Bitcoin Series 35: How I Stopped Worrying & Learned To Love Hashing Rates

 Bitcoin Hashing Rate over time (source: Blockchain.info)

Bitcoin Hashing Rate over time (source: Blockchain.info)

Every few weeks, someone involved in Bitcoin writes the following somewhat breathless blog post:

  1. Block rewards account for almost all mining rewards 
  2. Block rewards are going to drop (next drop in 2016/2017) and drop by 50% every four years forever
  3. There won't be enough rewards for miners leading to insufficient hashing power
  4. Therefore either (a) Bitcoin is in trouble (anti-Bitcoin camp) or (b) block rewards / inflation will have to be increased to keep miners incented (pro-Bitcoin camp)

Points 1 and 2 are true.  Points 3 and 4 are an exercise in ignoring all data that you can't model, like the old cliche of the drunk looking for his keys under the lamppost because that is the only place he can see...

First, let's recall how rewards in $ terms to miners are calculated because that is what matters for hashing rate since silicon and electricity are priced in $, not BTC.

[# of bitcoins in the block rewards] x [price/bitcoin] + [# of transactions] x [fees per transaction]

Things to ask the author when you read the inevitable Concerned-About-Hashing-Rate-In-The-Future-But-Sanguine-About-Hashing-Rate-In-The-Present Blog Post.

  1. What is your assumption for BTC price at the time of the block reward drop?  Also, in the long-future (2020s before it drops again!), what are your assumptions about transaction volume and fees?

    As a general rule, nobody has any defensible assumptions because if you can predict the BTC price in 2017 and 2021 (note, you can't), you would have more lucrative pastimes than predicting hash rates!

    I will declare upfront that I have no idea what BTC price will be 1/5/9/13 years from now (same goes for transaction volume) and I have never read any credible analysis suggesting anyone else has a clue either.
     
  2. Isn't it a bit odd to be predicting that BTC price certainly won't double at least once every 4 years (particularly since reduced block rewards mean fewer BTC to market)? If that not-particularly-dramatic-move-by-BTC-standards happens, rewards in $ terms stay stable or rise, even if transaction-related fees don't rise.
     
  3. How does your model account for the last 2 years?  The block reward in $ terms swung 10x over the last couple of years during the price run up/down.  During that period, hashing rate has gone consistently up (see chart above)   Why would a crisis emerge with a 2x swing in block rewards when it did not with a 10x shift in block rewards?
     
  4. Do we have [not enough][too much][just the right amount] of hashing power today? How about six months ago when we had less? Was it a crisis?  Or 18 months ago when there was much less?

    This is a very tough, very technical question but not nearly as tough as predicting if tomorrow's hashing power will be sufficient (since today all the variables like BTC price, transaction volume, etc) are known.

    So, without a credible model for "Is hashing sufficient now?" how can one have any opinion on "Will hashing be sufficient in the future?"

This is not to say that I have the answers to the questions above and/or I know it will all be OK.  I don't! 

My position is that:

  1. I don't know what will happen but neither does anyone else.   In these cases, worrying is premature since small shifts in other variables will render the block halving irrelevant.  In fact, if you want to worry about blockchain security, worry about colored coins, not native units (that is a different post though)
  2. The implied models of the pessimists don't account for the history of the hashing rate to-date very well at all if they were to be back-tested...
  3. It is hard to imagine too many cases where Bitcoin is a big success and either BTC price doesn't go up or transaction volume doesn't go up.  In which case, in the cases when you want the network to work, it will reward miners and in the cases where it won't need a lot of hash rate, it (correctly) won't have it.

For the full Bitcoin series:  ledracapital.com/bitcoin
Follow us on Twitter: @polemitis and @ledracapital
Facebook:  facebook.com/ledracapital
Email Newletters: ledracapital.com/subscribe
Important disclaimer: ledracapital.com/btcdisclaimer

Posted on August 9, 2015 and filed under Bitcoin.

Bitcoin Series 34: Oracles, Oracles, Oracles and Oracles again

One topic that I believe is under-studied, under-invested and generally under-played in the world of Bitcoin is that of oracles.   Bitcoin is a spectacularly smart global decentralized open programmable value network that can natively trade any asset without counterparties so long as that asset is bitcoins, a currency that is a unit of account in exactly zero economies.

On the other hand, once you want to trade in instruments related to the existing financial system (currencies, stocks, bonds, etc), you are back in the world of issuers, counterparties and slow moving centralized organizations.

Oracles (today, mostly at the concept level) propose an incredibly useful solution to this topic. Imagine an arbitrary number of sources (each with different level of user-determined/rated trustworthiness) broadcasting, for lack of a better world, financial, political, weather and state-of-the-world information in a format that could be incorporated into Bitcoin contracts.   Once that exists (and it will), you will have an explosion of decentralized financial instruments executing over the Bitcoin network.    

Here are some sample contracts you could design to work natively on the Bitcoin blockchain in a world of oracles:

(1) Call options on IBM shares where price is determined by the end-of-day values provided by the following 5 oracles: Google,  Bloomberg, Financial Times, Thomson Reuters and Wall Street Journal. If less than 4 of the 5 agree, cancel the contract.

(2) Weather derivatives that pay out based on the average rainfall in Maine as measured by the average of [the National Weather Service oracle] and [at least 100 Weather Underground Stations oracles]

(3) Election contracts based on majority [5,000 out of 7,000] oracles on who the President of the USA is in 2020

(4) Futures contracts on electricity prices from distributed solar generators in San Francisco for your self-flying Jetsons car 

And so on.  I summarize this concept in this chart below.

Side-note 1: Yes, you probably need some long exposure to bitcoin for this to work today and/or well developed ways to also hedge BTC-[your currency] risk during the contract period but either seems plausible by the time this exists 

Side-note 2: Yes, this is not totally decentralized but could be quite decentralized for things like "who won the Super Bowl" and reasonably decentralized for "what is the price of IBM stock EOD today?".  On the whole, probably at least as decentralized as mining is, if not much more so.

Side-note 3: Could be a nice monetization model for some sensor networks

For the full Bitcoin series:  ledracapital.com/bitcoin
Follow us on Twitter: @polemitis and @ledracapital
Facebook:  facebook.com/ledracapital
Email Newletters: ledracapital.com/subscribe
Important disclaimer: ledracapital.com/btcdisclaimer

Posted on May 12, 2015 and filed under Bitcoin.

Bitcoin Series 33: FedCoin, revisited

One of the final exam questions for the Open Financial Systems course I co-taught on the MSc in Digital Currency Program at the University of Nicosia was "If you were asked by the Federal Reserve to design a US Dollar cryptocurrency (a "FedCoin"), what would the design parameters be?"

Since the course is over, I will give my answer to this question.

Basic Design Parameters

  1. The wallets should be completely open to any human or machine user without pre-approvals, KYC or any other gating process. 
     
  2. No one party should have the ability to block transactions.
     
  3. It should trade at a 1:1 ratio to the US dollar (otherwise, why bother doing this?)

I understand one could design a FedCoin with other parameters but this is my FedCoin so I get to pick the parameters.  In my opinion, if you loosen #1 or #2, you might have a perfectly useful national digital currency but you no longer have a cryptocurrency or a FedCoin.

Execution

  1. A Bitcoin clone (alt-coin) would be fine with some changes in the issuance structure. 
     
  2. Pick a value in USD for how much FedCoin you want to start with.  Today Bitcoin has a market value/supply of $3B.  Let's be generous and say that, in time, demand would be for $100B of FedCoin (It does not really matter what value you pick as we will see below because you won't issue most of the coins unless needed.  In fact, this value could be arbitrarily large and would be better if it was larger rather than smaller).
     
  3. Target 1 FedCoin to equal 1 US Dollar for ease of use.
     
  4. Pre-mine, say, $80B worth of FedCoins and hold them at the Federal Reserve.
     
  5. Have Bitcoin-like mining rewards (new coin issuance and transaction fees) with new coin issuance diminishing over time under the principle that there will be more transactions in the future than in the present.  You won't have to get this exactly right because, if point 8 is well executed, there will always be sufficient transaction security due to Federal Reserve mining capacity.  
     
  6. Have an unlimited standing policy to buy any number of FedCoins for $1 or sell any number of FedCoins for $1 (you can use the premined FedCoins for selling).  

    This ought to maintain a stable exchange rate.   Nobody would willing sell for less than $1 if the Federal Reserve is buying for a $1 and nobody would buy for more than $1 if there is an effectively unlimited sell order at the Federal Reserve for $1, so long as there weren't onerous restrictions on buying/selling.  

    Note that the pre-mined coins are not an increase in the US money supply as a dollar is redeemed in exchange for a FedCoin.  The mining rewards are an increase, albeit trivial, in terms of US money supply, so I suppose the Federal Reserve would need to incorporate that into its overall money supply targeting (or just have the Treasury fund the fees).
     
  7. Presumably there would be dealers that aggregate buying/selling volume with the Federal Reserve but it should be allowable for individuals to come make a trade of FedCoins for USD or vice-versa.  The goal is to have sufficient openness for people to take advantage of the Fed buy/sell offer to avoid dealers adding significant markups for access.
     
  8. Commit to maintaining sufficient Federal Reserve mining capacity to prevent any other party from gaining 51% control of the network, but to not exceed 51% under the Fed's control. 

    In other words, adjust Federal Reserve mining capacity up or down as needed.   This is out of enlightened self-interest as FedCoin is more credible if the Fed is not monitoring/potentially blocking transactions.

    Be transparent with statistics about the Fed's mining power.

Objections:

  1. "But the Federal Reserve couldn't just let [drug dealers/terrorists/enemy states] use FedCoins so we need some account sign up process"

    So long as the Federal Reserve still issues cash, this is a specious argument.  Despite that, I can see that being potentially a mental hurdle.
     
  2. "What is the advantage of this vs. Bitcoin?"

    Creating a cryptocurrency whose value matches the unit of account in the largest economy in the world.  That has value.  

    To the degree that enough market liquidity & hedging instruments emerge in time in Bitcoin to allow one to trade in USD but using the Bitcoin blockchain (at reasonable hedging costs), then the value of a FedCoin diminishes substantially.   I think this will happen in time, it is just a question of how long that will take.
     
  3. "Shouldn't the Federal Reserve give individuals access to have depository accounts at the Federal Reserve instead?"

    Yes, the Federal Reserve should probably do that also, but that is something different.
     
  4. "Allowing competitive mining is wasteful of energy/resources"

    Operating a broad-based retail payment network (cash, credit, debit) at a cost of $0 is a fantasy.  In fact, it appears that many of the existing payment networks consume resources that are measurable in points of GDP so I think this is a silly parameter to add.
     
  5. "I want to conduct monetary policy with my Fed cryptocurrency"

    No, that is a bad idea.  Existing mechanisms are better.   In the context of an existing national currency, FedCoin is just cash for the 21st century (a payment mechanism) and should not attempt to serve a monetary policy also.
     
  6. "Bitcoin is 'hard' finite money, FedCoin is just the plain old US dollars"

    So what?  Anyone who thinks that bitcoins have greater appreciation than US dollars will hold bitcoins.  And vice-versa.

Other:

  1. "Should this happen?"

    It would be a neat preemptive move by the Federal Reserve if done early.

     
  2. "Will this happen early?"

    I doubt it.  A lot of people would have to sign off on such a move and "concerns would be raised" :)

     
  3. "Will FedCoins happen eventually for various national currencies?"  

    Yes, I think so, eventually.

     
  4. "Will that kill Bitcoin?"  

    Probably not, there would be co-existence, particularly if Bitcoin is given the field to itself for a while.

     
  5. "Who will do this first?"

    One of the smaller countries that has a decent financial services industry and would view this as a cool promotional tool (the way Estonia has implemented 'e-citizenship').

For the full Bitcoin series:  ledracapital.com/bitcoin
Follow us on Twitter: @polemitis and @ledracapital
Facebook:  facebook.com/ledracapital
Email Newletters: ledracapital.com/subscribe
Important disclaimer: ledracapital.com/btcdisclaimer

 

Posted on March 29, 2015 .

Bitcoin Series 32: I rented a NYC apartment with Bitcoin and I liked it

 

A few months ago, a family member needed to rent an apartment in New York.  I offered to help out and decided to use Bapple Real Estate as our broker since they were the first brokers in New York to accept Bitcoin. 

For those who have not had the pleasure of renting an apartment in Manhattan in 2015, it is something like applying to an elite university, but more expensive.   In this case, after an extensive application, we had to produce 6 (six!) certified checks (!!) to, among others, our broker, the landlord, the condo board and so on.  Bapple was willing to receive their payment via Bitcoin so this was the interesting part.

This is what the process looked like.

The Payments Made By Bitcoin

  • Bapple sent me their address
  • I sent them some bitcoins from my computer

Total Time: 5 minutes for me; 2 minutes for Bapple

The Payments Made By Certified Check

  • Bapple sent me the list of check recipients
  • I went to the bank branch during business hours
  • I waited in line for 35 minutes for certified checks
  • I realized the checks given to me mixed up the notes and the recipients
  • I stood in line again for 30 minutes 
  • I had the checks re-issued
  • I called the broker
  • She came to Midtown to meet me at the bank branch
  • She wrote out a receipt on paper for the certified check

Total Time: 90 minutes for me; 90 minutes for Bapple

Now, what is interesting is that both processes have one flaw.  Most of the payments were refundable in case the condo board rejected the tenant.  But in both cases, I was relying on Bapple and the other parties to behave correctly.

Bitcoin has the ability to do better than that.   The money could have been sent to a multi-signature (2/3) address with an independent arbitrator.  If both the landlord and the tenant agreed that all conditions were met, they could release the money by themselves.  If there was a dispute, the arbitrator could be brought in to break the tie.  This is how things should work in 2015...

For the full Bitcoin series:  ledracapital.com/bitcoin
Follow us on Twitter: @polemitis and @ledracapital
Facebook:  facebook.com/ledracapital
Email Newletters: ledracapital.com/subscribe
Important disclaimer: ledracapital.com/btcdisclaimer

 

Posted on February 16, 2015 and filed under Bitcoin.