithography, as wafer aerial savings will be insignificant cf investment. Negative cycle ending Moore’s law. https://t.co/t5VTBz3GcU— Satoshi Matsuoka (@ProfMatsuoka) August 28, 2018
News & Commentary
"Strong network effects and integration into financial infrastructure will most likely be reserved for the largest protocols. A quality development team will usually be the strongest moat that a protocol can have in early days. With the value of that team rises and falls the defendability of the rent-seeking mechanism that they created. Once network consensus turns against them, for example because the network is working fine and no more important features are required, the team can be forked out together with the rent."
"The very ideology that justifies the existence of Bitcoin Cash, also justifies the use of chain splits to settle any disagreements within the community. Its easy to see that this ideology, that a hard forked minority chain can be a legitimate successor to the original chain, is completely unstable."
Am I the only one who thinks fees are too dang high in crypto?— Patrick Dugan (@duganist) August 28, 2018
Kraken is charging 0.25% taker, 0.15% maker for starters, crazy!
0.075% taker on BitMex and Deribit perpetual swaps seems very high also. OkEx once had effectively 1.5bps fees (3 points to open), now 8 and 5.
"Coinbase’s mechanism for taking internet money off the internet begins with Martin’s security team pitching its tent and running a shielded power supply inside. This smooths out power fluctuations that can leak clues as to what’s happening. Then the team sets up a folding table with a printer on it, and gets to work."
"On Instagram, STD has posted a string of negative comments as part of the extortion attempt. 'Scam,' reads a comment from one user under a CheapAir photo. A screenshot SSPR shared with Motherboard shows more, seemingly now deleted, comments, including 'I was scammed by this fraudulent company!', 'Stay away from CheapAir,' and 'Used them once, never again.'"
"Here’s the central insight of Shannon’s work: information is measured by how much it changes your mind. That’s it. There is no Truth with a capital T to information. Information is neither true nor false. It’s just more or less influential. If a signal doesn’t make you see the world differently, then it has zero information. As a corollary, the more confident you are in a certain view of the world, the more new information is required to make you have a different view of the world and the less new information is required to confirm your initial view. As a result, the informational strength of any signal is relative."