Tuesday, November 30, 2021

A CBDC, eh?

David Andolfatto recently wrote a helpful article about whether Canada needs a central bank digital currency, or CBDC. I agree with David that a CBDC is "not an essential initiative at this point in time."

The way I see it, there are two big elephants in the room when it comes to introducing a central bank digital currency. Both of them suggest we should slow down any effort to issue CBDC.

But before I get to that, what is a CBDC? In brief, a CBDC is a new payments option that would allow regular Canadians to interact digitally with our nation's central bank, the Bank of Canada. For example, instead of having to use your Royal Bank account or Visa card to buy stuff at the supermarket or on Amazon, you could pay with a central bank digital tool, perhaps a Bank of Canada app or card. The Bank of Canada has been exploring whether the idea of issuing a CBDC for several years now, but so far hasn't pulled the trigger.

Here are what I believe to be the two big risks to rolling out a Canadian CBDC.

The most important one is white elephant risk.

In mature democracies like Canada, the provision of retail non-cash payments is a mostly-solved problem. Decent access to payments is already provided by a panoply of bank accounts, financial  apps, and cards. These existing options for connecting to the payments system are very safe. The $1,000 we hold in a checking accounts to make payments, for instance, is protected by government deposit insurance.

So if the Bank of Canada were to issue a CBDC, it's not obvious to me that many of us would bother using it. It would be just one more safe payments account among a sea of safe options.

That leaves the central bank in the position of having spent large amounts of money building and maintaining a payments network that none of us really needed or wanted in the first place.

Getting rid of an expensive and lacklustre CBDC could be difficult. Central bankers may feel like their reputations are tied to their CBDC projects. This, combined with the fact that central bankers don't feel the sting of a bottom line, may allow these white elephants to limp on for a very long time.

At the moment, most Canadians don't interact with Bank of Canada products (apart from holding cash, usage of which is falling). We mostly view the central bank as a competent technocratic body that does complicated stuff with interest rates. A CBDC project would suddenly put the Bank of Canada in direct digital contact with Canadians. But if the CBDC were to become a white elephant, this proximity could backfire on the central bank. It'll be know as the agency that runs a bloated payment system that the public dislikes. We don't want the Bank's brand to be hurt. We want the Bank of Canada to be trusted and deemed competent.

The second risk is black elephant risk. A black elephant is an obvious risk that people avoid discussing ahead of time.

Most discussions about CBDC centre on the technological hurdles involved in building one. But they often ignore the pesky sociological difficulties of running payments systems.

Fickle customers want to be able to reverse payments when they aren't happy with the products they buy. Users are frequently swindled out of their money by fraudsters and will expect restitution. If Jack accidentally send $500 to Alice instead of the $50 that he intended to send, he will want $450 back. What if Alice disagrees?

This requires that central bank constantly arbitrate disputes.

That's not all. Criminals will try to use a CBDC to sell illegal products. The central bank will have to start policing what is illegal and what isn't. Controversial businesses, say white nationalist publishing houses or kinky porn sites, will line up to use it. That means getting embroiled in politics.  

These are the not the sorts of issues I want my central bankers to get bogged down in. The risk is that demanding CBDC users distract the Bank of Canada from the vital task of conducting monetary policy.

The Bank of Canada might try to outsource the governance of a CBDC to the private sector. But that begs the question: if the central bank doesn't want the burden of running a payments system, why is it trying to get in the game at all? Why not just stick with the status quo, which seems to be working?

In sum, if I had any suggestions for Canadian citizens on how they should appraise a Canadian CBDC, it would be this. Given the above white & black elephant risks, the Bank of Canada shouldn't lead, it should follow. Watch the first few trail-blazing central banks to see how their CBDC projects pan out. (The Swedes, who are aggressively purchasing a CBDC, are a good candidate). If the Swedish CBDC succeeds, copy it. If the Swedes fail, continue on as before. There's no advantage to being the first to market.

Saturday, November 20, 2021

A dark world where bitcoin payments have gone mainstream

[This is an adaptation of an article I wrote last year for CoinDesk]

Satoshi Nakamoto's electronic cash system – Bitcoin – was originally intended for people to make online payments. But it never caught on as a mainstream payments option. Bitcoin's wild, and potentially lucrative, price changes have prevented it from developing into a popular substitute for Zelle, Visa, ACH, or PayPal. On top of that, the process used to run the bitcoin network, proof-of-work, is incredibly costly (by design).

What would cause bitcoin payments to go mainstream in America? That is, if ten years from now everyone was using volatile bitcoin tokens as their main medium of exchange, what major events would have gotten us to that point?

Unfortunately, the path to mainstream bitcoin payments is not an uplifting one. It requires that the U.S.'s reliable payments pipelines, the ones that have knitted Americans together for decades, stop doing their job. This unraveling of the payments system would be just one part of a broader decaying of American society. Only when these core payments systems are inoperational, and American society is on its knees, will a third-best payments rails like bitcoin be called into play.  

Here's a short story about how America's payments infrastructure slowly implodes and bitcoin payments go mainstream.

We all know that America is ideologically divided. This political storm has been spreading into commercial affairs with companies being required to take a stand on many polarizing issues. The payments industry in particular has become a major venue for conflict. (Think controversies over fundraising for Kyle Rittenhouse and card network censorship of sex workers.)

Imagine a world in which these divisions were to deepen.

In 2023, activists successfully pressure payment processors to make broad-based purges of businesses that are deemed too Republican. One casualty, the Wall Street Journal, is de-platformed by its acquiring bank. (An acquiring bank is the financial institution that hooks businesses into the Visa and Mastercard networks.) The Journal quickly gets a new Republican-friendly acquirer. Companies with Trump-supporting executives like Home Depot and Goya Foods are cut off by their acquiring banks, too.

Republican activists react by pressuring financial institutions to unplug Democrat-aligned businesses. In 2024, several large banks stop connecting abortion clinics to the Visa and MasterCard networks.

By the late 2020s a divided ecosystem of payments processors and acquirers has emerged. One half specializes in connecting businesses and nonprofits deemed Republican to the card networks. The other half specializes in connecting Democrat ones. Any bank or processor that tries to stay neutral is shunned – she who connects my enemy to Visa is my enemy.

Even at this level of divisiveness, Republicans and Democrats can still make payments with each other. That's because MasterCard and Visa remain neutral. The two networks allow both Republican- and Democrat-aligned acquiring banks, and the businesses that these banks serve, to connect to their networks. And thus dollars can flow across the ideological chasm.

But in 2029, Democrat activists succeed in pressuring Visa to end their neutrality and disconnect all Republican acquiring banks and processors. Suddenly, businesses that are deemed Republican can no longer accept Visa cards. The next year MasterCard is pressured to go Republican. All Democrat-leaning businesses are exiled from the MasterCard network.

America is now divided into two card fiefdoms. Apple (D) is Visa, Walmart (R) is MasterCard. Amazon (D) is Visa, Home Depot (R) is MasterCard.

But commerce can still occur across the divide. Any consumer who wants to shop at both Republican and Democrat stores need only make sure they have both a Visa and a MasterCard.

Getting both brands might not always be possible, however. Republican individuals may find it difficult to pass the increasingly politicized application process for a Visa card. Likewise, Democrat consumers find it challenging to make it through the application process for a MasterCard. 

That's when bitcoin might become a more useful payments mechanism. Since the Bitcoin network is censorship resistant –  anyone who want to use it can easily get access – it provides a means for Republicans to shop at Democrat stores and vice versa.

And so bitcoin finally becomes more popular for payments, but only because American society has moved backwards to a less civilized state. The easiest and most efficient option, cards, have degraded to the point that a back-up technology, Bitcoin, must be relied on. You can see that bitcoin isn't a progressive technology, it is a retrogressive one.

Up till this point in my story, the broad ideological upheaval between left and right has been reflected in a splitting-up of the card networks. Notice that the underlying payments plumbing on which America's entire private payments system runs, the Federal Reserve, has remained neutral throughout.

In 2031, that changes. The neutrality of the Federal Reserve, made up of 12 district Reserve banks, comes to an end. The CEO and directors of the Federal Reserve Bank of Kansas City, all staunch Republicans, decide to stop providing Democrat-leaning banks in their district with access to Fedwire. (The Kansas City district includes the states of Kansas, Wyoming, Nebraska, Colorado and Oklahoma.)

Fedwire, the Federal Reserve's real-time settlement system, is America's core payments utility. When anyone makes a payment from his or her bank to another bank, it'll eventually be settled by a movement of funds along Fedwire. By cutting off Democrat-leaning banks and their customers from this key utility, the Kansas City Fed effectively severs them from the U.S. payments system.

In retaliation, the Federal Reserve Banks of San Francisco and Boston disconnect Republican banks from Fedwire, in one swoop unbanking all Republican-leaning businesses located in their districts. The remaining ten district Reserve banks all pick sides, too.

If they haven't already done so, Republican leaning businesses rush to relocate to Republican districts. Otherwise they will not get banking services. Democrat businesses migrate to Democrat districts.

Those businesses brave enough to stick it out in a hostile district will need an alternative payments mechanism for connecting with their suppliers and customers. Cash will be one option. For non-face-to-face payments, however, bitcoin may be their only option. And so as America descends into partisanship and the Federal Reserve crumbles, an awkward bitcoin "cash system" becomes a way around an increasingly balkanized payments system.

Even in this hyper-factionalized America, inter-district trade between Democrat and Republican zones can still occur. A car mechanic in a Democrat district can buy tires from a part dealer in a Republican state. That's because Democrat-leaning Federal Reserve banks (such as the San Francisco Fed) remain connected to Republican-leaning Federal Reserve banks (such as the Kansas City Fed) through Fedwire. Fedwire continues to unite disparate parts of the country.

That stops in 2033. The San Francisco Fed halts all incoming payments from Republican Reserve banks including the Kansas City Fed, Atlanta Fed, and Dallas Fed. In reaction, Reserve banks in Republican enclaves such as Kansas City cut off Democrat districts. At that point there ceases to be a universal U.S. dollar. Money held in accounts in Georgia and Florida and Oklahoma can't move into accounts in California or Washington, and vice versa. The payment tissue that once connected all Americans has torn.

With the collapse of Fedwire, cross-border trade and remittances between hostile Democrat and Republican enclaves get very tricky to carry out. Society may have regressed far enough back that silver and gold once again become an international settlement medium, just like in the 1600s and 1700s. Or perhaps bitcoin would become America's preferred medium for making payments across enclaves. Unlike gold, bitcoin can be transferred remotely.

The collapse of America's payment infrastructure would be just one theatre in a much larger cleaving of American society along ideological lines. Other key bits of American infrastructure would also begin to fall apart: the courts, law enforcement, the education system. There would be large physical dislocations as Republican families flee Republican enclaves and Democrats to Democrat enclaves.

But if America's electrical and telecommunications infrastructure has crumbled, too, would it even be possible for people to use bitcoin, which is reliant on the internet?

It’s a stretch, but we can imagine distributed solar power solving the electricity problem. As for accessing the bitcoin network, tinkerers could try to connect old-fashioned ham radios to Blockstream's bitcoin satellite. If the remnants of AT&T and Verizon can only provide patchy internet service, so-called decentralized mesh networks might offer an alternative way to access the web.

This dystopian future probably isn’t going to happen. It's just a story. For now, bitcoin remains an unpopular payments system. Let’s all hope that it stays unpopular. No one wants to live in a country that has declined so far that bitcoin has become a vital way to make payments.

Tuesday, November 16, 2021

The dangers of stablecoin lending


These days I see many do-it-yourself investors comparing the huge yields they can earn on stablecoin lending to the tiny yields on bank accounts. Cryptocurrency influencers like to draw attention to this big gap, portraying crypto as the heroic replacement to stodgy regular finance, or "TradFi". The Celsius Network, one of the leading providers of high-yield stablecoin products, uses the slogan "Unbank yourself." The implication is that anyone who holds their money in a bank account is a chump.

Beware, DIY investors. These marketing pitches are wrong, indeed dangerous. 

In finance, a juicy yield is almost always associated with big risk. Shifting finance to blockchains doesn't change this truth. High-yielding stablecoin strategies are not a better sort of bank account. Rather, they're a potentially hazardous investment more akin with penny stocks and CCC-rated junk bonds.

Let me explain with a recent example:

The premise of this tweet and the attached chart is that you can make far more on your stable crypto dollars than on old fashioned dollars stuck in a bank account.

The problem with this comparison is that it's not contrasting equal things. It's comparing apples to oranges.

The true counterpart to a 0.06% yield on a bank account isn't the interest rate one can earn by on-lending stablecoins via protocols like Celsius or Compound. No, the proper analog is the interest rate one earns by simply holding a stablecoin such as Tether or USDC. And because stablecoin issuers don't pay interest to people who own stablecons, this rate is effectively 0%. Which is *ahem* below the 0.06% rate on a U.S. savings account.

Hardly a selling point. Unfortunately, the above chart forgets to mention the 0% rate on stablecoins.

Let me flesh this out further. When you own a stablecoin or keep money in a saving account, you are basically lending to the issuer of those dollars. If you hold 1,000 USDt (Tether stablecoins), for instance, you're a creditor to Tether Inc. That is, Tether Inc owes you $1,000. Likewise, if you keep $1,000 in a Bank of America savings account, you're lending $1,000 to Bank of America.

Think about this or a moment.

Lending involves risk. The borrowing party may not be able to keep its promise to you. Bank of America is a pretty safe entity to lend to. It'll probably keep its promise to you. But the firms that issue Tether and USDC are not safe borrowers. They are small. Not much is known about them. In Tether's case, it is entirely unregulated. And Circle, the issuer of USDC, is only lightly regulated. If you are acting as a lender to Tether or Circle, you should be getting *much* more than the 0% rate that they're offering you.

On top of that, your loan to Bank of America is protected by government insurance. Nothing protects your loan to Tether or Circle. Even worse, as a creditor to Tether and Circle, it's not apparent where you rank in terms of seniority. This ranking is important because in the case of a failure, senior creditors get paid first, junior creditors last. At least with a Bank of America account you're at the front of the line.

Far more prudent to lend to a government-insured bank and collect 0.06% than lend to a black box stablecoin and get 0%.

Of course, stablecoins aren't just held. It's what you can do with stablecoins that excites people. Which gets us to the massive crypto lending rates that are illustrated in the chart. Aave and Compound are decentralized lending protocols. If you on-lend your stablecoins via these two protocols, you can earn 2.69% to 3.14%.

Celsius, Nexo, and Blockfi are centralized marketplaces where rates for onlending stablecoins reach as high as 8.88%.

The thing is, you *should* be getting a high rate for onlending your stablecoins on these venues. You're taking a big risk by using them. These platforms could go broke, get hacked, or break. In the case of centralized platforms, there is very little information about how they are using your funds. Furthermore, you don't know where you stand in seniority among other Celsius, Nexo, or BlockFi creditors. These are black boxes, folks.

And remember, even though you've lent away your Tether or USDC on Celsius or Aave, you're still fully exposed to all the original credit risk of Tether or Circle. 

For instance, say you lend 1,000 USDt on Celsius's platform and Tether, the issuer of USDt, collapses. The price of USDt stablecoins falls from its $1 peg to $0.10. Celsius comes through, though. It keeps its promise to you and repays the 1000 USDt it owe you. Alas, now that amount is only worth $100.

So for DIY investors considering stablecoin lending strategies, any loan to Celsius or Aave involves a combination of two risks: the possibility that Tether or Circle (the issuer of USDC) go under and the chance that Celsius or Aave break. Add the two together and you're getting involved in a pretty dangerous strategy, one for which you should be well-compensated.

Rather than clapping your self on the back for getting 8% from stablecoins instead of 0.06% in a savings account, you should be asking yourself whether 8% is enough.