Wednesday, October 14, 2009

Index Futures Convertibility: A Constitutional Approach

On Cato Unbound, and on his blog, Money Illusion, Scott Sumner has described index futures targeting. The details of Sumner's proposal are somewhat fluid, and what I call index futures convertibility is one among many institutional frameworks within a broader set that also includes his schemes.

When talk turns to "schools," such as Keynesian, Monetarist, New Classical, or Austrian, my allegiance is closest to the Virginia School. Macroeconomics has never been a strength of the Virginia School, with Public Choice economics or Constitutional Political Economy being the central focus. However, the title of the 1962 Yeager-edited volume, In Search of a Monetary Constitution, summarizes the flavor of Virginia School macro.

When I was a student, first at Virginia Tech, and then at George Mason, the emphasis was on money creation as a revenue source and the various agency problems in politics that make unstable inflation a danger. Given James Buchanan's key role in the Virginia School, and his roots in Public Finance, that should be no surprise. However, Buchanan was trained in "old" Chicago, with Knight, but also Mints and Simons, at a time when the Great Depression still seared the memory of all economists. Constraining a political order inclined to excessive and variable inflation may have been the emphasis in the seventies, but developing an institutional framework that would provide the proper incentives for monetary stability was the fundamental goal.

Simons' framework for monetary reform was a feedback rule from deviations of the price level from target to changes in the quantity of money. The assumption of that approach was that velocity can change and some scheme for compelling offsetting changes in the quantity of money is necessary. Buchanan went so far as to describe a common brick standard. In a depression, there may be a surplus of most goods and services, but with a common brick standard, people all over the country could produce new bricks and new money.

However, the shadow of Milton Friedman loomed large over Virginia School monetary political economy. His "middle" Chicago School had marshaled impressive empirical evidence that independent fluctuations in velocity were an insignificant problem and that changes in some plausible measure of the quantity of money have been the source of nearly all instability in nominal expenditure. Imposing a money supply rule looked to be an obvious approach to a monetary constitution.

Unfortunately, imposing a rule of slow steady growth on the monetary base--currency and bank reserves--appeared at odds with Friedman's critique of Fed actions during the Great Depression. The M2 measure of the money supply fell because the Fed failed to increase the monetary base enough to offset a drop in the money multiplier. Slow steady growth of base money would have prevented the Fed from maintaining the slow steady growth in the M2 measure of the money supply that would supposedly generate slow steady growth in nominal expenditures.

Friedman's critique of the Fed during the Great Depression suggests that the monetary authority be given discretion over the quantity of base money, subject to the constraint that the M2 measure of the money supply remain on target. How could this restriction be enforced? Arrest of the open market trading desk personnel for treason should auditors discover that the money supply has deviated from target?

As the stable relationship between M2 and nominal expenditure broke down at the end of the 20th century, interest in motivating the monetary authority to keep some amalgam of assets more or less related to the medium of exchange on a stable growth path waned. Returning to Simons, the constitutional approach would be to give the monetary authority discretion to change the quantity of base money subject to the constraint that it keep some specified measure of the price level on target.

How could such a rule be enforced? Imagine the hearings before the House and Senate banking committees after a deviation of the CPI from target. The testimony should begin with an apology from the monetary authority for failing to adjust base money properly. Of course, excuse making would soon follow, starting with the most obvious one, that the models used to determine the proper amount of base money were in error. And then, more excuse making, explaining how such errors could not be avoided due to recent financial innovation, the aftermath of an asset bubble, or whatever.

Sadly, the actual testimony would likely skip all of the accountability and go straight to the fundamental excuses--the price level deviated from target because of financial innovation, or the collapse of an asset price bubble or whatever it is that in hindsight was not properly accounted for in the monetary authority's model.

A different approach to a monetary constitution allows for a very simple scheme of enforcement. The unit of account, the dollar, is defined as a specific weight of gold or silver. Then, the monetary authority can issue whatever quantity of dollar-denominated base money it chooses, subject to the constraint that its price remains at par. The monetary authority must adjust the quantity of base money so that it never trades at a discount or premium from face value. However, the historical record of the gold standard suggests that pricing is inevitably in terms of the medium of exchange. The rule then amounts to requiring that the monetary authority adjust the quantity of base money so that the market price of gold (or silver) remains at its official, defined price.

The restriction can be enforced by redeemability. The monetary constitution can obligate the monetary authority to stand ready to redeem the base money it issues for gold at par. In other words, the monetary authority can be required to sell (and buy) gold at the official, defined, price.

As long as the gold standard is maintained, it protects against the danger of the government using excessive money creation as a source of public finance. After the monetary authority creates the maximum amount of base money consistent with the price of gold remaining at its official price, the price level depends on the supply and demand for gold. The quantity of base money, and the quantities of all the more inclusive amalgamations of monetary assets must adjust to their demands at a price level fundamentally determined in the gold market.

The problem with the gold standard is that the supply and demand for gold are unlikely to result in a stable price level. If the supply of gold grows more rapidly than the demand for gold, the result in inflation. If the demand for gold grows more rapidly than the supply of gold, the result is deflation. A monetary authority that holds gold reserves that are large relative to the total stock of gold can release or accumulate gold reserves and so manipulate the gold market. Then, of course, it is evident that there is some other, more fundamental, macroeconomic goal than a fixed nominal price of gold.

What happens when the gold standard fails? It is a story repeated again and again. We left the gold standard because our reserves were depleted. We were busily stabilizing credit conditions, employment, prices--take your pick--then unforeseeable changes in international trade, perhaps including a speculative attack, depleted our reserves. And so, money is no longer redeemable.

Index futures convertibility is similar to a gold standard, but rather than tying money to gold, redeemability is used to tie money to an index futures contract on the macroeconomic magnitude whose stability provides the best environment for microeconomic coordination. The monetary authority can adjust the quantity of base money as it sees fit, but subject to the constraint that the price of money is fixed to an index futures contract. In other words, the monetary authority has discretion to adjust the quantity of base money subject to the constraint that the price of an index futures contract remains on target.

How is the constraint enforced? It is enforced through redeemability. Index futures convertibility enforces the obligation to fix the price of the index futures contract by obligating the monetary authority to redeem money with index futures contracts. Or more precisely, to buy the contract from, and sell the contract to, market participants at the official price

Index futures convertibility could be applied to any macroeconomic measure. The most obvious candidate is a measure of the price level, such as the Consumer Price Index. However, a much better approach is to create a stable growth path for aggregate nominal expenditure. The monetary authority should be obligated to maintain convertibility between the base money it issues an index futures contract on total final sales at a price that reflect a 3% growth path. With a 3% trend growth rate in productive capacity, this should result in a stable price level in the long run, but with deviations of productivity from trend resulting in opposite changes in the price level.

Like the gold standard, the scheme should work well to deter excessive money creation as a source of public finance. If the monetary authority gives in to pressure to create money to fund government programs, tax cuts, or debt monetization, the excess money will push nominal expenditure ever further beyond its target. Even if this new policy were sprung upon the market as a surprise initially, nominal expenditure will soon leave the target growth path behind. The monetary authority will be obligated to sell index futures contracts on nominal expenditure at the target price. With settlement values being ever further above this price, the speculators buying the contracts will be making ever greater profits at the expense of the monetary authority. The monetary authority will lose money, rather than earn an income from excessive money creation that it can transfer to the government.

Index futures convertibility also harnesses market forces to the goal of macroeconomic stability. While the monetary authority can create the quantity of base money it believes will best keep nominal expenditure on the target growth path, bear speculators, who believe that too little money has been created so that nominal expenditures will be lower than target, can sell index futures contracts to the monetary authority. On the other hand, bull speculators, who believe that the quantity of money is too high so that nominal expenditures will be above target, can buy index futures contracts from the monetary authority.

The trading of the bear and bull speculators provide a signal of market expectations about the impact of current money market conditions for future nominal expenditures. If the trades of the bears and bulls exactly match, then the monetary authority is fully hedged. The "market" and the monetary authority agree that the current level of base money is creating current money market conditions consistent with nominal expenditure remaining on target. When nominal expenditure is realized, if it is below target, the monetary authority transfers funds from the bulls to the bears. On the other hand, if it is above target, then the monetary authority transfers funds from the bears to the bulls. Those whose forecasts were more accurate profit at the expense of those whose forecasts were less accurate.

If the market is bearish, with more speculators selling futures at the target price than are buying, then the monetary authority is left with a long position on the contract. The monetary authority, along with all the bull speculators, is taking a risk of loss if nominal expenditure is below target.

In the opposite situation, if the market is bullish, more speculators buy futures at the target price than sell, leaving the monetary authority with a short position and exposed to risk that nominal expenditure will rise above target.

By adjusting the quantity of money through conventional monetary policy, such as open market operations in government bonds, the monetary authority can impact current money market conditions and so market expectations of future nominal expenditure to bring about a balance of bulls and bears at the target value of nominal expenditures. A risk minimizing rule would be for the monetary authority to change the quantity of base money and current money market conditions such that its own net position on the futures contract remains equal to zero. This criterion is met if no speculators trade the contracts at all, but otherwise, it requires a balance between the trades of bear and bull speculators.

Suppose nominal expenditure deviated from target. The leader of the monetary authority testifies before House and Senate banking committees. What happened? There are three possibilities. The monetary authority was hedged, with either no trades by speculators or else balanced trades. The monetary authority lost no money on the contracts. Its leadership explains that the quantity of base money was too high or too low. No doubt, the standard excuse making occurs. The models were wrong because of unpredictable changes in financial markets, etc. However, the monetary authority can point out that market sentiment was divided. While some did correctly forecast the deviation of nominal expenditure from target, others expected the opposite deviation.

Then there is the possibility that they did have a position on the contract and made money. The market signal was wrong. The monetary authority can report that while they were in error and base money was too high or low, market sentiment was even more in error. The intervention of the monetary authority limited the deviation of nominal expenditure from target.

Finally, of course, there is the scenario where the monetary authority took a position on the contract, and lost money. They must explain that not only did they make a mistake, it was apparently an avoidable mistake, as shown by the profits earned by those speculators who correctly saw that money market conditions were inappropriate. And then, perhaps it is time to look for new leadership for the monetary authority, perhaps among those who were successful speculators.

Sumner's version of index futures targeting aims at avoiding all discretion by the monetary authority. The general idea is that ordinary open market operations in bonds are tied by some mechanical rule to trading in index futures contract. The simplest one is dollar-for-dollar parallel trades. Every time a bear speculator sells an index futures contract, and so, the monetary authority buys one, the monetary authority also makes an conventional open market purchase of government bonds. The expectation by speculators that nominal expenditure will be below target results in an increase in base money. Similarly, if a bull speculator buys an index futures contract, and so, the monetary authority sells one, the monetary authority also makes an open market sale of government bonds. The expectation by speculators than nominal expenditure will be above target results in a decrease in base money.

I am skeptical of mechanical rules. It is not realistic that a monetary constitution can end all discretion by the monetary authority. What it needed is a framework that provides incentives to promote macroeconomic stability. Index futures convertibility shows promise as useful limit on the discretion of a monetary authority.


  1. Bill,
    Thank you for this primer on the mechanics of index futures targeting. The details have always been fuzzy to me. You and Scott Sumner are the chocolate and peanut butter of monetary bloggers...good on your own but better together.

    I don't understand why open market sales of government bonds is necessary. Presumably the monetary authority is creating money to buy a contract from a bear speculator and creating a contract to sell to a bull. The monetary base is increasing when people are bearish and decreasing when bullish. What am I missing?

  2. Aaron:

    Thank you for your comment.

    Index futures contracts are a committment to make a payment in the future based upon the difference beween the realized value of the index in the future and a "price" in terms of the index determined by a trade made now. Usually, the index is multiplied by a dollar amount, so it is a difference between the dollarized value of the index in the future and a money price determined now.

    Both parties to the contract are required to put up funds in a margin account, which is really a performance bond. The buyer and the seller don't really transfer money now. They are committing to do so when the contact is settled. With a futures contract for a commodity, like oil, the buyer and seller are committing to exchange money for the commodity, though they usual settle up with a cash settlement based on the difference between the market price and the agreed price. With index futures, it is always a difference beween a realized value and agreed "price."

    With index futures targeting or convertibiliy, speculators and the Fed are making promises to pay money in the future based upon the difference between the realized value of the targeted macroeconomic variable and the target value.

    Don't be too concerned that you didn't get this right. Many of the pioneers in this area, such as Sumner and Dowd, had problems as well. I think part of the problem is that with normal futures contracts, funds are moved between margin accounts continually as the price of the contract changes. Some people do get more money right now on existing futures exchange based upon today's trading.

    But with index futures targeting, there is no change in price. And, anyway, while some people do make money now from changes in futures prices, this is just a transfer from those on the opposite side of the contract.

    Trades in index futures contract, even with the monetary authority pegging the price, do not cause changes in base money. The do create an incentive for the monetary authority to cause such chnages, and Sumner, for example, favors a rule requiring such changes.

  3. Bill, I finally got the post set up. It is a bit longer than I expected. let me know what you think.

  4. I forgot the link:

  5. "In a depression, there may be a surplus of most goods and services, but with a common brick standard, people all over the country could produce new bricks and new money."

    Thats the most important thing. That "people all over the country could produce new bricks and new money" a good monetary system should make it so a large number of people and institutions can produce money directly. In a free banking system you have that, but then you have the problem of huge numbers of types of currencies.
    A commodity index (or future's system) has that benefit while also having the benefit of being a single currency.

  6. I forgot to mention:

    Second of all, I don't think it will work for the same exact reason the gold standard didn't work. I don't think any amount of monetary restraint will work so long as the government has the power to abandon that system, or force inflation to allow itself to borrow cheaply.
    The only way I see to resolve this problem is to create a separation between economics and state. I am not sure how this is workable, and it harms one of the state's primary purposes (to be able to make war, when necessary)

  7. Doc:

    Thank you for the comment.

    Index futures convertibility can be combined with free banking. Banks issue dollar-denomiated monies redeemable in index futures contracts. Interbank clearings generate inter-bank positions on the futures. Using ordinary asset and liability management to impact net clearings generates market forces that keep the expected value of nominal expenditures on target.

    While there are multiple monies, they are all denmoinated in the same units and should generally exchange at par.

    There is little doubt that government can always change the monetary regime. While free banking requires that they do a bit more than usual--issuing legal tender notes in place of T-bills and other bonds--the Revolutionary War shows that it is not all that difficult.

    As for the monetary regime, including index futures convertibility or a gold standard, the best that can be hoped for is an explicit political decision to suspend or end the standard. For example, requiring a 2/3 vote by house and senate on this issue alone. Think about various proposals for a balanced budget amendment.

    With a gold standard, devaluation or revaluation might be desirable sometimes. With index futures convertibility, if the growth rate of potential output changes in a persistent fashion, perhaps the growth rate of nominal expenditure should be changed. In the context of a monetary constitution, these should be separate, special decisions. Perhaps they should be implemented with a lag (I don't know about changes in the price of gold.) Perhaps they should require supermajorities.

    Constituional thinking separates operations within the rules of the game (open market operations that keep nominal expenditure on the target growth path) and changes in the rules of the game (changes in the target for the growth path, or moves to a fiat currency so that govenment has another source of revenue.) The decisions should be separate. And changes in the rules of the game should require more political consensus--usually supermajorities.

  8. With any stable unit of account, one entity (or group) will inevitably end up controlling the supply of money and ab-using that control for political and/or religious purposes. It is not therefore possible to sever the link between government and money. As a matter of historical fact "money" is what government says it is. Money is what the sovereign will accept in settlement of a tax liability. That is the difference between money and whale's teeth (or tobacco leaves).

    If you want a fair and honest monetary system then you must teach monetary economics in the K12 system and insure that there is an actual representative government as opposed to the sham we currently have.

  9. No idea if you're still reading comments on posts this old.

    I think this works for a single point in time, but how does it work for a full futures curve? What if setting the money supply so that 2011 NGDP growth is expected to be 3% (or whatever the target is) makes it very challenging to hit the same target in 2012?

    Metaphorically, say you're driving a car and trying to stay close to the middle of the road. If you encounter some shock and find yourself driving nearly perpendicular to the road, you'll swerve sharply to get back on track, but may oscillate around the middle of the road for some time. If those oscillations are predictable, the market will be able to consistently profit from you. Or perhaps you can make them unpredictable by increasing volatility, but that's probably not a great outcome. The optimal "steering policy" is probably to understeer in the short term and miss the target in order to more smoothly get back on track in the medium term.

    I suppose you could perhaps address this by allowing the fed to suspend trading at the normal target and start trading at a new target if they deem that optimal. To enforce discipline, you could also require that they issue a longer maturity contract that must have the normal target. In terms of the metaphor, if you're way to the right, admit that you plan to miss right in the next 10 seconds, but commit to being right in the next 2 minutes.