Thursday, June 7, 2018

Housing: Part 303 - The Fed Balance Sheet

Scott Sumner shared my recent Mercatus papers over at EconLog.  Scott generously supports much of my push-back against the bubble story.  Economist Bob Murphy saw the post and reacted with some chagrin that Scott or I could question the idea that the pre-crisis housing market should be broadly characterized as a bubble, or that the crisis could be blamed on tight monetary policy.

In support of his chagrin, he uses the measure of the Federal Reserve's balance sheet (the blue line in this graph).  The Fed balance sheet shot up in late September and October of 2008, and then continued to grow with the QEs.

I have probably gone over all of this before, in some form, but this seemed like a good excuse to look at it again.  First, it is worth taking a closer look at this graph.  Here is a close-up look at the Fed balance sheet in 2008 along with the Fed Funds Rate.

By mid-November, the Fed had added more than $1 trillion to its balance sheet.  Half of that rise came in September and October when the Fed Funds Rate was sitting at 2%.  Even Ben Bernanke admits that holding the rate at 2%was a mistake.  The initial burst in the Fed balance sheet was due to interest on reserves, which the Fed began to pay because the 2% target rate was so far above neutral at that point that they would have had to sell every single Treasury they had in order to hit their target.  As crazy as this sounds, this was the actual reason for the policy (see here and Bernanke's memoir).

Most of the rest of the increase came when the target rate was still at 1.5%, and the target rate was still at 1% when the balance sheet topped out.  Until QE3, most of the increase in the Fed balance sheet dates to this period that was before QE and was even before the Fed Funds rate hit the zero lower bound.  Heck, half of it happened while the Fed was maddeningly trying to keep the Fed Funds rate at 2%.  They couldn't keep the rate at 2% because they had induced a financial panic, so interest on reserves was intended to force a rate floor by sucking funds out of the banks.  So the initial increase in the Fed balance sheet has nothing to do with loose monetary policy.  It seems as though many people who use the monetary base as a measure of monetary policy haven't accounted for this at all.

QE1 really didn't add much to the balance sheet.  It was mostly swapping Treasuries and MBSs for emergency loans to banks.  The short hand that I use for what happened is that the Fed had policy so tight it was running out of ways to suck cash out of the economy, so instead it started sucking credit out of the economy by offering to borrow cash from the banks and then stick it in a vault so it couldn't be used (excess reserves).  And, looking back at the first graph, we can see that happening.  From September 2008 to the end of QE1, bank lending dropped way off while deposits continued to grow at somewhat normal rates.  So, the gap in bank lending roughly equals the amount of excess reserves.  The banks "loaned" cash to the Fed so that it could hoard the cash and do nothing with it instead of loaning it into the private economy.

Clearly the growth of the Fed balance sheet ceased to be a good measure of monetary accommodation at this point.

Considering conditions today, it seems as though what we should see happen as the Fed reduces its asset base is that bank lending should re-converge with the level of deposits.  There is a bit of a delicate balance here for the Fed, because if they reduce the balance sheet size too fast with monetary policy that is otherwise too tight, the net effect will be for that convergence to happen through declining deposits.  If they reduce the size of the balance sheet while being too accommodative (for instance, this might policy happen if they stopped paying interest on reserves and returned the Fed Funds Rate back to zero, though even then I'm not sure the net effect would be accommodative if it coincided with a reduction in the balance sheet) then possibly banks would lend at such a pace that deposits would start to grow more quickly.

Unfortunately, it appears that they may be erring on the side of reducing the balance sheet while maintaining policy that is too tight.  Here are the 1 year rates of change in deposits and bank lending.  Both are growing at less than 5% and are decelerating.  Lending had looked like it was stabilizing, but over the past 4 weeks, levels of Commercial and Industrial Loans and Closed End Residential Real Estate loans have both declined.

But, the more important story here is that clearly the monetary base is not a good measure of monetary policy under the current regime, and certainly it wasn't in 2008.


  1. - But Scott Sumner doesn't believe that bubbles can exist.
    - I have tried to convince him that Australia has a housing bubble right now. And that bubble is in the 1st stages of deflating.

  2. - Here's is another flaw in your thinking. Yes, the FED performed QE but that doens't help to aleviate the problem. Because QE increases the reserves of a bank and those reserves are NEVER lent out.
    - In other words: Even Bernanke doesn't know how the banks operate.


    worth a grim chuckle….