Sunday, December 28, 2014

Robin Hood central banking

Robin Hood, N.C. Wyeth, 1917

There were plenty of reports in the press this year accusing central banks of behaving like King John, stealing from the poor to help the rich. Rich people's wealth tends to be geared towards holdings of stocks and bonds whereas the poor are more dependent on job income. By pushing up the prices of financial assets, central bank quantitative easing helped rich people while leaving the poor in the dust.

There are a lot of problems with the King John critique of quantitative easing.

First, a good argument can be made that QE had almost no effect on prices. Insofar as purchases were considered temporary by market participants, then the newly created money would not have been spent on stocks and whatnot, its recipients preferring to keep these balances on hand in order to repay the central bank come the moment of QE-reversal. If so, the large rise in equity prices since 2009 is due entirely to changes in the fundamentals and animal spirits, not QE.

But let's say that QE was not irrelevant and can be held responsible for a large chunk of the rise in equity prices over the last few years. Even then, the real economy, and therefore the poor, would have been equal beneficiaries of QE. As I pointed out in my previous post, financial markets are not black holes. Newly-created money, insofar as there is an excess supply of the stuff, cannot stay 'stuck' in financial markets forever. For every buyer of a financial asset there is a seller, and that seller (or the next seller after) will choose to do something 'real' with the proceeds, like buying a consumption good, investing in real capital, or hiring an employee—the sorts of purchases that benefit the poor. So if QE succeeded in pushing up financial markets (thus helping the rich), then the real economy (and the poor) must have benefited just as much. The King John argument doesn't hold much water.

But wait a minute. If both financial markets and the real economy were equally inflated by QE, then why have wage increases been so tepid relative to equity prices? One explanation is that wages are sticky whereas financial prices are quick to adjust. The relative wealth of the poor, comprised primarily of the discounted flows of wage income, stagnates, at least until wages start to catch up at which point it is the turn of the the relative wealth of the rich to decline.

Can we right this short-term wrong? Even if we try to convert central banks from being King John central banks into Robin Hood ones, things wouldn't change. Say we change where central banks inject new money. Instead of conducting QE with a select group of banks, central banks now purchase directly from the populace. And instead of buying financial assets, they bid for stuff that regular folks own, like cars, houses, and wedding rings. The moment Robin Hood QE is announced, the same effect occurs as when King John QE is announced: the prices of financial prices will be the first to jump. This 'injustice' occurs even though the counterparties to Robin Hood QE are all too poor to play the stock market and the items being purchased from them are not financial assets. Consider that an impoverished recipient of new funds may use them to purchase something at a grocery store, and the owner of that store may in turn use the proceeds to buy new inventory, and the farmer producing that inventory could use the funds to buy seed, etc etc. Someone along this line will eventually purchase shares. However, stock markets participants don't wait for this excess money to flow into stock markets before marking up prices; they adjust their offers ahead of time upon the expectation of excess money being used to purchase stocks. Robin Hood QE or not, the relative wealth of the rich is the first to rise thanks to flexible financial prices, at least until sticky wages start to catch up. This isn't the fault of central bankers, and there's no way they can restructure their operations to promote short-term equality in wealth.

How long can this short-term inequality last? I can't see it lasting longer than a year. Maybe two. But we've seen so many years of QE now that I don't think we can attribute the gap between the rates of increase in stock prices and wages to stickiness. The most likely explanation is the one in my third paragraph: on the whole, QE has done very little to affect prices, whether they be financial or not. Wages have been stagnant because QE is irrelevant, or at least close to it, and the S&P 500's rise from around 700 to 2090 has been by-and-large achieved of its own accord. Without QE, where might the S&P be? Maybe 2060, or 2065?

Central banks aren't like King John, nor is there anyway we can turn them into Robin Hoods.


  1. What if the only assets that a central bank was allowed to purchase was a variant of Miles Kimball's FLOC, a standard NINJA loan extended to all adults that had a standard interest rate and a standard capped amount, say $2000 with a 15-20% interest annual rate, compounded weekly?

    Monetary expansion would be forced to go through, quite literally, the poorest and most desperate people. If the CB is undershooting its target, it reduces the interest rate on the loan, viceversa for overshooting. This policy, if it gets too close to the ZLB, can increase the loan amount without limit.

    Would this not be a robin hood policy?

    1. If the central bank did open market operations of NINJA loans, the effect would be the same as if they bought treasuries. Investors would anticipate this new money flowing into financial markets and quickly adjust asset prices higher -- sticky priced goods would be slow to adjust. We still get the same distributional effects.

  2. Distributional cantillon effects from Robin Hood QE? Financial markets would adjust yes - but in increasing expected inflation rate the NPV of the new money in financial assets would be less than the NPV of the money held in hand by jo public, because the higher inflation rate would decease the NPV of the financial assets. So by definition Robin Hood QE would have positive distributional benefits if an only of the new money was seen s permanent and permanently shofted NGDP without a CB inflation target offset.

  3. Exactly, and I would add that asset prices suddenly jumping only makes the rich richer if they sell their investments and spend the money. It doesn't immediately raise the returns they get from the assets, in fact, it lowers them relative to the new prices. The returns only rise after the gains have been spread to the greater economy and even there, low unemployment enabling a better bargaining position for labor, the rich may lose a big part of the gains to workers.

    So yes the initial gain (at least on paper) is to the rich, but the real long term gains is to everybody band likely skewed in favor of the worker.

  4. “And instead of buying financial assets, they bid for stuff that regular folks own, like cars, houses, and wedding rings.”

    Two reactions:

    a) I had to double check to make sure I wasn’t reading a Nick Rowe post :)

    b) Institutional architecture matters I think. The existing process is “trickle-down” in the sense that institutional portfolio managers are the first level recipients of QE money injections. They will rebalance their portfolios (primarily with each other at first) with some consequential flow of funds between those institutions and their clients (including retail). But the primary front end effect is predominantly financial and institutional rather than real and retail.

    Your alternative vision reverses that order. But I think it’s likelier that money remains trapped at low (financial) velocity in the banking system this way. The initial breadth of the QE money distribution is restricted in institutional scope – the money distribution will flow first through the back door of the banking system instead of through the front doors of broader financial system portfolio managers.

    For example, in bond QE, a sell order for bonds flows from an insurance company portfolio manager to an investment dealer. That starts an active portfolio allocation process and ripple effect at the institutional level.

    In automotive QE, a sell order for a car from an individual moves new money into a retail bank deposit account. The process by which that money starts to affect broader institutional portfolio managers may be slower – bottom up. As a first response, the bank reserve manager may just say “what – more reserves? Nothing our loan officers can do, and I’ve had enough of this hot potato business in a closed reserve system. Maybe we should just increase our deposit fees to offset the margin cost. Meanwhile, I’m going to the golf course.” For banks, that piece is not dissimilar from the case of bond QE, but the essential interaction with and broader exchange through other institutional portfolio managers is missing as a first order effect.


    “The newly created money would not have been spent on stocks and whatnot, its recipients preferring to keep these balances on hand in order to repay the central bank come the moment of QE-reversal.”

    Interesting, and weirdly Ricardian-equivalent-like.

    Somebody will “repay” by buying bonds as a discretionary alternative to holding money. But it’s not as if they’re in debt to the central bank as is the case with taxes.

    1. "...that institutional portfolio managers are the first level recipients of QE money injections. They will rebalance their portfolios."

      If QE is announced publicly, then this rebalancing will occur even prior to injection. Portfolio managers will anticipate the effects of the injection so that by the time it occurs, its effects will have been built into prices.

      In automotive QE, the same thing occurs. Portfolio managers will anticipate the new funds eventually arriving in financial markets (after progressing from car buyers to car dealers to car suppliers etc) and will reprice securities prior to the actual arrival of those funds.

      So even if they are first level recipients or back door recipients, upon the announcement of QE participants in financial markets will always rapidly mark up the prices of the financial assets they deal in prior to the actual injection.

  5. In automotive QE, the same thing occurs. Portfolio managers will anticipate the new funds eventually arriving in financial markets... and will reprice securities prior to the actual arrival of those funds.

    Yes, but under automotive QE**, households can immediately increase their real cash balances which they can spend or use to pay off loans. In comparison to bond QE, which must first trickle-down through the banking system, households benefit right away, don't they?

    **Or let's say a helicopter drop or FLOC, since I don't think anyone is advocating for a literal "QE for the people." However, Ashwin Paremesan, author of the Macroresilience blog, seems to be promoting a privatized version of FLOC/QE for the people at his site Seems like an interesting experiment.

    1. But households can already increase their cash balances by auctioning their cars to a non-Fed actor ie. they can already use a car sale to pay off loans. All that changes with automotive QE from the perspective of a household is the source of the funds. Whether they sell at a car auction or a Fed-organized auction, they've not been made any wealthier.

    2. I'm not advocating car QE, but let's try it as a monetary parable.

      But households can already increase their cash balances by auctioning their cars to a non-Fed actor

      Sure, but compare car-owning households now to car-owning households under car QE (let's imagine that instead of QE3, the Fed announced an open-ended pledge to buy $40 B worth of used cars every month).

      (Some quick estimates--annually about 40 million used cars are sold in the US at an average price of $15,000. So let's say monthly sales are around $50 B).

      Before car QE, households that hadn't sold off one or more of their cars valued the fundamental value of their cars more than the cash that they could have received at market prices. Since car QE almost doubles the amount of money flowing through used car markets, prices will rise and many households on the margin will now want to sell (making do with one car, Uber, or the bus).

      The additional sellers under car QE benefit by the extra amount they can sell their cars for relative to the no-car-QE world (and even car owners who don't sell benefit because they retain the option to sell at a better price than they could have without car QE). Also, since cars are now more salable, car owners would benefit from the slightly increased liquidity premium of used cars.


      1. Even if the prices of financial assets jump in anticipation of higher future money inflows into asset markets, under car QE, car owners would at least benefit from either higher prices on sales that wouldn't have otherwise occurred or the appreciation in the value of their cars. They can used their improved financial positions to increase spending (due to their increased real balances) or pay down debt. Isn't this outcome more Robin Hood than the actual QE3?

      2. Wouldn't buying tangible assets, such as used cars, be a more reliable way to ignite a hot potato effect than buying financial assets? If the NGDP increase and reduction in unemployment were faster under car QE instead of bond QE, wouldn't this also be a Robin Hood outcome?

      3. Re: "fairness"--under car QE, some people wouldn't benefit (relative to the non-car-QE scenario), namely people who don't own cars. In fact, people who needed to buy a used car would suffer, since they would have to pay higher prices.

      If we substitute "Treasuries/agency MBS" for "cars," what changes? Don't the holders of assets directly purchased under QE3 benefit in the same ways from capital appreciation and/or higher selling prices that car owners under car QE would? Furthermore, couldn't this "unfairness" be eliminated by helicoptering the new money equally to everyone?

      Again, I'm not necessarily pushing a particular policy, but I just want to address your claim that central banks can't act in more Robin Hood-like ways. It seems to me that Robin Hood-oriented policies are at least imaginable.

  6. Questions: under classical LOLR theory, the idea is to relax the survival/liquidity constraint for solvent but illiquid financial institutions. But from a purely theoretical perspective, why should this benefit only be available to financial institutions? What makes them more deserving than other firms or households?

    According to Minsky, all agents in the economy, including households, can be seen as banks. So what would be the argument against LOLR for everyone (e.g. FLOC for households)? If gov't debt is a concern, unpaid loans could be penalized with a longer waiting period to receive Social Security benefits (e.g. if I'm due $1,000/month in SS and I'm $10,000 in arrears on my FLOC, I have to wait 10 extra months to start receiving SS).

    1. "...why should this benefit only be available to financial institutions?"

      I have no problems with central bank liquidity insurance being only available to financial institutions with one caveat, it needs to be done at market prices. Central banks probably don't price this insurance product properly. Don't take away the product, just implement it better.

      Households can insure themselves against illiquidity to... see: