The return of Monetarism vs. Keynesianism?

Maybe it isn't the return of Monetarism. Or rather, it isn't the return of only Monetarism. Maybe it's the return of Monetarism vs. Keynesianism, 1960's style, applied to unorthodox monetary policy.

At the risk (or certainty) of caricature, the 1960's argument went something like this:

The Monetarists argued that it didn't matter (much) what the government and central bank bought; what mattered was how they paid for it, and whether it was financed with money. Money-financed tax cuts, money-financed increases in government expenditures, open-market purchases of bonds with money, were all essentially the same. Just different ways of increasing the stock of money. You might as well (theoretically) throw the money out of a helicopter.

The Keynesians replied that it did matter what the government and central bank bought; and it mattered less whether they paid for it with money or bonds.

Apply that same distinction to unorthodox monetary policy today.

A monetarist would argue that if you cannot increase the money supply by cutting the overnight rate, because it's already at zero, then find some other way to increase the money supply. Buy short bonds, buy long bonds, buy commercial bonds, buy stocks, buy real assets, buy goods and services, buy tax cuts, buy whatever; or just give the money away. But just make sure the money supply increases and stays increased. It has to be a permanent increase in the stock of money. Fiscal deficits (if money-financed) are just one of many devices to get the stock of money to increase.

A Keynesian would reply that it does matter what you buy. Buying short-term government bonds won't help at all, because they are perfect substitutes for money. Buying long bonds, or commercial bonds, or stocks, might help, because it would reduce the supplies of those assets on the market and might raise their prices. And fiscal deficits will help whether money- or bond-financed.

It also sounds very much like Willem Buiter's distinction between quantitative and qualitative easing. Quantitative easing means expanding the size of the central bank's balance sheet, and if its only liability is money, that means increasing the supply of money. Qualitative easing means changing the mix of assets on the balance sheet, so it depends on what the central bank buys. Monetarists look at quantitative easing; Keynesians look at qualitative easing.

And remember the old debates about the transmission mechanism? Again at the risk (certainty) of caricature:

Keynesians argued that the transmission mechanism from money to inflation went via the short rate of interest, which influenced other rates of interest, which influenced demand for goods. So if the short rate of interest is already at zero, the monetary transmission mechanism is broken at the first link in the sequential chain.

And monetarists argued that there were multiple parallel links from money to demand. An excess supply of money would spill over into the short-term bond market, but also into the long-term bond market, and the commercial bond market, and the stock market, and the supermarket. So losing just one link would not break the net of links between money and the economy.

And this debate does not even make sense in a model of frictionless Walrasian markets, where "liquidity" is an unintelligible concept.

If you believe that this crisis, at least in part, is a crisis of liquidity, and people are desperate for liquidity, then the monetarist position has an at least prime facie plausibility. Liquidity is the ease with which a good can be converted into other goods. In a monetary exchange economy, all goods are converted into other goods via the medium of exchange. Money by definition is the most liquid of all goods, since it is the medium of exchange. So increase the supply of liquidity by increasing the supply of money.

But then buying stocks or commercial paper with money ought to satisfy both sides, and might be the most prudent option, since we (or at least I) don't know which side is right.

Comments

[JKH]

Nick,

I think its a (US) credit or solvency crisis more than a liquidity crisis. That includes banks, businesses, and households. Asset and balance sheet deflation. The Fed is undertaking credit easing (Bernanke's term for qualitative easing) with an increase in the size of its balance sheet (quantitative easing) only as a necessary result. If it starts buying Treasuries, it will be to manipulate Treasury and mortgage rates, not because it really wants to print money as a result. I'm not sure what the Bank of Canada's intentions are, given that they appear all to be smoking something quite heavy for another type of easing purpose in their economic forecast meetings.

[JKH]

On the other hand, the fear of future inflation, which is rampant already, seems to be entirely a monetarist fear.

[westslope]

From the master himself:

" If all you want is stability of the general price level, apparently it's an easy job to do. I really must say I am surprised at how well all of these different countries have been able to keep to their inflation targets." -Milton Friedman in interview with Robert Kuttner, December 18, 2005.

Full interview: here

[Don the libertarian Democrat]

Westlope,

Thank you for the Friedman/Kuttner dialogue. I'm a big fan of both of them, but more Friedman theoretically and Kuttner politically. However, this quote is my position, which I've been arguing for years, and conservatives hate it:

RK: I couldn't agree with you more. We have the worst mix of government and private, I could not agree with you more.

MF: We ought to have much more private or much more government.

So thanks for the link. I also believe in a guaranteed income.

[Too Much Fed]

It does matter what is bought. It does matter whether currency is used. It does matter whether debt is used. It depends on the situation.

I also believe that wealth/income inequality matter. With a quick glance I do NOT see that mentioned here.

[Nick Rowe]

JKH: I disagree. I think it is both a liquidity crisis as well (or more than) a solvency crisis. The US was where solvency was (at least at first) most damaged. And yet the US$ rose against other currencies. Because the US$ is the most liquid of all monies (in an international context). The widening spreads between off-the-run and on-the-run bonds suggest an increase in demand for liquidity. And haven't trading volumes for many assets, from houses to ABCP declined drastically, which makes them much less liquid than before, and explains at least part of their declines in price, over and above risk, or changes in perceived fundamentals?

[JKH]

Nick,

Interesting. Here's a real disagreement, or at least a distinction:

Liquidity more than solvency, versus solvency more than liquidity.

I'll be back.

[bob]

"And haven't trading volumes for many assets, from houses to ABCP declined drastically, which makes them much less liquid than before, and explains at least part of their declines in price, over and above risk, or changes in perceived fundamentals?"

If we're talking about the bad assets like sub-prime MBS, lack of liquidity is actually keeping their value up at the moment, and perceived risk is driving it down. There are many buyers who wish to buy MBS, at a market price, but the big holders are refusing to sell a) because if the securities were cleared at a market price, the institution selling them would be revealed to be insolvent and b) these institutions think that they can get a better above-market price from the government. This is what Krugman has been saying about the bad asset buying: it only helps if the government drastically overpays, because the default rates (risk) dictate that the value is far lower than what the banks have them valued at on their balance sheet, and if they are valued based on standard models, the rate of default alone implies a price that renders the institutions insolvent. The banks can't acknowledge fundamentals, because they are fundamentally insolvent. Letting them value this stuff highly on their books lets them stay in existence, but the drawback is that it destroys liquidity in those markets, because they can't sell.

Take aMBS with a face value of $100, and if it declines below $50 the institution holding it becomes insolvent:

On the books it is written down to $80
The government is willing to pay $60
The default rates imply that it is only worth $30
In a liquidity constrained and risk averse market, buyers are only willing to pay $20 (the effect that you mentioned originally)

So I would say that the lack of liquidity is actually keeping the prices up in many of these markets, because it allows the banks to price their assets at face value, which is far far above fundamental value. Lack of liquidity is depressing the Bid by $10, but increasing the Ask by $50.

[Nick Rowe]

bob: OK. The bid-ask spreads are increasing. Let me restate what I perhaps should have said: the lack of liquidity is reducing demand for those assets. And perhaps there's a nasty negative feedback loop, in that lack of liquidity discourages both selling and buying, which worsens liquidity further.

[Nick Rowe]

BTW, I disagree with Paul Krugman (and others) on the idea that buying bad assets only helps if the central bank overpays. The key is, that when you are a central bank, the fundamental value of an asset is not exogenous wrt what you pay for it. By buying a (large) quantity of assets, the central bank may increase liquidity, help recovery (and prevent deflation) which makes those assets worth more (both in nominal and real terms, given we are off the LRAS).

[bob]

1. yes, poor liquidity definitely discourages both. the point I was trying to make was that the sellers are far more liquidity constrained (due to solvency issues that are related to liquidity: if they could borrow enough money for to cover the losses from liquidating an MBS portfolio and have good enough business prospects to pay it off over the subsequent years, they no longer have to worry about solvency) than the buyers. Lack of transactions is driving down the bid below fundamentals, but is allowing a much greater over-valuation on the sell side. The fundamental liquidity-neutral price is much closer to the bid than the ask. So I'm saying that lack of liquidity increases the spread, but also skews the net price upwards because the sellers are more liquidity & solvency constrained than the buyers, and for them every sale involves acknowledging a loss.

2. yes, I actually do agree with this point, although I think that the increase in the value of these assets probably won't rise above the price that the banks need to get for them to stay in business, although it is definitely debatable. In my above example, a successful government purchase in conjunction with other stimulus might raise the fundamental value from $30 to $40, but the government would still have to pay at least $50 to keep the bank afloat. A lot of this stuff really is pure garbage, that nothing will help with, so I'm skeptical of the potential for long-term gain. There's only so much the government can do to improve the viability of a whole bunch or sub-prime and even criminally fraudulent loans. Mish does some good breakdowns of Alt-A MBS pools if you want to see how truly awful this stuff is.

IMO estimates that show both the banks staying solvent and the government making money from the MBS purchases drastically over-estimate the fundamental quality of the loans. No matter how liquid the market, and healthy the housing sector is, many of these loans are bad and can never be made good again. the problem is with the terms and lending standards of the original mortgages. That's why I don't mind the BoC getting into decent Canadian mortgages, but I think that the US government shouldn't buy the bottom of the barrel non-conforming garbage. If the loans were any good, they would have already been sold to Fannie Mae. I forget who said it, but I found this quote hilarious and true: "In Tim Geithner's world there are no bad assets, only misunderstood ones"

[Nick Rowe]

Good comment bob.

[reason]

Nick,
maybe you need to define solvency for us to decide this. The trouble, as I see it, is that insolvency is defined in nominal terms so expected inflation could potentially solve it, so a shortage of liquidity makes insolvency more obvious. In a Ponzie scheme, you are (publicly) solvent as long as you are liquid. But the amount of inflation needed keeps rising.

So I think you are technically right, but really wrong. The point at which the crash happened was in a sense arbitrary, but the fundamental insolvency (in real terms - that the real NPV of the debt exceeded the real NPV or the assets) was still the fundamental source of the problem.

[Nick Rowe]

reason: someone on the econ blogosphere, I can't remember who, about a couple of weeks ago, gave about 4 different definitions of bank solvency/liquidity. In part, they matched my post several months ago on balance sheet vs income statement solvency.

There's: if you sold everything right now, at firesale prices, would you be able to cover your liabilities? If not, you are balance sheet illiquid. If you can't cover your flow of obligations right now, you are income statement illiquid. If you can eventually cover your debts, as a going concern, when they come due, you are solvent. Something like that.

A Ponzi scheme is insolvent, but liquid on the income statement (as long as the money keeps flowing in).

Inflation only matters insofar as some assets and liabilities are fixed in nominal terms; and inflation may help general economic recovery.

[westslope]

I'm not sure that 'liquidity' for private agents and 'liquidity' for a central bank mean the same thing. In ordinary markets, is liquidity a problem in itself or simply the outcome of mismatched expectations?

We can try to explain markets that go from being 'liquid' to 'illiquid' through a disequilibrium model (that may describe more than explain as such) and behavioural models that emphasize sunk cost valuations or endowment effects. I suppose one could also argue that markets take a long pause as agents thoroughly invest in research and information gathering. The expectation of deflation by itself explains nothing because it only focuses on one side of the market.

As Nick points out, asset valuations will clearly be affected by the success or failure of policy, or perhaps unrelated events that overcome market coordination failure. As many economists and other pundits have pointed out since this sudden contractionary phase started, hopefully the infinitely wealthy insurers of last resort--government and the central bank--should make significant returns on purchased distressed assets over the longer term.

Buiter's distinction between qualitative and quantitative easing may be useful but to be complete I would like to examine empirical evidence where a central bank literally kept the size of the balance sheet constant while changing the riskiness composition of the balance sheet. I would guess--but do not know--that there is a significant assymmetry at work here. Qualitative easing never occurs in the absence of quantitative easing. Qualitative firming may occur when the size (or growth rate) of the balance sheet remains steady.

[reason]

Nick,
the Principal is ALWAYS defined in nominal terms! You have never addressed the issue of possible cash settlement and principal. Just extrapolating nominal interest rates to infinity gives a misleading view of debts. Higher rates of inflation are also less certain rates of inflation so it makes sense for the risk averse to pay off their debts in cash (also borrowing rates are higher than deposit rates).

[reason]

Nick to make it clearer, consider the problem with deflation. The principal can never be paid off, and nominal interest rates don't go negative.

[reason]

And add the tax advantages of paying off debts as against receiving interest from deposits or other investments.

[JKH]

Nick,

Here are two people who agree with you that it’s a liquidity crisis more than a solvency crisis – John Hempton and Warren Buffett:

http://brontecapital.blogspot.com/2009/03/fools-seldom-differ.html

And here is the post that you referred to above, where the same John Hempton delineated various definitions of solvency:

http://brontecapital.blogspot.com/2009/02/bank-solvency-and-geithner-plan.html

Now I can cover myself and say that it’s a crisis in the definition of solvency.

Finally, stay tuned for the possibility of critical adjustments to mark-to-market accounting in the roll out of the full Treasury plan over the next few weeks. MTM treatment also relates to the appropriate definition of solvency.

[Nick Rowe]

Good find JKH! Yes, that's the post about definitions of solvency I had in mind, but couldn't remember where I had seen it.

Apart from definitional issues, when solvency of a firm, financial institution, or household, becomes questionable, this makes its liabilities less liquid. ("Am I buying a lemon? Do I need to do costly research into the value of this asset? If I need to sell it again quickly, will the person I sell it to also need to do costly research?")

If it were only a solvency crisis, and if insolvency did not cause illiquidity, what's the big deal? Some stuff isn't worth as much as we thought it was, but it's hard to see how that alone would spillover into the markets for all assets and goods. But a liquidity crisis (whether or not caused by a solvency crisis) would spillover into an excess demand for money, and because money is the medium of exchange, an excess supply of nearly everything.