The Case For Slowing Reinvestment Before Fed Hikes Rates by David Schawel CFA, Economic Musings

With Friday’s unemployment report showing a clear acceleration in both job gains and wage growth, the conversation turns to how and when the Fed should tighten policy. Since the onset of the financial crisis the Fed has loosened policy through both a zero interest rate policy (ZIRP) as well as large scale asset purchases (QE). The optimal way of tightening is commonly thought to be first a rate hike, and then a reduction of the Fed’s balance sheet, but I am going to argue that tightening in the reverse order might be worth considering. Here’s why:

  1. The liability repricing cycle has mostly exhausted itself:

A major benefit of lower long term interest rates and tighter spreads is that a large liability repricing cycle has occurred for both households (via mortgage refinancings) and corporations (refinancing corporate debt & issuing new debt). Between late 2011 and summer of 2013 a giant refinancing cycle occurred where 30yr fixed mortgage rates stayed under 4% and hit as low as 3.375%. In fact, as of 12/31/14 ,the US effective rate of interest on mortgage debt outstanding stood at 3.83%, down from 5.25% at the end of 2008. With mortgage debt outstanding at ~13.3Trillion, the annual savings works out to ~189billion. The same goes for corporations which have refinanced hundreds of billions of corporate debt into lower rates lowering than interest expense and freeing cash flow.

  1. Reversal of portfolio balance channel

One can debate the impact that the portfolio balance channel has had, but there’s no denying that trillions of dollars worth of Agency MBS & USTs have been removed from the system forcing bond investors into other asset classes.

The rationale for buying these bonds at the time was important & serves a great purpose. In 2009-2010 the market for lower rated credit securities was frozen with loss adjusted yields for non agency MBS, CMBS, and HY comfortably in the double digit range.

The removal of so many safe bonds drove spreads lower and forced investors to take credit risk again. This was good, it was healthy, and it worked. But today is different, and it’s not necessary to maintain a multi-trillion dollar balance sheet. The credit markets are roaring and fully thawed & if anything some moderation might be necessary.

  1. Scarcity effect & the Impact of draining reserves via runoff

In my opinion, draining reserves through the slowing of reinvestments (tens of billions of $’s of mortgages prepay each month which the Fed reinvests into new bonds to keep the balance sheet flat) would result in a few things in my opinion. The first is higher long term interest rates and credit volatility. I’ve discussed how & why that QE is the “greatest volatility sale ever” here on Ello https://ello.co/schawel

The second impact would be through the steepening of the long end of the curve. Currently the Fed holds the vast majority of the long end duration. Slowing reinvestment would drain reserves and reduce the scarcity effect that is currently keeping longer UST yields lower than they would otherwise be.

4. Draining reserves is necessary to eventually have a “normal” fed funds market

The lack of trading in fed funds is well documented and is easy to understand given the flood of liquidity in the banking sector from trillions of excess reserves. A rate hike without draining reserves requires the use of facilities such as RRP, TDF, and others to temporarily drain reserves. The permanent draining of reserves through letting the bonds runoff would start the process towards a system where a normal supply / demand market could exist for fed funds. Maybe these facilities they are proposing will work seamlessly & the quantity of reserves in the system will not matter, but I am skeptical of that.

  1. Slowdown reinvestment & the hike

In my mind the majority of risks that should be concerning are a bigger result of the size of the Fed’s balance sheet versus ZIRP. In terms of bank lending, the quantify of reserves in the system is so large that banks would likely keep loan yields the same if the fed were to hike (thus lower loan spreads). This is a result of QE which has added trillions of dollars of deposits to the banking system. If those deposits were gone the fight for loans would be different & maybe 10y 3% fixed rate CRE loans wouldn’t exist. So an overnight rate hike wouldn’t slow/tighten the channel of lending as much as a gradual balance sheet reduction. Overall, the things that need to be cooled down are impacted more from the size of the balance sheet than the overnight rate.

  1. There’s no established playbook for the exit

Although I feel confident this is the optimal way to proceed, we all need to realize this is a giant experiment and we don’t know the implications of an exit. The unknown doesn’t mean calamity but we should realize that this is unchartered territory.

The Case For Slowing Reinvestment Before Fed Hikes Rates