Sunday, December 10, 2017

Commentary: My fun but ultimately disappointing foray into cryptoland

Okay, so despite being a crypto skeptic, I don't live under a rock.  With bitcoin's insane surge this week, and the launch of a futures market later today, I finally fell into the rabbit hole and studied the economics and market structure of cryptocurrencies.

First, there is nothing novel nor particularly innovative about bitcoin itself. A distributed ledger (the block chain) is a horribly inefficient way to record and process payments. The main advantage bitcoin has over traditional digital money is that it is not a liability of a financial institution, so in theory it carries no counter-party risk. However, the downside is that it relies on a brute force method to keep track of payments which uses a tremendous amount of electricity and computational resources.  The block chain completely unscalable. And to be frank, except inside the dark musings of libertarian fantasy, with modern deposit insurance, the counter-party risk of holding deposits denominated in major currencies issued by governments with stable politics is essentially zero.  Bitcoin proper is a solution in search of a problem. Speculators might run the price up some more, but it has no staying power.  Bitcoin is nothing more than a digital beanie baby. 

But let's explore two very popular platforms which attempt to solve the scalability problem, Bitshares and Ripple. Both of these systems reintroduce some counter-party risk in order to make block chain technology more practical for widespread use. 

Ripple processes payments between gateways, essentially digital transfer hubs. Block chain technology processes payments between gateways, rather than everyone having to keep an updated copy of the ledger. Gateways themselves are de facto financial institutions. They issue liabilities which are used to settle payments. These liabilities (called issuances or deposits) can be denominated in anything, but for the most part in are denominated in national currencies or crypto-currencies. Liabilities can be redeemed on demand at gateways. 

The result is that end users of the Ripple network must take on a non-negligible amount of counter party risk. Any balance that one holds on the Ripple network is by definition the liability of a gateway. (Except for XRP, Ripple Lab's own crypto-currency which is the only token which exists natively on Ripple) 

Technically speaking, traditional digital money (aka bank deposits) also has this 'flaw.' However, in most countries depositories have the backing of deposit insurance combined with a theoretically unlimited liquidity backstop from their respective central banks. The system of gateways is essentially a banking system operating with no liquidity backstop, limited regulation, and no government guarantees.  Nothing stops gateways from investing in dubious assets and sticking it to their depositors when things go south.  Again, this might tickle the fancy of readers of Reason magazine, but it's a system which has failed spectacularly over the course of history. One only needs to study the free banking era in the United States. When banks operate with no liquidity backstop, the failure of one prominent institution can set off a wave of panic which leads to system wide collapse. This happened time and time again during the 19th century, and I have no doubt that failure of a popular Ripple gateway would bring the whole system down. 

 Finally, forcing depositors to impose market discipline on banks has its own cost.  It means that while there may be a common unit of account, there is no common currency. Because a US dollar held on the Ripple network is not really a dollar, but a promise by a gateway to pay a dollar, these assets do not trade at par. In other words a dollar denominated deposit at one gateway won't be worth the same as a dollar denominated deposit at another. This was how banking worked in the United States until the 1860s, when Congress passed the National Bank Act. It regulated banks on the asset side of their balance sheets, while simultaneously requiring nationally chartered banks to accept each other's notes at par value. Imagine if Wells Fargo dollars were not worth the same as Capital One dollars.  That was the reality of free banking and it's the current state of affairs on Ripple. 

At first, I was intrigued by Ripple. (I even applied for a job there, one I am now very unlikely to get!) It claims to be a real time payments system which uses actually currencies.  Then I realized it was essentially a resurrection of a deeply flawed system which has rightly been relegated to the dustbin of history. 

Government guarantees are one way to address counter-party risk. Collateral is another. Bitshares attempts to use the latter to solve the counter party risk issue. In theory, so long as one's collateral is good enough, any debt could be considered safe. I might willingly lend a homeless man one million dollars if he pledges 1.2 million in Treasurys. This means that unlike on Ripple, any user can issue liabilities on the Bitshares network.  But unlike shadow banking system, the form of collateral is not US Treasurys or Agencies, but BTS, a token native to the Bitshares network.  Current margin rules require 175 percent collateralization. That is, to short one US dollar, one must pledge USD 1.75 BTS equivalent of collateral. This may seem conservative, but consider the extreme volatility realized this year in virtually all crypto currencies. The Bitshares platform pledges to close out positions which don't meet collateral requirements, closing out the least collateralized positions first.  The claim is that this means collateral would be automatically liquidated before default could occur. However, its unclear if the platform could handle a daily swing of say fifty percent in the price of BTS, something which is hardly hypothetical given historical realized volatility in all crypto-currencies including BTS.

Bitshares is just the crypto equivalent of shadow banking. But unlike shadow banking in the real world which uses rock solid collateral (US Treasurys, Agency MBS) combined with over-collateralization, Bitshares relies on shaky collateral which is almost guaranteed to experience high levels of volatility. (Can positions really be closed out fast enough? What happens when a big price decline occurs and BTS liquidity dries up?)  For all its flaws, shadow banking is at least backed by government issued collateral. Bitshares uses untested collateral whose volatility can and will put enormous strain on the platform's systems.  Again, this is not hypothetical. Many currencies brokers went bust when Swiss Franc soared in value after the SNB ended its exchange rate floor for EUR/CHF. The CHF surged in price, liquidity all but evaporated, and lots of big players couldn't close out their clients' short CHF positions fast enough and got caught holding the bag.  One silver lining may be the platform's rather strict margin requirements. But again, given the level of price swings seen in the crypto-space, putting faith in that alone seems foolishly optimistic. 

There's nothing new under the sun. Both Ripple and Bitshares are based on flawed systems. Ripple is free banking 2.0 dressed up with fancy technology. Bitshares is shadow banking based on questionable collateral. Both systems have existed at various points in history and at times have failed spectacularly. Far from being a financial paradigm shift, cryptoland is just a collective of bright eyed neophytes eager to repeat the financial follies of the past. 

    






No comments:

Post a Comment