Topic: What network hash rate is finanicially sustainable long term? (Read 849 times)

Okay, so lets say roughly 9 years out.

Total dart at a dart board here, but lets say we triple the total power consumption to 30,000 mega watts. 
Lets say somehow that gets consolidated into 3.3 cent power.
Network cost in power alone is now $1,000,000 per day. 

Currently, the revenue is a little over 3x the expense, but that can't last forever.  Lets stick with the theme of 3, and say that the network will scale back if it ceases to generate 133% of power costs in mining revenue.

At 468.75 coins a day, we need price to be around $2800 a coin.  Lets call it $2000 and assume the remainder is transaction fees. 

Okay, that seems plausible, although I suspect the 133% is more like 200% in terms of revenue vs cost.

But lets say the price stabilizes at $1000.

Now we need to come up with ~$800,000 a day in transaction fees. At the current estimated transaction volume ($172 million per day) that means an average tx fee of 0.5%.  Is that competitive?  Granted, by that time, one would hope transaction volume is much higher than todays numbers.

Reality is this probably isn't going to affect the mining sector until 4-5 halvings out.  And by that time, who knows.


Frankly my biggest concern is what happens when some large entity shutters their doors and the hash rate drops significantly.  Especially at the start of a difficulty cycle.  Confirmation times skyrocket, and who knows what happens to the public sentiment/market.   

What I have always found kinda confusing is how people view the "difficulty adjustment mechanism". My view is that it is primarily a means to control the creation of the currency over time. Come hell or high water it's going to do it's best to make sure that until the next halving we add 3600 BTC to the "money supply" every day. It makes this adjustment every 2016 blocks based strictly on how long it took to "mine" those blocks. Folks have thrown incredible amounts of hash rate into mining with associated expenditures in electricity and infrastructure. This is what has led to a multi-thousand fold increase in difficulty in the last 2+ years.

Other folks seem to think that is central to "securing" the network, along with "decentralization" (whatever that means to you). I have no real idea as to what is required to "secure" the network, since there are a variety of threats that you could be concerned about. All of which might have different answers. I can't believe that "more expended hash" per transaction is the answer in general. It's also NOT obvious to me that the infrastructure we have today, and it's expected growth, is truly required for the current transaction volume. Right now, and probably for the next decade or more, it's all about the "block reward", and the transactions are just a sideshow.

I think the difficulty is way to high for the purposes of securing the network. It's hard for me to see what would have to happen in order to actually reduce the difficulty in any meaningful way given the current investment in infrastructure. It will have to be some gut-wrenching reduction in infrastructure for that to happen.

Well lets look at this very simply.
rewards  will drop to 12.5 coins
then 6.25 coins
then 3.125 coins
then 1.5625 coins

transactions are 200,000 per day so 200,000 x 0.0002 = 40 coins  say 150 blocks a day at 25 = 3790 coins a day and 420 a coin = 1,591,800 was earned today   and what was spent on power ?

at best 1100 ph used .4 watts or 440 mega watts x 24 = 10,560 mega watts per day. at 4 cents  means 422400 was spent

that would mean more then 1 million usd  was made world wide.  very small number as we are a small industry.

frankly my guess is power cost was more like 6 cents worldwide and eff is closer to .45 world wide

but the ½ing is coming and coins will drop to 1895 a day and no mechanism has come along to expand transactions.

200,000 a day is very small.

Perhaps I fail to see the direct correlation between this fact and the question my OP is asking. 


All fair points.  I don't really view this (my OP) as a question of miners impacting prices.  I do suspect that there is some relationship between the perceived security of the network and the price of bitcoins. 


I tend to agree with this conclusion, although I hesitate to label it a flaw.  I simply see it as a factor.  Mining presented an opportunity to make money, specifically for early movers.  We are past that phase where the early movers make all the money, and into a phase where the ASIC producers are the great beneficiaries of mining rewards, by and large.  There will come a time, inevitably, when we reach homeostasis, and that final, steady state phase is what my OP is asking about. 

I understand the question I am asking is rather difficult and nebulous, and that any potential answer or calculation is nothing but a semi educated guestimate. 
Still, there exists some upper limit for transaction volume globally. There also exists some lower bound for both price of energy and for silicon based hashing efficiency.
I wager there is also some upper bound for tx fee %.  The huge unknown is obviously also the price.  But I think that unknown sort of drops out.  You have energy costs in dollars, you have operational costs in dollars, you have some dollar/day volume maximum, and some % of that which miners see as revenue, again in $.   

Again for the sake of discussion, I assume price volatility to be drastically reduced by this time, otherwise I simply don't think the widespread adoption is likely.  \
So maybe a decade in the future, you pay for power with BTC, you pay rent in BTC, and you calculate global commerce in BTC/Day, again then you slap that same tx % on to get revenue.  The point is, the mining revenue, in whatever currency is most relevant, will have to come nearly exclusively from tx fees at some point.   So there HAS to be some upper bound to long term sustainable network hash rate, or more specifically, network power consumption. 

I think you are missing the critical point.  Being that Bitcoin is a decentralized p2p digital currency, that means there is no central authority.  

Next point, how is the price determined and who determines it?   Answer: The market determines it.  

Question:  How is the price determined?    Answer: The price is basically a mesh of the 4-5 highest liquidity exchanges, personally I used a blended price that averages them all

Question:  Do miners control the price of Bitcoin?   Answer: No

Question:  Did the miners ever control the price of Bitcoin?   Answer:  Not fully but at a time when there were many more Bitcoins to be mined and the price was much higher than the cost of production, the miners had GREAT influence on the price.    

Why?:  Because they could CHOOSE to hold back supply and pay bills from past earnings and that would LIMIT supply and that tends to raise prices.

What does that mean?:  It means that because the prices are basically marginal (break-even) on the most efficient miners, miners that are not paying for expenses from sources outside of their earnings and are forced to sell Bitcoins as soon as they can and that puts constant downward pressure on the price (not good for miners).   Also there is really nothing that would FORCE the Bitcoin exchange market to play nice so we could get into a situation where the market could get upside down and the miners would just have to take it.

Conclusion: Satoshi was a great visionary that we owe a great debt to, but he had one flaw in his project.  The flaw was securing a trust network with human greed.  This will likely get consensus as time goes on.  I know its the hard truth but it needs to be said.  I hope for the best but plan for the worse.

Centralization, the intended vs realized nature of the network, these are certainly factors that will influence the acceptance/adoption/future of Bitcoin.  

I still get the feeling that (myself included) everyone is focused on the short term, where everything is mining reward centric.  Granted, this makes sense for anyone with money at stake, it is the short term that is making or breaking our investments.  

But I am curious about the long game here.  Lot's of folks talk about "to the moon" and "Bitcoin can change the world". From a financial perspective, some seem to feel like selling bitcoins at any price that is foreseeable in the near future would be foolish, because it is bound to appreciate greatly as the network grows and becomes more heavily utilized by the general population.   I am interested in a technical analysis of mining from that perspective, one whose primary source of revenue stems from transaction fees, not mining rewards.

Certainly, price increases will prolong the effective relevance of mining reward revenue.  But eventually, it will be transaction volume/fee percentage that caps mining revenue.  If we ever get to that point, a point of widespread adoption, I think it is a reasonable assumption that price will have had to come to some level of stability. I doubt merchants will accept bitcoin on a globally significant scale without price stability, barring some instant, free/low cost, exchange mechanism for converting BTC to fiat.  

So with the assumption that you reach a point where the price ceases to increase drastically.  And also assuming that by that time, the network has shifted to operate on the lowest cost power available, you are left with two basic variables.  Revenue from network usage vs Cost of operating the network at a given hash-rate.   This is obviously an oversimplification, as the hash rate will probably influence the confidence in the network, and hence tie back into its usage/overall revenue.  

One of these evenings maybe I will do some research and try to come up with some data based numbers, but basically:

1) Assume Bitcoin grows to handle some X % of global economic transactions (X needs to be derived from some realistic source)
2) Define some upper limit Y of % Fee that miners can expect to be paid for handling the transactions (compared against current fees for other payment methods)
3) Assume power costs are condensed down to some minimum, perhaps we call it $0.01/kWh
4) Assume we start coming up against some asymptotic efficiency limit of silicon/efficiency (this number would probably overestimate efficiency, but that leads to a more conservative model)

It should be possible to come up with some model that gives upper and lower bounds to theoretically profitable network hash rate, given a total $ amount of network usage.

If this model was put into some sort of iterative simulation, it might even be fun to integrate other factors in adjusting the bounds.
For example, if globally the network has a total potential hash rate of 10 Exa, but only 4 Exa is predicted to be viable, the existence of the remaining 6 Exa powered down represents a security risk, and potentially negative impact on network usage.

Long winded, but I am just really interested in the long term aspects of mining.  I have my $5k tied up in my little apartment experiment, and it may or may not ROI.  But whats really interesting to me is if the mining sector is really aligned with the future of Bitcoin, or if mining is simply a gold rush right now, with near sighted goals neglecting to account for the long term viability of the network.  

It is my fear aswell it seems the golden ticket is having good chips.  And I do realize these companies spend a ton in R/D so they deserve a profit.  But I think companies right now are making a killing on current gen.  And they also can put in their own data centers.  And it's hard to tell cost for data center there are some miners I suspect.... cost a lot cheaper then what we pay as regular people.

China manufactuers with low electricity will take control of the backbone of the network if price does not go up more, or difficulty does not stop big jumps.  The question is what will happen when market see's it's not really distributed near as much as most would like..... I'm not sure.

I think were in for a heck of a price jump, or eventually a heck of a decrease.  All depends on variables I cannot anwser.

What you will see is more centralization for the super-majority of the network.   We are at a point where cost efficiency and scale matters.  You just need too much hardware and it requires and certain amount of cooling and power to keep it stable that far exceeds homes and now even commercial spaces that are not specialized for Bitcoin Mining. 

My fear is that in the end it will, for the most part, be just the ASIC Chip manufacturers competing directly with their customers even while selling hardware.   If that is their plan, they should be ethical and come out and say it, if they don't it will become obvious and we will figure out regardless. 

Unsustainable as in, if the network continues to grow and reaches, lets say 8 Exa, that hash rate may be unsustainable long term.  Meaning perhaps it will have to power down and run at 4 Exa.  Etc Etc.



Even if both price and volume do increase, I can't imagine that the chip makers are going to just halt operations?  I would guess we see at least 2-3 more generations of chips hashing into the 12.5 era.  The network is going to continue to grow and grow, reducing everyone's margins.  Eventually it just feels like that bubble has to pop.

There is no such thing as "unsustainable". The system will balance itself out such that those that are borderline profitable will continue to mine whatever value there is in mining. If the value of bitcoin continues to rise, and the number of transactions continues to rise (both of which seem likely), then it will shift that borderline profitable point towards more hashrate and vice versa.

at this ½ ing the hash rate will lower or price will jump.

we have reached a point were we are not sustainable right now.

Correction

the diff level is not sustainable after the 1/2 ing at current btc prices and transaction levels


  people do not see the problem so we carry on ward.

we do 200,000 transactions a day as I type.

the net work is over 1000 ph   so that would be a little more then 200,000 s-7s  maybe 215,000 s-7s

so we use 1 s-7 to process 1 transaction a day.

a trans action fee of  0.0002 is common   the means an s-7 earns 0.0002 a day without rewards.

we are running on empty as i type.

unless price rises and transactions increase .

So eventually mining rewards drop out and mining is funded entirely through transaction fees...

Does this not generate some sort of (dynamic) ceiling for the networks sustainable hash rate?

There will be a minimum transaction revenue required to pay the networks power bill, in aggregate. 

Transaction fees can only increase so much before the market resists paying higher fees.

Transaction volume can increase, but there must be a realistic upper limit to that as well.

After that you can argue the bitcoin price can/will increase.

I don't have the time to pencil this out, but I'm curious if the network can sustain itself out past a couple more halvings? 
Has anyone done any calculation into this showing what the set points might look like for the variables in play? Price, Transaction Volume, Energy Consumption, TX Fee %, etc.