In the blocksize debate one thing is really beeing overlooked: security=cost

[BNO]

Hello,

i enjoyed the videos of the scaling bitcoin conference. But one thing that is all the time NOT talked about is what i call the economic law of bitcoinsecurity:

the cost of  the network = the fees for the consumers

The security of the network depends on Hash-Power which costs money. Right now the consumers of bitcoin don't "see" those true costs because they are masked by the block reward mechanism. If you sum up all the fees for transactions for a day then its on average not more than 25BTC which is the reward for one block. Or in other words:

if you compare the 3600 BTC that come per day through mining with the transaction fees of 25btc/day, one can see that 99.3% of daily income of miners consists of block rewards.
Even if you take the next halving of the block reward to 12,5btc/block into account you arrive at 1800BTC/day vs 25 btc.

On the other hand its obvious that one can not maintain the hashpower of the network, with investment for the gear, electricity cost, personnel etc. for 25 btc per day for the whole network:



that means longterm when the block reward halves itself out of existence, either the hashing power needs to go down, or the price of the transactions need to go up.
I don't want to make predicitons about the future here, just have a look at bitcoin today, how it would look with 0 blockreward.

The total income of miners a day of 1825BTC would have to come from transactions.

To not bore you with the calculations, here just the result what it means for the fees (if you are more interested how it was calculated, you can find the calculation below):

0,025btc/KB

From a customer perspective if he says i want to pay a fees of max 0,3% (for a european bankwires nationally are for free and arrive within 1,5 days) that leads to the fact that a customer who wants to pay with bitcoin needs to send for an average transaction:

2,5BTC/transaction !!!

or he needs to accept a higher fee percentagewise per transaction, which might be to much to ask for many transactions. Besides this there are only 2 other possibilities:

• Hashing power of whole network goes down
• cost per GH needs to go down

One might argue that hardwarecosts in the future will fall and that will will lead to lower cost /GH from alone. This forgets, that the hardwarecosts are the security of the network - the network needs to be "expensive" that no one can easily aquire 51% of the networks hashing power.

Conclusion:

The fee market as discussed today is at best only solving the problem that miners need to be able to collect a certain amount of btc to maintain their operations.
It does not adress the underlying deeper problem, that the transaction fees which maintain the network, are from consumers perspective costs. Therefore a real solution can only be cost per GH need to go down. This can be achieved with much more transactions (which means a larger block size), or more efficient mining operations. both lead to more centralization in one way or another. I think therefore that it would be more reasonable for the bitcoin-community to come to terms with a reasonable form of decentralizaition, to much decentralization means otherwise simply that consumers will turn to other payment options, since prices for a transaction are to high.

Calculation:

The typical block is about 50% full (pre-Stresstest-Levels see: https://tradeblock.com/bitcoin/historical/1d-f-txval_per_tot-01071-blksize_per_avg-01071, so 144 Blocks with 500KB each, leads to 72.000 KB/day with transactions.

1825BTC/72.000KB=0.25BTC/KB

the mean transaction is: 0.3KB in size, therefore the average cost is: 0.3*0.25BTC=0.0075BTC

Transactionsize in BTC to achieve 0,3% cost threshold= (0,3*0,025)/0,003=2.5BTC

[1714121876]

1714121876

[1714121876]

1714121876

[1714121876]

1714121876

[Come-from-Beyond]

the cost of  the network = the fees for the consumers

[BNO]

care to elaborate more. Or are you just saying: When we will need hashing power there will be hashing power. The lord will provide for us   

[Come-from-Beyond]

care to elaborate more. Or are you just saying: When we will need hashing power there will be hashing power. The lord will provide for us   

There were moments in history of Bitcoin when a single pool operator controlled 51%+ of hashing power. Security of Bitcoin was near zero at those moments because it cost nothing to that operator to conduct a successful attack. And this was even done with GHash. Does your hypothesis take this nuance into account?

[yayayo]

In broad terms I agree with you that a rise in fees is probable and also necessary to maintain network security. However you left one variable out of the equation: Bitcoin's purchasing power.

When the block rewards drop, this means that the monetary supply of new Bitcoin drops. This will very likely lead to an increase in purchasing power. So calculating transaction costs based on today's Bitcoin values makes not much sense, since it does not take into account the probable development that miners need to spend less Bitcoin to pay for their expenses.

ya.ya.yo!

[BNO]

@come-from...:sorry but please read the posting, the scope of my posting is more on the FACT that the Hashingpower of the network needs to paid via fees in the future (when Block reward diminishes) and what problems arise from that with a simple calculation. Please argue against the calculation an not bringing random stuff up, which is only marginally related...

[BNO]

So calculating transaction costs based on today's Bitcoin values makes not much sense, since it does not take into account the probable development that miners need to spend less Bitcoin to pay for their expenses.

If you read it carefully you can see that the beauty of the approach i took was to calculate everything in BTC the conversion rate to any currency does not matter in this regard. 0,3% of the value of btc is in btc only.

It doesn't matter how much BTC/$ rate is, because even if hardware is denominated in $, the network security must protect the value of the transactions of the lets say 21 mio. BTC. The higher the bitcoin price the sweeter the honey-pot what an attacker could steal, that balances the cost for hardware needed to attack.
In other words very much simplified: If you can steal trillions of dollars you need another security level than if you want to steal billions.

So one could say that the security cost are a constant% of the Bitcointransaction volume of a year. For example 20%

[Come-from-Beyond]

Please argue against the calculation an not bringing random stuff up, which is only marginally related...

Sorry, I didn't notice that you analyzed a spherical cryptocoin in vacuum.

[BNO]

Please argue against the calculation an not bringing random stuff up, which is only marginally related...

Sorry, I didn't notice that you analyzed a spherical cryptocoin in vacuum.

 The purpose of my post is to make a -not much talked about- connection between the Level Security and the transaction prices in an environment with ever diminishing block rewards.

And i honestly don't get what point you are trying to make: "Because at one time Ghash.io had 51% of network so security doesn't matter anymore and screw it??
What does this have to do with the calculations presented that show that a todays security level  (measured in BTC/year) will lead to drastic fee increases in the future 4-8 years  

[Come-from-Beyond]

The purpose of my post to make -not much talked about- connection between the Level Security and the transaction prices in an environment with ever diminishing block rewards.

And i honestly don't get what point you are trying to make: "Because at one time Ghash.io had 51% of network so security doesn't matter anymore and screw it??
What does this have to do with the calculations presented that show that a todays security level  (measured in BTC/year) will lead to drastic fee increases in the future 4-8 years 

You said that "the security of the network depends on Hash-Power which costs money." First of all, it's far from obvious that security depends on hashpower, mining is an open system, we don't know how much power can be brought overnight. If it's 10-fold then no matter how much power you already have you are in trouble. Second, hashpower doesn't cost money, GHash case proves that.

You oversimplify the problem, economic incentive is not enough to assess how secure a cryptocoin is.

[BillyBobZorton]

By the time all Bitcoins are mined, if all things went as it should, the price of 1 Bitcoin will be insane, so even if the people processing transactions only get rewards via fees, it should be enough. Small amounts of BTC will be worth a lot more in the future, it's a self sustained system. We just need that demand coming in and everything will turn out ok.

[gentlemand]

It's adoption or death. The old Satoshi quote about the volume being enormous or nil in a couple of decades applies to this aspect of things.

Lots of people in Bitcoinland have a giant blindspot about little of a shit the rest of the world gives about this thing. If it costs several dollars for the honour of making a transaction then that's an absolutely superb reason to give even less of shit than before.

The folks steering the ship really need to start looking beyond their own little techno bubble and think about how things look to the billions of potential users out there.

[BNO]

In broad terms I agree with you that a rise in fees is probable and also necessary to maintain network security. However you left one variable out of the equation: Bitcoin's purchasing power.

When the block rewards drop, this means that the monetary supply of new Bitcoin drops. This will very likely lead to an increase in purchasing power. So calculating transaction costs based on today's Bitcoin values makes not much sense, since it does not take into account the probable development that miners need to spend less Bitcoin to pay for their expenses.

WTF !! People don't you get it, that the calculation presented does not have/need any BTC/$ Rate!!!!!! It is totally irrelevant for the calculation if BTC costs 1$ or 1Trillion $ The cost for the network will be pretty high and its really the question how to handle this costs. It does not help if bitcoin goes to the moon, the security cost will equally go to the moon.

Read my post again. Read it again and understand the calculations. Now understand: a higher btc price is not helping this problem.

[Come-from-Beyond]

It is totally irrelevant for the calculation if BTC costs 1$ or 1Trillion $

It does matter for security. Look at this - https://blockchain.info/charts/network-deficit

[johnyj]

Nice post, I think people are overlooking many aspects of bitcoin when they are focusing on most unimportant thing: Fee

There are more important things than fees:
Survival is the top priority
Security comes after
Stability is the third concern

People would still use bitcoin if the fee is 1% or even 10%, given that bitcoin's value rise 50% per year. However, if bitcoin's value drop by 50% each year, then even bitcoin transaction becomes free, people will drop it one by one: They don't lack of cheap means of doing transaction

The bitcoin's value is destined to rise long term wise because of its anti-inflation nature, so the fee will never be a problem to stop adoption. However if one of the fundamental issues like survival, security or stability is threatened, the adoption will stop

[Q7]

The folks steering the ship really need to start looking beyond their own little techno bubble and think about how things look to the billions of potential users out there.

This something that I really agree. We always expect when there is no more block reward, price of bitcoin will rise accordingly so that the fees are enough to cover. But what if it doesn't because there is no demand and bitcoin is still used as a hoarding tool and pay for tips and cups of coffee. Obviously i didnt put it well into words but OP definitely did a good job to elaborate in details.

[Quantus]

BNO I understand your post. I have known this for years. But I have utterly failed in trying to convince anyone of this issue.


Let us hope mining pools stop pumping out empty blocks; let us hope the user-base grows substantially to fill up those blocks and be willing to pay substantially more in fees. Let us hope we reach consensus on a reasonable block size when that time comes.  

Because like you said it dose not matter the price of bitcoin or the cost to buy mining hardware. As the price of mining hardware falls it becomes cheaper for all actors good and evil to participate in our ecosystem. Because as the block rewards fall on each Halfing-Day so dose the investment in our security.

[Kprawn]

We are hoping a lot of things will happen... when the halfing occur, we hope the price will go up and the block reward will become more valueble. We also hope that the adoption would increase for people to make

more transactions. In my view, when the price goes up, more people adopt the currency in the hope of getting a piece of the pie. This does not mean that they would make more transactions, I would rather say that

they would hodl more. This is most probably the the reason why most refer to Bitcoin as a experiment.  There will always be people who would be going after the pot of gold, no matter what the reward might be. 

[BNO]

Let us hope mining pools stop pumping out empty blocks; let us hope the user-base grows substantially to fill up those blocks and be willing to pay substantially more in fees. Let us hope we reach consensus on a reasonable block size when that time comes.  

Because like you said it dose not matter the price of bitcoin or the cost to buy mining hardware. As the price of mining hardware falls it becomes cheaper for all actors good and evil to participate in our ecosystem. Because as the block rewards fall on each Halfing-Day so dose the investment in our security.

Yes exactly. I was thinking about the number 70fold. That`s what's needed on how much the daily fees collected need to increase to just cover the 1800 BTC of block reward AFTER the halving (todays security level needs 3600 btc/day). So still it would be a big fat -50% on security after the halving day, but to keep the security at least at the level after next halving. An increase of 7000% in daily fees till Jul/2016 is totally delusional, it not going to happen. Not for this halving, for the next halving in 2020 many things would have to work out. We should get some countries in latin america to use btc on a daily basis for everything instead of their high inflation crap currencies. That would give momentung to the whole thing.

If bitcoin doesn't arrive there the halving in 2020 will be super centralizing from its effects only the large mining farms will survive. It doesn't even help if coinbase totally exploedes since they have a large percentage of ofchain transactions. But unfortunately ofchain transactions pay no fees to the miners...

It's adoption or death. The old Satoshi quote about the volume being enormous or nil in a couple of decades applies to this aspect of things.

So if its not working out security in 2020 would be roughly 75% less than today and at the same time we would try to convince everybody to store things of even higher value on our blockchain, the smart property toppic. Like convincing the nasdaq to make their settlement on the blockchain while security is shrinking and shrinking hope this doesn't come to a chicken and egg situation which hinders the adoption of new markets...