Implicit debate about Keynesian stimulus

[jon_leonard]

I've been musing about an opinion piece in Monday's Wall Street Journal, entitled "Cash From Clunkers". It's fairly short, pointing out that a lot of the "cash for clunkers" program consists of dubious wealth transfers, from taxpayers to people who might buy cars

Comments

[jon_leonard]

I suppose it depends on what exactly you mean by "Keynesian trap".

As I understand it (and I really need to finish reading his General Theory), it involves a situation where the total propensity to consume is less than the amount produced, to the point where people who want to work can't find work. In the liquidity trap version of this, the interest rates are also low enough that a central bank can't stimulate the economy by lowering them any further.

But the weirdness in the current situation is that interest rates have very high spreads, not that they're too low in general. They're certainly absurdly low for Treasuries and the Fed rate, but interbank loans were higher, and formerly respected companies have been paying extremely high rates. Much depends on what credit risk the borrower presents, and this current crisis is peculiar in that all the of the banks presented a large credit risk -- and there is no other lending system.

Essentially, I don't think a crisis should count as a Keynesian trap unless his model's prescription would work to escape it, and I'm far from convinced that simple stimulus spending is the answer. (I'm also not really convinced that it was the answer in the 1930s.)

I agree that the crisis has deep roots in the loan portfolios, to go with any number of other roots.

[jon_leonard]

I haven't read the original, but I have the general description of Keynes' theory from my economic classes, and the economic blogs I read (Krugman, etc.). The interest rate spread may be the result of the unusual actions the Fed is taking in the markets, because it can't lower the overall interest rate any more. Also, since the banks themselves pose a large credit risk -- where does the money go? Commodities, corporate and Treasury bonds...

Better regulation may also be part of the answer, as was in the 1930s, to improve investor confidence. But what do you think the answer is other than stimulus spending?

[jon_leonard]

I'd describe the problem at its core as a misallocation of resources: Lots of effort was spent on activities that were unuseful and/or unsustainable.

So the first correction is to stop or slow down those activities, which inescapably shows up as a fall in GDP. The trick is to avoid panic (or the regulators looking panicked), so as to limit the spillover effect.

If the forward looking business climate were good, then I could see stimulus, but at the moment trying to plan ahead runs into the problem that a fairly small number of powerful people are making sort of random decisions. If, for example, I need to borrow money to start up a business, I'll need to know what interest rates will do, which translates into forecasting when the Federal Reserve will shift from stimulus mode to crush-inflation mode.

So I'd really like a medium to long term plan that doesn't inflate the currency, doesn't inflate the debt, and includes regulation that demonstrates that they understand where the problem came from, and will tend to limit a recurrence.

As a practical matter of regulation, I'd:

Adjust "conforming" home loans so that they require at least 20% equity, and fixed interest rates. If ordinary borrowers really want floating interest rates (systemic risk!), they can do interest rate swaps.

Adjust the tax code to stop favoring overleverage. Debt is favored over equity in corporate structure because money paid to cover debt is pre-tax, but money paid to equity (dividends) is pre-tax. Similarly for home ownership: Deductible mortgage interest doesn't promote home ownership, it promotes renting from the bank.

Some GAAP adjustments, to discourage the sort of off-the-books nonsense that made a lot of the big losses look like surprises. (The possibility of a big loss should always be listed in company reports.)

A shift away from Value At Risk measurement: It assumes that the underlying distributions are gaussian, and in times of stress, they aren't. It'd be better to look at what would happen if all the relevant variables shifted 2 or 3 sigma unfavorably, at the same time.

A move to exchange-traded derivatives for more (all?) of the new financial instruments. The structure for things like exchange-traded options dramatically reduce counterparty risk, and as a matching benefit, makes it clear how large the aggregate positions are. That way systemic risk is much easier to judge.

But no, I don't see any magic way out of the current recession. I just worry that piling more money into various parts of the system will delay necessary adjustments, while doing little or nothing as Keynesian stimulus.

It's probably too late to do TARP right (but it might not be): It should really have been structured as a one-time offer to buy various assets at fire-sale prices, with a promise of strict regulatory investigation for any bank that didn't take them up on the offer. That way all the banks are left with known balance sheets (the uncertainty problem diminishes), and the cost to the taxpayer is more controllable.