Tuesday, April 26, 2022

Where are the customers' rate increases?

U.S. banks are at it again. Inflation is at its highest level in decades. At the same time, interest rates on deposits at the Fed, Treasury bills, bonds and mortgages are rising rapidly to compensate. Yet banks are still in a holding pattern when it comes to the interest rates they pay to customers on savings accounts, certificates of deposit (CDs), and interest-checking accounts.

Here's the chart, which uses FDIC data from FRED. Note how customer deposit rates (in red) have hardly budged, despite the Fed beginning to raise rates last summer. This same sluggishness also occurred in 2015, the last time the Fed began to hike rates. Banks didn't boost savings accounts rates till two years later, in 2017!


Historically, rates on CDs seem a little more responsive. But they're still sluggish. The average 6-month CD still only yields a scrawny 0.09%, whereas the yield on a 6-month Treasury bill is now at 1.4%.

This stickiness wouldn't be such a big deal if banks were also slow to reduce interest rates on customers' accounts win some, lose some, right? But take a look at what happened when the Fed began to cut rates in mid-2019. Banks didn't hold off. They immediately started to pass lower rates on to their customers, and only became more aggressive when COVID hit in March 2020.

So for bank depositors, it's all lose. U.S. banks are slow to increase rates on checking accounts, CDs, and savings accounts, but quick to reduce them. I wrote about this sad lack of symmetry back in 2020. Do go back and read it.

This observation isn't something that economists have ignored. In a paper entitled "Sticky Deposits", Federal Reserve economists John Driscoll & Ruth Judson found that rates are "downwards-flexible and upwards-sticky." This stickiness has consequences for regular Americans. If rate stickiness didn't exist, the authors estimate that U.S. depositors would have received as much as $100 billion more in interest per year!

Friday, April 15, 2022

A sound debasement

An imitation English half noble issued by Philip the Bold, Duke of Burgundy, 1384-1404 [source]

[This is a republication of an article I originally wrote for the Sound Money Project. When we look back at old coinage systems, our knee-jerk reaction to the periodic debasements that these systems experienced is "ew, that's gross." But things were considerably more complex than that. This article tells the story of a healthy, or wise, coin debasement Henry IV's debasement of the gold noble during the so-called "war of the gold nobles" between England and Burgundy in the late 1300s and early 1400s.]

A Sound Debasement

In his excellent article on medieval coinage, Eric Tymoigne makes the seemingly paradoxical claim that “debasements helped preserve a healthy monetary system.” A debasement of the coinage was the intentional reduction in the gold or silver content of a coin by the monarch by diminishing either the coin’s weight or its fineness. I’m going to second Tymoigne’s paradoxical statement and provide a specific example of how a debasement might have been a sound monetary decision.

First, we need to review the basics of medieval coinage. In medieval times, any member of the public could bring raw silver or gold to the monarch’s mint to be coined. If a merchant brought a pound of silver bullion to the mint, this silver would be combined with base metals like copper to provide strength and from this mix a fixed quantity of fresh pennies — say 40 — would be produced. These 40 pennies would contain a little less than a pound of silver since the monarch extracted a fee for the mint’s efforts.

The merchant could then spend these 40 new pennies into circulation. Coins were generally accepted by tale, or at their face value, rather than by weight. Shopkeepers simply looked at the markings on the face of the coin to verify its authenticity rather than laboriously weighing and assaying it. This was the whole point of having a system of coinage, after all: to speed up the process of transacting.

As long as the monarch of the realm continued to mint the same fixed quantity of coins from a given weight of silver or gold, the standard would remain undebased. Sometimes, however, “coin wars” erupted between monarchs of different realms, the aggressors minting inferior copies of their victims’ coins. Since these wars hurt the domestic monetary system of the victim, some sort of response was necessary. One of the best lines of defense against an aggressive counterfeiter was a debasement.

John Munro, an expert in medieval coinage, recounts the story of the “war of the gold nobles,” a coin war that broke out in 1388 when the Flemish Duke Philip the Bold began to mint decent imitations of the English gold noble. Flanders, comprising parts of modern-day Belgium and northern France, was a major center of trade and commerce on the Continent. Both the weight and fineness of Philip’s imitations were less than those of the original English noble. According to Munro’s calculations, by bringing a marc de Troyes of gold (1 marc de Troyes = 244.753 grams) to Philip’s mint in Bruges, a member of the public could get 31.163 counterfeit nobles. But if that same amount of gold were brought to the London mint, it would be coined into just 30.951 English nobles. Given that more Flemish nobles were cut from the same marc of gold than English nobles, each Flemish noble contained a little bit less of the yellow metal.

Philip’s “bad” nobles soon began pushing out “good” English nobles, an instance of Gresham’s law. Given Philip’s offer to produce more nobles from a given amount of raw gold, it made a lot of sense for merchants to ship fine gold across the English Channel to Philip’s mints in Bruges and Ghent rather than bringing it to the London mint. After all, any merchant who did so got an extra 0.212 nobles for 244.753 grams of the gold they owned. By bringing the fakes back to England, merchants could buy around 1 percent more goods and services than they otherwise could. After all, Philip the Bold’s fake nobles were indistinguishable from real ones, so English shopkeepers accepted them at the same rate as legitimate coins. English nobles steadily disappeared as they were hoarded, melted down, or exported. Why spend a “good” coin — one that has more gold in it — when you can buy the exact same amount of goods with a lookalike that has less gold in it?

Philip’s motivation for starting the war of the gold nobles was profit. By creating a decent knock-off of the English noble that had less gold in it, though not noticeably so, Philip provided a financial incentive for merchants to bring gold to his mints rather than competing English mints. Like all monarchs, Philip charged a toll on the amount of physical precious metals passing through his mints. So as throughput increased, so did his revenues.

The health of the English monetary system deteriorated thanks to the coin war. With a mixture of similar but non-fungible coins in circulation, there would have been an erosion in the degree of trust the public had in the ability of a given noble to serve as a faithful representation of the official unit of account. Nor was the system fair, given that one part of the population (people who had enough resources to access fake coins) profited off the other part (people who did not have access). Finally, when Gresham’s law hits, crippling coin shortages can appear as the good coin is rapidly removed but bad coins can’t fill the vacuum fast enough.

The English king’s efforts to ban Philip’s nobles had little effect. After all, gold coins have high value-to-weight ratios and are easy to smuggle. One line of defense remained: a debasement. In 1411, some 20 years after Philip the Bold had launched his first counterfeit, King Henry IV of England announced a reduction in the weight — and thus the gold content — of the English noble. This finally resolved the war of the nobles, says Munro. By reducing the noble’s gold content so that it was more in line with the gold content of the Flemish fakes, the English noble lost its “good” status. Merchants no longer had an incentive to visit Philip’s mints to get counterfeits, and English nobles once again circulated. The health of the English coinage system improved.

We shouldn’t assume that all medieval debasements constituted good monetary policy. There were many coin debasements that were purely selfish efforts designed to provide the monarch with profits, often to fight petty wars with other monarchs. These selfish debasements hurt the coinage system since they reduced the capacity of coins to serve as trustworthy measuring sticks. As Munro points out, each medieval debasement needs to be analyzed separately to determine whether it was an attempt to salvage the monetary system or an attempt to profit.

Thursday, April 7, 2022

The hoopla over the EU and self-custody wallets

There's been plenty of anger this week in the crypto world about a EU law that, if passed, would require European crypto exchanges to collect and verify information about so-called unhosted wallet users.

What the new rule boils down to is that if you have an account at an exchange like Coinbase, and you send some crypto off of the exchange to an unhosted wallet (i.e. a self-custody wallet or personal wallet), then Coinbase will have to verify the European owner of that unhosted wallet and store their information.

Messing around with self-custody wallets is verboten among crypto fans. But is the EU's proposed rule as unprecedented and unfair as the crypto press is making it out to be? I think the angst is overdone, and I'll show why.

One of tests we can perform to determine the fairness, or neutrality, of any new crypto-targeted anti-money laundering (AML) regulation is to ask the following question: does the same regulation already apply to cash and cash-remittance providers like Western Union? If so, then it's fair.

The "Western Union test" works on the assumption that cash and crypto function in similar ways, and so the principle of technological neutrality dictates that they should be regulated similarly for AML purposes. Both are transferable on a person-to-person basis, they can each be self-custodied (i.e. they don't require a bank), and they provide a degree of privacy. And so institutions that deal in crypto and/or cash should face the same AML requirements. Put differently, whatever Western Union is already obligated to do with cash, Coinbase must do with crypto.

So what does the test tell us? A quick glance at Article 5 of EU Regulation 2015/847 indicates that payment service providers are already required to identify anyone from whom they receive cash, or to whom they pay out cash. Which means that if a customer of Western Union asks an agent to disburse 1500 in banknotes to a European, it must verify the recipient's ID.

And so the "Western Union test" indicates that it's only fair that Coinbase be required to verify the ID of the owner of an unhosted wallet to which it pays 1500 worth of crypto. Like cash, like crypto.

AML regulation often exempts service providers from ID requirements for small transfers. But the potential EU law on verifying the identity of unhosted wallet owners does not come with an exemption. Even if Coinbase only pays out a tiny amount (say 50 in crypto) to an unhosted wallet, identifying the owner would be necessary. I don't like this aspect of the law, but it does pass the "Western Union test," as I'll show.

I'm not a fan of a lack of thresholds for transfers to unhosted wallets because thresholds allow those without an ID -- say refugees and homeless people -- to make payments. Thresholds also afford licit users a window for privacy. We should try to preserve a small privacy safe haven.

While I'm not the fan of  lack of a crypto threshold, the measure does pass the "Western Union test." In the EU's case, there is no verification exemption for transfers received or paid out in cash. That is, even if Western Union disburses a tiny 50 in cash to someone, identity verification is still required. (See Article 5, Section 3 of  EU Regulation 2015/847). And so its fair to subject Coinbase and a €50 transfer to an unhosted wallets to the same stringency.

Even though the EU's new crypto regulation passes my "Western Union test," crypto advocates are unlikely to be fans of this legislation. But they can still protest. Don't want to go through an AML process for unhosted wallets? Then don't use payments service providers like Coinbase. Always transfer coins bilaterally through self-custodial wallets.

P.S. There is one way in which the EU's proposed law doesn't pass the "Western Union test." It would require that for every transfer received from an unhosted wallet, the payment services provider inform the competent authorities (see Article 16, section 4a). This requirement doesn't exist for cash and cash-based payments providers like Western Union. Yes, a cash payment requires ID verification, but as far as I know there is no notification requirement. So in this one respect, the new law doesn't treat cash and crypto in a consistent manner.

P.P.S. I am open to the idea that traceable crypto like Bitcoin should be subject to less stringency than cash, in the form of a higher threshold, and that untraceable crypto like Zcash and tumbled crypto would be subject to the same stringent threshold as cash. So for example, if cash enjoys a $1000 exemption, then Zcash should also get a $1000 exemption, and so should bitcoins mixed by Wasabi, but unmixed bitcoins should get a less stringent $3000 exemption.

Tuesday, April 5, 2022

Crypto’s privacy/commingling dilemma

Abandoned panopticon prison in Cuba [source]

[This is a re-post of an article originally published last week at the AIER's Sound Money Project. Leave a comment at the bottom and we can hash out our differences.]

Crypto’s privacy/commingling dilemma

People need financial privacy. (Yes, even regular people!) Alas, blockchains are panopticons. The entire history of one’s bitcoins or ether are indelibly etched on the blockchain for all to see.

Mixing is one way to get necessary on-chain privacy. In short, a bunch of people use an automated tool such as Tornado Cash or Wasabi Wallet to combine their crypto together and then withdraw it, obfuscating the trail of each person’s crypto.

But mixing is a less-than ideal way to gain privacy. In solving for privacy, mixing introduces another problem; commingling. When you use a mixing tool, you’re introducing your licit funds into the same pot as potentially criminally-derived funds. That is, you run the risk of serving as a laundering entrepot for criminals.

This is bad for three reasons: 1) Money laundering is a criminal offense, and if you’re a normal person you’d probably prefer not doing anything illegal. 2) It’s also morally questionable. By serving as a counterparty to criminals you are helping them get away with theft or fraud. 3) Commingling may be financially risky for you. Because mixing is visible on the blockchain, other people may see that your crypto is mixed and may not want to take on the risk of accepting it. Put differently, mixed crypto might not be worth as much as unmixed.

That, in a nutshell, is the privacy/commingling dilemma. If you’re a licit owner of crypto and you mix, you’ve swapped your lack of privacy for a potential felony indictment, a bad conscience, and deteriorated finances. This is likely not what you were looking for.

The risks of commingling crypto are heightened due to the fact that mixing is not turned on by default. And so only the most motivated people – criminals and privacy advocates – end up gravitating to mixing tools. This only increases the odds that a licit crypto user who enters a mix unintentionally serves as a counterparty to a crook.  

I’ve written about the risks of commingling with respect to Tornado Cash, a non-custodial Ethereum-based mixer. But the same risks apply to bitcoin-based non-custodial options like Wasabi. (A custodial mixer is run by an individual who takes possession of users’ tokens and then mixes them. A non-custodial mixer like Tornado Cash is a purely automated protocol. No third party takes possession of the crypto that is to be mixed.)

For licit crypto users who want privacy but are wary of the privacy/commingling dilemma, centralized exchanges are probably the safest option. Centralized custodians like Coinbase screen out criminals. So you, a licit user, can seamlessly get your coins into Coinbase’s big mix (without worrying about the risks of engaging in money laundering), and then withdraw them to a new address, safely obfuscating the history of your coins. It’s a relatively safe version of mixing.

But centralized exchanges don’t provide you with the full set of privacies. While they temporarily absolve you from the panopticon of the blockchain (i.e. everyone else’s eyes), they don’t shield your personal information from the exchange itself. For many people, that degree of privacy is probably good enough. For others, it’s not.

Native privacy is the best alternative. Imagine if your bitcoins or ether were private by default, sort of like cash, or gold. You would never have to endure the risks of commingling via a service like Tornado Cash or coming under the watchful eye of Coinbase. But native privacy is only in its infancy. (I wrote about it here).

Another way to tackle the privacy/commingling dilemma is an approach recently adopted by Wasabi Wallet, a non-custodial provider of bitcoin mixing services. Historically, Wasabi’s owner zkSNACKs has allowed everyone to access Wasabi’s services. But, going forward, it will screen out bad bitcoins by analyzing the blockchain for known nefarious actors:

"We were always against using [CoinJoin] for illicit activities, and as far as we could see from the news, lots of actors started to take advantage of the software, and this created really bad press for us… Wasabi is for people to preserve their privacy, and not for hiding illicit activities. [source]"
zkSNACKs hasn’t provided precise detail on how it will determine which addresses are to be censored. Would its blacklist include addresses of known hackers, scammers, sanctions evaders, and those subject to injunction or freezing orders? It remains to be seen.

In any case, introducing filters could finally provide licit crypto users with an answer to the privacy/commingling dilemma. Wasabi will provide much needed privacy while reducing the risk that they are laundering funds for the mob.

Anecdotal evidence suggests that there is demand for a clean mixing service. Wasabi developers had expected its announcement to anger its customers and lead to a reduction in mixing activity. Instead, the company says that demand tripled.

Gaining a reputation for clean mixing will attract not only licit users, but criminals keen to get their coins into Wasabi’s mix in order to launder them. A cat and mouse game will develop as Wasabi evolves new methods to screen these bad actors out, and these bad actors react. It won’t be an easy battle.