Indeed, I didn't look into the numbers, but I have the impression that analysis only considered the high-frequency part of the volatility (say, daily or hourly or so).  As you correctly point out, the low frequency volatility (weeks, months, years) of most big fiat is pretty small, and that of bitcoin is still huge.
For instance, over its 17 years of existance, the EURO and the USD never deviated more than something like 50%, going from an exchange rate of 0.8 to 1.5 or so.

Over its 8 years of existence, bitcoin did 6 orders of magnitude.  Of course, that won't happen any more (1 billion a coin I'm hodling in case ).  But in 2 years time, same vol, don't think so.  It doesn't contain any regulating mechanism !

I think it is a misconception that authorities hate bitcoin.  In fact, it allows them far more control over finances than has ever been, because of the transparent block chain, and the fact that they can apply strict AML/KYC rules on exchanges, knowing the inputs and outputs of the chain.

Moreover, bitcoin being essentially a speculative asset, I think banks are gonna LOVE this once they are allowed to speculate on it.

Bitcoin moreover being a collectible, somewhat similar to gold, central banks will be able to steer entirely its market, like they do with gold, as they can corner it, and dump it at essentially no cost (just the cost of printing fiat to buy it, keeping it as a reserve asset like they do with gold).

I think that if ECDS is broken, we have more worries than bitcoin.  Essentially, everything which is based upon it, which is A LOT, is broken.

Bitcoin has an accidental protection of addresses that were never spend, by the fact that an address is a hash of a public key.  However, as pointed out, that protection is gone from the moment that an address is used more than once, and I consider this as a kind of design error in bitcoin to ALLOW for more than one usage of an address (in the same way that double *spending* is impossible on bitcoin, one could have made double crediting impossible - with each address, there would only have been one possible UTXO, and hence could only be spent once too).

That said, it is not too late, and people owning coins could at any moment, before ECDS is broken, decide to strictly adhere themselves to such single-spend policy, by transacting all their coins to new addresses, of which they will never reuse anything.  For that, however, they should also avoid people to credit their same addresses multiple times, and because that's not forbidden in the bitcoin protocol, that can always happen.

Well, call me conservative, but I'm not buying the idea that whole international pans of economists are in a kind of conspiracy or group think erroneously putting aside a former correct theory of money to replace it with gobbledegook.  Economy is a complex science, because it combines different fields and sociological aspects, but I don't think that world-wide scolars are deluded to the point of defending totally erroneous views on things.  Our state of knowledge advances of course, and what was once thought to be more or less right, turns out to be more subtle than imagined, but I won't just say that most of late 20th century monetary theory is bogus, and that some ideas from the beginning of the century are "obviously" closer to the truth.  In fact, the "classical" story of how money came to be (namely, a generally used commodity that was more and more used as an intermediate good of exchange) is historically proven *entirely wrong*.  Nowhere, ever, a generally used commodity became money.  This is nevertheless on what older theories, such as sound money doctrine, are based on.

And the "this time it is different" story, I don't buy either.  Bitcoin is a monetary asset, just like any other monetary asset, that is, not of direct utility, but only of value to be exchanged for something of value.  It doesn't fall outside of the scope of any theory regarding monetary assets, not more than most other things of which the direct utility is small or non existent compared to its exchange value.  What sets bitcoin apart is its form.  Not its nature.

There is not much difference in principle between private money (as it has existed often in the 19th century) and bitcoin.  The issuers of the private money are the miner pools ; for the moment they stick to a pre-determined emission curve (the famous 21 million coins) - in principle they could change that any moment.  And there is a public auditing system of their ledger.  So the concept is quite well known ; the form is new.


Well, that's a very radical way of "regulating velocity".  You can hardly call that a monetary policy.  It is an emergency (panic) decision to do so, to try to block a run-away condition.  Things like what happened in Greece at a certain point, was just a measure to avoid Greek banks which were essentially bankrupt, to collapse, the time needed to bring the Greek government back to reason, or to have it decide to step out, so as not to have the whole of Europe pay for Greeks' jokes.


In the modern fiat system, a bank run doesn't exist any more, because there's no fractional reserve banking any more.  Banks are 1-1 covered in principle by non-monetary assets, which they can use to obtain as much central bank money as they want - the only thing is that it costs them interest, and hence they only keep a small reserve handy ; but normally banks can now obtain as much central bank money as their assets allow them.   It is because, exactly, this was locked for the Greek banks, that they had to restrict money movements at a certain point.  But a healthy bank cannot undergo a bank run any more in the modern system, because it is fully covered.


Yes, but they don't control velocity directly.  By setting interest rates, they influence the desire to exchange non-monetary assets for new money, so essentially, indirectly the monetary mass, because they change the incentive for people to want to borrow (= exchanging non-monetary assets for freshly invented money) and of course, this also has an indirect effect on velocity.


In fact, it is difficult to talk about velocity for a speculative asset, because it doesn't buy goods and services, but it exchanges in and out with other monetary assets, so there is no "Q" that has any meaning.  If I exchange 1 bitcoin once a day for fiat, and back to bitcoin, my contribution to Q is 2 x $1700, and my contribution to the number of coins exchanged is 2 (two trades).   If I exchange this coin 100 times a day, my contribution to the number of coins (velocity) is 200, but my contribution to Q is also 200 x $1700.  So this doesn't modify anything.  Fisher's formula had as an idea that Q was some genuine "irreversible" acquisition of goods and services.  It is still mathematically valid for exchanges, but as Q augments with V, it doesn't change anything.

The deflationary spiral resides rather in the fact that people hoard (even if they quickly trade in and out) coins for the rise of its value, and not because they use it as an intermediate between earning and spending value.  In other words, the deflationary spiral is the monetary equivalent of the speculative bubble run-away but seen from the other side.

Bitcoin's price is not pushed upward by the demand for bitcoin to be used as a currency.  It is mainly pushed by the desire to hold bitcoin when it rises, with the idea to make benefit over it, by "buying low and selling high".  Not by earning for goods and services (say, "salary") and by spending it on goods and services (say, buying stuff).  This happens, I don't deny this, but it is not the demand for salary that pushes the price of bitcoin.  It is the hope for higher prices that makes one buy bitcoins, not the need to use it as a currency. 
If there had been an upper cap on the value of a coin, then this speculative nature wouldn't be there, and bitcoin would *essentially* be used as a currency, because there's no hope for "rise".  Yes, there is volatility, and normal speculation as a side effect, but it wouldn't make its market cap.

"Centralized authority, please save us !"

The economic invention of the century would regulate itself, and wouldn't need an intervention from the board of gouvernors, don't you think ?

This is an interesting article, but I would like to see the volatility definition he's using.  Of course bitcoin cannot be that small circle.  Maybe he's only taking, say, 24 hr volatility or something, and not the entire RMS of the whole curve, in other words, only the high-frequency spectrum of the price curve ?

Clearly, bitcoin's dollar price doesn't look like a FLAT LINE with some noise on it at all, which is what low long-term volatility would mean.  

The point is that it is economically well-known that collectibles have no stable price, which was the problem with the booms and busts of gold-based currencies (which was essentially "solved" by fractional reserve banking, putting again elasticity in the monetary offer).

Fisher's formula is essentially tautological.  It goes like this:
Q = the amount of economic goods that were bought with the currency, in a unit of value (say, a Big Mac, but it doesn't matter) in a given period
P = the price in units of our currency, of that unit of value (presumed stable enough during the period)

Clearly, Q . P = the amount of our currency that went over the counter to buy Q

M is the total amount of coins of our currency in existence (presumed fixed during the period)
We split that amount in different "kinds": m0 ; m1 ; m2 ; m3 .... mn so that their sum is M.

m0 are the coins that didn't move during this period (were held)

m1 are the coins that were spend once during this period

m2 are the coins that were spend twice: from A to B, and from B to C.

m3 are the coins that were spend three times

...

mn are the coins that were spend n times.

Clearly, the total amount of spendings, is:

m1 + 2 m2 + 3 m3 + ... + n mn

The total amount of spendings must be equal to the total amount of coins that went over the counter:

P . Q = m1 + 2 m2 + 3 m3 + .... + n mn

Now, define V = m1/M + 2 m2/M + 3 m3/M + ... n mn/M
which is the weighted average number of times a coin was spend during this period.

Then we have: P . Q = M . V


You shouldn't consider my critique of bitcoin as a critique of the idea to have a free currency.   My critique of bitcoin considers the bad design of it, not its pretended reason of existence.

M is monetary mass: the amount of coins in circulation.

That was explicitly pointed out to Satoshi in 2010, and he waved that away, when he introduced the 1 MB limit.

It comes down to the users, if they can refrain from transacting themselves the time needed to put miners sufficiently under pressure that some start hardforking, and then have the market decide.

As long as the miners stay on their single chain, nobody can do anything else but follow them (and be able to transact) or refuse to use their chain (and be locked out of bitcoin).

If miners fork because users keep themselves locked out of bitcoin for a long time, then users can vote with their money on the two chains that will appear, and we are in the ETC/ETH scenario.

And, again, in all of this, Joe's full node in his basement means zilch.

From what distribution were the times between discoveries of the same pool drawn ?  I have the impression it are UNIFORM random distributions, and not EXPONENTIAL distributions.  That's crucial.  Uniform distributions, as I said before, increase the probability of winning if you have already worked a lot without success.

The question simply was: will B find solutions to blocks at any other rate than A ?

The answer is that B will find solutions to blocks AT EXACTLY THE SAME AVERAGE RATE than A, but it would have been great if you saw this yourself.

How so ?

Suppose all miner pools agree to keep the old protocol.  Suppose most non-mining nodes upgrade to the new protocol.  These nodes don't see a block chain that is according to their rules, so they stop.  The miners continue to make the old chain and that's the only chain that is around.  What happens next ?

What will USERS do ?  If users connect to one of these new full nodes, they don't see any transactions, and they cannot transact.  Their wallet simply doesn't see new stuff coming.  If these users connect to one of the miner nodes, they can.  They see their transactions, they can transact, like in the old days.  The block chain is still growing, it is simply not copied on the non-mining nodes who refuse to do so.

Exchanges, what will they do ?  Upgrade their node to the new protocol, and have no deposits, no withdrawals ?  Or keep the old software, and sync with the sole block chain that is out there, and allow their customers to deposit and withdraw ?

I have argued this many times, but non-mining full nodes can impose ZILCH on miners.  Users can vote with their money, but still need to be able to transact to vote with their money, so still need to use the miner's block chain.

Suppose that users would like the new protocol too, but miners keep on making the old chain.  What can they do ?  Refuse to transact ?  Refuse to access their funds ?  Sure.  So what ?  

What are the possible scenarios when miners remain united on the old protocol and still only make one block chain, while by far most of the full nodes have switched to the new protocol ?

=> if some exchanges, and some users, still want to use bitcoin, they have no choice but to revert to the old node software and/or connect directly to miner nodes.  All the others cannot transact.  The bitcoin volume will go down, because only determined by those users and exchanges that accept to use the old protocol.  Price may go up or down, depending on offer and demand in this restricted market.  All the others exclude themselves from bitcoin transactions.  Miners won.

=> if all exchanges and all users still use bitcoin as before, they will connect directly to miner nodes.  Everything like normal.  Nobody gives a shit about all the stopped full nodes.  It is as if they left the network.  Miners won.

=> if no exchange, and no user, wants to use bitcoin the way it was.  Nothing happens on bitcoin.  No transactions.  Empty blocks, with just coinbase transactions on the old chain.  At a certain point, miners will want to sell their coins, but they can't: no exchange sees their block chain or their coinbase transactions.  --> ONLY HERE the true power game starts.

What will happen first ?
A) some miner pool accepts to switch to the new protocol, and mines slowly on his own, the new chain.  Users can finally transact again on the new chain but blocks are rare.  We have a hard fork.

Aa) exchanges see the opportunity in having two bitcoin, and start an old node and a new one, and allow people to exchange old bitcoins for new bitcoins and vice versa, like with ETC/ETH.

The market will determine the relative value of the new and the old bitcoin, and miners will follow ==> we have now two bitcoin, according to the market forces, just like with ETC/ETH.

Ab) the exchanges don't consider the old chain, and keep with the new chain only.  Sooner or later, the miners will have to give in, to sell their coins.  One after the other switch to the new protocol.  ==> new protocol in bitcoin

B) some users accept to buy OTC coins from miners at reduced price.  Miners can continue mining on the old chain, get paid less per coin, and some users think they are getting real cheap bitcoin.  

Ba) more and more users start to buy OTC "cheap bitcoin" and exchanges give in to their lost market share -> back to old bitcoin, miners won

Bb) miners stop considering bitcoin because they are mining at loss, the block chain stops.  -> Bitcoin is over.


==> in all this, we see that miners can be forced out of their mutual agreement to continue the old chain ONLY if simultaneously:

1) no exchanges and no USERS are willing to deal any more with the old chain, resisting all temptation to want to transact
2) some miners start hardforking
3) depending on exchanges, the hard fork leads to two bitcoins, or only to the new one.

In this whole story, the full nodes didn't matter.  What mattered was the attitude of the USERS (could they resist wanting to transact), the EXCHANGES (could they resist winning fees on transactions) and the MINERS (could they resist forking to sell their coins quickly).

Indeed, it is somewhat delusionally to think that if some asset in a liquid market has high volatility as expressed in a major fiat, such as the US dollar, it is the dollar that is fluctuating and not the asset.  In the relatively short term, big fiat are very stable units of account (unless something quite spectacular happens) in the sense that a quantity of this fiat represents a constant market value of commodities, whether it is bread, computers, cars, bananas, coffee, theatre tickets, or... other stable currencies.

Of course, and that was I think the essence of cryptoanarchists' remark, this is somewhat self-referential.  But in the end, it doesn't matter.  If you can buy about the same amount of Big Macs, bread, tooth brush, cars, bananas etc... with, directly, or after exchanging it for another currency, it HAS a stable value.  Because in the end, value, that is an amount of Big Macs, bread, tooth brush, cars and bananas.

In the very long run, fiat currencies are less good units of account ; but then, not much is, because the notion itself of "same value" in remote and totally different economies is ill defined.  How do you compare the value of, say, a dagger in the 7th century with something today ?

But a thing even used as a major currency, that doesn't have a price-stabilizing mechanism, will NEVER stabilize, simply because of the variable economic activity, and the variable velocity of money in Fisher's formula:

Q.P = M . V

From which: P = M . V / Q, the price level depends on V and on Q.   V depends heavily on the hoarding habits of people, and Q depends on the economic activity.  If M is a hard number, it will never be stable.  Especially if a large part of M is being hoarded, and V is very sensitive to the small amount of non-hoarded coins.

If there's a feedback mechanism from P to M, which is what central banks do, then this can stabilize P.  If no such mechanism is known, P will be very dependent on V and Q.

I wholeheartedly agree with you.  In fact, the limited emission of bitcoins to a finite number, creates many fundamental problems, macro-economical, but also technical/economical.

My macro-economical complaint with bitcoin is that it is designed to be highly speculative, will never stabilize and will hence never be a unit of account, which will always make it a marginal means of payment at best, and a gambler's/speculator's/investor's toy dominantly.

But its reducing emission causes a lot of problems on lower levels too.  The whole fee market battle by miners comes about because their block rewards will dry up in the future.  Who says fee rewards, says scarce transactions, and hence artificially obstructed transaction system.  The reduced block rewards have also pushed the competition in mining to the extreme.

An emission curve that emits less assets when more users come in, gives HUGE seigniorage advantage to early adopters, and pushes speculation to unseen heights.  Even if the idea was to come to a limited amount of coins, the largest emissions shouldn't have happened when nobody was around.  This turns bitcoin into rare paintings of which a few early collectors have HUGE amounts, leading to a very unequal distribution of coins (not by the normal functioning of a market but by early bird adopters), which give rise to big whales, which can then play the market like they wish with their significant stash (read: pump 'n dump, corner, crash, ....).

The reducing block reward will also make the security by PoW *relatively speaking* less and less secure.  After all, the PoW securization is financed by the rewards, which reduce.  An attack has thus a cost, in relationship to the total amount of reward per block, which is decreasing in BTC.  It can be increasing in $$ of course, but in BTC, it is decreasing.  The *incentive* for an attack, however, will remain formulated in BTC: it is the ability to undo a transaction a few confirmations ago.  As such, the ratio between attack cost and attack incentive, is decreasing with effective block reward decrease.

I hope you understand that the probability of winning a block doesn't depend on what block you are mining, or how long you were mining on that block.  Each hash you calculate, on each thinkable block, has exactly the same probability to "win the block" as any other.  I hope you understand that.

You should first answer this:

Of course it is going to take on average 1 hour and 40 minutes.   That's really so trivially true that I can't wrap my mind around you not understanding that, unless you know ZILCH about probabilities.

If you have one chance in 1000 to win each time you play, and you have a "hash rate" of 100, meaning, you can play 100 times per second, it will take you on average 10 seconds to win ; if you play for, say, 5000 seconds, you will have won about 500 times.

If you play with 9 other peers in the same way, and all of you have a "hash rate" of 100, meaning, ALL of you play 100 times per second, it will take each of you on average 10 seconds to win, like you ; but every second, on average, SOMEONE will win because all 10 of you won on average once in 10 seconds.

So if all of you play 5000 seconds, each of you will have won about 500 times, and in total, you will all have won together, 5000 times of which, you have one out of 10.  If the 9 others stop playing, this will not influence your winning rate, which is 500 times in 5000 seconds, but the 4500 other winnings by the others are of course not there, because they didn't play.

That's so elementary trivially true, that I really don't see how you can't get that.

Indeed, the true centralization of crypto in general is linked to who can decide about the protocol.  MOST crypto is heavily centralized in this respect, because most crypto has a lead developer that decides about future evolution of the protocol.  We saw this clearly with ethereum where the unthinkable in crypto happened: modifying the past, not by modifying the block chain data, but by modifying the protocol that tells you how to interpret that block chain data.  Many other coins have such single dev team: DASH with Evans, LTC with Lee, of course ETH with Vitalik, ...
If you can change the code, you can change all of it, no matter all the "crypto".

In fact, bitcoin is decentralizing by not letting Core just do what it wants on this side ; but it is centralizing in the mining industry.

The point, of course, is that if there's a limited power structure in crypto (that is, if the deciders can sit in one room and come to an agreement, like they did on litecoin), it is a kind of "governors of the central bank".  Governors of the central bank also don't want to run the fiat they command, into the ground.

This is not neo-Keynesian at all, it is rather neo-classical, in Friedman's tradition.  Keynesians want to influence economic activity with monetary policy ; Friedman considers just the monetary value stability.  Keynesians think that one should print more money if the economy depresses, and that one should restrict monetary supply if the economy booms.  Friedman's monetary view is rather that the monetary supply should keep money's value as constant as possible:

https://en.wikipedia.org/wiki/Monetarism


You can hardly think of Friedman as a neo-Keynesian.

One should regulate the money supply so as to keep as stable a monetary value as possible, in the sense of Nash's ideal money.

https://en.wikipedia.org/wiki/Ideal_money


Nash proposes something of the kind in principle as the Big Mac index, but smarter.  The idea is to have *constant economic value* of the money unit even though that notion itself is difficult to define precisely.


He considered the Euro as a good candidate for asymptotically ideal money, exactly because the European central bank has only a single objective: inflation of 2%, and no neo-Keynesian targets to meet.


That is true, but velocity is something that is hard to control, is a function of people's habits, mood and so on, and is especially a function of the perception of whether an asset is speculative or not.  When you look at monetary velocity, it is not something that has uniform behaviour, and remains grossly within some boundaries.  You cannot "regulate" velocity.  You cannot make people spend faster or hoard more.  In fact, velocity is at the origin of two instabilities: the deflationary spiral, and hyper inflation.   The deflationary spiral happens when people speculate on the strongly rising value of a monetary asset: they hoard it more and more, lowering as such, the velocity, and hence increasing even more the market value of those few coins on the market, confirming the speculation of rise.  This is bitcoin's behaviour.  
On the other hand, hyper inflation is when people speculate on a strongly falling value of a monetary asset: they try to get rid of it as quickly as possible, increasing as such the velocity to very high values, and hence, decreasing even more the market value of the mass of coins chasing goods in the market, confirming the speculation of drop.  This is what has happened to some famous hyper inflations like the Reichsmark.

The knowledge of a stabilizing mechanism avoids both instabilities, but one doesn't have any handle on people's spending decisions which determine velocity ; as such, the only thing one has a handle on, is the coin emission.


But then it is not in a state to become a currency !  If the argument against why it is not behaving like a currency, is: "because it is not a currency yet", then that's not very convincing as an argument of why it is a good currency, no ?  In fact, the deflationary spiral is even worse for a non-essential asset, because in as much as the deflationary spiral of the principal currency is tempered because after all, you HAVE TO BUY FOOD, here, nothing stops one from hoarding all of bitcoin, because you don't have to spend it to get food.


Apart from ransomware and dark markets, I wonder what ?


You mean that someone was offering an exclusive bitcoin tooth brush for 0.1 BTC 2 years ago when it was around $200,- (so a brush for $20,-), that same tooth brush will still be sold for 0.1 BTC ($180.- right now) ?  I would hurry cashing out my BTC into fiat and buy 9 toothbrush in the supermarket with it, no ?


Absolutely.  I don't claim that "bitcoin is dead".  It has a bright future !  I think it will go sky high - but in the same way that complex derivatives could go sky high.  Bitcoin is a hugely speculative asset, it is BY FAR its principal usage.  Look at this forum.  Look at what people say about it.  Look at the volumes.  This is all speculation.  But speculation is a HUGE market.  Hell, it is MUCH MUCH bigger than the fiat market.  The whole financial world is the market.  Not the "payments" application of fiat.  It is a market that is about 10 times the world's economy all by itself.  This is the true nature of bitcoin and most crypto.  It is purely speculative, and once big finance will get into it, it will be HUGE.  But entirely speculative and if it is well done, totally unpredictable (efficient market hypothesis).  

In fact, I think crypto finally invented, what the financial world has been looking for: abstract speculative tokens.  No more need to speculate on derivatives linked in the real economy, like oil, real estate, food, metals, technology or whatever: just abstract tokens to gamble on.

That's not what one pretended, but that's how it was designed.


I think that this is again this fallacy of taking a desirable objective as an actual consequence.  It is not because "bitcoin would crash if there wasn't any merchant usage" that "merchant usage will emerge".  But moreover, I don't think that it is true, that bitcoin has what so ever to do with merchant usage.  Merchant usage was a story told in the beginning, to kickstart it, and there IS some marginal merchant usage, but it is not what bitcoin is about, nor what sets its price.  Pure speculation is.  And I think it has a monstrously bright future in that.

My current own understanding of bitcoin is the following: it has *economic* incentives that centralize it, into a separate industry of "block chain providers" (miner pools / miner hardware owners) and its customers (bitcoin users that want to transact).

The customers need the block chain providers to be able to use bitcoin (that is, to transact) ; while the chain providers sell the chain to the users who pay for it by buying the coins that these chain providers can obtain or invent (fees and block rewards).

This is not very different from a central bank + commercial banks providing payment services to their customers, but there ARE notable differences, nevertheless: there's no organized decision hierarchy amongst the industrials, and as such, these competing entities are for the moment locked in into an immutability of the protocol they use to make the block chain.  None of them, individually, can deviate from it ; and there's no global decider for them (for the moment).  There's no "board of governors" of the industry of block chain providers.

As a whole, this system works quite reliably.  The customers send their transactions to the industrials, who have some incentive to include it in the block chain they provide to the customers and it is almost impossible for any one to deviate from the established rules of functioning, without risking to lose a lot of money.  The customers use the transactions to obtain value, to speculate amongst themselves and so on, and this establishes a market price of the tokens they transact ; this market value is what the industrials can obtain from their customers by selling them back the coins they obtained from them or they were allowed to invent from thin air.

This split between industrials and customers is a basic design "feature" of bitcoin, which I consider a big mistake.  The other feature, which I also consider a big mistake in bitcoin is its emission curve which turns it into a speculative asset, and doesn't allow it to become a genuine currency.  But who am I to consider these features "mistakes" ?  They are only mistakes if we wanted to make a decentralized currency.  They are not a mistake if we wanted to make a highly speculative asset: in fact, the design for that is near perfect.

However, the distinction between "currency" and "speculative asset" plays a huge role if one were to change bitcoin's protocol.  If one doesn't change it, whatever the "original idea" was, doesn't matter, it will do whatever it will do (it IS a highly speculative asset and nothing indicates it won't stay that).  If we want to modify it, however, we would need to know where we are heading.

I think that if we want to turn it into a genuine currency, there's no point in twiddling the technical aspects of it, as long as the two basic aspects, industrials vs customers and emission curve, aren't modified to turn it into something that might look more like a decentralized currency: no more "industrials selling block chain", but users making their own block chain (say, something like PoS) ; and a flexible coin emission that stabilizes the coin's value.  But this is entirely against the religion of bitcoin.  So this is not possible. 

As long as that is not possible, bitcoin cannot turn into a genuine currency and will remain a speculative asset.

I haven't fully studied all aspects of segwit, but it contains a lot of very smart improvements in the technical details of transaction processing.  Purely on the technical side, segwit is a good improvement. 

However, segwit got entangled with something totally different, which is the LN.  The LN can be a smart idea too.  It is however, a totally different concept of transaction, with a totally different trust model. What has screwed up the whole story, is the fact that segwit proponents are LN proponents who seem to want to FORCE bitcoin usage onto the LN, and for that, OFF the main chain.  But that is fundamentally problematic in my eyes.  The fact that one uses a totally bogus argument, namely the "importance of the non mining full node in Joe's basement", while one ignores entirely the actual semi-centralized true power structure of bitcoin, which is industrials making block chain versus customers using the block chain and paying the miner's tokens, and *keeps a tight limit on on-chain transactions* on the basis of this bogus argument, sounds strange.

Because in order for the LN to be trustless, one needs to be able to settle easily.  If not, if ever the on chain settlements are *not in principle capable of settling the entire LN quickly at any moment*, the LN is like "fractional reserve banking" but not in coins, but in transactions.  In principle each one of us can settle, but we can't, all of us.  In a fractional reserve bank, each one of us can withdraw his money, but not all of us.

In other words, the LN is only trustless, if at any moment, it can settle FULLY on chain.  Like full banking: at any moment, EVERYBODY can withdraw their money.

I would be entirely favorable for the LN, if the possibility to settle, for everyone, at any moment, were possible.