Fed to Shrink Bloated Balance Sheet Gradually

The massive size of the Fed’s balance sheet naturally raises the question:  what is the optimal or target size of the Fed’s balance sheet in the long run?  All the asset purchases and QE since 2010, after the economy was out of the recession and growing again, do not appear to have helped get U.S. real GDP growing any faster than 2% and has not helped to raise inflation above 2%.  Our conclusion is that QE basically failed in meeting its growth and inflation economic objectives, even if QE did depress Treasury yields, support stock price increases, and reduce equity and bond market volatility.  So, it seems reasonable that the Fed’s balance sheet at its current size is unnecessarily large and represents considerably more intrusion into the market’s price discovery process than makes sense.

We do not know what the optimal size of a central bank balance sheet is, but we do estimate that the Fed will be aiming to split the difference.   Our estimate is that the Fed will seek a target balance sheet size of about 12% of GDP – twice what is was before the financial panic and half what it is today.

There are other reasons to shrink the balance sheet that the Fed may prefer not to discuss.  From our perspective, not shrinking the balance sheet by always re-investing 100% of every dollar received from maturing Treasury securities or principal payments from mortgage-backed securities (MBS) make little sense when the Fed is also raising interest rates.  Because of the bloated balance sheet, banks hold $2.2 trillion of excess reserves on deposit at the Fed.  These excess reserves were created as the Fed bought securities through QE programs and paid for them by crediting the accounts of the sellers, namely banks and primary dealers.  These deposits are federal funds.   It is useful to note that while banks can trade their federal funds among themselves (i.e., at the federal funds rate), banks cannot destroy these deposits.  Only the Fed can do that by not re-investing principal received or by outright sales of securities, and the Fed is not going to sell anything.

And, here comes the rub – with rising interest rates, to enforce its federal funds rate target range, the Fed pays interest on required and excess reserves at the ceiling of the target range.  So, if the Fed is targeting 0% to 0.25% as it was from late 2008 until December 2015, the Fed would pay 25 basis points or 0.25% annually on excess reserves.  Every time the Fed has raised its target funds rate range, it has raised the interest it pays on reserves by the same 0.25%.

Now, let’s do the arithmetic.  Every time the Fed raises rates, it increases its own interest expense by about $5.5 billion.  This reduces the Fed’s earnings on it portfolio of MBS and Treasury securities.  Note that the Fed contributes its earnings to the U.S. Treasury to help reduce the budget deficit, and over the last few years this has been about $90 billion a year.  This means, as the Fed raises rates, its earnings will decline.  Shrinking the balance sheet will help get the Fed back to its appropriate role in the economy –  it will pay less in interest expense which some view as a subsidy to banks, it will have a smaller balance sheet, and will have a lower but more stable contribution of its earnings to provide to the U.S. Treasury.

We estimate that over the next 12 months about $300 billion of Treasuries will mature and the Fed will also receive about $400 billion of MBS principal payments.  That is $700 billion of MBS and Treasury security purchases to keep the balance sheet from naturally shrinking.  Of course, the Fed will not go cold turkey and simply stop all reinvestment activity.  The Fed does not want to disturb markets, especially the U.S. home mortgage market.  Consequently, the Fed will take this whole process in stages and go very very slowly.   The Fed will put a cap on reinvestment activity, delay the start, and then gradually over the next 4-5 years, it will adjust the cap until the balance sheet stabilizes around 12% of GDP, in 2022 or so.

This means, interestingly, that Fed Watchers will now have two decision items to watch for at each FOMC meeting:  1) Will the Fed raise or lower rates?  2) Will the Fed adjust the cap and reduce its reinvestment activity?  These decisions are likely to be staggered and may (or may not) occur together.

Our conclusions are, first, that there will be no impact on short-term rates.  The Fed anchors the short-end of the yield curve with its federal funds rate target and the payment of interest on federal funds – bank reserves held at the Fed.

Second, we think the impact on the long-end of the Treasury yield curve will be minor to imperceptible.  Other factors, such as a rising budget deficit, rising debt issuance, inflation patterns, ups and downs of GDP growth will swamp any affect from balance sheet shrinkage.

Third, we do expect an impact of an upward move of 0.25% to 0.5% in the 15-year and 30-year home mortgage market as MBS reinvestment slows.  Or more precisely, we expect the spread between mortgage rates and the Treasury yield curve to widen by about 0.25% to 0.5%.

Why the 15 and 30 year?   These are the two MBS maturities the Fed buys.   The Fed is a large and steady buyer of these mortgages under the current 100% reinvestment approach.  As the Fed withdraws its support of this market, some impact may be felt.

All in all, it seems well be past time to start shrinking the balance sheet.  Now, with this decision to shrink the balance sheet reinforces the desire of the Fed to move the federal funds rate target higher to be more or less in line with long-term inflation.