We won’t know the final lessons of QE until it’s over

Published in Real Clear Markets.

A justly famous line of John Maynard Keynes is:  “Soon or late, it is ideas…which are dangerous for good or evil.”

The first lesson from ten years of Quantitative Easing (QE) is that the ideas of those who run fiat currency central banks, as these ideas change over time and go in and out of central bank fashion, are extremely important for good or evil, on a very large scale.

Contrasting the ideas of QE to earlier governing ideas of the Federal Reserve is instructive.  In the 1950s under Chairman William McChesney Martin, the Fed adopted the “bills only” policy.  That meant the only investment assets from the Fed’s open-market operations were short-term Treasury bills.  The theory was that the Fed’s open market interventions should not operate directly on long-term interest rates and should never try to allocate credit among economic sectors.

This was clearly the opposite of the theory of QE.  It was not a policy directed at financial crisis, as QE originally was.  But the last crisis has now been over for a long time, and QE still amounts to $4.2 trillion on the balance sheet of the Fed.

How many Treasury bills does the Fed own today?  The answer is zero.

So over 50 years, the Fed has gone from believing in all Treasury bills to no Treasury bills.

Another lesson of QE:  Do not look for the ideas of central bankers to be eternal verities—they aren’t.  The times call forth the ideas, for better or for worse.

The Credit Crunch of 1966 was created by the Fed’s regulatory ceilings for interest rates—the Fed used to believe in those, too.  In that year, mortgage lending funds dried up, the savings and loan industry (then politically powerful, believe it or not) was unhappy, and many in Congress wanted the Fed to buy the bonds of Fannie Mae and the Federal Home Loan Banks in order to support housing and housing finance.

Chairman Martin did not agree.  Correctly pointing out that this would be credit allocation by the Fed, he found it a bad idea “to divert open market operations from general economic objectives to the support of specific markets for credit.”  This would “violate a fundamental principle of sound monetary policy, in that it would attempt to use the credit creating powers of the central bank to subsidize programs benefitting special sectors of the economy.”

In my judgment, Martin was right about this, but Congress loves nothing better than to subsidize and overleverage real estate.  Here was the consequent threatening message to the Fed sent by a future Banking Committee chairman, Senator Proxmire, in a 1968 hearing:

“You recognize, I take it, that the Federal Reserve Board is a creature of Congress?

The Congress can create it, abolish it, and so forth?

What would Congress have to do to indicate that it wishes the Board to change its policy and give      greater support to the housing market?”

If Proxmire were still alive, he would presumably be a fan of QE forever.

The new Fed Chairman, Arthur Burns, who arrived in 1970, decided that the Fed should “demonstrate a more cooperative attitude.”  So by the 1980s, the Fed’s bond portfolio came to include the debt of Fannie Mae, the Federal Home Loan Banks, the Farm Credit Banks, the Federal Land Banks, the Federal Intermediate Credit Banks, the Banks for Cooperatives, the United States Postal Service, Ginnie Mae, the General Services Administration, the Farmers Home Administration, the Export-Import Bank, and even the Washington Metropolitan Area Transit Authority.  In other words, the Fed was helping fund the Washington DC Metro system!  That was along with funding Fannie Mae when it was insolvent on a mark-to-market basis, as well as the Farm Credit System when it was broke.

Still, at their peak, all these totaled about $9 billion—or about 0.2% of the current size of QE.

Let’s review where the Fed’s QE-dominated balance sheet is now.  As of May 30, 2018, it includes:

Treasury bills:                                                     zero

Longer term Treasury securities:               $2.4 trillion

Long-term mortgage-backed securities:   $1.8 trillion.

These MBS are funded with floating-rate deposits, which makes the Fed in effect the biggest savings and loan in the world.  Even if we needed the world’s biggest S&L in the crisis—do we now?

Total assets:                                                   $4.3 trillion

Total capital:                                                   $39 billion

This means the Fed is leveraged 110 to 1.

As we know, the immense QE portfolios are very slowly running off—not being sold.   Among the reasons for not selling is that the Fed does not want to face the very large losses in its super-leveraged balance sheet that it would probably realize when selling any meaningful part of its huge, unhedged QE position.

Another lesson: It is easier for a central bank to get into a QE portfolio than to get out.

Thus, the exit strategy is constrained to being very gradual, accompanied by the Fed’s intense hope that the ultimate adjustment in inflated house prices, and inflated stock and bond prices, will also be gradual.  This is especially true for house prices, which affect 64% of American households.

Here is a further QE lesson, this one fundamental:  In principle, a fiat currency central bank can make unlimited investments in anything, financed by monetization.

The Federal Reserve is the champion investor in mortgages.  The European Central Bank has invested in corporate bonds, government agencies, regional and local government debt, asset-backed securities, and covered bonds (which include mortgages). The Bank of Japan has bought asset-backed securities and equities, in addition to vast amounts of government debt.  The Swiss central bank has a huge portfolio of foreign currency bonds, a big position in U.S. equities, and also makes loans to domestic mortgage companies.

The Swiss central bank, by the way, is required by law to mark its investment securities to market, a discipline the Fed sedulously avoids.

All of these versions of QE involve credit allocation.  In the case of the Fed, its two favored allocations are housing and long-term financing of the government deficit.

The only limits to what a fiat currency central bank can finance and subsidize are: the law, politics, and the ideas of central bankers.  There are no intrinsic financial constraints.

Whether there will be new legal and political constraints in the future depends, I believe, on how the end of the QE experiments ultimately turn out.  In other words, will the correction of the QE-induced asset price inflations be a soft landing or a hard landing?  If the latter, you can easily imagine a legislature wanting to enact future constraints.

Ten years into QE, what should the Fed be doing now?  The distinguished expert on central banking, Charles Goodhart, recently wrote that it is “generally agreed” that “Where the QE involved directional elements, to support credit flows through critical but weak markets, e.g. the mortgage market in the USA, such assets should be entirely run off, and the assets left in the central bank’s balance sheet should be entirely in the form of government debt.”

With all due respect to Senator Proxmire, this seems correct to me.

But, as Goodhart continues, it does not answer a further question QE makes us ask: What is the optimal size of central bank balance sheets in normal times?  They have become so large—how much smaller should they get?

Other related questions include:  Will the central banks’ credit allocations become viewed in retrospect as misallocations?  And how should we understand the respective roles of the Treasury and the central bank?   Are they essentially one thing masquerading as two, as QE tends to suggest?

I conclude with a final lesson:  We won’t know what the final lessons are until after the exit from QE has been completed.  It ain’t over till it’s over.