Tuesday, March 18, 2014

The Downside of Monetary Policy

Many times in recent years I've seen liberal commentators (<cough>Kevin Drum</cough>) say things like, "we need looser monetary policy ... it can't hurt". Unfortunately, that's not true.

Monetary policy is really a very blunt tool.  The central bank can raise or lower interest rates, but the effects are unpredictable.  If we rely on monetary policy to repeatedly boost the economy over a secular period where the economy needs repeated boosts due to globalization, we're likely to see interest rates fall to 0 and stay there, while overinvestment in interest rate sensitive sectors such as housing results in repeated booms and busts in these sectors.  Indeed, that is exactly what we have witnessed over the last 30-some years.

I ran across an article with data and commentary along this line discussing the housing market in the U.S. over the last several years:  http://www.alhambrapartners.com/2014/03/18/departing-science/.  Excerpt:
New York Fed president William Dudley said in an interview with the Wall Street Journal this month that “persistent headwinds” to growth explained why the economy would be unable to bear much higher interest rates in the years ahead...  Headwinds are nothing more than the economy doing something other than modeled...   
Monetary policy’s most fervent channel lies through mortgage finance... Unfortunately, mortgage issuance is off nearly 60% in less than a year since the word taper entered the mainstream... there is something very wrong when a relatively small increase in mortgage rates leads to such a dramatic decline
Real estate construction has a macro component to it that has already seen some reversal in the GDP figures (a tailwind turning headwind, as it were). It looks like that is deepening still further, but more importantly, there is the looming possibility of a second, albeit smaller, housing bust forming in such close proximity to the first.
Headwinds are not some exogenous factor, they are monetarism put into practice.
Here's another post from Alhambra Partners demonstrating the recent ineffectiveness of monetary policy... 

Sunday, March 16, 2014

Money Supply -- Bank Lending and Government Spending

The post below is inspired, and to some extent taken literally from my comments at http://www.winterspeak.com/2014/03/bank-of-england-goes-mmt.html.

Here's a simpler proof that government spending accounts for more money creation than does private credit creation:

1. All government spending transfers money to private sector (by definition).

2. Private credit creation may or may not create new money in any particular time period (as repayments are greater than new loans in some time periods).

3. Empirical data exists for 1 and 2 above and shows that government spending generally results in almost twice as much money creation in a typical year compared to net private credit creation.

4. Government bond issuance withdraws money from the economy in exchange for the interest bearing bonds.There is no evidence that this withdraws any purchasing power, since the T-bonds are guaranteed by U.S. government and are the most liquid investments in the world. The inconceivability of default is one of the few things that Dems and Reps agree upon. Even if bonds were worthless (a ridiculous assumption) and subtracted from the impact of government spending, the amount of money created by government spending would be larger than the amount created by net private credit.

5. Taxes come out of the total money supply without regard to how the money was created.

What are the implications?

1. People like Cullen Roche who say that private credit creation accounts for 90% on total money creation are wrong.

2. Conclusions based upon #1 above are without foundation.

3. Fiscal policy is the main factor in money creation, outweighing all private banking activity.

4. Adding in the fact that interest rates are only one of many factor in private credit creation, monetary policy has a trivial impact on the macro-economy.


Here's another way of looking at it...

Those who compare government spending and private bank money creation imply that we can keep track of money inflows and outflows from one time period to another, and in this manner determine the relative impact of various sources of money in the economy.

This seems reasonable enough to me, so let's look at this sort of model.    In the beginning (first time period), money is injected into the economy via government spending and private loans.  This gives us our initial partition of the total money supply into government generated (say G$) and bank generated (say B$).  

In each successive time period, money is injected into G$ via government spending, and removed from G$ via taxes.  Note that all government spending goes into G$, but not all taxes are collected from G$ as the money in B$ is also subject to taxes.  

In each time period, money is injected into B$ via private bank lending, and removed from B$ via repayment of private loans and via payment of taxes.

In each period, the size of B$ changes by bank lending - repayment of loans - tax payments.  The size of G$ changes by government spending - tax payments.  Assuming tax payments are the same proportion in the two sectors, the relevant measures are net bank lending versus total government spending.


The common mistake is to compare the effect of private sector lending before taxes with the effect of government spending after (incorrectly computed) taxes.

Take the following example of a new monetary system.  In the first period there is no government deficit and net $1 T private credit creation.  The government spends $1 T collects $1 T in taxes using a 50% marginal tax rate.  So at the end of the first period private bank lending has resulted in $500 billion new dollars after taxes.  Government spending has also generated $500 B after taxes.  


Here's another way of looking at it:

When discussing the impact of various sectors, you may think at a superficial level that one sector's impact is zero because it has various components that cancel each other out.  However, that is shown to be incorrect when you look in more detail and see that the apparent cancelling out is caused by reducing the impact of another sector, and not just cancelling out the same sector.