Saturday, November 8, 2014

A Workable Commodity Reserve Currency

I seem to have developed a reputation as being a "market monetarist" but not being an Austrian economist according to one corner of the web. I am far more interested in the principles behind policy rules than the rules themselves. In fact, I have been considering the nature of a proposal different than a nominal income target, but which is derived from the same principles. This proposal came to my attention when David Glasner decried the possibility of the Swiss National Bank adopting a fixed ratio of gold reserves as this issue will be decided in a vote at the end of the month.
There they go again. The gold bugs are rallying to prop up the gold-price bubble with mandated purchases of the useless yellow metal so that it can be locked up to lie idle and inert in the vaults of the Swiss National Bank. How insane is that?
Being intimately familiar with the gold standard in both its historical and theoretical forms, I noted in the comment that having a fixed reserve ratio without a fixed-exchange rate for gold itself may actually make for a good monetary policy rule. As the price of gold rises, the bank will sell its gold. When the price falls, the bank will purchase more gold. There is a problem though. Employing only gold as a reserve commodity adds noise as the price of a single commodity is unlikely to reflect underlying market conditions. I suggest this can be amended by broadening the basket of commodities.

The functioning of a commodity reserve currency is straight forward (Hayek 1943; Graham 1944). Hayek and Graham suggested that base money be fully backed by a basket of commodities and that authorities peg the price of the basket, selling the basket when the price rose above the target price and purchasing it when its price moved below that price. I suggest some slight deviations from this formulation. Under a regime with fractional reserves, the monetary authority agrees to back a particular fraction of its notes with a basket of a broad array of commodities. When the price of these commodities rise, the central bank must automatically sell its holdings of this basket until the ratio returns to its statutory level. If the value of the basket falls, the central bank will purchase shares of this basket to maintain the statutory reserve ratio. As prices are procyclical during the business cycle, this policy automatically acts as a stabilizing force in the economy, as it stabilizes prices and provides liquidity when money is dear. This rule is convenient because, once the basket of commodities and operating structure are chosen, it is simple to execute. That the base money stock is responsive to changes in prices means that this rule promotes a money stock that is adaptive to changing market conditions. That is, it tends to facilitate changes in demand for money. Furthermore, since not all base money must be backed by the basket, expansion and contraction will also target credit markets and, therefore, regulate them according to market conditions as reflected by the value of the basket of commodities.

* I'm still working out the details, so I would be more than happy to read your critiques in the comments.
** The above is taken from an amendment to "Wither Gold". The new draft is up online!

Late: The 20% reserve requirement for the Swiss National Bank is a minimum, not a fixed reserve ratio.

9 comments:

  1. James,
    Here is why I think a nominal target is preferable to a fixed basket of commodities, even though I support a fixed anchor or tenet which is not necessary central to market monetarism. The primary fixed variable in both a micro and macro sense is the use of our time. A nominal target implicitly follows this in a complex economy under normal conditions, such as prior to 2008 before time aggregates began to be discounted. I have my own issues with that of course and seek to return time aggregates to normalcy.

    A nominal target recognizes how asset formations and commodity flows coalesce around the use of time aggregates. Which is why a nominal target already has a fixed component which is more reliable than other variables because it has a constant, dependable relation to money use, unlike other resources and commodities which gradually change. Granted, a commodity basket catches some of this but there are times when it would still create hardships for aggregate time use with certain shocks, just as happened with oil.

    ReplyDelete
  2. Becky,

    You make good points. The trouble is several fold with nominal income targeting. 1) The accuracy of the nominal income estimate and the weighting of prices are very tricky. 2) A futures market must be set up to estimate expected nominal income. 3) Choosing a level is not a clear cut exercise, and would likely require a clause for revision. That said, I think that a nominal income level target is far superior to our current system. If it was adopted, I would only harp on those issues to the extent that they were causing problems.

    Having a broad basket of goods greatly limits price volatility of the basket. Instead, the price of the basket will reflect the underlying conditions of the market more generally. Nicely, it's implementation is much simpler than a nominal income target and need only market prices to function. Once established, its operation is straight forward.

    ReplyDelete
  3. I think your proposal is similar to what we have now in a sense. If commodities in the basket go below the fix level the fed buys these commodities to expand liquidity. This is similar to inflation going below target and the fed reacting with QE.

    IMO what is missing here is the realization that counterparties are key because firms have more unstable money demand and it can become extremely elevated or even infinite. Buying commodities will mostly expand liquidity to firms which have this unstable money demand. If liquidity is expanded to general public with a lower and more stable money demand then we have something new.

    ReplyDelete
    Replies
    1. The rule is the important part. Right now, policy is given to discretion. Central bankers decide policy based on any number of factors that they think are important.

      There is no such thing as an infinite demand for anything. Wants are unlimited, but demand is constrained by one's budget. Sometimes, a shortage of money at given prices leads to sharp economic contraction.

      The purchase of commodities would channel the expansion into the labor market during busts, and pull it out during booms.

      Delete
  4. "There is no such thing as an infinite demand for anything. Wants are unlimited, but demand is constrained by one's budget. Sometimes, a shortage of money at given prices leads to sharp economic contraction."

    Maybe extremely elevated is more representative that infinite.

    "The purchase of commodities would channel the expansion into the labor market during busts, and pull it out during booms. "

    It would channel money to commodity producers. How much of that would go to the labor market Im not sure. Channeling money to commodity firms is similar to expanding liquidity to firms through QE. Firms have very elevated money demand at zlb and unstable money demand due to higher speculative demand.

    Heli drops on other hand channel money to general public with high MPC and low stable money demand.

    ReplyDelete
    Replies
    1. Firm money demand is largely a function of risk during times when it increases in large spurts. This would be alleviated by this policy, especially for firms that produce two modifies in the basket. They would be more than willing to hire new employees than otherwise. If commodity producers increase output to meet meet an increase in demand, they have to increase employment, at least until robots can do everything more cheaply than humans.

      I have never seen a defense of the unstable money demand hypothesis that I have found convincing. If you find one please let me know.

      Also, if you don't start using your name, I will delete your posts. This is your only warning.

      Delete
    2. I cant login using the options available. Check out my website if you want to know who I am. cmamonetary.org You need to use a better ID system.

      Money demand is a function of risk and return. This policy would alleviate money demand about as effectively as QE. There is some degree of effectiveness but limited.

      "If commodity producers increase output to meet an increase in demand, they have to increase employment, at least until robots can do everything more cheaply than humans."

      Are commodities producers going to increase output because the central bank is buying commodities to hoard? Maybe not maybe a little. How much of the economy will be stimulated by stimulating commodity production anyway? Its better to stimulate the entire economy broadly and evenly through heli's.

      Its just common sense that money demand is key in monetary policy and that firms and people have a different money demand.

      Delete
    3. Be sure to leave your name with each post.

      The prime motivation behind the proposal is money demand. You seem to be claiming that a predictable monetary policy does debit reduce uncertainty and that firm activity is not a function of expected demand. That is absurd. QE is anything but predictable. The Fed has expanded its discretionary power during the crisis. I'm having a difficult time finding coherence in your objection.

      Delete
  5. Please leave your name with your comments.

    ReplyDelete