Economists vs What Works: Lessons from the ETF Market

By Brian Andersen

In my previous post I outlined the strategy used by Exchange Traded Funds (ETFs) to create financial assets that offer both price stability and high liquidity. I defined price stability as the ability to hold the asset without gaining or losing purchasing power. And liquidity as the ability to buy or sell the asset on demand. Since this mechanism has been shown to work very well in the equity markets, I think it could be instructive for issuers of fiat currency, because fiat currency are also financial assets that are supposed to offer price stability and liquidity.

To that end I extracted four guidelines for a currency issuer to follow, based on what is already known about successful ETF operations:

  • The issuer should only spend/create assets for those willing to earn/buy them.
  • The issuer should only tax/redeem currency while earning it itself.
  • The issuer should remain responsive to users desire for more or less currency in circulation.
  • The issuer should not attempt to increase or decrease the supply of currency of its own volition.

The purpose of this post is to show how these guidelines differ with the recommendations of economists, which have served as guidelines for actual currency issuance until now. I will discuss each rule in turn.

  • The issuer should only spend/create assets for those willing to earn/buy them.

ETF issuers do not issue new shares of the ETF unless you bring them them a basket of stock that perfectly matches the basket held by the fund. If they were to give away newly issued shares for free, the value of those shares would be diluted. But economists insist that holders of currency require free assets in the form of interest payments. Holders do not have to do anything other than register their desire for free assets by purchasing a special type of currency called government bonds. Their justification for these free assets is elaborate, but the net result is a belief that if you hold currency, the currency issuer should give you more of it as a reward for holding it. Needless to say, ETFs would be far less popular if the ETF issuer diluted its assets just to entice people into holding ETF shares. Furthermore, the existing popularity of ETFs indicates that no such enticement is necessary. For centuries, economists have argued that if government were to stop paying interest, no one would want to hold the currency they issue. However monetary policy experiments in the wake of the financial crisis have proven this theory incorrect. It turns out that there is strong demand for currency assets even without the promise of compensation to the holders. Just as there is strong demand for ETFs.

  • The issuer should only tax/redeem currency while earning it itself.

In the world of ETFs, redemptions are what destroy shares of the ETFs. You approach the ETF issuer with shares in hand, and they give you back the fixed portion of the basket that your shares represent. In other words, they earn the shares back by delivering a fraction of the basket to the user. A consequence is that the basket controlled by the ETF shrinks along with the quantity of shares. Of course, the currency analog of redemption is taxation. And the currency analog of the basket is the real economy, which is just a pool of real assets that can be purchased in exchange for currency. For a currency issuer, this means that taxation should be tied to events that cause the pool of real assets to shrink! But in economic theory, taxation has nothing to do with shrinkage of the real economy. Instead, they believe that taxation is required to fund every single dollar of spending by the currency issuer, even in the presence of an expanding pool of real assets and increasing demand for the currency! This amounts to a forced “redemption” of assets with nothing offered in exchange. Needless to say, if ETFs operated this way, no one would want anything to do with them. You could even call it theft. I am not saying that all taxation is theft. I am saying that taxation beyond what is required to avoid dilution of the currency could be called theft. This means that any activity that shrinks the pool of real assets should be associated with taxes in kind. This includes war, which is the outright destruction of the wealth that markets create. It also includes consumption of non-renewable resources. The point is that the issuer should only tax what makes us permanently poorer. What leads to shrinkage of the real basket. It should never try to tax the growth of the basket because that means the currency issuer is putting the brakes on the creation of real wealth.

  • The issuer should always remain responsive to users desire for more or less currency in circulation.

In the world of ETFs, users express the desire for a larger or smaller quantity ETF shares in circulation via the create/redeem process. In the world of currency, unemployment is the expression of desire for more currency. Inflation is often thought to be the expression of a desire for less currency. I don’t believe that but we will save that discussion for another post. The view of most economists, however, is that currency issuers should not respond to unemployment with increased production of currency. Instead the currency issuer should sit on its proverbial hands while wages and prices fall. Keynesians will tolerate some currency issuance in the short run, but only with the caveat that it must be bought back later. This is an extraordinarily unhelpful position because it makes the issuer look irresponsible later on when it cannot get its currency back, and actually needs to issue more. These ideas are in clear conflict with the notions of liquidity and price stability. Unemployment is fundamentally a failure of liquidity; it is the inability to buy/earn currency. And a currency that appreciates in value is a failure of price stability. It may be great for savers of currency who get a free ride, but it discourages production and consumption hurting both earners and spenders. Why invest in a productive business when you can get a real return from sitting on your hands? Remember, the operational innovation behind ETFs that enables them to be stable and liquid is that the quantity of ETF shares is allowed to float. If we want our currency to offer price stability and liquidity, the quantity of that currency must be allowed to float. And that requires an issuer that is responsive to the demands of its users. It simply is not compatible with the contention of mainstream economists that the currency issuer should avoid interacting with and responding to signals from the private sector.

  • The issuer should not attempt to increase or decrease the supply of currency of its own volition.

ETF issuers always stand ready to redeem shares in response to demand from users, not because they believe that the ETF is too big and therefore needs to be made smaller. ETF issuers want more assets under management, which means they want to issue more shares of the ETF. The ETF shares are liabilities to the ETF issuer, just as currency are liabilities to their government issuers. In both cases, the increased quantity of liabilities is necessitated by an increased quantity of real assets in the basket. To do otherwise would be inconsistent with price stability and liquidity. But in offering advice to currency issuers, economists believe that increasing the quantity of currency is dangerous and unsustainable. They want the currency issuer to become an active, net purchaser of currency. They believe it is possible for the issuer to run a perpetual surplus of the currency that only it can create. But the role of an apple orchard is to be a net seller of apples in the market. The role of an ETF issuer is to be a net seller of ETF shares in the market. And the role of a currency issuer is to be a net seller of currency in the market. But under the advice of economists, currency issuers try to become active, net buyers of their own product. It is hard to overstate how detrimental these policies have been to the goal of creating a stable and liquid currency. The result is that government is always active in the market, competing with households and firms to earn the currency. If you want to know why it is easier to spend money than to earn it, it is because you are in competition with government when you try to earn, but not when you try to spend. If only the currency issuer would restrict itself to a passive role in the marketplace, where it create/redeems currency only in response to the expansion/contraction of the real basket, then the currency would be equally earnable and spendable. The currency would not be subject to dilution or inflation; it would be stable in value.

The conclusion is that we can create currency that is both liquid and stable, or we can follow the recommendations of mainstream economists, but we cannot do both. That is because their advice is in conflict with the strategy for creating liquid, stable financial assets. Specifically, they want to spend too little, they want to tax too much, they do not want to respond to currency user demands for more currency, and they want the currency issuer to be an active market participant where it should be a passive one. We have tried the recommendations of economists without success. May we now try what works?

20 responses to “Economists vs What Works: Lessons from the ETF Market

  1. golfer1john

    Still good, but despite the advice of mainstream economists, the US has been, except for short intervals, a consistent net seller of its currency in the market, just like the apple orchard. But the apple orchard wants to sell as much as it can possibly produce, and that’s not an appropriate strategy for a currency issuer, as its production is unlimited. The ETF issuer would like to sell as many ETF shares as it can produce, but is limited by its business model to the amount of baskets of shares offered to it by buyers. A sovereign currency issuer has no such limitation, and should not pursue the strategy of buying as many real goods and services as are offered for sale in the market. To do so would shrink the private sector to zero, as there would be no output available for it to buy with all that currency, as all the output would be consumed by government.

    There has to be some other unique guiding principal for the currency issuer, and that is MMT’s full employment policy. Full employment is what needs to limit the currency issuer’s activities, and JG is needed to prevent that high level of aggregate demand from causing (additional) inflation.

    You still haven’t explained why “free assets in the form of interest payments” cause inflation, but free assets in the form of Social Security payments (for instance) do not.

    From the point of view of the holder of Treasuries, interest is his reward for temporarily relinquishing his purchasing power, the actual currency. The grocery store doesn’t accept Treasuries in payment for apples. In order to own treasuries, you have to forego buying groceries until the Treasuries mature and you get your cash back. It’s just like a time deposit at the bank. It’s not a demand deposit that you can spend, so the bank has to pay you interest in order to get you to do the deal. That’s not an “elaborate justification”, it’s how these things work. None of that has anything to do with the notion that the monetarily sovereign government doesn’t need to issue Treasuries at all, doesn’t need to borrow the currency that only it can issue.

    • financial matters

      Yes, the same would be true of bonds offered by the Fed for public projects. They could take capital out of the economy from say pension funds and money market funds and give them a chance to really put that capital to work productively. Not to asset manager fees and people churning away at repo desks. And with the same fairly risk free nature of Treasuries. (Actually probably better as it would have a true labor backing.)

      They could pay a reasonable amount of interest and also put people to work. The Fed after all is really the fiscal agent of the Treasury.

  2. The currency cannot be considered liquid by my definition without full-employment because liquidity implies that the currency can by earned/bought by offering labor in exchange. A JG is they only way to guarantee liquidity of the currency. So I would rather call it a liquidity guarantee.

    Regarding Social Security, its payments are funded mostly via payroll tax, which cannot dilute the currency because it is a transfer from current earners to past earners. The rest is funded via interest on government debt, which I have already said leads to dilution of the currency in the long run. Having interest-rates permanently at zero will lead to a shortfall in the trust because it was designed to be partially funded using interest payments. As MMT has said for years, the government can easily make good on those claims anyway. Yes it leads to the same kind of long run currency dilution that you are concerned about. But it is no more or less of an issue than other interest payments.

    What I cannot tolerate are people who say we cannot afford to satisfy all of the Social Security claims because it will cause inflation, while they concurrently argue that higher rates are needed to prevent inflation! But that is the argument economists have been making.

    Now for your final point. The way I see it is that currency is like any financial asset. There are three things you can do with it. Buy, Sell or Hold. Buying currency always carries a real cost. Selling it always carries an equally valued real benefit. Assuming the purchasing power of the currency is reasonably stable, holding it allows one to avoid the risk associated with other financial assets like stocks or bonds or real-estate. In a world full of risky financial assets, the ability to store purchasing power is an enormous benefit which the holder enjoys without cost, in the same way that people enjoy free use of public roads and schools. These things all have costs, but to the public, not to the user. That is why people will hold currency when its financial return is zero, and even when it is slightly negative.

    You are trying to argue that holding and not spending represents a cost which deserves remuneration. But it doesn’t. Preserving purchasing power for the future is a benefit to to the user and a cost to the public.

    • financial matters

      It’s not so much that it has a cost that is being remunerated but it can help direct investments (it’s ok for governments to pay interest to the private sector as long as that deficit spending is productive) and it does take money out of the system which is anti-inflationary.

      The payroll tax is regressive and isn’t needed to pay for social security. Taxes are useful if we do reach full employment to take money out of the system to avoid inflation and they are useful to also direct investments particularly away from rentier type activity. The main guarantee of social security is full employment generating goods

      • The short run effect of free interest payments is that they attract capital away from productive activity which then sits on the sidelines while it collects a handout. The long-run effect is that new nominal wealth is created that was never justified with new real wealth. Eventually, this currency gets used to purchase things and it drives prices to levels they could not have reached otherwise. This is a different thing entirely from money that is spent to hire teachers and firefighters, where real assets in the form of labor are received in exchange for the newly created nominal assets.

        I do not understand how they get away with the claim that printing up new money to pay teachers causes inflation, while printing money to pay interest to bondholders prevents it. That is absurd. The currency that was created by Bush’s tax cuts will eventually cause the same type of inflation. When it does they will blame it on healthcare, infrastructure, and education spending. But the true cause will giving away free money.

        • golfer1john

          You’re not really saying that the 0.25% interest on reserves and short-term Treasuries is the reason for the lack of investment by business today, are you? Maybe you should think about that again.

          And MMT completely refutes the idea that tax cuts (or any other budget action, or private sector action for that matter) in 2001 will cause inflation in 2020. Whatever inflation occurs in 2020 will be because of economic conditions in 2020, not 2001 (and not QE in 2014 either). In 2060 will President Malia Obama still be running on “It’s Bush’s fault”?

          • We are not discussing the amount of inflation that will occur in 2020 or 2060. Indeed, that will be determined by economic conditions at that time. We are discussing the cumulative amount of inflation that will occur between now and then. I am trying to take you up on the conversation you started in your first post where you said that MMTers can answer to hyperinflation and wage inflation but not long-run inflation in the 2-3% range. You said that this inflation causes people to work longer and harder than they should need to. Of course not everybody agrees that this inflation is harmful. One could argue that that is simply the “expense ratio” of the currency, keeping to my ETF analogy. After all, price stability is a feature and a benefit of the currency that doesn’t come “out of the box”, it requires careful administration and that has costs. Also, I could say that you have equities as an alternative to holding currency and equities have natural protection against this sort of inflation.

            But nonetheless, if the currency were an ETF with an expense ratio of 3%, I would want to know why it is so much higher than other ETFs. Needless to say, the currency issuer has responsibilities that go far beyond what an ETF issuer has. It has to ensure both the physical and legal integrity of private property and that could be one legitimate source of additional costs. But if I found out that the ETF issuer was printing up free shares for certain individuals and passing the inevitable costs down to the remaining holders, I would say that the ETF was acting improperly. Wouldn’t you?

            It seems that no matter where you look on the political spectrum, everyone is in love with the idea of free money on some level, whether it takes the form of welfare payments, interest payments, subsidies or overpaying for goods and services. Maybe there are good reasons for some of these things that trump the needs of very long-term currency holders. I’m ok with that possibility. But you can’t argue for free money out of one side of your mouth while the other is acting surprised that the currency doesn’t hold it’s purchasing power for centuries, or significant fractions thereof.

            • golfer1john

              I see what you’re saying, and if the economy were as simple as an ETF, I could agree with it all.

              I think, though, that whether government hands out free money or not, the thing that influences the value of the currency is the difference between spending and taxing.

              Looked at another way, what is the difference in economic effect of these two actions:

              1. A general tax cut of $200 per person
              2. A “free money” helicopter drop of $200 per person

              You seem to be saying that 2 is inflationary and 1 is not. I don’t know of any economic justification for that position.

              As for the harm, yes, people disagree because many benefit from it. Anyone with a big fixed-rate mortgage and a wage that keeps up with prices makes out like a bandit. Inflation is great, give me more. In the aggregate, though, the private sector is a loser. If government truly controls the inflation rate, it has no business deliberately and covertly penalizing one group of people to reward another.

            • golfer1john

              What you said was

              “The currency that was created by Bush’s tax cuts will eventually cause the same type of inflation.”

              Implying that it is not yet causing inflation, but will cause it at some unspecified time in the future.

              Now you say future inflation

              “will be determined by economic conditions at that time”

              and not by the Bush tax cuts. Which is it?

              • Number 1 and 2 are both identical in terms of their consequences on balance sheets. I would say they are both dilutive, (but not necessarily inflationary) to the currency because in both cases there is an increase in nominal assets on the balance sheet without any increase in the real assets.

                The distinction between dilution and inflation is that dilution is a change in the relative quantity of nominal and real assets, but it doesn’t translate into inflation until the excess of nominal assets gets injected INTO things, thereby inflating their market values.

                When the wealthy get tax cuts, they are more likely to invest those additional funds (or use them as collateral against a trading strategy if they invest in hedge funds) in attempt to acquire even more money. It’s only when those “investments” generate losses that the additional currency that was created becomes inflationary. So yes dilution leads to inflation but only with an unspecified time delay as new currency that was never earned into existence trickles back into the market.

            • golfer1john

              The dollar did hold its purchasing power for more than a century, from 1791 to 1913. Government paid interest then, as it does now, and it gave out “free money” in other ways.

              If interest and free money cause inflation, why did that relationship fail for 122 years?

    • golfer1john

      FICA doesn’t “fund” Social Security. Monetary sovereigns do not need to tax or borrow in order to spend.

      Money is fungible. If you can say that Social Security payments are not inflationary because of the FICA tax, then I can say that interest payments are not inflationary because they are funded by the corporate income tax, and it is defense spending that is not offset by taxes, and is the cause of the deficit/dilution. There is no economic relationship between any particular tax and any particular expenditure or budget category. The amount of the deficit matters to the economy, but it makes no difference that you, or even Congress, can associate any expense with a certain tax. That’s completely arbitrary and economically meaningless.

      At full employment, any current expenditure that does not purchase real goods and services must be offset by leakages or taxes. And expenditures that do purchase real goods and services that otherwise would have been consumed by the private sector will also only raise prices, unless that private sector demand is reduced by taxing.

      Any deficit “dilutes” the currency. But there are other factors that negate that dilution, which are called “leakages”: domestic savings, and foreign savings (aka the trade deficit). Some amount of “dilution” is necessary to offset the leakages, else there would be deflation. Actually, MMT claims that the deficit is “largely endogenous”, meaning that it is determined (by accounting identity) by the other factors, and not by the will of the government to target any specific size deficit.

      It is not that holding and not spending “deserves remuneration”. All it deserves is what the ETF issuer provides, a reliable store of value. Currency doesn’t provide that these days, and the fact that it depreciates drives people to hold their savings in other forms. There are risks associated with holding assets in any form. Stocks and bonds may lose value, maybe even all their value. These days, holding currency is a guaranteed loss, because of inflation. But holding currency also provides liquidity, and that has some value. Everyone holds some currency, enough to conduct their transactions. But because of inflation, people are reluctant to hold much more cash than that. They invest in other assets with less liquidity that have a fairly sure or even risk-free expectation of a return in excess of the negative return of holding cash (bonds, including government as well as corporate bonds), or a possibility of much larger gain (stocks, real estate, collectibles, etc.). Zero is probably the appropriate risk-free overnight rate, but there would still be a yield curve even on Treasuries. Even though there is no credit risk, there is interest rate risk and lesser liquidity. The only way longer-term Treasuries would pay 0% is if the Fed bought them all. Anyone else looking for zero credit risk would prefer demand deposits.

      • Any deficit “dilutes” the currency.

        No. You are missing a very important point. I use the word “dilution” to refer to a situation where new currency is created without any new wealth being created. Just as ETF holders would be diluted if the ETF issuer created new shares without receiving the underlying stock. All deficits are not this way. Only the portion of deficit spending attributable to free money in the form of interest payments and subsidies and so on.

        When government spends on projects that contribute to the real wealth basket, such as education, infrastructure, healthcare and defense, it does not create dilution. This is the analog of the ETF issuers creation trade, where new ETF shares are issued as compensation for an increase in the stock holdings of the ETF itself. So deficits (net issuance of shares) are only justified in one way, and that is by growth of the underlying basket. That growth might have already happened in the private sector, leading to unemployment which creates a need for increased spending via a JG. Or that growth might be in the future in the case of a speculative government that wants to create something new, like the internet or a space program. Liberals and conservatives can bicker of which of these two growth paths are better. I happen to believe in both.

        But the important point is that net issuance of currency can only be justified in one way that is consistent with liquidity and price stability. That is by an expansion of the real wealth basket. And by the same type of reasoning, a net reduction (surplus) in the quantity of currency can only justified by a contraction of that same basket.

        Or, as I keep saying over and over, the quantity has to float to keep the price fixed.

        • golfer1john

          “That growth might have already happened in the private sector, leading to unemployment which creates a need for increased spending via a JG. ”

          No, growth does not cause unemployment. Unemployment is caused by excessive taxation.

          Mosler’s story of the origins of a monetary system are instructive. Government issues the money to buy the labor of the citizens, and institutes a tax to give value to its money. If it wants to hire 100 citizens, and taxes enough to hire 110 citizens, then 10 citizens will be unemployed: they need to work to earn their tax liability, but government is not hiring, and nobody else has extra government money to give them: all of it is due in taxes. If government only taxed enough to hire the 100, there would be no unemployment, because there would be enough money in the system for all the people to work and pay their taxes.

          At this point the simple economy is enough like the ETF for the analogy to hold very strictly.

          As the economy develops, the people will want to save, so that they can pay their taxes even if they are sick and can’t work, or if they are old and want to retire, or whatever. In that case the government must make “free money” available to satisfy the savings desires, either by reducing taxes or raising pay or hiring workers it doesn’t need, or just giving it away. Which one they choose makes no difference. ETF’s cannot do any of them.

          If people took ETF certificates and buried them in the ground, or in safe deposit boxes, or burned them or lost them, that would be the analog for savings in the economy. New ETF shares would have to be issued to replace the ones that had disappeared from the system, else the ETF would have more underlying shares in its vault than there were ETF certificates in circulation, and an ETF certificate would be worth more at liquidation than the basket of shares that purchased it: deflation.

        • golfer1john

          Yes, if the price is to be fixed the quantity must float. That means government must either take it away or give it away.

  3. financial matters

    J.D. Alt addresses the issue of Treasuries..

    “We have added a “Treasury Bond Account Chamber” which holds the Dollars that have been transferred into Treasury Bond “accounts” for a period of time. We have also added an “INTEREST” spigot to the FG pot—the Dollars paid as interest on the Bonds flows from this spigot into the PS pot. We can see that this plumbing arrangement, in effect, prevents a very large number of Dollars from chasing goods and services in the Private Sector economy—something that would push up prices (inflation)—and replaces those Dollars with a much smaller purchasing power (the interest payments.)”

    If the Fed issued bonds for public purpose projects it would have a similar effect. It would provide a relatively risk-free investment vehicle for pension funds and money market funds while providing useful employment. Not a bad use for the Feds Congressional mandate. See the conclusion of Wray’s second paper on the Fed.

    One of the main problems of money sitting on the sidelines is that it does not see the demand that would cause it to increase production.

    I think Godley’s sectoral approach is the correct way to look at this where private wealth comes from government spending. Healthcare and education are two examples right now where we are creating private debt where the government could come in and take this expense off of individuals and provide a much more efficient socially productive use of resources which would have positive feedback for the economy. Right now students can’t afford to buy homes because of outrageous student loan debt. And medical bankruptcy and overcost of new medicines, etc is a national disgrace.

  4. I would like to point out that a significant portion of the world holds that interest on money should not be paid and that idle money should be taxed which encourages its use and discourages “hoarding”.

    • I agree with the first idea, and that is one of my major points. No free money. No compensation for the mere act of holding the currency.

      But on the second point I have to disagree. I can’t see how a free market can work with a currency that does not have a sufficient shelf-life. You need to be able to accumulate value in the form of currency over time to be able to invest and participate in the entrepreneurial process that defines capitalism. It isn’t hoarding. It is capital formation. Of course not everyone believes in capitalism but I do.

      • golfer1john

        If the value is chronically decreasing, there is no compensation for holding currency, there is a penalty. No “shelf life”. Something has to offset the penalty, or people will hold more risky assets in an attempt to sustain their savings. Without savings, people are dependent on “free money” to sustain themselves after they can no longer earn the money.