Believe those who are seeking the truth. Doubt those who find it. Andre Gide

Thursday, May 5, 2016

Why the Blockchain should be familiar to you

From L2R: Michael Casey (MIT Media Lab), David Andolfatto (FRB SL),
Simon Johnson (MIT) and John Schindler (FSB)
I'm freshly returned from Consensus 2016: Making Blockchain Real where I participated in a panel on "Digital Cash for Central Bankers." Michael Casey did a stellar job in crafting the session. It was fun and informative to have Simon Johnson and John Schindler as co-panelists. As we didn't get booed off the stage, I think maybe the audience enjoyed what we had to say as well. (I left the session with almost a kilogram of business cards--odd that paper is still so widely used in this capacity. By the way, some of what I had to say can be found in my blog post here.)

Today's post is more about marketing the idea of blockchain. The word sounds intimidating to many people. That's probably because attempts to explain it often make use of a highly technical trade language that few people understand. My goal here is to think of ways to communicate the idea of blockchain in a manner that will make people feel like the concept is familiar to them. Indeed, I believe that the broad conceptual idea of blockchain should be familiar to us all.

Renowned Bitcoin expert Andreas Antonopoulos writes here:
It will take time for the idea of decentralized trust through computation to become a part of mainstream consciousness, and until then, the idea creates cognitive dissonance for those accustomed to centralized trust systems. With thousands of years of practical use, centralized systems of trust are accepted unconditionally and without much thought as the only model of trust.
It's an excellent article and I highly recommend you read it. What I want to do here is push back a little on the notion that decentralized trust systems should necessarily create cognitive dissonance. In particular, I should like to point out that we've had tens of thousands of years of experience with decentralized trust systems. Alright, so let's get started.

Consider the following scenario. You are attending a cocktail party with dozens of people present and you are asked by your hostess to deliver a short speech. Now suppose you utter something outrageous, e.g., "I think the Fed should buy the existing stock of bitcoin and store it as a foreign currency reserve!" The audience will stare at you, mouths agape (especially if you're a central banker, or a renowned Bitcoin enthusiast). You wake up the next day and regret your rash public remark. You wish you could take back what you said, but how? The only way this could be done is if you could somehow persuade the group to forget what you said. But just think about how difficult it would be to do that. Especially if the number of people in attendance was large.

What has just been demonstrated (I hope) is the power of a distributed database validated through a communal consensus algorithm. The database here is your silly statement above together with the time you made it (a timestamp). The information in this database is shared on a distributed network of brains (what you said and when you said it is imprinted forever in the memories of all who witnessed the event). The consensus algorithm here is "let's all agree to remember what was actually said (as opposed to some alternative, fabricated statement)."

A database in this form is extremely secure. It will survive intact even if some brains holding the database are destroyed. The database can be communicated to other brains (who can confirm the validity of the statement by seeing how it squares with the memories of others). If one or more people tried to fabricate an alternative history, the attempt would almost surely fail (we cannot rule out the possibility entirely, however). If your remark instead lived only as an electronic recording in a central databank, the task of re-writing history would be much easier.

Now imagine living in a primitive village. Relevant elements of the database would include observations like: [1] John had his wound tended to by Bob at date t, [2] John killed a wild pig and shared it with the village at date t-1, etc. The database in this case can be organized in a sequence of time-dated blocks X(t) = {x(t), x(t-1),...}, where x(t) is the database (block) at date t, and X(t) is the "blockchain." So, the blockchain is just a communal databank recording some relevant aspects of villagers' activities. In village economies, this communal memory typically exists in a virtual state (written records are a much more modern invention).

Notice how the blockchain described above could serve a very useful economic purpose. In particular, note that the act of consumption (medical services) in [1], John is effectively using [2] as currency. At least, this is how things work in what anthropologists describe as "gift-giving societies." And if you think about it for a while, you'll notice that the same principle is at work in the various groups you interact with on a daily basis (your friends, your family, coworkers, etc.). Much, quite possibly most, economic exchange occurs via such localized trust networks.

The problem with this ancient blockchain technology is that it doesn't scale very well. There's only so much data we can fit in our brains.  So as populations grew and as people started forming large communities, a new type of record-keeping system was needed. The model that came to dominate is one in which databases are collected and maintained by trusted third parties. Much effort is expended in keeping these private databases secure (not always successfully). It is often difficult for these agencies to communicate and reconcile their databases (as in when you try to send money from your bank account to your friend's foreign bank account overseas).

And so enter the "new" technology, blockchain. I hope I have convinced you what is new here is not the principle of the blockchain. The new technological developments are: [1] bigger brains (increased capacity for data storage and processing via computers); [2] better communications (the Internet); and [3] computer-based algorithms to serve as communal consensus mechanisms (e.g., proof-of-work).

These innovations will permit a revolution in the truest sense of the word: we are traveling back to where we began--but with planet earth as our village.


PS. Please let me know if this was helpful or how it could be improved. After writing this post, I came across this short video: Blockchain for Dummies. Some of the comments are critical of it, but I thought it communicated the idea in a nice way.

Monday, May 2, 2016

On Cochrane's dream of equity-financing banking

John Cochrane has a dream where the banking sector is financed entirely with equity. The dream is premised on the notion that debt-financed endeavors--especially those using short-term debt--are prone to runs. Run-prone structures can cause, or contribute to, financial crises. The possibility of crisis invites government regulation. Government regulation leads to regulatory arbitrage, much of which occurs in the shadows of the financial market. Which leads to more (now harder-to-monitor activities), leading to ... well, you get the picture. Why not just structure a regulatory framework that permits equity-financed banking ventures, like SoFi, to weave their (run free) magic? It's a good question. I'm not sure what the answer is. But I wonder if it's all as easy and straightforward as Cochrane makes it out to be.

I don't want to nitpick here but, I don't think I'd classify SoFi as a "bank" in the legal or economic sense. True, SoFi is acting as a financial intermediary. But insurance companies and pension funds are also financial intermediaries, and we do not think of them as banks. SoFi is more like a venture capital fund. It sounds like it's doing a wonderful job matching savers to borrowers. But matching savers to borrowers is not (the main) business of banking. Even a simple bond market matches savers to borrowers. We do not need banks to do that.

So what do banks do? We can get a grasp of their business model by comparing the structure of their balance sheets to other financial intermediaries. In many respects, the asset side of these balance sheets looks broadly similar: they consist of cash reserves, bonds (government and corporate) and other securities. Retail banks also hold personal and small business loans, which they typically originate as a part of their business. The differences on the liability side of financial intermediaries are much more striking. Pension funds issue time-dependent liabilities. Insurance companies issue state-contingent liabilities. Banks issue demandable liabilities. In all three cases, one wouldn't be so far off in forming the impression that financial intermediaries are fundamentally just "asset-transformers" (they transform a set of assets into structured liability products that people find useful).

Banks, in other words, are in the business of supplying a particular structured liability product: the demand deposit liability (DDL). It is Cochrane's worst nightmare. That's because the DDL is a "fixed-value promise." Specifically, a DDL promises redemption at (or close to) par for cash. But it's even worse than this because the promise is to redeem on demand (rendering the DDL a form of short-term debt). Worse still, the banking system does not possess enough cash in reserve to honor these short-term obligations in the event that all DDLs are presented for redemption at once (this is what it means to be a fractional reserve banking system). And as if things could not get any worse, they actually do. These bank-created DDL products -- they're used widely as payment instruments. That's right, banks are in the business of creating money (out of their assets).

Before I go on, I want to say something about the manner in which "deposits" is used in discussions of banking. It is sometimes said that "banks take deposits." But what does this mean? Even a Las Vegas slot machine takes deposits (and issues a very unattractive state-contingent liability in exchange, I might add). Well, yes, I can make a deposit of cash at my local bank. And my employer "deposits" my paycheck in my bank account (in reality, just a debit-credit operation on a ledger). But this is probably not the best way to think about "deposits." I sometimes like to say that banks don't take deposits--they create deposit liabilities. Related to this notion, the banking system does not "lend out cash." The banking system funds its assets (including loan creation/acquisitions) by creating DDLs. (At the individual level, banks need to acquire cash to fund their operations only to the extent they want or need to meet some reserve requirement). Cash finds its way into circulation whenever the owners of DDLs exercise the redemption option embedded in the DDL contract. Alright, with this out of the way, let me continue.

It's not been easy to discover the fundamental economic (or social) rationale for banks (defined here as intermediaries that fund their assets, including their loans, through DDLs). Economists have struggled to understand debt, never mind demandable debt. Probably the best theory of bank debt we have is still the Diamond and Dybvig (1983) model. Like any model of debt, certain "financial market frictions" need to be present; else, the Modigliani-Miller theorem holds, in which case we should all be living in Cochrane's dream world (assuming no bad government fairy, of course).

The root frictions appear to be what economists label "private information" and "limited commitment." Among other things, limited commitment renders all sorts of assets, like our human capital, illiquid. In a frictionless world, there is no reason why I shouldn't be able to buy my Starbuck's latte by peeling off a slice of my house or my future earnings. It just doesn't work. That's what banks are for. They measure the value of my house, my future earnings, and they create DDLs that are backed by their assessed value of the collateral I have to offer. They would be performing an equivalent service by acting as licensing agents whose job is to verify the quality of the promises I issue (imagine an Andolfatto-IOU stamped as "BoA approved.") What's not entirely clear is why banks couldn't just get me the money I need by the way SoFi does--by first acquiring state-created money from willing lenders?

To put things another way, if banks are primarily in the business of payment services, why are they not limited to that business? Why are banks permitted to create money? (Why should banks help render my illiquid assets liquid?) Why not make banks hold 100% cash reserves? And then let the financial market handle matching lenders with borrowers, a la SoFi? This is the line taken by those who favor "narrow banking" proposals (see, e.g., Musgrave, 2014).

I have yet to digest all the arguments made by Musgrave and others. But they make enough sense to be taken seriously (so I plan to continue reading). I have a lot of questions. I am not bought into Cochrane's claim that equity is a run-proof security. The equity traded on junior exchanges, for example, does not appear run-proof (one can "run" to your stock broker and scream "sell, sell, sell!" just as easily as you can "run" to your bank to ask for your money). Moreover, there's a lot of evidence to suggest that equity makes lousy money. Gorton and Pennacchi, 1990 claim this is the case because equity is "informationally sensitive," and (senior tranches of) debt is not. Like it or not, most contracts are drawn up in nominal terms. In such a world, it would be terribly inconvenient, I think, to have floating NAV MMMF shares used as an exchange medium. People seem to like fixed-exchange rate systems (which is what DDLs are, after all).

What I would really like to see is how their claims stack up in a formal model. After all, the Diamond and Dybvig model does suggest the possibility of a trade-off. Maturity transformation enhances risk-sharing (when conventional markets are absent or too costly to operate), but potentially exposes the bank sector to self-fulfilling bank runs. A narrow banking regime kills risk-sharing but enhances financial stability. So in some jurisdictions, the switch to narrow banking might be worth making (although, there may be other ways to enhance the stability of fractional reserve banks, like central bank lender-of-last resort facilities, etc.).

I suspect that narrow banks might work relatively well in low-inflation environments, but possibly not so well in high-inflation regimes. The reason is because high inflation imposes a big tax on cash reserves (unless they pay interest, I guess). In such an environment, fractional reserve banks may be preferred as a way to escape the inflation tax by offering a higher rate of return on their DDLs. (Of course, it would be better to encourage a low-inflation regime, but that's not always possible).

So, these are just a few of the thoughts that came to mind after reading about Cochrane's dream. It's an interesting debate and I look forward to reading a lot more about it.

Sunday, May 1, 2016

Monetary policy implications of blockchain technology

As I'll be at Consensus 2016 event speaking in a session on "Digital Cash for Central Banks" (agenda available here), I thought this might be a good time to gather my thoughts on what central bankers should be thinking about as a new wave of financial innovation comes crashing on our shore. (Warning: the views I hold presently are subject to change. And, of course, my personal views do not necessarily represent the official views of any central bank anywhere!)

Before talking about policy, what is a "blockchain technology?" Like a lot of new terms that are bandied about, it means different things to different people. But for my purpose, I'm just going to think about it as a different way to keep account of information. The Bitcoin blockchain, for example, is a distributed public ledger that records the entire history of bitcoin transactions (the movement of BTC credits from account to account), where the ledger is updated, maintained, and kept secure by profit-seeking accountants (miners) who are incentivized through a clever algorithm to act in the interests of the Bitcoin community (their actions are also publicly observable so any shenanigans, should they occur, are likely to be short-lived.) There are many possible variations of this basic idea.
Now, it just so happens that money is just a type of ledger, as I explain here: Money and Payments, or How We Move Marbles. The notion of money as a record-keeping device goes back at least to Ostroy (1973). We worry about record-keeping systems because people are opportunistic and cannot be trusted. This is what Kiyotaki and Moore (2001) meant when they quipped that Evil is the Root of All Money. A well-designed record-keeping system constitutes a solution to a social problem (the existence of people willing and able to fabricate information for their private benefit at the expense of the community). Similarly, money should be viewed as a solution to a social problem.

Needless to say, none the solutions that have emerged over time have been perfect although, a Darwinian might claim that there is a time and a place for every species (see also: The Byrds). And now, as the technological environment evolves, a mutation threatens the prevailing order. What are the implications for monetary policy?

In what follows, when I speak of cryptocurrencies, I'll I focus on Bitcoin, first, because of its relative popularity, and second, because it's designed to compete directly with central bank money and payment systems. But what I have to say pertains more broadly to all innovations in this space. I'll also sometimes refer to the Fed but, of course, feel free to substitute in your favorite central bank.

1. Currency competition. To a domestic central bank, Bitcoin looks just like a foreign currency which, of course, it is (since its monetary policy is governed by an entity that is outside the domestic government's jurisdiction). Viewed from this perspective, Bitcoin presents central banks with an old and familiar threat: currency competition. Americans traveling abroad are familiar with the phenomenon--one is often presented an opportunity to exchange USD for local currency at unofficial exchange rates. People in these countries are often just trying to avoid a very high inflation tax. 

Annual Inflation Rate
The willingness and ability of domestics to substitute into a competing currency with a more stable value will put limits on the ability of a government to use the inflation tax as a revenue device. Governments sometimes go to great length to restrict the use of currency substitutes. The new threat poised by Bitcoin is that it's likely going to be much more difficult to enforce domestic currency controls. Anyone with a phone and access to the Internet will have access to an alternative digital bearer asset to use as an exchange medium. Bitcoin, or even just the threat of Bitcoin, will put much stricter limits on the amount of revenue governments can extract through the inflation tax. 

2. Maturity transformation using a foreign currency. While Bitcoin is unlikely to displace a major world currency any time soon, it's likely to play a prominent role in certain niches. I am reminded of the role the USD plays in some countries. An issue that arises in those jurisdictions is the creation of USD denominated bank deposit liabilities by foreign-based banks. Fractional reserve banking can be problematic in the best of times, but could you imagine U.S. banks offering loans denominated in BTC and, more importantly, redeemable on demand for BTC? This type of arrangement is not fantasy--it happens all the time in the so-called Eurodollar market and elsewhere. How should regulators respond to such an activity? How can a central bank act as a lender-of-last resort when, in a crisis, people are wanting their BTC bank deposits and not USD? What role, if any, might the treasury in these circumstances? Lender-of-last resort interventions are not limited to central banks, after all.

3. The safe asset phenomenon. A safe asset is not a risk-free asset--it's an asset that people flock to in times of crisis. (They are more accurately described as "flight-to-safety" assets.) In the 1970s, real estate was a safe asset, and investors ran away from the USD/UST (hence, inflation and high interest rates). In the late 2000s, the USD/UST was a safe asset (hence low inflation and low interest rates), and people ran away from real estate. The set of assets that investors perceive to be "safe" evidently varies over time. Could BTC be the next great safe asset? Maybe yes, maybe no. But monetary policy is all about formulating contingency plans. What if BTC denominated deposit liabilities are a significant source of financing, like CUF denominated mortgage loans in Hungary prior to European crisis? And what if BTC is regarded a safe asset in our next crisis, the way CUF is perceived to be in Europe? If this happened in the U.S., it would mean a large depreciation in the USD/BTC exchange rate, price inflation (measured in USD), price deflation (measured in BTC) and, of course, all of the other wonderful things that accompany financial crises. Except that the Fed would have no direct control over the supply of BTC (i.e., for the purpose of expanding its supply to accommodate the elevated demand for BTC, thereby alleviating the BTC deflation). To the extent that the UST is not a safe asset in this event, the Treasury's powers would also be greatly diminished. 

4. Securities exchange. The standard macroeconomic model typically assumes that securities are exchanged in frictionless financial markets, where trade is instantaneous and property rights are enforced at zero cost. Needless to say, this abstraction is ill-suited for the purpose of understanding monetary policy. Most bonds are thinly-traded on over-the-counter (OTC) markets and so, are highly illiquid. Even the most liquid of bonds, like the 10-year on-the-run U.S. treasury, is prone to unsettling "liquidity events" (e.g., Oct 15, 2015). Despite improvements over time, it can still take days to settle and clear securities transactions. This delay, along with other frictions, generates a huge demand for collateral, largely in the form of USTs to guard against counterparty risk. Improvements in securities exchange brought about by the application of blockchain (or other) technologies has the potential to release billions (or more) of dollars in collateral assets into the market place. The effect of this is likely to lower the liquidity premia on USTs, leading to higher interest rates. The implication for the treasury is obvious, but clearly any force that is likely to impinge on the structure of interest rates is also relevant for monetary policy.

5. Financial stability. There are some who claim that blockchain applications will one day render fractional reserve banking (or maturity transformation in general) obsolete. Maybe. But I am not so sure. One way this might happen is if every asset, our homes, our human capital, can be somehow transformed into perfectly liquid bearer instruments. This won't be happening any time soon. Proponents of blockchain technology point out that it has the potential to remove opacity in financial markets, something that would surely lead to a more stable financial system. However, it's worth pointing out that the leading economic theory of bank sector fragility, the Diamond and Dybvig model, does not rely on the existence of opacity in the financial market. In that model, the portfolios of banks are perfectly transparent. A bank run may nevertheless be triggered by the expectation of a mass redemption event, which subsequently becomes a self-fulfilling prophecy. It is also interesting to note that (in the same model) bank-runs can be eliminated if banks adopt a credible policy of suspending redemptions once they run out of cash (this commits the bank not to firesale assets to meet short-term debt obligations). The perception of perfect credibility is essential for the result and, needless to say, the degree of credibility needed here is frequently lacking. If the suspension clause could somehow be made to trigger automatically and mechanically--perhaps a smart contract could be employed--then depositors would never have an incentive to run a bank and the contract would never be exercised. Of course, this solution relies on the common knowledge assumption required in MAD (see footnote 1 below). I'm not sure what implications for policy this has, but it's fun to think about and, well, who knows where all this might lead. 

6. Central bank digital cash. The existing structure of money and payments (including central bank design) was built for the pre-Internet world. The world is now changed and we must deal with it. Among other things, there is no reason why, in principle, central banks could not offer online digital money accounts for the public. I'm thinking here of a basic utility account, a place to keep your money safe and pay bills. (Private banks could still compete by offering full service accounts). There is a sort of precedent for this: the U.S. Treasury, for example, offers online digital bond accounts. And while that system is not specifically designed to make payments, it could be (again, in principle). There are a number of advantages to consider. First, there would be no need for deposit insurance since the central bank accounts have no default risk (they can just print the money, after all). Second, cash managers at large corporations could simply park their money overnight at the central bank, rather than seek collateralized lending arrangements (repo) in the shadow banking sector. Third, the cost of maintaining the paper money supply can be eliminated. Fourth, it is easy to pay interest (possibly, negative) on digital money accounts, leaving central banks with an additional monetary policy tool. There is the issue of how such an arrangement may impact the funding of private banks. But such an object, if it was to exist, could I think, compete favorably with Bitcoin and other cryptocurrencies, assuming that monetary policy is conducted responsibly, of course. (I discuss a more radical form of central bank digital cash--one designed to compete more directly with Bitcoin--in this blogpost: Fedcoin.)

There are so many more things to discuss, but I think I'm at my limit for blog post length. If you have ideas to share, or papers to link to, please feel free to comment below. Thanks!


Footnote 1: Naturally, one would never want anything to trigger a mutually-assured-destruction clause in a contract. And such an event would never occur, theoretically at least, if everyone is perfectly rational. Few people need to be convinced that this assumption is rather extreme. However, if the collective punishment cost is not too large (well, at least finite), then one might be able to live with the occasional "mistake" and subsequent punishment. I am reminded of the "contract" that governed Roman legions. Good behavior was rewarded (after 20 or so years of service) with land to retire on. While bad behavior in battle was easy to identify collectively, it was sometimes hard to identify individually (a legion consisted of thousands of men). To discipline group effort, a credible threat of group punishment is needed (Holmstrom 1982). Credibility (the ability to commit) seems to have posed no problem for the Romans (how they ever became Italians, I have no idea). A legion deemed to have performed in a cowardly manner was punished by having each soldier draw lots, with a 1 in 10 chance of winning the lottery. The "winners" were then summarily clubbed to death by their colleagues. (Incidentally, this is where we get the word decimation--to reduce in number by one-tenth.) The punishment was not carried out very often, suggesting that the credible threat of the punishment worked reasonably well.

Thursday, April 28, 2016

On the want of U.S. government debt

In a recent article, Narayana Kocherlakota lays out the case for why, under present conditions, the U.S. government should be issuing more debt, using the proceeds to cut taxes, finance infrastructure spending, or both. It's a policy that many economists, including yours truly, have been advocating for some time. And while I generally support the policy, I thought it would be useful, nevertheless, to reflect on some possible counterarguments. It's not a slam dunk case, one way or the other, I think.
Kocherlakota does a good job explaining why a deficit-financed tax cut, or deficit-finance infrastructure spending is a good idea. I want to make it clear that the argument in favor of the policy hinges critically on the presumption that we can rely on Congress to manage the public debt over time in a responsible manner. Let's accept this assumption, provisionally at least, in order to understand the economic argument. I will come back to the political argument later.

While the debt-to-GDP ratio (D/Y) is presently high by historical standards, it's not unmanageable. The key is not the D/Y itself, but its trajectory over time. Clearly, D/Y cannot grow forever. And fortunately, market signals are available to monitor how the public perceives the likely path for D/Y over time. These market signals are: (1) the yields on U.S. treasury debt (at various maturities), and (2) inflation and inflation expectations. So what are these market signals telling us? The yield on U.S. treasuries is presently very low. Both inflation and inflation expectations are presently running below the Fed's 2% target and have done so for years now. So far, so good.

The large increase in D/Y since 2008 together with plummeting yields and low inflation may seem puzzling, but it's not really. Usually, a bad event that triggers a large increase in the public debt also triggers higher bond yields and the prospect of inflation. We can expect this to be the case in any experiment where the supply of debt increases in the face of a stable (or diminished) demand for the debt that is being issued. Think Zimbabwe or Venezuela.

But the U.S. is not Zimbabwe or Venezuela, or the Weimar Republic, for that matter. Rightly or wrongly, the U.S. treasury security is viewed by investors around the world as a safe haven asset. So when the financial crises hit in the U.S. and Europe 2008-10, investors moved en masse into U.S. treasuries (and other sovereign debt instruments viewed to be relatively safe). In short, while the supply of U.S. debt spiked up, the demand for U.S. debt increased by even more. We can infer this from the behavior of bond yields, which went down (the price of debt went up) at the time.

So the economic argument is simple. The U.S. government can presently borrow at essentially zero interest (more or less) even 10 years out and more. This effectively gives the fiscal authority the ability to print money (low-interest debt), so there's no need to rely on the Fed. To the extent that domestic real economic activity is still not firing on all cylinders, why not offer temporary tax cuts to stimulate demand? Why not re-build that crumbling infrastructure, putting people to work, all financed at zero-interest? It sounds like a no-brainer.

Alright, now for a couple of counterarguments, one economic and one political.

An economic argument against temporarily increasing the public debt further (and indeed, taking measures to reduce it) could be made on the basis of the Triffin Dilemma. The economist Robert Triffin noted back in the early 1960s that world reserve currency/debt status is a double-edged sword. On the one hand, it's great that the U.S. can just print paper that is coveted around the globe. If foreigners are willing to export their goods and services to us, expecting only paper in return, then we are extracting wealth from the rest of the world (in exchange for what ever financial service our paper is providing them).

One implication this power, if exercised, is that the world reserve currency issuer is likely to run persistent trade deficits. Triffin worried that the huge amount of U.S. currency held by foreigners exposed the U.S. to foreign risks. What might happen, for example, if foreigners suddenly decided they no longer wanted to hold USD or USTs? This could result in a sudden and dramatic change in the exchange rate, leading to domestic inflation and sharply higher bond yields.

There is also the trade-related argument that persistent trade deficits kill domestic industries and domestic employment. After all, if we can make the rest of the world work for us in exchange for paper, where is the need for us to work at all? The implied boom in domestic leisure consumption sounds good theoretically. But of course, in reality, the gains are not evenly shared. The rich gain by purchasing cheaper foreign goods. The poor are out of their jobs.

A political argument against more government debt could be made by challenging the assumption that it will be managed responsibly. This "we can't trust future politicians to do the right thing" argument is (sadly) not without empirical merit. I am reminded of the following quip by P.J. O'Rourke,
"The Democrats are the party that says government will make you smarter, taller, richer, and remove the crabgrass on your lawn. The Republicans are the party that says government doesn't work and then they get elected and prove it."
I can't help but note a certain irony here. There seems to be a strong presumption among people (Americans in particular) that the government should run its finances in the manner of a household. Economic theory is quite clear that this sentiment, however noble, is just plain wrong. The irony is that to the extent that this sentiment finds its way to being represented in Congress, it proves to be a very valuable "anchoring" device for the fiscal authority.

That is, I sometimes wonder whether US treasury debt is valued around the world the way it is precisely because it is known that Congress is impregnated with a large number of genetic "debt-ceiling" algorithms. It may not be an ideal situation from the perspective of pure economic theory, but then again, it's not hard to think of worse scenarios.

Friday, April 22, 2016

Interest Rates and Aggregate Demand Revisited

Nick Rowe has a nice post (written some time ago) that frames an old macroeconomic issue in a very nice (teachable) way.

In macro policy discussions, one often hears something like "lower interest rates stimulate aggregate demand.'' Many people view such a statement as self-evident. It's only when you think about it for a long time that you realize it's not self-evident at all (few things are when we are left to ponder them long enough, it seems).

The purpose of this post is to add a bit of formalism to Nick's discussion. (Sorry for the wonkish display, but I think it's necessary at this point to make things clear.) To this end, let's begin with an off-the-shelf bare-bones macro model. There is a representative agent (this is not necessary, but makes things easy) with additively-separable log preferences defined over consumption sequences {c(t), t = 0,1,...,∞}, with discount factor 0 < β < 1. Let R(t) denote the gross real rate of interest (risk-free) earned on a bond held from date t to date t+1. Assume that all individuals can borrow/lend freely at the risk-free rate.

Now, consider the cost-benefit calculation associated with the consumption-savings choice. Suppose an individual refrains from consuming one unit of consumption today. The marginal utility cost of this sacrifice is given by 1/c(t). This one extra unit of saving delivers R(t) units of extra consumption tomorrow. The marginal utility benefit of this extra consumption is R(t)β/c(t+1). Individual optimization requires equating marginal cost to marginal benefit:

[1] 1/c(t) = R(t)β/c(t+1) for t = 0,1,...,∞

Condition [1] is sometimes called the consumption-Euler equation, or just the Euler equation, for short. (Noah Smith has a nice post on the Euler equation here.)

One can do a lot with the Euler equation. Here is how it is used to derive "aggregate demand." First, assume that all output is (for simplicity) in the form of nonstorable consumer goods and services. Let N denote population size. Then C(t) = Nc(t) denotes aggregate consumption or GDE (gross domestic expenditure). Now rearrange [1] as follows,

[2] c(t) = [ 1/(R(t)β) ]c(t+1)

Thus, if we hold c(t+1) fixed, then equation [2] traces out a negative relationship between c(t) and R(t). That is, an increase in R(t) results in a decrease in planned present day consumer spending (aggregate demand). This negatively related locus of consumption and interest rate pairs is sometimes called an IS curve (IS = "investment-saving" where investment is fixed at zero in this model). The economic intuition is simple: raising R(t) makes it more attractive to save (lower current consumption). [Never mind for now that any extra saving is likely to boost c(t+1).]

Let me consider an endowment economy where each individual is endowed with a deterministic sequence {y(t), t = 0,1,...,∞}. Usually, y(t) is thought of as an individual's output or income at date t, so that Y(t) = Ny(t) represents GDP (gross domestic product) or GDI (gross domestic income). But more generally (and this is what the General refers to in the General Theory) we can think of y(t) as a maximum production capacity. "Full employment" refers to the special case where Y(t) is the GDP.

The standard neoclassical assumption is that the economy is always at full employment. Maybe calling this property an assumption is not quite accurate. We can derive the property as a result of some deeper assumptions relating to the ability of individuals in an economy to coordinate their activities in an efficient and socially desirable manner (this is the force behind Says' Law, that "supply creates its own demand.") In any case, the upshot is that Y(t) represents the GDP. And since GDE = GDP, we have C(t) = Y(t), or c(t) = y(t), at every date t. Each person consumes his value-added, the economy consumes what it produces.

Suppose that real income grows at rate α, so that, y(t+1) = αy(t). Since c(t) = y(t) for all t, condition [2] can be used to deduce the equilibrium real rate of interest:

[3] R*(t) = α/β

The real interest rate is predicted to be high when growth (α) is high. The real rate of interest is low when growth is low. The intuition here is as follows. An increase in α means that people are expecting higher levels of future income. People will want to bring some of that future income forward in time. They will try to do so by borrowing, or saving less. Either way, the effect is to put upward pressure on the interest rate.

Alright, it's time to do some "textbook" aggregate demand analysis. Actually, I don't like the way textbooks usually do this. The usual assumption is a "sticky wage" that mucks up the labor market (here is my critique on that idea). This idea is certainly not Keynesian:
"There is, therefore, no ground for the belief that a flexible wage policy is capable of maintaining a state of continuous full employment..." [General Theory, 1936 Chp. 19]
Indeed, Keynes (1936) wrote that flexible wages could make things worse, not better (consistent with my Figure 2.12 here.) The best representation of "what Keynes actually meant" is, in my view, expressed formally in the game theoretic notion of multiple Bayes-Nash equilibria (a tool that was not available to Keynes in his lifetime). See Cooper and John (1988), Howitt and McAfee (1990) and Roger Farmer for example.

How to proceed? There are many ways, but I don't want to get bogged down in the details here (although I should stress that the details are critical for other questions). One way to proceed is to embed my static high/low equilibrium model into the model above. In that model, aggregate demand C(t) can be either high or low, and the equilibrium level of output can correspond to the high or low level of demand as a self-fulfilling prophecy.

Here's another way to think about it. Peter Howitt would explain it to me this way. Imagine that the people in our model do not like the smell of their output (any reason to motivate intratemporal trade here will do). So they will want to swap their goods with others. If everything works well here (the neoclassical assumption), then all goods will be traded at par. The real GDP is Y(t).

Now, suppose that trading is costly. Suppose that it is prohibitively costly to sell any output beyond some level k(C), where C is aggregate demand. Assume that k(C) is increasing in C. The idea here is that it is easier to sell larger quantities of output when demand is high. In fact, we could just assume k(C) = C/N. Next, consider an arbitrary 0 < C/N < y. Then the most anyone can expect to sell (and buy) at a given date is c = C/N. In this model, there is a continuum of equilibria, each indexed by an expectation defined over C/N. If everyone expects a thickly traded market, it is individually rational to trade a large volume and, collectively, this is what transpires. "Animal spirits" determine which of these equilibria actually prevail.

Alright, back to Nick's point. Assume that in the "long run," the economy returns to full employment forever. For simplicity, assume that the long-run is expected to occur tomorrow. In this case, c(t+1) = y(t+1) and R(t+1) = α/β for every date t going forward. Now let's take a look at today, t = 0, using condition [2].

[2a] c(0) = [ 1/(R(0)β) ]y(1)

Assume that c(0) is determined by an animal spirit (as described above) such that c(0) < y(0). Then condition [2a] can be used to solve for the equilibrium interest rate,

[4] R(0) = (1/β)y(1)/c(0) > R*(0) = (1/β)y(1)/y(0)

That is, the economy is presently in recession and the interest rate is too high. And if the interest rate is too high, well, then, why not take policy actions designed to lower it? The recession is like diabetes and low interest rate policy is like insulin, as Kocherlokota argues here.

And the argument makes sense IF full employment lives somewhere in the foreseeable future. Lowering the interest rate in this model has the effect of stimulating consumer demand as people try to bring future output closer to the present. But future output here is fixed at full employment. So, to the extent that lowering the real rate of interest increases C(0), the effect is felt entirely in the contemporaneous depressed period in the form of higher real GDP.

But what justifies the assumption that the economy will somehow find its way back to full employment? This is the missing piece in our conventional models.

This leads Nick to ask: what if people do not expect a return to full employment in the near future? Indeed, what if? As it turns out, there are many, many other equilibria in the model above. One such equilibrium path satisfies

[5] C(t+1) = R*βC(t)  where C(t) < Y(t)  for all  t = 0,1,...,∞

That is, the economy can be permanently stuck in a "secular stagnation." Moreover, the equilibrium interest rate is exactly where it should be: it is neither too high  nor too low. Consumption and GDP are growing at rate α. It's just that the level of GDP is permanently below its full employment level.

The real interest rate measures the relative price of output across time. In the equilibrium described by [5], the relative scarcity of output across time is just right. Its the contemporaneous level of output that's off at each date. How is a change in the interest rate supposed to fix this problem?

The short answer is that it can't. In fact, it's easy to construct examples where attempting to lower the interest rate could make things worse (perhaps this is an overdose of insulin, in Kocherlakota's example).

Suppose we're in a situation described by condition [2a], for example. In that exercise, I assumed that y(1) is fixed at full employment and that c(0) is depressed. This is what justified lowering R(0) to stimulate contemporaneous consumer demand. But suppose that animal spirits keep contemporaneous consumer demand fixed, and that the effect of lowering R(0) is to reduce future consumer demand to c(1) < y(1)? There is no a priori reason to expect c(0) to do all the "equilibrating" here. And so, in this manner, the effect of low interest policy could be to cause future recessions, possibly a secular stagnation.


I think most of what I said above can be shown in a conventional 2-period economy (a current and future period). Here are some diagrams.

Consider first the neoclassical general equilibrium (full employment at both dates). Condition [1] states that the slope of the indifference curve is the same as the slope of the intertemporal budget constraint (the real rate of interest). The full employment assumption means that the equilibrium lies on the budget constraint. This is point A in the following diagram.

Suppose now that the economy is expected to be at full employment in the future; i.e., fix c(2) = y(2), but that the economy is presently depressed; i.e., c(1) < y(1). The interest rate is too high, R' > R*. This is point B in the following diagram.

Lowering the interest rate in the diagram above (making the budget line flatter) moves the economy from B back to A. But suppose we instead forecast a future recession, so that c(2) < y(2). Then condition [5] prevails, and the economy moves to point S (secular stagnation) in the diagram below.

And finally, here's how to depict a future recession caused by an artificially low interest rate  policy (point F).

Thursday, April 14, 2016

How old were the inventors of major inventions?

I came across this fun column the other day listing a number of Famous Inventions, like the airplane, the camera, electricity, the car, etc, along with their inventors. A thought crossed my mind: how old were these inventors when they invented these inventions? Were they young like Marconi, who invented the radio in his early 20s? Or were they old like Gutenberg, who invented the printing press in his early 50s? In short, is there an age demographic that is responsible for producing major innovations?

Let's take a look at the data based on 34 major inventions listed in the article I cited above (thanks to Michael Varley for organizing the data).

Here is what the data looks like for the full sample:

I have to admit, I was a little surprised--the median age is 40 (I was expecting younger). In what follows, I report the age distribution for different centuries. I'll save any commentary for another time and let you draw your own conclusions. Feel free to send me links to any literature related to this issue.

Postscript: The demographics of innovation in the United States. Reports that the median age for an innovator is 47 years. It's also interesting to note the disproportionate share of innovation attributable to immigrants and the children of immigrants. 

Postscript April 16, 2016. 
Interesting comment thread here

Sunday, March 27, 2016

Is Bitcoin a Safe Asset?

You're probably thinking no, of course not. The dollar price of bitcoin can be quite volatile (see here). One can easily gain or lose 50% over a very short period of time. So if we're talking about an asset that offers a stable rate of return, Bitcoin ain't it.

Except that this is not what I mean by a safe asset. 

I'm not even sure how to precisely define what I mean by safe asset. Loosely speaking, I'm thinking about an asset that people flock to in bad or uncertain economic times. In normal times, it's an asset that is held despite having a relatively low rate of return, perhaps because of its use as a hedge, or because of its liquidity properties.

U.S. dollars (USD) and U.S. treasuries (UST) are examples of safe assets today. Now, you might think that they're safe because they're close to risk-free in terms of what they promise in the way of a nominal rate of return. A paper USD promises a zero nominal interest rate and you'll be sure to get that if you hold on to the note over time (USD in the form of central bank reserves presently earn 1/2%, but only depository institutions get this rate.). A UST bill also promises zero nominal interest and you can be sure to get that with full principal repayment upon maturity. The coupon payments associated with a UST bond are virtually risk-free.

But that's not a complete way to think about the risk associated with a security. First, economists (rightly) focus on the real rate of return on an asset. Investors don't care how many paper dollars are promised to them in the future. They (presumably) care about the purchasing power of those future dollars. If inflation turns out to be high, that future purchasing power will be low. The opposite holds true if inflation turns out to be low.

As for the "risk free" UST bill, its market price will generally fluctuate between the issue date and maturity date. This is sometimes called "interest rate risk." If you buy a bill that promises $100 a year from now for $99, you will make about 1% if you hold the bill to maturity. But if market interest rates spike up in the interim, and if you are forced to sell your bill to raise cash, you're likely to realize a substantial loss.

That's the thing about a safe asset. It's return can appear to be stable for long periods of time and then--bam--something happens. (Something always happens.) Interest rates may spike up--a sudden sell-off in bonds may occur. What might trigger such an event? All sorts of news. Foreign banks may need to liquidate their foreign reserves consisting of USTs for political or economic reasons. A sudden increase in inflation expectations would lower the expected real rate of return on nominal bonds, inducing a sell-off. A bond sell-off might even be triggered by a good news event. An increase in productivity growth increases the expected return on private capital investment, inducing portfolio substitution out of bonds, for example.

Another thing to keep in mind is that the asset classes that constitute safe assets can change over time. In my recent piece on secular stagnation, I noted that a "flight to safety" seems to occur near regime changes that imply productivity slowdowns. In 1974, investors flocked to gold and real estate--they ran away from USD (rapidly rising price-level) and UST (rapidly rising nominal interest rates). In 2008, the situation was quite a bit different--both USD and UST were highly sought after safe havens (with investors fleeing real estate).

The observations above suggest that the monetary policy regime matters a great deal for whether a fiat currency is perceived to be safe or not. When Nixon and his advisers chose to abandon the gold standard (against the recommendation of Fed chair Burns) in 1971, monetary policy appeared to lose its nominal anchor. So when the oil price shocks and productivity slowdown hit in the early 70s, investors ran away from cash. Gold is often credited as being a safe asset because of its supply "policy." But there must be more to it than this because, like gold, the supply of real estate is not very elastic. And yet real estate was not a safe asset in 2008.

Patience, Grasshopper. I will get to Bitcoin soon. Before I do, I want to ask "what makes an asset safe?" According to Gary Gorton, it has a lot (perhaps everything) to do with information asymmetry:
A "safe asset" is an asset that can be used to transact without fear of adverse selection; that is, there are no concerns that the counterparty privately knows more about the value of the asset. (Safe Assets, Working Paper, March 2016).
In other words, a safe asset is an object with attributes that traders can mutually agree on very quickly and at little cost. Objects with this property tend to become monetary instruments or, to use a more broad term -- exchange media (which includes objects commonly used as collateral to support lending arrangements). Safe assets tend to be "simple" assets. Historically, commodities such as salt, precious metals, or coined tokens. It's easy to verify your salary in salt (just taste it). It's a bit more difficult to assay gold. The whole purpose of coinage was to make objects easily recognizable without much effort. 
It goes without saying that most financial instruments are complicated objects. Consider your life insurance policy, which is relatively simple as far as financial products go. The reason you can't buy your morning latte with a slice of that asset is because it's simply too costly for the vendor to do the necessary due diligence. So you pay in cash. Everyone knows what cash is. Cash may be "junk" (i.e., unbacked), but at least everyone can agree that it's junk. There's nothing complicated about cash. (The same principle holds true for UST, which are used extensively as collateral in overnight lending arrangements.)  
Cash and gold are "simple" objects. The fact that they pay no interest makes them even simpler. In particular, there's no need to spend time investigating the reliability of a dividend paid by "barren" asset--everyone can agree right away that the dividend is zero. This type of informational symmetry appears to be in high demand in times of financial uncertainty (when nobody knows for sure what other people know about the securities they're selling.) Of course, the situation is somewhat more complicated when counterparties (intermediaries) are involved, but this is true of any asset.

This brings me to Bitcoin. I think that Bitcoin could be the world's next great safe asset. At least, it certainly seems to have all the properties that are desired in a safe asset.

Importantly, it is a "simple" asset. It's simple in the sense that it's a pure fiat object--the monetary objects (called bitcoin) constitute no legal claim against anything of intrinsic value. Bitcoin is simply a record-keeping technology (and economists have known for a long time that money is memory). It pays no interest. Possession corresponds to ownership (unless counterparties are involved). The ledger has proven itself secure (not a guarantee that is can never be compromised, of course). 

Now one might object that Bitcoin is not that simple, not to the average person on the street, at least. Bitcoin consists of 30MB of C++ code. And the algorithm that governs the accuracy and security of the ledger can be hard to understand. But I liken this to the way most people understand how their car engine works. We have a vague notion of how internal combustion works, how power is transmitted through the drive train, blah, blah, but all we really know for sure is that our collective experience with the technology has proven useful. We also know that there are mechanics out there that do know how a car engine works. Because the Bitcoin code is open source software, attempts to modify the code for personal gain at communal expense are easily detectable through expert eyes. And we trust that there are many expert eyes on the watch.

Finally, Bitcoin has a very simple monetary policy. Essentially, the policy is to keep the money supply fixed (actually, it will grow asymptotically to a fixed number, 21 million units). Although this money supply rule could potentially be modified by communal consent, there are reasons to believe that this is unlikely to happen. And even if it does happen, it can only happen if it somehow serves the community of Bitcoin users in some broad sense.

As is well known, there's been a bit of a civil disturbance in the Bitcoin community as of late. The issue, as I understand it, concerns a proposed amendment to the Bitcoin constitution (see blocksize controversy). People fear that if the amendment does not pass (and it does not look like it will), then Satoshi Nakamoto's original vision of a low-cost, high-speed, high-volume P2P payment system may fail to materialize. Others are confident that a solution, in some form, will eventually be found. (These people breathe optimism, remember. It's the fuel that powers entrepreneurship.)

But suppose that the original vision doesn't pan out. Suppose instead that Bitcoin hits a hard limit on the volume of transactions it can process (presently far below what Visa can accomplish). Suppose further that as the subsidy on block rewards (the seigniorage revenue used to finance book-keeping costs) becomes negligible. Then a fixed transaction fee (and possibly a substantial one at that) will have to be paid, since someone has to finance the book-keeping costs. If this were to happen, then it would only make sense to hold Bitcoin for large-value transactions (the fixed cost associated with each transaction would make small-value transactions uneconomical.) 

This "Bitcoin as a large-value transfer system" does not destroy my thesis: Bitcoin can remain a desirable safe asset. (Smaller players could presumably get involved by investing in Bitcoin ETFs, although doing so would introduce counterparty risk.)

I've argued before that Bitcoin makes for lousy money. I still believe this. If it isn't the unit of account, users are subject to extreme exchange rate volatility. In a world where it is the unit of account, a "flight to safety" event would cause an unexpected and severe deflation. We have the experience of the early 1930s to show us what a Bitcoin monetary policy can lead to. (And while a Bitcoin monetary system may free people from the inflation tax, it won't free them from more general forms of taxation.)

However, even if Bitcoin is not, in my opinion, a particularly ideal monetary instrument, this does not preclude it from serving as a safe asset or longer-term store of value. Once market penetration is complete, its return behavior is likely to mimic the return behavior of any other safe asset. Safe assets generally earn a low expected return (that is, they are priced dearly). Investors can expect to earn unusually high returns in a crisis event. But if you buy at the top, you can expect to realize unusually high losses when the crisis subsides. In short, it's a great investment -- assuming you can predict when a crisis will occur and when it will end!

There are a host of issues related to safe assets that I think deserve some attention. Let me offer a few that come to mind here. First, it's not even clear that safe assets are socially desirable. Bryant (2005) demonstrates that the existence of a safe asset can induce coordination failure. Is this an argument to be taken seriously? Second, I think that policymakers should be aware that the class of safe assets may change over time. Should policy be conditioned in any way on the existing set of safe assets? Third, how should we think about "close-to-safe-asset" substitutes that seem to proliferate in periods of prolonged economic tranquility? Barren assets like cash, gold and Bitcoin generate no income. It is evidently very tempting to construct "safe senior tranches" of private interest-bearing debt to compete with these low-return barren assets--a practice that sometimes gets out of hand--and with disastrous consequences. Should a central bank issue its own interest-bearing digital cash to discourage the practice?

Pasqua 2016