Guan’s blog

Matt Levine has a post on the limitations of Bitcoin as a way of making payments. I think he makes some factual errors. For example, he states that “neither the buyer nor the seller normally pays any transaction fees” when making Bitcoin payments, citing the Bitcoin wiki page on transaction fees. Now, I’m not an expert on Bitcoin, or anything else for that matter, but I believe that 0.0001 BTC (1 mBTC, or millibitcoin), is, in practice, the minimum transaction fee, and that it would be hard to get your transaction processed for free these days.

He also says that the transaction fees that are actually paid are “for, essentially, odd-lot transactions”. Here’s the comment I left about this, which hopefully clarifies things a little:

Bitcoin transaction fees are, by default, a function of transaction size (in kilobytes), not transaction value (in bitcoins). The default appears to be 0.0001 BTC per thousand bytes, and that’s also the minimum.

Transaction size is only vaguely related to transaction value, and depends on the number of inputs (how many different addresses you’re funding the transaction with) and outputs (how many different addresses you’re sending too). If you have an address with a large number of coins—say, 1,000 BTC—and you're sending them all to a single recipient, then the transaction will be relatively small, and as far as I understand, the network will process it at a cost of 0.00001%.

On the other hand, if you have received a lot of small amounts over time and you want to consolidate them in a single address, or if you have a complicated script (rule for when the coins can be released), you could end up with a big transaction of small value, and a large percentage fee.

For example, here’s a fairly large transaction in block 278,331. I have no idea what the purpose of the transaction was, but it could have been consolidating coins from a few dozen addresses into a single address. The total value was 100 BTC, or about $78,000. The size of the transaction was 4,482 bytes, and the transaction fee was 0.0005 BTC, or 0.0005%.

The fee will typically be proportionately higher for small value transactions. In the same block, we have this transaction that paid 0.00011 BTC in fees. The value appears to be 0.07419827 BTC, but some of that was change, and I believe the actual value transferred was only 0.01109714 BTC, or $8.68. Since the fee was $0.09, the percentage transaction fee was a whopping 1%.

Now, this whole discussion was trying to compare Bitcoin to the payments infrastructure for US Dollars. I am not a huge Bitcoin booster and don’t think it will revolutionize everything, but I am a big critic of the USD payments infrastructure. I am biased towards thinking that anything would be better than what we have right now.

There are several different kinds of payments an individual might want to make:

1. Small point of sale and online store payments to merchants. Mostly handled by credit cards and cash.
2. Small payments to merchants, to pay bills. In the United States of America, done by check, recurring credit card transactions and ACH debits.
3. Small payments to other individuals in the same country. In the US, mostly done by check, or services such as PayPal.
4. Small payments to other individuals in a different country.

In my opinion type 1 and 2 payments basically work fine in the United States. It’s absurd how many checks are still mailed around the country, but the US financial sector has lowered the cost of check payments to close to the cost of a stamp, especially now that many checks are only transported once—from sender to recipient—and the recipient will scan the check in an online banking app, if he or she is an individual, or convert it into an ACH debit, if the recipient is a business.

Some of these payments are instant, or at least can be verified instantly, even if they are settled a few days later. Very few of them are irrevocable: a merchant receiving a credit card or ACH payment can never be entirely sure that there won’t be a chargeback later, with very high fees added on. That’s horrible for small businesses, but it makes things very easy and secure for consumers, and larger businesses can handle the risk.

It’s the US individual-to-individual payments market that I curse at more frequently. Here’s how it works in a moderately civilized country:

That’s my Danish online banking system. I put in the recipient’s routing number and account number, and amount, and click OK, and I can transfer up to 100,000 kroner (about $18,000) without interacting with a human or filling out a paper form. It’s free for me, and most people would pay no more than 1 kroner (about 18 cents) in transaction fees. Settlement takes one business day, with the money showing up before 6am, and can be withdrawn from an ATM or such immediately after settlement. Starting in 2014 there is a same day option.

(I can also make foreign payments this way, which either fall under SEPA rules and are very cheap, or are routed through SWIFT and are slightly less cheap. With foreign payments, the only additional piece of information I typically have to enter is the recipient’s name and address.)

Most European countries now settle this kind of payment several times a day, and a few settle it instantly. The payment is final and irrevocable, and you’d normally have to sue the recipient to get your money back. The fees are low enough that people would use it for very small payments, for example splitting a dinner check or small gambling debt.

Another benefit is that anyone can do this. The details vary from country to country (the UK is particularly bad in this area), but in practice, almost anyone can get a bank account, and so anyone can make payments to anyone else.

What are the options in the US?

1. Travel to the recipient’s home and hand over cash.
2. Mail a check.
3. Use your bank’s “bill pay” service.
4. Use an annoying service such as PayPal.
5. Use a slightly less annoying service such as Square Cash or Popmoney.

The first two options are obviously slow, but they are effective. Cash payments are irrevocable.

There is no equivalent to putting in a recipient’s ABA routing number and account number and using the existing banking infrastructure to pay the recipient using the ACH system, which has very low transaction fees. For credit unions that use the ItsMe247 online banking system, you can fake it (but you’d probably violate the Computer Fraud and Abuse Act and could end up in federal prison for 5 years) by doing the following: add the recipient as a “Person”, but put in your own email address. Then enter the recipient’s account information, pretending to be the recipient, and finally make the payment.

Of course giving your account number to a stranger can be dangerous in the US because of the way ACH is structured, where an account number is the only thing you need to steal someone’s money. (Don’t get me started on this.)

This kind of payment will typically take 2 or 3 business days. There is a system for same day ACH, but approximately zero financial institutions participate in it. And the payments are still not irrevocable.

What if you want to make an instant, irrevocable US dollar payment? You’d have to use Fedwire. The system transaction fee is less than a dollar, but most banks will charge you between $20 and $50 to do it. I don’t know of any US bank that allows individuals to do Fedwire transfers online, so you’d have to fill out a paper form and/or interact with a human being, who then has to input all your information into a computer. The payment is instant within Fedwire’s opening hours, but most US banks don’t have straight through processing for Fedwire on the receiving end, so you’d have to wait for a clerk at the receiving bank to manually credit the recipient’s account. If you’re lucky it could be done in an hour; most likely it will take a business day.

That leaves us with third party services such as PayPal. (Or really fifth party services, because the first four are the sender, the recipient, and their respective banks.) PayPal has a history of stealing people’s money without explanation. PayPal is incredibly annoying, at least in my personal experience: it can be very hard to get your account verified, it takes days to verify a bank account, and my US Social Security Number is rejected by most of these services because it has some kind of dirty foreigner flag attached to it. Many of these services will not let you mail documentation such as passports or social security cards to verify your identity: they will only accept the word of credit bureaus like Experian. It’s perfectly reasonable for a private company to reject me if I am too much hassle, but it doesn’t bode well for a payments system.

Square Cash is a lot better. I’ve been able to use it without any problems. Who knows what it will cost or what kind of hassle will be added when Square’s investors get tired of subsidizing it though.

I do think there’s a fundamental problem with these services. The recipient has to be signed up to actually get paid, so you end up having to maintain several different accounts simply to receive payments. Access is not guaranteed: these are private companies, so they can decide who they want to allow to become a customer, and a rejected customer can’t exactly appeal that decision. (You have the same problem with banks, but there are a lot of different banks and credit unions you could try, and in practice it’s very easy to get a bank account in the United States.) You are also subject to fairly arbitrary limits that you have little control or influence over.

If you want to make instant payments, most of these services, such as PayPal, only allow you to pay with USD balances that you’ve already settled with them, and it can take time for the recipient to get money out. This means that what you pay isn’t really a US Dollar in the traditional sense—cash or a checking account balance. Rather, you are paying PayPal balances.

There’s nothing wrong with innovation and competition in the payments market. Maybe you want to pay with your phone through text message. Maybe you want instant payments, and that’s not judged practical in an ACH like system because it would be harder to settle. Maybe you’d like to pay telepathically or telekinetically. There’s no way that the culturally conservative banking system is going to be first in those markets.

That doesn’t change the fact that the payments infrastructure for US Dollars—ACH in the United States of America—should provide a basic level of functionality for individuals.

Getting back to where we started, Bitcoin has obvious flaws that are well described by Matt Levine. But I would argue that US Dollars, while a good store of value and unit of account, are also terrible for making payments. Right now, for the typical American, it doesn’t do that: he or she cannot make payments to other typical Americans using the most common form of US Dollars, namely checking account balances.

Update: Shamir Karkal has a great reply with some corrections.

1. Tirole: The Theory of Corporate Finance
2. Campbell, Lo and MacKinlay: The Econometrics of Financial Markets
3. Cochrane: Asset Pricing
4. Björk: Asset Pricing in Continuous Time
5. Gollier: The Economics of Risk and Time
6. LeRoy and Werner: Principles of Financial Economics
7. Cramer: Jim Cramer’s Mad Money

I wrote a lengthy post in Danish yesterday following a Twitter discussion with Kenneth Praefke, a journalist who wrote what is probably the worst Danish article on economic policy in 2013 (not that I made a top 10 list or anything). The background is that since the 1980s, the Danish krone has been pegged in a fairly narrow band to the Deutsche mark, and to the euro since 1999. This means that unless the peg is abandoned, Denmark does not have the ability to carry out its own monetary policy.

Such a peg is not unusual for a small country. Personally, I don’t think it would be deep heterodoxy to suggest that the peg has advantages and disadvantages for Denmark. What has actually happened is strange: monetary policy has not been a factor in Danish economic policy for so long that Danish politicians, economists and reporters have completely forgotten that it exists, and that in most countries, it is a natural part of the overall economic policy.

Behold, now, Mr. Praefke’s article. The headline is: “Economists: Sweden is not an economic role model”. It starts with the premise that various Danish politicians have pointed to the fact that Sweden has weathered the 2007–? economic crisis better than Danmark, and suggested that perhaps we should learn something from the Swedes. Here’s a chart of Swedish and Danish real GDP (indexed to the 2000–5 average):

He then quotes two economists, Torben M. Andersen and Martin Madsen, who state that the Swedes have not carried out any significant structural reforms that can explain the difference. Labor market reforms in Sweden lag behind Denmark and top marginal tax rates are not lower. Torben M. Andersen goes so far as to say, “In my opinion there are no magic tricks that one could say they’ve pulled out of the hat in Sweden.” We don’t have a transcript of the interview, so it’s possible that it was very clear that the conversation was only about structural reforms, but that’s not the impression given in the preceding paragraphs.

The fact that Sweden has had an excellent monetary policy carried out by the world’s foremost monetary policy magician is not mentioned at all. On Twitter, Kenneth Praefke explains the omission by claiming that he set out to write an article on structural reforms, and good monetary policy is not a structural reform.

Praefke’s very successful attempt at ignoring the monetary elephant in the room is only a symptom of a syndrome of monetary policy amnesia in Denmark. It’s telling that the Danish politicians he quotes in the beginning would like to emulate Sweden’s handling of the crisis. They’re not stupid, so they probably thought about what the Swedes might have done better, and then they realized: we can’t talk about that.

While money has been entirely forgotten in Denmark, there is still some countercyclical fiscal policy. Most of the economic policy debate, however, deals with structural reforms. Reform, reform, reform. Sweden hasn’t reformed as rapidly, so they suck. All anyone can talk about in policy circles these days is the government’s new Productivity Commission, whose members are charged with boosting productivity. Meanwhile, the country is left with some excellent (though sometimes excessive) reforms and some odd articles about economics.

Jeg deltog i længere diskussion på Twitter, som er værd at læse i sin helhed. Baggrunden er en artikel af Kenneth Praefke, hvori han citerer to danske økonomer og diverse politikere.

Når man umiddelbart læser artiklen, skulle man tro at den handler om den økonomiske politik i henholdsvis Danmark og Sverige, og især om hvorvidt svensk økonomisk politik har hjulpet Sverige bedre gennem krisen. Overskriften er for eksempel “Sverige er ikke et økonomisk forbillede”, og den første sætning omtaler påstande om Sverige “som et økonomisk foregangsland”. Men det er ifølge Kenneth Praefke en forkert læsning af hans egen artikel.

»Jeg har lidt svært ved at se, hvad det er, man præcis mener, når man siger, at man skal gøre som svenskerne,« siger Torben M. Andersen.

»Der er efter min mening ikke nogen magiske trick, som man kan sige, at de har trukket op af hatten i Sverige,« siger han.

Politikerne har ellers igen og igen peget på de initiativer, der tages i Sverige.

Økonomi- og indenrigsminister Margrethe Vestager var tidligere i december i en række avisinterview ude med budskabet om, at Danmark skal lære af svenskerne, der har klaret sig bedre gennem krisen.

Hvis man spørger enhver der ikke er født og opvokset i Danmark hvilket magisk trick svenskerne har trukket op af hatten, og hvordan Sverige klarede sig bedre gennem den økonomiske krise, vil de typisk pege på pengepolitik, og på en konkret troldmand: Lars Svensson.

Det mest nyttige mål for hvor godt pengepolitikken er tilpasset konjunkturforholdene er udviklingen i nominel BNP, som er den samlede værdi af den økonomiske aktivitet i samfundet, i kroner og øre. Her er svensk nominel BNP fra 2006 til 2012 som blå linje (i milliarder svenske kroner per kvartal), hvor den røde linje er en trendlinje baseret på 2000-2006:

Krisen kan ses som et stort fald i nominel BNP. Herefter begynder den svenske pengepolitik, bl.a. med en negativ indlånsrente i Riksbanken, og det lykkedes hurtigt Lars Svensson at få nominel BNP tilbage på trend.

Her er så nominel BNP i Danmark (milliarder danske kroner per kvartal):

Der er ikke så katastrofalt som de tilsvarende grafer for Spanien eller Grækenland, men det stadig ikke ret god pengepolitik. Resultatet er tydeligt. Her er BNP i faste priser for de to lande, som et er mål for hvor stor økonomien er (indeks 2000–2005 gennemsnit=100):

Vi kan også se på arbejdsløsheden:

Sverige har klaret sig langt bedre end Danmark og er nu, som man skulle forvente, et mere velstående land end det var i 2007. Danmark er fattigere end i 2007. Det er ikke så stor en tragedie som de tilsvarende begivenheder i Spanien eller Italien, men det er stadig tragisk.

Udenfor Danmarks grænser er den konventionelle visdom om svensk økonomisk politik, sammenlignet med andre europæiske lande, at Sverige har haft stor gavn af en lempelig og korrekt pengepolitik, og af at have haft Lars Svensson i Sveriges Riksbanks ledelse. Svensson er et pengepolitisk geni som fortjener at være centralbankdirektør i en af verdens største økonomier, ikke et lille ubetydeligt skandinavisk land.

Hvem fører så pengepolitikken i Danmark? I Danmark har man siden 1980’erne haft en såkaldt fastkurspolitik, hvor den danske krone har en mere eller mindre fast vekselkurs overfor euroen. Det betyder at euroens pengepolitik automatisk gælder for Danmark. Euroens pengepolitik føres af den Europæiske Centralbank, hvis målsætning er at skabe stabil vækst i nominel BNP for eurozonen Tyskland. Det går faktisk meget godt for dem:

Den store tragedie er altså at Danmark har en helt forkert pengepolitik fordi man blindt kopierer den tyske pengepolitik. Men der er en grund til at vi ikke bare mekanisk kopierer den tyske økonomiske politik på alle mulige andre områder: beslutninger der er gode for Tyskland er ikke nødvendigvis gode for Danmark. Det skyldes blandt andet at Danmark og Tyskland er to forskellige lande. Danmark har ikke beskæftiget sig med pengepolitik i så mange år at man helt har glemt hvad det er, og at det i de fleste lande udgør en væsentlig del af den økonomiske politik, især når det kommer til hvordan man skal reagere på økonomiske kriser og økonomien på kort sigt. På længere sigt er produktivitet vigtigst, men det er der jo styr på i Danmark—måske lidt for meget.

En lidt mindre tragedie er hvordan Kenneth Praefke vælger at dække den økonomiske debat i Danmark. Han fremhæver på Twitter at han har fundet nogle skikkelser som har lyst til at snakke om strukturreformer, blandt andet på arbejdsmarkedet og skatteområdet, og at sammenligne Danmark og Sverige på det felt. Han har valgt at agere mikrofonholder for dem, og det sådan set også fint nok: Torben M. Andersen er en klog fyr og der er ikke noget galt i at høre hvad han har at sige.

Problemet er at Praefke, både i artiklen og på Twitter, giver indtryk af at debatten handler om økonomisk politik. I overskriften fremhæver han påstanden om at Sverige ikke er et økonomisk forbillede. Han har ret i at danske politikere ikke har lyst til at snakke pengepolitik. Han er dog selv medskyldig i det danske establishments forsøg på at ignorere elefanten i rummet: Vi ved allesammen at Sverige har klaret sig bedre i krisen, og det er et faktum som han selv citerer politikere for af og til at bemærke.

Men ingen har lyst til at snakke om hvorfor Sverige har klaret sig bedre; i stedet har vi Praefkes artikel og en masse udenomssnak om alle mulige områder hvor Sverige ikke har gjort det bedre.

Det er fint nok at skrive en artikel der sammenligner strukturreformerne i de to lande. Det er acceptabelt, men skuffende, at sige: “Sverige har klaret sig bedre gennem krisen fordi de har ført en fornuftig pengepolitik. Det er ikke muligt i Danmark fordi ‘pengepolitik’ er et tabu,” og så stoppe der. Det er mærkeligt at påstå at man sammenligner økonomisk politik i Danmark og Sverige og så ignorere den vigtigste forskel på de to landes økonomiske politik, også selvom årsagen til den forskel findes i dybe problemer i dansk politik.

PS: JJ opfordrer mig til at tilføje en graf over SEK-kursen 2007–13. Her er den effektive valutakurs for DKK og SEK i den periode:

(Men glem ikke at det mest handler om indenlandsk efterspørgsel.)

Medium is—well, I’m not quite sure. A website? In any case, scrolling is broken on Macs on the site when you use page down or the spacebar to scroll. Instead of the smooth scroll that we’re used to in every other context, it scrolls an entire page in one frame. This Safari extension fixes that.