Tuesday, July 7, 2015
Statistical agencies employ data collectors who walk up and down aisles with hand-held computers gathering sticker prices for things like frozen french fries and bicycles. The data they collect gets amalgamated into an index and passed on to central bankers who use it as the basis for rate change decisions. It seems simple enough, but what happens if the source material has been corrupted? Might central bankers be reacting to mere shadows on the wall?
Here's how prices might go bad. Start with the U.S. and Canadian payments systems. For each credit card payment, a North American merchant must pay 1-2% in fees to the card networks Visa and MasterCard. Retailers in both countries have very different strategies for coping with this burden. In the U.S., retailers are permitted to offload network fees onto customers by asking them to pay a surcharge on each credit card payment. Because Canadian retailers are prohibited from surcharging customers, they react by marking up every sticker price in their store by a percent or two, the extra margin they collect sufficient to cover the card network fees. (Canadian retailers almost never offer cash discounts.) For a more complete explanation, see here.
This arcane difference in payments habits has the potential to result in a divergent evolution of prices, vastly different monetary policies, and an uncoupling of North American economic growth.
Consider what happens when the Visa and MasterCard networks decide to offer North American customers a universal 5% cash back reward. The networks fund this bonus by requiring merchants to submit a 5% fee on each card transaction. U.S. retailers cope with this levy by boosting the surcharge they apply on each card transaction to 5% or so, forcing the card-paying customer to bear the cost. Those Americans who pay with cash i.e. banknotes continue to get the sticker price, which stays constant. Without the ability to surcharge, Canadian retailers cope by boosting sticker prices by 5%, thereby indirectly passing the costs of the cash-back bonus onto the customer.
Marching up and down the aisles, U.S price collectorsdon't notice a thing. As a result, the U.S. consumer price index stays the same and the Fed doesn't lift a thumb. Canadian price collectors, however, find that prices have risen. Upon reception of this data, the Bank of Canada anxiously raises rates. This is because Bank officials believe that the rapidly rising price level indicates that the economy-wide rate of return (the natural rate of interest) has risen above the Bank's market rate, breeding inflation. A rate hike is necessary to bring the two rates back in lane, thus choking off the incipient inflation that seems to be developing. The natural rate hasn't budged, of course. All that has changed is the card networks' fee policy. Instead of bringing the market rate of interest in line with the natural rate, the Bank of Canada has been fooled into moving the market rate of interest above the natural rate.
If the card networks again increase the cash back reward, say to 10%, Canadian prices will rise even more. U.S. prices stay flat. The Bank of Canada once again confuses the effects a new cash-back policy with a rise in the natural rate of interest, tightening while the Fed stays pat.
Assuming that prices react rapidly and fluidly to central bank policy, then the Bank of Canada's tightening will simply drive consumer prices back down and restore equality between the natural rate of interest and the market rate. But if changes in monetary policy have effects on the real economy, then we've got problems. It could be that certain prices are sticky downward so that markets can't clear, the result being inventories of goods going unsold. Or perhaps the fact that debts are nominally denominated creates a Fisherian debt deflation. If so, then the Bank of Canada could end up unnecessarily driving Canada into a recession. The Federal Reserve, reacting to the very same set of stimuli, does not.
Now of course I'm exaggerating things. In real life such large increases in cash-back policy are unlikely. Nevertheless, we have seen a progressive increase in network fees over the years, enough to have probably inspired Canadian retailers to ratchet up sticker prices. As a result, Canadian CPI may be slightly over-stating actual inflation. On the U.S. side, consider that American retailers only recently gained the power to allow surcharging credit card payments. As U.S. retailers roll out surcharge policies and reduce sticker prices, CPI will be pressured down. This may fool Fed officials into believing that the economy is slowing and reducing rates, when in fact all that has changed is credit card pricing habits. One wonders if the monetary authorities take into account these arcane features of our payment system when they set their monetary policies.
This answers a question I asked more than a year ago.
Tuesday, June 30, 2015
|This 2 euro coin is issued by Monaco, which is not a member of the Eurozone|
Grexit isn't what people take it to be. The standard narrative is that Greece is approaching a fork in the road. It must either stay in the euro or adopt a new currency. I don't think this is an entirely accurate description of the actual fork that Greeks face. Over the next few months, Greece will either:
- A) stay a member in good-standing of the institution called the "Eurozone" and continue to legitimately use that institution's currency, the euro, or
- B) leave the Eurozone while continuing to use the euro 'illegitimately.'*
The probability of a new drachma emerging is awfully low. The widespread idea that a sick country can rapidly debut a new currency and, more importantly, have that currency be universally adopted as a unit of account is magical thinking. Greece has been using the euro as a universal "language of exchange" for fifteen years. Switching over to a new unit is about as unlikely as Greeks suddenly beginning to speak German, network effects and all. Consider too the fact that Greeks don't want the drachma—they have consistently voted for euros. Syriza has no mandate to bring a new unit into existence. ***
Option B isn't an odd one. All sorts of countries 'illegitimately' piggy-back off the currencies managed by others. Zimbabwe, Ecuador, and Panama use the U.S. dollar without being card carrying members of the Federal Reserve System while Andorra, Kosovo, Montenegro, Monaco, San Marino, Vatican City use euros without being part of the Eurozone. Nor can the Eurozone can do anything to prevent de facto adoption of the euro by Greece. It's a decision that Greeks get to make themselves.
If Greece leaves the Eurozone on a de jure basis while staying a euro user, what will it be giving up?
The Greeks would NOT be losing the price stability and commonality already provided by Eurozone membership. These are presumably the features that lead most Greeks to declare in polls that they want to stay in the euro.
However, Greece would no longer get access to Eurozone lender of last resort facilities. One could argue that the nation would be better off without these facilities, given the discipline that a true 'no bailout' policy would enforce on both the banks and the government. Greece also loses direct access to the monopoly supplier of euro banknotes. A Greek banker can currently ask to have their Eurozone account be debited and a batch of freshly printed paper euros trucked over to their vault. Gone is that functionality. Panama has survived, even prospered, for decades without access to the Fed's discount window or Fed cash facilities.***
Greece would also lose its seat on the ECB Governing Council and therefore any say in determining monetary policy. Greece's one seat probably never gave it much influence anyways, especially compared to Germany's dominant influence in ECB decision making. Nor would Greek data be considered as an input into Eurozone policy decision should Greece leave. However, as it clocks in at just 2% or so of the Eurozone's total size, Greece's data could never have had much influence on the aggregates that ECB policy makers watched to begin with. Official user of euros or unnoffical, Greece will always lack an independent monetary policy.
Another concern is that Greece might not be allowed to use the ECB's Target2 real time settlement mechanism anymore. However, Denmark, Bulgaria, Poland, and Romania all connect to Target2, despite not being Eurozone members. Surely Greece would qualify. If not, it wouldn't be too complicated for Greek banks to set up their own payments system.
Lastly, Greece would forfeit all seigniorage revenues. Eurozone members currently get a share of the profits that the ECB earns on its monetary monopoly. I don't see this loss as being a big deal. Seigniorage has long since been eclipsed by taxes as the key source of a modern government's revenue.
The upshot is that whether Greece remains a legitimate member of the Eurozone or an unofficial user of the Eurozone's chief monetary product, the implications are about the same. There is no fork in the road, at least not from a monetary policy perspective; just a continuation of the same euro path as before.
I've left two features out. If Greece leaves, the claims and liabilities it has on the Eurozone must be unravelled and settled. Having invested around 200 million euros in the ECB when it was formed, Greece would have to be bought out by remaining Eurozone members at a reasonable price. Counterbalancing this would be Greece's obligation to unwind the debt that it has amassed to the Eurozone in the interim. This debt, known as its Target2 deficit, currently clocks in at around 100 billion euros, far in excess of the capital it is owed. It would take an incredible outlay of resources to pay this amount off. The advantage to the Greek population of staying in the Eurozone is that their debt need never be settled. After all, Target2 debts are by nature perpetual. Only by leaving do they face a final day of reckoning.
However, if Greece puts little-to-no cost on squelching on its debts, it may as well just leave the Eurozone without paying back the 100 billion it owes. It gets to continue to ride piggy-back on top of the euro, enjoying (almost) all the same benefits of being a Eurozone member, without being on the hook for anything. Why not perpetually bum cigarettes rather than pay for them?
Which is why Greece has a certain degree of power over the remaining Eurozone members. Should it shrug and leave, the remaining members are on hook for its unpaid tab. And once Greece goes down the de facto euroization path, how long before the next largest debtor to the rest of the Eurozone decides to shrug and leave? As Nick Rowe says, the last one holding a common currency is the sucker since they'll be left with everyone's bad debts. To keep the system going, the Euro project's architects need to do their best to ensure that Greeks aren't incentivized to just shrug and bum a free ride on the euro. I don't envy them their task, it's a difficult one.
The other aspect I've left out is the Greek banking system, which is probably insolvent. Once cutoff from the central bank that prints the stuff, Greek banks simply wouldn't be able to meet the rush to convert deposits into euro banknotes. The only way to return the banking system to functionality would be to chop the quantity of Greek bank deposits down to size so that the banks' asset base would be sufficient to absorb the run into cash. We're talking a multi-billion euro "bail-in" of depositors. The prospect of such a hit certainly tilts the decision between A and B back towards A.**
* Having been cut off from additional ECB lending, one might argue that Greece has already gone halfway towards exiting the Eurozone.
** Paragraph added July 2.
*** Added cash facilities on July 2.
**** Added last two sentence to this paragraph on July 3.
Note: apologies for the constant additions, but this subject is complex and the situation getting more complex by the day, so rather than writing two or three posts I'm adding bits to the original.
Thursday, June 25, 2015
The free banking era and the "multitude of dollars" problem
There's a precedent for this sort of dollar heterogeneity. During the U.S.'s so-called "free banking era" that lasted from the 1830s until 1863, hundreds of different types of banknotes circulated, each issued by a unique bank. Notes were universally redeemable in a certain quantity of gold. Varying physical distances between a note's final resting point and its birth at an issuing bank (often in another state) led to widely disparate redemption costs across note brands. A merchant in Philadelphia who was paid in local bank notes need only take a short walk in order to redeem the note. If that same merchant was paid in a note issued by a bank in Chicago, however, redemption was much more onerous due to the time and distance required to travel from Philadelphia to Chicago.
So in the same way that an American Express dollar is the most costly of the modern day dollars, a distant bank's notes were the most costly of the free banking era's dollars.
Here are two interesting problems. How can a merchant establish a single set of sticker prices while still accommodating a wide range of disparate dollar payments media? Second, consider the fact that shoppers paying with a premium card like American Express (or distant bank notes) enjoy the widest range of benefits, but should also face the highest costs. How can the system ensure that the person who enjoys the marginal benefits associated with a given payments medium also bears its marginal cost? Put differently, how is quid pro quo ensured?
The solution: surcharging
In the free-banking era, the "multitude of dollars" problem was solved by a form of discriminatory surcharging. To begin with, merchants displayed their sticker prices in terms of a single unit; standard U.S. dollars (defined as 1.5 grams of gold). When the customer arrived at the till, the merchant determined what sort of surcharge to apply to each of the banknotes proffered according to its distance-to-redemption. With hundreds of note-issuing banks in existence, this was a cumbersome task. In order to speed up the process, the merchant would consult what was called a "bank note reporter." This handy publication, which was compiled by professional bank note analysts, provided merchants of a certain location, say Philadelphia, with the rates for all bank notes that circulated in Philadelphia adjusted for their shipping costs.
The image below is a section from Van Court's Bank Note Reporter and Counterfeit Detector, which I've snipped from Gary Gorton's introduction to the subject. It shows the the recommended price at which a merchant in Philadelphia should accept notes from Vermont, Maine, Pennsylvania, and elsewhere.
After a merchant had consulted his bank note reporter and tacked on the appropriate surcharge, it was time to redeem the note. Merchants didn't actually ship the notes themselves but sold them to a local note broker at a discount to face value. In fact, the numbers that Van Court published would have been sourced from this broker market. The broker in turn shipped the note back to the issuer, getting full face value upon redemption. The gap between face value and the initial purchase price covered the broker's shipping costs.
Thus the "multitude of dollars" problem was solved. By surcharging relative to a benchmark dollar, merchants succeeded in setting a single array of prices while accommodating a much wider range of heterogeneous payments media. This also allowed them to efficiently pass the marginal cost of using distant notes onto those customers who chose to pay with them, while rewarding customers who used low-cost local notes by not applying such surcharges.
So why not implement this same technique of surcharging today?
Consider that in our modern era, credit card networks recoup much of the cost of the services they provide (which are many, but include perks like rewards and the right to dispute a transaction) by requiring that card accepting retailers collect a fee from customers on the network's behalf. This parallels the free banking era, in which banks required third-parties to bear the cost of transporting notes for redemption.
The best way for a modern retailer to establish prices in this heterogeneous dollar world would be to replicate the solution settled upon by their free banking ancestors: set sticker prices in terms of the most basic unit, paper dollars, and exact an appropriately-sized surcharge on each card transaction at the till.
Rather than an old-fashioned Van Court's Recorder, a merchant could go about this by installing card-reading software that would quickly determine both the card being used and the appropriate surcharge. Thus, consumers who paid with premium cards, those cards that offer the most bundled services per transaction (and therefore incur the highest costs), would have to bear the largest surcharge while those with bare bones cards would pay a minimal surcharge. Anyone paying in cash, much like those who payed with local banknotes during the free banking era, would not incur any surcharge whatsoever. This system would ensure that each customer bears the marginal cost of their chosen means of payment. The retailer, who routes all of the surcharges they collect back to the card network, doesn't eat any costs and therefore succeeds in preserving their margins.
If only things were that easy. Though surcharging would be a great way to deal with the "multitude of dollars" problem, card networks like Visa and Mastercard have typically "legislated" against surcharges.* The networks can successfully impose this no-surcharge obligation on retailers because as an oligopoly, Visa and Mastercard can banish offenders from the network, putting a huge dent in the offender's sales. Why prevent surcharges? One reason the card networks probably do this is because they don't want the card-paying public to feel that they are being penalized in any way. If the feel put off, consumers might choose alternatives like cash and debit that aren't subject to surcharge.
Another solution: discounting
The no surcharge rule puts retailers in a bind. They are obligated to collect fees on behalf of the card networks, but without the ability to surcharge they're left absorbing the costs imposed by the networks while the customer enjoys all the benefits.
There's a neat way that that retailers can get around this hurdle. All they need to do is to mark up all their sticker prices to the level of the highest cost credit card, and then offer discounts to everyone who uses lower cost credit cards, debit cards, or cash. Discounting allows the merchant to collect the appropriate fee from each customer, funneling these fees back to the networks. As before, a given set of prices can accommodate a wide range of dollar payments media. Each party who enjoys a given marginal benefit also bears its respective marginal cost.
So as not to leave our analogy hanging, if this solution had been chosen by free-banking era retailers (perhaps because the free banks insisted that merchants avoid bank note surcharges), then the price level in Philadelphia would have been marked up to the value of the most-distant bank notes in circulation, say those from Chicago. Those paying in less-distant bank notes, say New Jersey notes, would have received an appropriately-sized discount.
An anomaly: we don't see discounts
Having outlined how to solve the modern "multitude of dollars" problem without surcharging, what happens in the real world? An odd phenomena tends to play out. While retailers have certainly marked up prices to a premium card standard (or thereabouts) in order to cover their costs, for some reason they rarely offer their customers discounts on cheaper payment options. Try asking for a cash discount at Walmart the next time you visit. This means that anyone who purchases something with cash, debit, or a bare bones credit card is being forced to pay for a juicy set of benefits associated with usage of an American Express card, namely fancy rewards and dispute rights, without actually getting to enjoy those benefits. Put differently, the merchant is effectively overcharging their customers by collecting more network fees than the networks actually require, keeping this excess to pad their bottom line.
Why this predatory behavior? Briglevics and Shy note that merchants may be wary of discounting if it creates confusion and distrust among customers. Potential delays at checkout counters might impose an extra set of costs on all parties. They also point out that merchants may not find it profitable to offer a cash discount to consumers who would use low-cost payments anyway. Schuh, Shy, Stavins, and Triest report that merchants may lack complete information on the full and exact merchant discount fees for their customers’ credit cards, and therefore can't implement a policy of accurate discounting.
Could it be that the right set of tools to provide discounts hasn't yet been created? Perhaps we need a modern version of Van Court's Bank Note reporter. Such a technology would allow merchants to rapidly determine the proper discount to apply to each disparate dollar type and clearly inform customers about the saving they have enjoyed.
Lack of technology may explain why cheap credit cards don't receive discounts relative to expensive ones, but it doesn't explain why cash discounts has never been adopted by retailers. One theory is that even if certain retailers start to offer discounts, the public may be too overloaded with information to switch, thus allowing the practice of predatory pricing to remain the status quo. Supporting this view is the following observation: while discounting for cash and debit payments is rare in most sectors of the economy, it is quite common among gas stations, as the image below shows.
Other retailers, say department stores, sell a wider variety of things than gas stations, many of these items only being purchased from time to time. This makes comparison shopping more costly. Brand loyalty only increases the hassles of switching. Department stores may find that a policy of cash discounts is simply not worth the effort as discounts get lost in the morass of data that a consumer is bombarded with on an hourly basis.
That being said, the online world's ability to provide faster cross-retailer comparisons than are possible in a bricks & mortar world could shake things up. Surely some smart fintech entrepreneur can create a way for online merchants to rapidly measure the appropriate discount (or surcharge in those jurisdictions that allow it) to apply to each card before consummation. This same tool would provide a user-friendly format for online shoppers to "see" competing card discounts across a number of merchants prior to hitting the buy now button. Just like they'll cross the street to hit the cheapest gas station, they may divert their purchase to the lowest cost website. If this sort of thing caught on, we'd see long forgotten free banking customs replicated in our modern era.
*This is currently the case in Canada. In Australia, merchants have been allowed to surcharge since the early 2000s. US merchants recently won the right to surcharge, although it is probably too early to know what effect these rule changes will have.
Tuesday, June 16, 2015
On Twitter, Noah Smith asks:
He answers his own question on his blog. I pretty much agree. It's interesting to imagine science fiction worlds where people do use volatile instruments like stock as their medium of exchange. Why would people in these worlds be willing to adopt volatile media while people in our world don't?@ClicheGuevara But why don't we use stock shares as our medium of exchange?— Noah Smith (@Noahpinion) June 13, 2015
Liquidity, the world's best insurance policy against uncertainty
First, we need to understand why our world has a preference for stable media of exchange. As Noah points out, people don't know exactly when they are going to need to spend money, or how much. Any individual faces the dizzying fact that any of an infinite number of events could hit them at any point in time. People have many ways to cope with this chaos, one of which is to build up an inventory of assets that can be deployed to help deal with surprises as they occur. Liquid assets—those that can be sold quickly, at low cost, and spent along multiple pathways, will do a better job of this than illiquid assets—those that take time to sell, incur transaction fees, and lack multiple pathways.
A buffer stock of liquidity offers people the same set of services as actual insurance, say a policy issued by GEICO. A home insurance policy immunizes against a range of disasters that might befall someone's residence. So does liquidity, since it can rapidly purchase materials and labour. If home insurance is cheaper than holding an inventory of liquidity (the cost of which, as Noah says, is an inferior expected return), then people will skimp on liquidity. Unfortunately there are only a limited range of future disasters against which people can purchase insurance. Good luck getting GEICO to insure you against a zombie outbreak, for instance. If a zombie scenario is something that you put a non-zero probability on, then staying a little more liquid than you otherwise would may alleviate some of your concerns. Come outbreak, the ability to rapidly dispatch liquid assets in all directions will come in handy.
Back to the question of volatility. If people are trying to build a moat against uncertainty, what use would insurance be if it offered $20,000 in protection on Tuesday, $17,000 on Wednesday, and $23,000 on Thursday? For the same reason that people require a fixed amount of insurance rather than a floating amount when constructing their moats, they will want to invest in stable liquid assets rather than volatile liquid assets. You can't solve for uncertainty with more uncertainty.
Liquidity is a virtuous circle
Liquidity is a virtuous circle; as an asset gets more liquid it becomes more attractive as an insurance policy, which brings in more buyers, which makes it more liquid, which increases its value as insurance, and so on. Once everyone holds the most-liquid asset(s), they have joined what in essence is an economy-wide mutual insurance scheme. At this point it makes little sense for a merchant to accept less-widely held assets as payment. Not only will it be a nuisance to set up the infrastructure, but the merchant runs into the coincidence of wants problem. Because the merchant's employees are already paid in the most-liquid asset, should the merchant accept volatile assets as payment he/she will have to bear the cost of converting them back into the standard unit. To avoid these inconveniences, only widely-held assets will be accepted by the merchant. A payments standard has developed.
Merchants will also try to please customers by setting their sticker prices in terms of the liquid asset. This reduces the calculational burden imposed on customers. At this point the asset has become a unit-of-account, and a pricing standard has developed. Adoption as unit-of-account provides all sorts of extra benefits to owners of the standard unit. As a courtesy, grocers and other retailers will typically keep their prices fixed for a few days, or weeks, which means that anyone who owns the standard unit knows ahead of time approximately how much food they'll be able to purchase. The tendency for prices to be sticky in terms of the unit of account only increases the standard unit's usefulness as a universal insurance policy.
The fact that shares aren't the unit-of-account means that consumers who own them miss out on all the uncertainty-alleviating benefits of sticky retail prices. They have no clue what their purchasing power will be one minute hence, let alone the next day.
Worlds with volatile money
If we lived in a world where we didn't need liquidity to shelter us from uncertainty, then we might be more receptive to using volatile media of exchange.
For instance, I sometimes wonder if the demand for dollar-denominated liquidity is less in Canada than in the U.S. since we Canadians have universal health care. With the nagging concern of how to pay for potentially life saving medical services solved for, we can economize on inventories of stable liquidity and seek out higher returns. Americans, who don't have such a product, probably need to hold larger dollar-denominated hoards in order to alleviate their queasiness about how they'll have to deal with future bodily harm.
Taking this idea to its limits, if an insurer were to introduce an "everything" insurance product (yes, this is science fiction), and it was cheaper than holding the standard liquid unit, then people would no longer need to self-insure against uncertainty by depending on the standard unit, typically central bank money and its banking derivatives. Instead, everyone would migrate their savings over to more volatile instruments like stocks, ETFs, or bitcoin, enjoying what Noah refers to as higher drift, or long-term returns. As long as GEICO is providing a low-cost universal salve against uncertainty, than people will be willing to shoulder all the inconveniences of fluctuating purchasing power insofar as it offers them a bit more drift.
With ownership of low volatility units (dollars, yen, etc) far less ubiquitous than before, and volatile asset (stocks, ETFs, bonds, etc) ownership much more prevalent, it might now make sense for merchants to incur the set-up costs of receiving volatile assets in payment. Furthermore, the fact that their employees will accept these volatile assets as salary, thus solving the coincidence of wants problem, will only accelerate the willingness of merchants to install the requisite infrastructure. After all, merchants can now pay their employees with the same ETF units that they receive from their customer, thus saving both commission expenses and the cost of incurring the difference between the bid and ask price. 
Information as a moat against uncertainty
In addition to liquidity and insurance, people can build their moats by acquiring more information. This is something I learnt from David Laidler. (If you want to learn about money via a fascinating detour through the history of monetary thought, you can't go wrong with these three books.) Knowledge allows individuals to better anticipate the future and plan accordingly, thus reducing the need for either liquidity or insurance contracts as hedges. If some cheap technology were to emerge that offered an infinite amount of free information (i.e. the internet?), then maybe people would cease amassing stable liquidity altogether. In this world, people's preferred solution to uncertainty would be to costlessly inform themselves and hold high return/high drift assets like stocks rather than stay uninformed and hold inferior, low return liquid assets.
Of course, we could also argue the opposite; that the internet has solved for one set of risks only to bring in a new set (identity theft, viruses etc), thus perpetuating the necessity of owning both GEICO insurance and liquid assets.
In sum, because our world lacks both infinite information and universal "everything" insurance policies, stable liquidity remain one of the cheapest ways to inoculate against uncertainty. The widespread prevalence of this monetary insurance across all strata of society has encouraged the development of a payments and pricing standards based on these assets. These standards further militate against the emergence volatile assets like bitcoin and shares as widespread media of exchange.
 I added this paragraph on June 19.
Friday, June 12, 2015
Bitcoin was supposed to result in the bitcoinization of the world; instead, I'd argue that the world of bitcoin is being dollarized.
A successful medium of exchange will be used by four types of actors: retailers, consumers, financial intermediaries, and speculators. In bitcoin's case, the inherent volatility of the cryptocurrency militates against its adoption by anyone other than speculators, leaving dollars as the default option.
Let's start with the first bit of the equation; retailers, or merchants. Entrepreneurs who have been trying to bring bitcoin to the mainstream have discovered that while merchants like the idea of allowing consumers to pay with bitcoin, the merchants themselves refuse to deal in the stuff. Instead, upon receipt of bitcoin, a merchant's bitcoin payments processors—usually an intermediary like Bitpay or Coinbase—will instantly convert bitcoin into U.S. dollars on behalf of the merchant. Retailers choose to dollarize rather than bitcoinize because they are afraid of bitcoin's volatility, and justifiably so.
The next bit of the equation is the consumer. I argued in my previous post that it is dubious whether paying in bitcoin offers mainstream consumers any benefits relative to dollar payments. People who buy with bitcoin must incur added costs in the form of trading fees if they wish to move from dollars into bitcoin. They must also bear the burden of bitcoin's volatility until the moment of making a purchase.
To drive mainstream consumer adoption of bitcoin, those intermediaries who are servicing buyers will have to begin offering the same sort of volatility-shielding services that bitcoin payments providers currently offer to merchants. A permanently shielded wallet, for instance, would allow consumers who want to make a purchase the opportunity to store value in U.S. dollar terms until the very last moment, at which point the intermediary takes on the bitcoin risk and consummates the deal. Of course, this only pivots things further towards dollarization, not bitcoinization.
The inevitable product that emerges will be a just-in-time bitcoin solution. Buyers have the benefits of holding dollars up until the moment at which they press the Buy Now button, at which point the intermediary takes over and sells their dollars for bitcoin. The switching of the payment from the dollar rails onto the bitcoin rails is only momentary. Upon receipt of the bitcoin payment an instant later, the merchant's payments provider will immediately sell the bitcoin and deliver dollars to the merchant. With both buyer and seller choosing to dollarize, neither has to suffer from bitcoin's volatility. However, they still get to enjoy whatever cost savings are provided by rapid just-in-time usage of the bitcoin rails. Only speculators and intermediaries, who have now taken on the responsibility of dealing in bitcoin from consumer and retailers, have avoided dollarization.
But hold on. If all parties to the transaction only want dollars, why not just cut bitcoin entirely out of the equation? Instead of a just-in-time swap of bitcoin, the intermediaries involved—the buyers' bitcoin wallet provider and the merchant's bitcoin payments processor—can get together and agree to exchange a dollar IOU instead, saving them the hassle of dealing in bitcoin. Gone are the exchange fees, the obligation to pay bitcoin's bid-ask spread, and slippage that might occur if bitcoin's price weaves dramatically as the transaction is going through. To spare readers the gritty details, the footnote below describes how bitcoin intermediaries might fashion a U.S. dollar IOU trading network.
This puts these bitcoin intermediaries in the rather odd position of no longer being part of the bitcoin universe. Instead, intermediaries have become like interlinked Paypals, offering U.S. dollar accounts to consumers and U.S. payment solutions to merchants.
Thus, in an effort to promote mass adoption of bitcoin, we've somewhat perversely arrived at the opposite, an all-out dollarization of what was supposed to be a bitcoin retail payments network. Buyer and merchant hold only dollars, and so-called bitcoin intermediaries like wallets and payments processors no longer deal in the stuff. That leaves only speculators to hold the bitcoin bag. This system of individual PayPals will be built on top of the very infrastructure that bitcoin was designed to tear down, namely the existing dollar rails run by incumbent banks and underpinned by the Federal Reserve.
This isn't to say that bitcoin is a failure as a retail payments option. But I have troubles seeing it ever going mainstream. Even if bitcoin continues to exist as an arcane niche payments system for a community of like minded consumers and retailers, that's still constitutes quite a success, albeit one that has not lived up to many people's dreams.
In writing this, I stumbled on an earlier post by Guan that already arrived at a similar conclusion. If you've already read his post, my apologies for wasting your time and making you read mine.
The gritty details: Rather than trading bitcoin to settle payments between consumers and retailers, intermediaries can simply trade dollar IOUs. Costs should be lower than settling in bitcoin. To begin with, bitcoin intermediaries will have to maintain U.S. dollar clearing accounts with all the other bitcoin intermediaries. Over the course of a trading period, dollar payments will flow into an out of these clearing accounts. At the end of the day, each intermediary's account will be netted and cleared against all other intermediaries' accounts. The result is that some intermediaries will be owed dollars, others will owe. These balances will all be settled that very evening on the books of an underlying commercial bank, say Citi, where all intermediaries also maintain accounts. Since accounts are settled on the books of Citi, intermediaries needn't incur expensive inter-bank wire transfer fees. Citi has, in effect, become the central bank for a bitcoin based payments system, sans the bitcoin.
Friday, June 5, 2015
It's pretty simple, really. For any medium of exchange to displace another as a means for buying stuff, users need come out ahead. And this isn't happening with bitcoin.
We can break any exchange medium's user base into consumers and sellers. Now we know that sellers love bitcoin—they've been adopting it at a blistering pace, from Amazon to Microsoft to CVS. No wonder when we consider the cost savings they enjoy. A merchant is required to pay around 1.5-2.0% for each credit card transaction. Bitcoin payment processors like Coinbase, Bitnet, and Bitpay charge just 0.5% while simultaneously absorbing all of merchant's forex risk. A retailer with $1 million in sales that converts all of its shoppers from Visa/Mastercard payments to bitcoin has just earned themselves $10,000. It's a no-brainer.
While sellers are jubilant, consumers aren't. Tim Swanson shows that bitcoin payments haven't budged in over a year with bitcoin processor Bitpay's transactions volume amounting to a piddling $57.5 million or so in 2014 (not including precious metals and mining). Bitpay controls at least a third of the payments market. That's what failure looks like, folks.
I think that this aborted takeoff can be blamed on the fact that the dominant consumer payments medium, the credit card, leaves the consumer with significantly more resources after each payment than bitcoin does. Consider the fact that consumers are always paid in U.S. dollars (or whatever their respective national currency happens to be). At the same time, sellers price their wares in dollars and accept payment in that unit, the dollar being both the dominant unit of account and medium of exchange. This is highly convenient to consumers. If someone wants to buy an annotated hard cover edition of War & Peace for $100, they never have to leave the dollar ecosystem.
Paying in bitcoin, however, means that the consumer must endure the cost of exiting the dollar ecosystem and entering the bitcoin ecosystem. One portion of this cost is comprised of the fixed non-recurring expense of learning how to set up the dollar-to-bitcoin portal. The next portion has to do with the commission that a bitcoin exchange will extract from the consumer for buying bitcoin, around 0.5%. At the same time, a consumer will have to pay an additional cost as they reach across the spread between the bid and ask price in order to amass the requisite bitcoin. Finally, consumers must bear the cost of coping with the incredible volatility of the stuff. In order to preserve the U.S. dollar purchasing power of the bitcoin up to the point of purchasing the $100 edition of War & Peace, the consumer needs to buy insurance. Either that or bear ghastly bitcoin exchange rate risk.
You can see why credit cards come out ahead. They are easy for the consumer to setup, they do not extract a foreign exchange commission, nor do they force users to bear any exchange rate risk. Let's work out the numbers. If a consumer earns $100 in salary and want to buy War & Peace for $100, a credit card provides them with enough purchasing power to consummate the deal. However, if they try to buy that same item with bitcoin, they won't be able to afford it. Assuming it costs 50 cents to buy bitcoin and 50 cents to hedge the price risk until the point of consummation, they need to earn at least $101 to afford War & Peace. It's out of reach.
There are ways to modify this setup so that War & Peace is brought back into the reach of the bitcoin paying consumer. Let's assume that bitcoin advocates are right and that the total resource cost of maintaining a bitcoin-based payments network is cheaper than running the credit card network by a significant wedge. The above calculations show us that, at the moment, consumers don't enjoy any of this wedge. In order to induce consumers to make the leap from credit cards, bitcoin sellers and payments processors have to share the savings with them.
Sellers can provide part of this inducement by introducing a lower U.S. dollar sticker price for bitcoin payments. Here's how it works. Our seller maintains their offer to sell War & Peace at $100 for credit card users but drops the price by seventy-five cents to $99.25 for bitcoin users. Let's further assume that the seller pays $2.00 in fees to the credit card network but only 50 cents to its bitcoin payments provider. Despite having discounted War & Peace's bitcoin price, the seller still comes out ahead for each switch from from credit card to bitcoin payments. Each bitcoin sales nets them $98.75 ($99.25 minus 50 cents), but each credit card payment only nets them $98.00 ($100 minus $2). Since they earn an extra 75 cents if they use the bitcoin payments ecosystem, sellers still have an incentive to adopt bitcoin payment.
The subsidy provided by the retailer reduces the consumer's overall cost of using the bitcoin ecosystem. As before, our consumer earns $100. Given the reduced $99.25 sticker price, a 50 cent fee to buy bitcoin, and a 50 cent fee to buy insurance, their net cost has fallen to $100.25 from $101. Its still out of their reach, but not by as much.
The bitcoin payments processor can join the merchant in providing consumers with an inducement. Say that for each transaction the processor pays the consumer a cash reward of 25 cents out of the 50 cents they earn from the retailer in fees. Let's rework the numbers. Given the $99.25 sticker price, a 50 cent fee to buy bitcoin, a 50 cent fee to buy insurance, and a 25 cent rebate, the consumer's net cost has fallen to $100. Paying for War & Peace with bitcoin is now competitive with a credit card. Only now does it makes sense for a consumer to make a leap from the credit card rails onto the bitcoin rail. If merchant and processor can afford to add even more inducements, consumers will switch to bitcoin all the faster.
As an aside, some readers may have noticed I haven't included credit card rewards (i.e. points, air miles, and cash back) into my calculation. I'm making what I think is a pretty fair assumption that no one gets something for nothing. Those running the credit card system fund the rewards they pay to consumers by charging merchants a higher fee. To preserver margins, merchants will build this fee into the U.S. dollar price of the products they sell. This means that the value of rewards that the average card payer earns is entirely canceled by the higher price premium, effectively driving their benefit to zero.
Back to bitcoin. Inducing participation from the consumer isn't a technical problem, it's coordination problem, one that bitcoin entrepreneurs haven't seemed to figure out yet. As far as I can tell, retailers are not providing visible bitcoin price discounts in U.S. dollar terms, nor are payments processors like Bitpay and Coinbase providing consumer's with rewards. By focusing on offering merchants a superior product and omitting the consumer side of the equation, bitcoin entrepreneurs are trying to lure the cat into the door whereas a true tipping point requires going after the tiger.
Alternatively, they may not be going after the tiger because they can't. The ability of bitcoin payments processors and retailers to induce participation from consumers depends on the size of the wedge. If a bitcoin payment system does not provide any resource cost savings, then there is no kitty from which to buy consumer participation. In which case, long live Visa and Mastercard.
There is a misguided view out there that the problem of coaxing consumers into the bitcoin loop will solve itself as bitcoin's volatility disappears and trading costs fall, thus reducing the average consumers' costs of engaging with the cryptocurrency. See the founders of Coinbase here, for instance. This view is wrong. Take trading costs. Even if bitcoin trading commissions fall to zero, there will always be a bid ask spread that consumers will have to endure in order to get bitcoin, and therefore some disincentive to switch away from cards.
As for volatility, the only way bitcoin will ever shake it's toing and froeing is if it is pegged to the dollar by some powerful organization. Not likely. Nor will increased participation flatten out bitcoin's screaming ups and downs. Unlike stocks, gold, or U.S. dollars, bitcoin lacks a non-monetary stabilizer (see here and here). Put differently, its price is indeterminate. More buyers and sellers participating in bitcoin markets will not change this fundamental fact. So contrary to hopes that Bitcoin will become more cuddly, its future is destined to be a frenzied one. Unless consumers are compensated for bearing this volatility, or shielded from it, they'll keep using cards. If they can, retailers and payments processors should be trying their best to subsidize these costs. Without such subsidies, bitcoin is unlikely to ever achieve liftoff.
I started to write this post a few days ago. This is a snippet of the original: "Bitcoin's inability to achieve mass consumer adoption is a good indicator that it will never take off." So you can see that I've changed my mind in writing this post.
Tuesday, May 26, 2015
This post is about the Alberta Prosperity Certificate, one of the world's stranger monetary experiments. Issued in late 1936 and early 1937 by the newly-elected Alberta government, these monetary instruments are the largest-scale example of Silvio Gesell's "shrinking money," or stamp scrip, in action. Gesell, a German business man and self taught economist, had written a treatise in 1891 in which he described a currency that depreciated in value, thus preventing hoarding and encouraging spending.
To make this more interesting, let's jump forward in time. In 2014, Greece's Finance Minister Yanis Varoufakis wrote a blog post that described a new Greek financial instrument that could be used to make payments while circulating in parallel with the already-existing euro. Varoufakis's post, combined with constant rumors that Greece may be planning to issue its own parallel currency in order to make internal payments,* means that a revisitation of Alberta's early dalliance with scrip, which circulated concurrently with Canadian dollars, is more relevant than ever. The attempt by Albertan authorities to issue scrip 80 years ago would end in failure; most of the paper refused to stay in circulation. Understanding why this happened provides some insights into what sorts of conditions might promote the success of a Greek parallel currency—or its downfall.
Virginius Frank Coe
The best source on Prosperity Certificates is a 1938 survey by Virginius Frank Coe, an American economist who visited Alberta in August 1937, five months after the program had been abandoned. Coe's life is interesting enough to deserve its own tangent. An economist educated at the University of Chicago, Coe would go on to hold a number of important positions in various U.S. government institutions both during and after World War II, including monetary research director at the Treasury Department. This brought him into the orbit of Harry Dexter White, then the Assistant Secretary of the Treasury and the architect of the Bretton Woods agreements. Coe himself was a representative at Bretton Woods and would go on to become secretary of the International Monetary Fund in 1946, nine years after having written his Prosperity Certificate paper.
Readers of Benn Steil's The Battle of Bretton Woods will know that much of the evidence incriminates Harry Dexter White as spying for the Soviets, an accusation White himself denied. The same sources who named White as a Soviet agent also fingered Coe, and in 1952 Coe was forced to resign from his post at the IMF. He would appear in front of the McCarran Committee later that year, pleading the fifth in response to all questions posed to him, and would later face Senator Joseph McCarthy. His passport revoked, and unable to find work in the U.S., Coe headed to China to serve as an adviser to Mao until his death in 1980.
Coe's Prosperity Certificate paper betrays the author as someone with a strong interest in alternative monetary systems. While we can't know for sure if his interest in alternative systems extended as far as being a Soviet mole, we shouldn't let this possibility detract from what is otherwise an excellent account of this early Canadian monetary experiment.
Alberta and Social Credit
Coe describes an Alberta electorate that is facing the same economic backdrop as Greece's voters did prior to the recent election of Syriza. Just as Greeks had endured seven years of famine prior to the 2015 election, Albertans going into the 1935 election had been beset by seven years of distress associated with low farm prices and bad crop yields. The incumbent United Farmers of Alberta government was not willing to implement the more drastic policies that the Albertan electorate demanded, says Coe. Into the void stepped William Aberhart, a pastor and newly-recruited believer in the tenets of Social Credit. Dreamt up by British engineer C.H. Douglas, the idea behind Social Credit was to create a more equal society by augmenting consumers' purchasing power via the payment of a national dividend. Aberhart formed the Alberta Social Credit party in 1935 and won the election a few months later. In electing Syriza, the Greeks, like the Albertans before them, have entrusted their future to a party of political novices.
Reading Coe, one gets the sense that the Aberhart government stumbled into Prosperity Certificates rather than purposefully selecting them as a policy. Gesell's dated stamp scrip was a rival monetary reform to Social Credit, not a complement. Why turn to a non-Social Credit policy? It seems that several months after coming to power, the new Social Credit government was already splintering as one faction had grown impatient with Aberhart's inability to implement economic changes. Coe, speculating that the decision to implement dated stamp money was a token gesture to demonstrate forward momentum and heal internal rifts, says that "any one of a number of plans would have done as well." If a non-Social Credit monetary scheme such as Gesell money were to fail, at least a Social Credit policy option still had a kick at the can. The implication that the government didn't put much thought into the design of the certificates finds some confirmation in the fact that the Free-Economy League, an organization formed by Gesell, published a criticism of the Alberta government's procedure for creating Prosperity Certificates and predicted their failure.
How the certificates worked
Here's how Alberta's stamp scrip worked. In early August 1936, when the program debuted, an unemployed Albertan was paid, say, a $1 certificate for each $1 worth of road maintenance work rendered. This certificate was to be redeemed by the Alberta government two years hence, or in August 1938, for $1 in Canadian dollars. However, redemption required that the certificate have 104 stamps affixed to it (see figure above). Each week during that two year period, the owner of the certificate was to buy a government stamp for 1 cent from an approved stamp dealer and glue it to the note.
The necessity of buying stamps created a fairly onerous fee on cash holdings. As such, any laborer who received the scrip from the government was unlikely to hoard it, preferring instead to spend it on, say at a retailer, who in turn would only accept scrip as payment for goods and services if the correct number of stamps has been affixed. In order to avoid the cost of buying the next weekly stamp in order to keep the scrip current, the retailer themselves would quickly offload it to their suppliers and so on.
The 1 cent stamp fee was collected by the Alberta government and held as a reserve for redemption in two years. With 104 cents being collected over each $1 certificate's life time, this meant that the scheme was entirely self financing. The extra four cents represented a profit to the government.
We know that the Prosperity Certificate scheme didn't work. The certificates began to be paid to unemployed Albertans in August 1936 for roadwork rendered in July. According to Coe, the maximum amount of outstanding certificates in circulation in August and early September was $239,391 (around $9 million in current dollars). However, by mid-September 1937, just one month after the program's debut, over 60% of the certificates outstanding, or $144,280 out of $239,391, had ceased to circulate.
Where had they gone? The government now held them. The reason for this development was a last minute decision by Aberhart to offer monthly redemption of certificates at par in Dominion currency (i.e. $1 in certificates for $1 in Canadian bills). This short-circuited the original two-year life of the certificates. Rather than continuing to pass the scrip along to the next Albertan, Albertans leapt at the government's offer and converted en masse when the first redemption date presented itself in early September.
In the end, the government might as well have paid for work rendered using Canadian dollars, since the net effect of paying in either Certificates or Canadian dollars was the same. As Coe says, "the dated stamp scrip was in the end little more than a small nuisance." Subsequent issues of scrip were small relative to the original August 1936 issue and the government officially ended the program in April 1937.
"The problem of the wholesalers"
In the planning stages of the program, government officials ran into what Coe refers to as the "problem of the wholesaler." The first to receive the certificates would be farmers on relief, who in turn would make payments to retailers. The payments by retailers would primarily flow to Albertan wholesalers whose dominant payments were to manufacturers and others outside the province. However, those outside the province would not accept Prosperity Certificates, requiring instead hard currency, or Canadian dollars. The Albertan wholesaler would be left holding the bag, so to say, having acquired the entire issue of Prosperity Certificates with no outlet. According to Coe, wholesalers and large retailers were vocal in their opposition to the plan, which they expressed through trade associations and in the press.
One way of solving the wholesalers' problem would have been to establish an exchange market such that wholesalers could sell certificates in order to buy the necessary hard currency and thus fund out-of-Province imports. Banks would normally be an important party to the creation of such a market. Irving Fisher, who wrote a book on stamp scrip, entitled one paragraph "Have at Least one Bank." But the banks who operated in Alberta refused to participate in the Prosperity Certificate scheme—no wonder given that one of the Social Credit party's planks advocated the removal of the "banking monopoly" on the issuance of credit. The tenets of Social Credit thus interfered with the execution of Gesell money, impeding the latter's success.
Even if such a market were to be created, chances are that it would have priced the Certificates at a large discount to Canadian dollars given the onerous fee on certificates relative to Canadian notes and the inferior credit of their issuer. After all, by then the Alberta government had defaulted on its international obligations whereas the Federal government's credit was still good. Such a discount would have been at odds with the Alberta government's policy of using a dollar's worth of certificates to buy one Canadian dollar's worth of labour. If the certificates were trading at 69 cents on the dollar in the wholesale market, workers paid in scrip would be loath to accept them at face value, for if they did, they would probably have problems passing them off at retailers for that amount.
In the end, the government's solution to the problem of the wholesalers was to allow wholesalers (and even retailers) to benefit from free monthly redemption at par. As I noted earlier, this resulted in most of the certificates being returned for redemption just a few weeks after having been issued.** Rather than bad money driving out the good, a garbled version of Gresham's Law had taken hold in Alberta, which Coe describes thusly: "Bad money obviously does not drive out good money when the government is willing to redeem the bad money in good money."
This garbled version of Gresham's law is a phenomenon I've described before to explain a number of monetary puzzles including the failure of the Susan B. Anthony dollar, the European Target2 bank runs of 2011-12, the proliferation of credit cards, and the zero-lower bound problem. See here and here.
What about Greece?
Alberta in 1936 and Greece in 2015 are in similar situations. Both are non-currency issuers within a larger monetary zone, in Alberta's case the Canadian dollar zone and in Greece's case the Eurozone. Both have awful credit. Neither is part of a larger fiscal union. In Greece's case, the mechanism hasn't yet been created whereas in Alberta's case, the Social Credit party was at such odds with the Federal government and the rest of Canada that it could not expect much help.
I'd argue that anyone planning to introduce a Greek parallel currency to circulate alongside euros faces the same problem that Alberta faced; the so-called problem of the wholesalers. If the Greek government starts to pay employees and contractors in Greek parallel IOUs denominated in euros, and employees buy stuff at stores with those IOUs, and stores purchase inventory from wholesalers, these wholesalers will need a mechanism to offload their parallel note surpluses in order to get euros to buy foreign imports. The IOUs can either find their own price, in which case they will most likely trade at a large and varying discount to euros, or the Greek government can offer one-to-one convertibility. They can do this by either redeeming IOUs directly for euros or allowing one euro worth of taxes to be paid with an equivalent number of IOUs.
Neither solution is ideal. If the IOUs trade at a variable discount to euros, then their ability to serve as a competing medium of exchange will suffer. People always prefer to trade using the medium in which a nation's prices are expressed, or, put differently, the medium which functions as a unit of account. For example, people see benefit in the fact that one euro will always discharge a euro's worth of Greek debt or a buy a euro's worth of Greek olive oil. But as long as Greek IOUs trade at a varying discount to euros, it is impossible to know ahead of time how many IOUs will discharge a euro's worth of debt or buy a euro's worth of oil, given that the euro will surely remain Greece's unit of account. This would hinder the IOU's ability to function as a currency. The fact that people prefer to accept stable exchange media in trade to unstable media is one of the reasons that bitcoin hasn't caught on.
So rather than serving as a competing medium of exchange, the parallel IOUs will probably function as illiquid and highly risky speculative fixed income securities. In order to compensate recipients of IOUs for this lack of liquidity, the Greek government will have to issue the IOUs at a larger discount to par than they would for an otherwise liquid equivalent, thus increasing the government's financing costs.
This lack of liquidity militates against one of the key selling points of a Greek parallel unit, which is to finance the government by displacing some of the existing circulating medium of exchange, euros, from citizens' wallets. Preferably, unwanted euros would trickle back to the European Central Bank to be cancelled, reducing the ECB's seigniorage but augmenting the seigniorage of the Greek state as Greek IOUs rush in to fill the void. However, if the new Greek parallel unit cannot compete with the euro's liquidity, then there will be very little 'space' for Greek IOUs to occupy in Greek portfolios, and little relief for beleaguered government finances.
If the Greek government tries to promote the liquidity of its parallel currency by having the units trade at a fixed one-to-one rate with euros, then the same garbled version of Gresham's Law that took hold in Alberta would overwhelm Greece. In Coe's words, the Syriza government's willingness to buy bad money, or parallel currency units, from the public with good money, or euros, will promote mass conversion into euros and thereby drive all the bad money from circulation. Greek parallel units will cease to exist.***
In sum, anyone planning a Greek parallel currency faces a conundrum. In order to pay its bills the government can do little more than introduce a volatile asset that trades at varying discount to euros. This asset's volatility and relative illiquidity won't make it very popular with its recipients. An attempt to render that asset more acceptable in trade by setting a one-to-one conversion rate to the euro will result in a short-circuiting of the scheme as everyone races to redeem IOUs. The issuance of parallel currencies seems like a hard battle to win.
*There are a number of plans including that of Biagio Bossone & Marco Cattaneo, Thomas Mayer, and Robert Paranteau
** Compounding the problem was that redemption at face value put a premium upon redemption, says Coe. "The holder who redeemed received face value; the person who did not redeem ran the risk of losing 1 per cent of the face value if he failed to pass the certificates within the next few days, and more for longer periods. This premium placed upon redemption could only have been eliminated by redeeming the certificates at a discount of more than 1 per cent-say, 2 or 3 per cent." So the government accidentally created an even greater incentive for certificate owners to redeem.
*** This is particularly damaging in Greece's case at will result in a perpetual draw down in the state's euro balances. These reserves are vital since the Greek government needs to service its (existing or renegotiated) Euro debts to the IMF and pay external suppliers, and can only do so with hard currency.
Thursday, May 14, 2015
The FT makes a hullabaloo out of Greece using special drawing rights (SDR) to pay the IMF earlier this week, referring to the step as "unusual." Zero Hedge predictably grabs the baton and runs as far as it can go with the story.
It's a good opportunity to revisit the SDR, a topic I last wrote about back in 2013.
The FT claims that the payment of SDRs to the IMF is "the equivalent of taking out a low-interest loan from the fund to pay off another." Here the FT has committed cardinal error #1 when it comes to understanding how SDRs work—SDRs are not lent out by the IMF.
I like to think of the SDR mechanism as comprised of 188 lines of credit issued to each of the IMF's 188 members. These lines of credit are denominated in SDR and apportioned according to each countries' relative economic size. Any line of credit needs a creditor. In the case of SDRs, who fills this role? Why, the 188 members of the IMF do. The SDR system is a mutual credit system, or what I referred to in my older post as the world's largest Local Exchange Trading System, or LETS. Where does the IMF stand in all this? It is simply an administrator of the system. So by paying the IMF in SDRs, the Greek government isn't taking out a low-interest loan from the IMF—rather, it's drawing down on the credit provided to it by 187 other countries. As for the IMF, it isn't getting another Greek-issued debt instrument. Rather, it is getting a mutual liability of 188 nations.
The second sin in the FT's article is the assumption that SDRs are "rarely tapped" and that therefore, Greece is doing something unusual in "raiding" its SDR account. As a quick glance to the data shows, that's simply not the case. The chart below (apologies for its extreme height, but it's the only way I can visualize the data) shows that over time, countries have tended to spend down their SDR lines of credit. Any nation to the left of the 100% line (and illustrated in light blue) has drawn down on their credit line while those to the right (illustrated in darker blue) have accumulated SDR surpluses. Most countries lie to the left of the line. Greece, which after this week's transaction has just 5% of its total line of credit undrawn*, joins Macedonia, Iceland, Hungary, Serbia, Ukraine, and Romania near the low end of the range, many of whom drew down their balances to deal with the after-effects of the credit crisis.
Nor is the FT article right in implying that it is unusual for countries to pay the IMF in SDRs. Consider that since the SDR's inception in 1969, 204 billion SDRs have been issued to 188 member nations. Logic tells us that each of these 204 billion SDRs must be owned by some combination of member nations, right? Not quite. The 188 nations collectively own only 189 billion SDRs. Who holds the missing 15 billion SDRs? Fifteen institutions, or proscribed holders, have been granted the ability to buy and sell SDRs in the secondary market, including the Arab Monetary Fund, the Bank for International Settlements, and the European Central Bank. Together they own about 1.2 billion SDRs. But the real sop here is the IMF itself, which owns around 13.5 billion SDRs. Because IMF members can use SDRs in transactions involving the IMF, namely the payment of interest on and repayment of loans (see here), the IMF has become the second largest owner of SDRs (after the U.S.).
So in general, the SDR mechanism has been characterized by steady drawdowns of SDR lines of credit by member nations, with surpluses accumulating to the IMF. Far from being unusual, Greece's decision to pay the IMF in SDRs is pretty much par for the course.
One thing I find interesting is that the SDRs that Greece used to pay the IMF are the property of the Bank of Greece, Greece's central bank, and not the Greek government (see here). This means that BoG Governor Yannis Stournaras had to willingly open his pockets to the Greek government to facilitate the IMF payment. In doing so, the central bank has accepted a Greek government-issued liability to pay back SDRs rather than the actual SDRs. As a claim on 187 nations, the latter is surely preferable to the former, which is a claim on a failing nation.
So what about the BoG's other larger unencumbered asset, its gold? According to its most recent balance sheet, the Bank of Greece now owns €5.4 billion of the yellow metal, or 3.62 million ounces. For more on Greece's gold, Ronan Manly has the details. Having just given up his SDRs, would Stournaras be willing to render this gold up to the Greek state in return for a gold-denominated IOU with finance minister Yanis Varoufakis's signature on it? If so, the Greek government could sell this gold on the market for euros to pay the IMF. Settling scheduled June and July payments would be a breeze. This would no doubt be a stain on the BoG's independence, but with the Eurogroup turning the screws, all chips may be in play.
*I'm assuming that Greece paid 517 million SDRs to the IMF, worth 650 million euros at current SDR-to-euro exchange rates.
Monday, May 11, 2015
Measuring liquidity is a pain in the ass.
The value of a good, say an apple, is easy to calculate; just look at the market price for apples. Unfortunately, doing the same for liquidity is much more difficult because liquidity lacks its own unique marketplace. Liquidity is like a remora, it never exists on its own, choosing instead to attach itself to another good or asset. For instance, a bond provides an investor with both an investment return in the form of interest and a consumption return in the form of a flow of liquidity services. Since the price of this combined Frankenstein reflects the value that an investor attributes to both returns, we can't easily disentangle the value of the one from the other.
Here's a symmetrical (and equally valid) way to think about this. If liquidity is a good then illiquidity is a bad, where a bad is anything with negative value to a consumer. This bad doesn't exist on its own but, like a virus, infects other goods and assets. While all goods and assets are plagued by a certain degree of illiquidity, determining the price of of this nuisance—the amount that people will pay to rid themselves of an asset's illiquidity—is difficult because the price of the compound entity combines both a flow of illiquidity disservices as well a flow of positive investment returns.
The only technique we currently have to back out liquidity valuations (or illiquidity penalties) from market data is to find the price or yield differential between two similar instruments, this gap indicating the value that the market ascribes to liquidity (or the negative value of illiquidity). Think identical twin studies in the life sciences. To get a clean differential, the two financial instruments must be "twins," issued by the same entity and having the same maturity. That way any differential between them can't be attributed to credit or term risk, the lone remaining factor—liquidity (or illiquidity)—being the culprit.
The best example is the on-the-run vs off-the-run Treasury spread, the difference in yield between newly-issued 10-year Treasuries and 30-year Treasuries that have 10 years left till maturity. The credit quality and term of these two issues is precisely similar, yet the yield of a newly-issued 10-year Treasury is typically 10 basis points below that of an "off-the-run" equivalent. This gap represents the extra bit of value that investors will pay to enjoy an on-the-run Treasury's liquidity (or, alternatively, the negative value of the illiquid "bad" embedded in an off-the-run Treasury). Assuming that a new bond worth $1000 has a 1.9% yield while an equivalent off-the run issues yields 2.0%, investors are valuing the extra bit of liquidity provided by the on-the-run issue at around $1 per year for each $1000 that they invest.
The first problem with this technique will be familiar to anyone who has tried to conduct studies using twins separated at birth; its very difficult to find twin assets. The second problem is that even if we succeed in locating twin assets, a comparison of them will only reveal the degree to which investors prefer, say, an on-the-run bond's liquidity to that of an off-the-run bond. In other words, it provides us with a relative value. But if we want to find the absolute value that investors place on an on-the-run issue's liquidity (or the absolute disvalue that they place on an off-the-run issue's liquidity), we're left empty-handed. Sean Connery may be cooler than Johnny Depp, but what if we want to calculate Sean Connery's total amount of coolness?
Here's an out. Unlike human identical twins, financial twin assets can be easily manufactured. Create a market in which identical duplicates of existing assets trade. This solves the first of these two problems; rarity.
As for the relative value problem, we can solve it by manufacturing these twin assets in a way that allows us to measure absolute liquidity (or absolute illiquidity). Just create an infinitely liquid twin. An infinitely liquid good can be traded frictionlessly and instantaneously for any other good. The premium at which the manufactured twin trades above the original asset represents the penalty applied to the illiquid original. We thus have a measure of absolute illiquidity; specifically, we have backed out the total amount of compensation that investors require for bearing the illiquid "bad" bound up in a given asset.
In practice, what would these infinitely liquid twins look like? Imagine that a risk-free institution, say a government-backed bank, creates deposits that are denominated and redeemable in Microsoft shares. The bank would pay interest at the same rate that Microsoft pays dividends. Since the purchasing power of these deposits would fluctuate in line with the price of underlying Microsoft shares, the Microsoft deposit would be an exact replica of a Microsoft share. One difference remains: Microsoft shares trade on just one market—the stock market—whereas Microsoft deposits, like bank deposits, have the potential to trade in all markets. Imagine buying an ice cream cone with 0.055 Microsoft deposits. The premium at which Microsoft deposit will trade is an absolute measure of the penalty investors expect to incur for enduring the illiquid "bad" attached to Microsoft shares. Problem solved, right? We've go a clean measure of illiquidity.
Not quite. While infinite liquidity is a nice idea, it's impossible to create. Bank deposits are highly liquid, but not infinitely liquid. Just try purchasing something at a garage sale with a bank card. Second, even if a bank begins to offer Microsoft deposits, there's no guarantee that merchants who already accept dollar-denominated deposits will accept Microsoft-denominated deposits. The upshot is that Microsoft deposits won't be able to serve as an ideal benchmark since they themselves are destined to be tarred by the same illiquidity as Microsoft shares.
We need a cleaner foil against which to compare Microsoft shares. Fortunately, there's an alternative to manufacturing an infinitely liquid twin—just fabricate its exact opposite, a perfectly illiquid twin. A term deposit is a great example of a perfectly illiquid asset; its owner keeps the instrument in their possession until it reaches maturity. During the interim they cannot trade it to anyone else. The difference in price between the original asset and its completely illiquid twin is a measure of the absolute value that investors ascribe to the liquidity embedded in the original asset.
In practice, imagine that our risk-free banks creates 1-year Microsoft term deposits. One deposit represents an irrevocable commitment to earn Microsoft dividends over the course of a year, the deposit maturing in one year with the paying-out of a Microsoft share. Investors facing the choice between purchasing a Microsoft term deposit and an actual Microsoft share will earn the same dividends and capital gains, but will have to weigh the disadvantages of being locked into the deposit versus the benefits of easily liquidating the exchange-traded share. As such, investors will probably only purchase Microsoft term deposits at a slight discount to the price of a fully-negotiable Microsoft share. After all, if you're going to commit yourself to owning Microsoft for one full year, you need to be compensated for your pains. This discount represents the absolute value of a Microsoft share's liquidity.
Voilà, we've unbundled the value attributed to an asset's flow of liquidity returns from its value as a pure financial IOU. We can do this for all sorts of assets. But it's a pain in the ass to do, since it requires the creation of an as-yet non-existent class of financial assets.*
Why bother decomposing an asset's financial return from its liquidity return? Assets provide both an investment return and a consumption good in the form of liquidity, but no one is entirely sure how to apportion prices among the two. Liquidity is static, it muddies many of the supposedly clear signals we get from market prices. Unbundling the liquidity return from the investment return could make the world a much more efficient place. People would be able to see how much they are paying for each of these two returns, thus potentially improving the way that they choose to allocate their resources. What was once static becomes just another signal.
PS: Apologies to long-time readers, who will have already read much of the above points in previous posts. I'm hoping a restatement may provide a different approach to thinking about liquidity.
PPS: The post resolves the problem mentioned in the last three paragraphs of Liquidity as Static.
*Interestingly, a limited market in twins already exists. In addition to providing chequing accounts, banks also provide term deposits. The yield differential between the two represents the absolute value of the liquidity services provided by a chequing deposit. The majority of assets, however, have not yet been twinned---think equities, bonds, bills, mortgage-backed securities, derivatives, and more.