Topic: Is it possible that bitcoin will become unaffordable to use for micropayments? (Read 14204 times)

Technical solutions have been proposed, but not implemented.  See Scalability. Without them, the Bitcoin network maxes out around 7 transactions per second, and lately traffic has bumped against that limit during busy periods. Scaling up requires cooperation from the big mining pools. Higher fees benefit miners. Whether they want to do more work for less money is a big question.

Some people have suggested using unconfirmed Bitcoins for small transactions, but that's a headache. Coinbase only buys "confirmed" Bitcoins. Read their blog as to why. It's not so much a rip-off problem as a problem with downstream transactions getting stuck because there's an unconfirmed input.

Coinbase's business is to provide a way for Bitcoin retailers to accept payment without taking a risk on the price of Bitcoins dropping. Bitcoin dropped 25%-30% today, then partially recovered. If you're a goods retailer with a 15% markup, you don't want to be exposed to a big drop like that. So if Coinbase insists on confirmed coins only, that sort of sets the pattern for Bitcoin retail.

So what might be a good solution to this? 

Bitcoin is starting to hit scaling problems. Look at Mt. Gox Bitcoin transactions unconfirmed after 2 hours. 155 in the queue right now, but numbers from 400 to 1000 have been seen. Also look at average transaction confirmation time. During busy periods, it's been spiking up lately. There's now enough transaction volume to temporarily saturate the network. This isn't happening all the time, but it is happening frequently now.  Mt. Gox has been getting complaints about how long it takes to confirm transactions, so they've bumped up their fee to the point that theirs should get through.

Many of the mining pools are only generating 250K blocks as a default. The Bitcoin client can be configured for bigger blocks, but it doesn't have to be. There are negative incentives to miners in making their blocks bigger - their confirmations slow, and there's a risk of an orphan block.   This is slowing confirmations. A bigger fee can get your transaction confirmed sooner.  See https://bitcointalksearch.org/topic/blocks-are-not-full-whats-the-plan-339505

Then ask MtGox why they are ripping you off charging 20x what is necessary.  Then look at how much they are actually paying.   MtGox ripping customers off is nothing new it also has nothing to do with Bitcoin.

The transaction fee is 0.001 BTC at MtGox.  It is near the end of the press release.
https://www.mtgox.com/press_release_20131120.html

At 0.0005 fee, 2 thousand transactions will make 1 bitcoin for miners each block, and that is approaching the block size limit

Anyway the micro payment is not the main purpose of using bitcoin, the store of value and international transaction are most important. For daily purchase of grocery,  people already have credit card and mobile payment solutions from banks

You sure it wasn't 0.0001.  If it is then it is more like $0.10 not a dollar.  If it isn't then you should ask MtGox why they are charging fees 10x what is required by the network.

Well, this morning, Mt.Gox informed us that they will require a 0.001 BTC transaction fee to "increase the speed of transactions in the network."
At current prices, that is over a dollar (US).  That's getting expensive.

I am whining about this for some time

Why not use an effectively dead  bitcoin address.An another payment system would originate there. With different rules I don't know what  kind of  properties are needed to enable  dust operation but I think it is feasible. Lost coin for bitcoin network could be  the root of value in this  parallel "world"     

bitpay doesn't wait for a block to confirm your payment why should a newspaper.

If there is no rush, then send without any transaction fee. I feel bitcoin wouldn't work for a pay as you read system or alike for online newspapers. The confirmation times are too slow, I want to click and read, it click, wait upto 20minutes them read...

The problem of fees for micro-payments only comes up when you want to send/receive many of them (when even Bitcoin's lowish fees become a burden).

If everyone you are dealing with is sending/receiving very few micro-payments (maybe you own and run some form of online game) then just have them pay the fees.

If many of the people you are dealing with are also sending/receiving many micro-payments (maybe payments for bandwidth on a darknet or for a torrent; something organised but decentralised) then you could use probabilistic payments to reduce your fees down to the cost of sending/receiving a few packets of data to/from a peer.  Probabilistic payments have the drawback of variance but this will be drowned out by the volume of payments you expect to make.

Highly unlikely I would think but must confess that I would also like a comment on this from somebody more knowledgeable

Good concern to have

Campaign for 0.000001 fee for bitcoin.

I can already +1  and +1 again (me, myself and I)

G:)

Now, that one, I don't know. However, this guy might: [email protected]

Other contact info (including a UK phone number) here.

Yeah how to get a blockchain.info/MyWallet to permanently stay in GBP 

We do have a single simple unit that is easy to correlate with the $/€/£. it is the BTC.
Currently the exchange rates are:
1BTC =
1BTC =
1BTC =

Tell me what wallet application you use, and I may be able to provide the steps necessary to achieve your goal in your application.

I don't use the default bitcoin wallet and neither will new adopters or the masses.  Calling them micro/milliBitcoins and working in different values requires more effort and thought for new adopters.  Just having a single simple unit that is easy to correlate with the $/€/£ would be simpler and ease adoption plus use.   

I think you're missing a 0, good sir.

Exactly  Bitcoin does have a very low price:  ~$0.01 per mBTC and there will be a high number of units ~21 billion mBTC.    

Settings > Options > Display > Unit to show amounts in > µBTC

Problem solved.

I think bitcoin would be better off with a very low price but a very high value.  Moving the decimal place would help that.

so far ,alt currencys have all been crap

the usual complaint about bitcoin is that the first miners got hundreds of thousands of coins almost for free
(milllions of dollars worth at todays rates )

people making alt currencys want to get that position for themselves but i dont think it will happen

Yes the block chain still generates fine. But people may start feeling insecure if it's getting a lot easier for some entity to launch 51% attack. Thus the network would be under pressure to raise transaction fees to increase hashrate.

Even if people don't think 51% attack would become real, I think if there is a competing altercurrency gaining hashrate users may start migrating to that currency so there is going to be market pressure for bitcoin to raise transaction fee so as to maintain its lead in hashrate.

And when it does, difficulty will go down to match it. No worries.

Just to make sure it is clear how it works.  If there is more network hashing power than is supported by the current level of the block reward compensation plus the transaction fees, then that excess capacity will go away (i.e., unprofitable miners will stop mining), and an equilibrium will be found.

Transaction fees may need to go up in the long run to maintain network hashing power.
But micropayment is a non-issue since if there is a need to reduce fees for micropayment then there would be micropayment processors forming in the marketplace. You don't have to do micropayments within bitcoin network natively.

You can do that now...  mBTC or even µBTC. Makes me feel better about my balance.

Seriously, to those of you who think "21 million isn't enough", or "micropayments will be too expensive", I suggest you start thinking in terms of µBTC1 as equal to $1. That puts (unless I've miscounted somewhere) the satoshi at parity with the US cent, and since transaction fees are adjustable down to a single satoshi, and completely voluntary, there's nothing to worry about.

FYI, at the protocol level, there are only Satoshis, no bitcoins. So moving the decimal place is a simple client-side update. You'll confuse other people, but it's your choice if you want your client to tell you that you have 10 times more than you really do.

This is exactly how credit cards work.

Someones stores get screwed because people pay for gum with a credit card.

Some stores make it so you have to spend over $5 or more USD in order to use a CC/DC to purchase, so that the processing fee doesn't eat as much into their profit.

In my opinion, it is a non-issue.

Well if we ignore transaction fees (and they are very small). 

50 BTC per block * 6 blocks per hour * 24 hours per day * 365 days = ~2.63M BTC @ ~$9 ea = ~$24 mil annually.

Any idea what is the current revenue for miners?

I get what you are saying I just happen to disagree with you.  At $0.02 per tx and VISA sized network we are talking about $2.5B in revenue for miners.  More than enough to support the costs of the network.  CC fees are high not because of tx volume they are high because there is a cartel of players who artificially keep the fees high.  There is free market because of artificial barriers to entry.  In Bitcoin mining no such artificial barriers exist.  If fees are too high more miners will join, and if fees are too low miners will leave.


They will.   Currently fees are so low and block subsidy so high it distorts the risk vs reward dynamic.  


That is false.  You don't need an unpruned blockchain in order to mine.  There is no security risk from using a pruned blockchain.  Pruned transactions are "dead ends" they will never (not in a year not in a trillion years) ever be involved in a future transactions.  The merkle tree structure allows trust-less verification of pruned blocks.

+1. Moving the decimal place doesn't change anything that isn't already easily handled by people familiar with the metric system. And it would seriously futz with the data types in the client.

allten both of your posts are false.  No reason to jump from one fallacy to another.  Can Bitcoin be free forever and for 100% of transactions?  No.  Obviously not the network has a real cost.  The idea that Bitcoin has to be expensive is equally wrong.

Two examples:

PayPal scale network:
The Bitcoin network reaches 50 tps (roughly PayPal sized); that is roughly 1.58 billion transactions per year.  At an avg tx fee of $0.01 USD that would generate ~$16 million in revenue for miners.   Now obviously it will take some time for Bitcoin to grow to this level so the established history means that Bitcoin will be seen as less risky relative to today.  Risk determines return on capital.  Lets assume miners are willing to accept a 10% ROI and that hardware costs makes up 80% of lifetime mining costs.   That puts the network hardware cost at ~$200 million.   Remember it is the cost of the network (not hashrate) that determines security.

VISA scale network:
The Bitcoin network reaches 4000 tps (roughly VISA sized); that is roughly 130 billion transactions per year.  At an avg tx fee of $0.02 USD (higher utility given Bitcoin is now as popular as VISA) that would generate ~$2.5 billion in revenue for miners.   Now obviously it will take some time for Bitcoin to grow to this level so the established history means that Bitcoin will be seen as less risky relative to today.  Risk determines return on capital.  Lets assume miners are willing to accept a 7% ROI and that hardware costs makes up 80% of lifetime mining costs.   That puts the network hardware cost at ~$30 billion.   Remember it is the cost of the network (not hashrate) that determines security.


While $0.01 or $0.02 isn't "free" (remember that was just the hypothetical average cost, some tx would still be free) it is hardly "high cost".  The idea that Bitcoin can only scale under high transaction cost is simply FUD.

Lets compare that to some other payment networks:
Money Order - $0.50 to $1.00
Cashier's Check - $5.00 to $10.00
GreenDot MoneyPak - $5.00 (instant, limited deposit options max of $500)
ACH - ~$0.20 per tx (3-5 days)
Bank Wire - ~$10.00 (4-5 hours)

WU - 4% to 10% (10 to 30 minutes)
Credit Cards - $0.30 + 3% (seconds  chargeback risk for 120 days)
Credit Cards (micropayment) - $0.05 + 5% (seconds  chargeback risk for 120 days)
mPESA - 5% to 15% (instant)

Bitcoin - $0.00 to $0.02 per tx (instant to hours depending on risk profile)

Yup Bitcoin is horribly expensive at $0.00 to $0.02 per tx.

Your right about the way bitcoin was advertised: literally free transaction fees. That has really annoyed me the more I learn about the technical aspects of bitcoin. It was not designed to be scalable or cheap and I think that is ok because services can be built on top of it that are scalable and are cheap. Stephen Gornick did a great job at pointing this out.

Bitcoin is high power money!

Is it possible that bitcoin will become unaffordable to use for micro payments?

Yes! and it should be.

Remember, it's tx fees that are supposed to carry the mining operation forward in the years to come.
If everyone expects the tx fess to be dirt cheap the the original design of bitcoin will have to be abandoned in one form or another.
Also, Bitcoin isn't scalable and I believe it shouldn't be. Bitcoin should be considered as "high power money" not a payment processor for the masses.
a micro payment system needs to be built on top of bitcoin, but not change it.

Open yes.  Zero fees.  No.

Though to most retailers, for instance, who pay $0.15 per-transactions plus 3%, a fee in the range of a U.S. penny is the same as free.

Now, if there is someone who really wishes to support the ability for zero-fee transations make sure your miner has a connection to the node listed in the Free Transaction Relay Policy wiki page:
 - http://en.bitcoin.it/wiki/Free_transaction_relay_policy

With that, any transaction -- fee paid or not, microtransaction or not, can be included in the next block that miner solves.

if free transactions are not processed people will pay a small fee. Possibly the cost of tx will make bitcoin transactions less desired, but otoh perhaps it will raise demand as people need coins to pay the fees.

15M plus whatever coins are worth in that scenario times the subsidy.

I always include a fee to help the network but i have tested it without a fee and it still works

If we want bitcoin to succeed we shouldnt try and dodge and fee but miners NEED bitcoin to succeed

Its still a long way off being able to refuse transactions tat dont have a favorable fee IMO but of course

the ones with the nice fee will be picked up ,usually inside 10 minutes

I believe miners can chose which trasnactions to mine, so I'm sure some miners wont care and some miners aim for the higher transaction fees. The spam prevention is up to the miners but the official client helps with that spam prevention by hardcoding an annoying message to include a tx fee

Miners dont have much choice to refuse tranactions yet if they want the population  to use bitcoin ,nobody is going to wait a few days on transaction

(if transactions are not processed ,all the coins will be worth zero )

I would say "low transaction fees" although remember it is very likely free tx will remain viable for a long time.  It won't be realtime but if you are willing to wait hours or days I am sure someone will include it for free.  Still low/free can be very very low.  If Bitcoin scaled to PayPal sized network (~50tps) that would be roughly 1.5 billions transactions per year.  If avg fee was $0.01 USD it would be the cheapest payment platform by far and would still generate ~$15 mil a year to fund miners and protect the network.

It bothers me to read bitcoin propaganda that harps strongly on the "no transaction fees" bit.

Hasn't Gavin said that bitcoin is not supposed to be particularly useful for micropayments in the long run?

It wouldn't require a hard fork.  Just make the change on the client front-ends.  1 BTC = 10 BTC on the client-side display.  No one would know the difference besides the programmers.

Dropping the fee is trivial.  It has been done once already.  It is simply changing some lines of code in the client as fee are enforced at the node level and nodes don't have to agree.  There is no issue of compatibility or causing a fork.  If you wanted to you could publish a "fork" of the client which has a 0.00001 BTC mandatory fee and it would work just fine on the network (albeit other nodes might drop your txs).


When 1 satoshi is >$0.05 (2012 dollars) then a fork may be needed still I doubt even then the 21M will be expanded to 210M to 2.1B.  Just make 1 satoshi 1E-12 instead of 1E-8 and keep the same number of BTC.  Then again if 1 satoshi = $0.05 then the Bitcoin money supply is on the order of $105 trillion and it is the large economy in the world and likely the reserve currency of the central banks for the few remaining fiat.  Worrying about how to solve that problem seems like putting the cart before the horse.

Maybe that hard fork won't be implemented until the transaction fee of 1satoshi is too much then.  With bitcoin value set to rise with it being deflationary people are going to be dealing in smaller and smaller amounts of coins so the Tx fee will have to drop and keep dropping as long as bitcoin remains to grow.  Or bitcoin market-cap will stop growing. 

You keep bring up this moving the zero nonsense.  It would change anything also it will NEVER (I don't mean unlikely I mean NEVER) happen as it is a hard fork on a unnecessary change.

Bitcoin can handle micro transactions just fine.  The scenarios describe above are talking about years (maybe decades) down the line.  Even if more advances services are built on top the platform is still open and that allows cross compatibility and open free markets for those "gardens".  Hardly walled, more like picket fences.

So bitcoin for micro-transactions in the future is a no go unless you stay within a walled garden.  I thought bitcoin was all about being open and zero transaction fees.  Those are two of its biggest sellers.  Cut them out too early and slow down bitcoin adoption.  The whole bitcoin market capitalisation is still under $100Million yet they sound expensive to buy at nearly $10 each and too expensive to transfer for frequent open micro transactions.  Revalue bitcoin once it's spent twelve months plateaued over $10 to 210,000,000 coins from 21,000,000 and revalue everyone's wallet by *10 keeping the transaction fee at BTC0.0005.  Plus keep doing that until the bitcoin market-cap is worth trillions of dollars and they are mainstream worldwide.  So take this bitcoins are worth $100 each and a transfer costs $0.50 yet the whole market-cap is only $1Billion.  Sounds like bitcoin would be getting too big for its boots when you look at how many dollars are in circulation in bills and coins.  Plus those in circulation as bills and coins are only the tip of the ice-berg.

Exactly.  We already have examples of these p2p micro-banks.   Most of us already use them without knowing it.  Mt. Gox is a p2p micro-bank that offers the ability to transact micro-payments with no fee, using their redeemable codes (denominated in BTCs, USDs, EURs, GBPs, etc.).   These redeemable codes / vouchers / coupons etc are today offered by many exchanges (Bitstamp. CryptoXChange, Bitcoin-24, and more) and now some eWallets are introducing them (e.g. EasyWallet.org) so there is already a method for a service to use bitcoin for micropayments where there are either no fees or the frees are ridiculously low (compared to any other alternatives).

Additonally, each service that has its own EWallet is internally able to support bitcoin-denominated micropayments without a fee for its users.  Ogrr.com is a great example of how this is done today.

Also, each micropayment doesn't need to be an individual transaction on the blockchain.  The transaction fee can be shared among multiple transactions, lowering the per-transaction code.  This is how the Bitcoin Faucet works, for instance, when it hands out its pennies-worth of transactions throughout the day.

So eventually it probably won't be feasible for microtransactions themselves to cross the blockchain.  Once that day arrives there are already alternative approaches that work nearly as well.  In the meantime, there are services like SatoshiDICE and others that have the opportunity to take advantage of this period where ultra low cost microtransactions that cross the blockchain are possible, and build a customer base while following a strategy that is compatible once the day that the "nearly free, ultra low cost" feature goes away.

Though there is a wide gap between fees that are "free or nearly free" and those that are "multiples of pennies" per transaction, the difference means little to most every normal commercial and P2P transaction so this issue is isolated solely to the use of microtransactions that themselves are amounts measured in pennies.

DOH.  Don't post while sleeping.  Your right. I should have said "I can send you a bitcent (0.01 BTC)".

Of course that treshold (0.01 BTC) is adjustable and if BTC were to rise to a significant value (i.e. $100 USD:BTC) it could be lowered without risking the network.

Yes, and it's not a problem. The security requirements for micro-transactions are not as high, so using payment processors or p2p micro-banks would be fine.

In fact, it has the advantage of being much faster. Your savings would still be safely in your own hands, and if tx volume at any given time is small compared to the cost to gain peoples' trust, micro-transaction fraud yields no profit.

Really? Don't dust outputs require a certain minimum fee regardless of priority?

When the block reward is cut to 6.25 or 3.15 the miners will figure out which transactions get priority......quite quickly id imagine 

I thought of maybe shifting the decimal place so we would end up with 210,000,000 bitcoins instead of 21,000,000 bitcoins and keeping the transaction fee at BTC0.0005  So everyone's wallet goes up by *10.  Then we can still easily deal with whole integers called a bitcoin instead of a milliBitcoins once the BTC/EUR/USD value plateaus over $100 a coin.

Right now, the fees paid by SatoshiDICE are not attractive enough to all the miners as some are now excluding those transactions.  The miners gain greater value by getting smaller blocks to propagate across the network a little quicker than the value the fees from those transactions brings.

So essentially, it is already happening that a fee of a half a cent isn't always enough to cause a miner want to include the transaction.  

When the amount of transactions grows even higher, the upper limit to the per-block data size will start to get hit.  The only prioritization method built into the system is the price paid for the transaction fee.

So bitcoin wasn't architected to be a micropayments system (defining micropayments perhaps as a dollar or less of value).  I wouldn't build on the presumption that fees will always be below some threshold, like maybe a few cents.

oh wow.  that's pretty snazzy.. answers my question!  I felt like the fees would get higher and higher as the years go by and eventually I would feel like Paypal with my mining rigs dealing transactions..

thx guys!

1) The 0.0005 BTC fee is a spam prevention mechanism.  You could use Bitcoin everyday for your entire life and never pay a fee.

2) The spam prevention fee only applies to low priority transaction.  Period.  If your transaction is high priority then there is no fee.  I could send you a single satoshi right now without a fee.  The input not output size/age is what determines priority.

3) The spam prevention fee can be adjusted.  It was adjusted downward from 0.01 BTC when Bitcoin went above $20 USD.  If necessary it could lowered again to sub cent range as BTC continues to gain value.  I would say at >$100 USD:BTC it might be worthwhile to consider lowering the spam prevention fee to 0.0001 BTC but not until then.    Someday the spam prevention fee may be a single satoshi

So ... no.

The fee can be changed.  It was already changed from 0.01 BTC to 0.0005 BTC in 2011, it can be changed again.

Regarding transaction fees, right now the fee is pretty reasonable at the price of a bitcoin...  But if transaction fees go up (more sites like SatoshiDice flooding the network?) and the price of a bitcoin goes up (eventually, to like $100+ theoretically), then the transaction fees might get really high.  To send a BTC right now, you might pay like 0.0005 BTC I think, doesn't seem like much, but what if you wanted to send someone a really small integer, like 0.0005, you would pay like double in fees.  If a BTC is trading for $1000, that makes 0.0005 tx fee cost $0.50...  That's not much either, but you pay that on tiny transactions.  Sometimes I go to send a BTC, and I guess if there are multiple small bitcoin addresses that it needs to send from, and I pay a 0.01 fee.  If the network gets very congested and the BTC price goes up (like it's destined to do), won't it be unaffordable to send small amounts of money?