The Fed Would Be Wise To Exit Sooner Rather Than Later

October was dramatic indeed, the list of events includes: the government shutdown, and its crescendo, rollout of the Obamacare website, the official nomination of Janet Yellen, continuing resolution, publication of delayed data releases by BLS and other furloughed government agencies, and finally the Fed statement. The price action was generally heads I win, tails you lose. In particular, given occasional signs of strength the response was "risk on," as markets welcomed growth. More frequently, however, we were confronted by signs of weakness, which the market also welcomed,in this case as harbingers of the Fed deferring the taper decision and keeping rates lower for longer. Meanwhile, broader perceptions of an easier Fed drove the dollar to weaken.

Though the immediate prognostications on Fed Taper timing are important, I want to turn to broader topic of the increasingly fragile nature of our monetary framework. Eugene Fama, who was awarded the Nobel Prize this month for his contribution to the efficient markets theory, opined while on CNBC that monetary policy, or the Taper, "is basically a neutral event." Though he was broadly criticized as being out of touch with reality, the reality is that episodes of QE in the U.S. have been more correlated with yield rises than any casual observer could imagine. Meanwhile, another academic, Ronald McKinnon, a professor at Stanford University, opined that the Fed could better begin to raise rates before tapering, as that would be easier to manage, given that the Fed need not mark its balance sheet to market. Both dovetail into my views on scale and scope of the Fed Balance Sheet.

The Fed is currently buying $85 billion a month of long term Treasuries and Mortgages, which is a rate of $1.020 trillion per year. Meanwhile the fiscal year 2013 deficit, which needed to be financed through Treasury net issuance, was $680 billion and falling. It follows that the Fed purchases amount to 150% of Treasury net issuance and that percentage is growing!

To understand what this means, I start by looking at the Fed as if it were an independent institution. In particular, it is a central bank, with stockholders (money center banks), liabilities (mostly currency in circulation), and a monopoly license from Congress to issue currency used by citizenry for cash transactions. It also has, like FNMA and FHLMC had, an implicit guarantee of the U.S. Treasury, backed by the Treasury's taxing authority.

As an independent institution, the Fed has been extremely successful. It doesn't literally print money; rather it has a "monopoly" license to issue currency as a liability that it "gives" to primary dealers. I put "gives" in quotations, because to get this "gift" the primary dealers must "give" the Fed Treasuries OF EXACTLY EQUAL VALUE. So there is no "gift" or "money printing"...rather there is a "trade" of one highly valued piece of paper, for another highly valued piece of paper.

As important, what this means is that our currency in circulation, and "reserves" deposited at the Federal Reserve Banks are backed by assets, Treasuries, that the Fed gets in exchange for the liabilities it issues. This monopoly license to issue currency, basically allowing the Fed to fund itself a 0% (i.e. currency doesn't have a coupon or yield) on the liability side of its balance sheet, while investing on the assets side of its balance sheet, has yielded huge seigniorage profits, most of which it upstreams to the U.S. Treasury. In addition to seigniorage profits, small additional Fed profits have been earned by a maturity mismatch, with currency having (conceptually) an overnight maturity, and assets' maturity historically averaging a year or two. With a few exceptions, the Fed has not taken material credit risk or experienced material credit losses.

This arbitrage the Fed conducts obviously has been generally consistent with its stated public purpose of ameliorating bank runs, managing inflation (to be low and stable), and keeping unemployment low. In addition, due to the low duration of the balance sheet, the profits have been relatively consistent, stable and growing!

From 2000 to 2007 when the Fed Balance sheet was less than $800 billion in size, and its assets were mostly T-bills, these profits averaged $25B per year. They grew to $88B for 2012. Arguably the size of this arbitrage profit would have shrunk in recent years, because as inflation and interest rates fall, the spread between T-bill assets, and 0% yielding currency liabilities fall. Similarly, the purchase of long-term Treasuries has flattened the yield curve and pushed down the curve premium (such that now the spread between 1 and 2 year treasuries is just 0.17%). However, instead of shrinking, the profits have grown for two reasons.

First, profits have increased because the balance sheet has more than quadrupled in size, from $800 billion, to $3.5 trillion. (It's on its way to $4.0 trillion.) Second, because the way fixed income investing works is that when interest rates fall, especially on longer maturity securities, the INITIAL effect is to mark up the value of assets, even though tautologically the long term effect is for income to be less.

I therefore see huge problems going forward with this Fed balance sheet. In particular, for over a year now, the Fed has been buying very long dated, fixed coupon Treasuries and mortgages with very low yields, and financing them in the money market. That is akin to what the S&Ls, banks, and hedge funds did prior to the great crisis of the past, right? Wrong.

It is wrong for two reasons. One reason it is NOT wrong is because the Fed can print money. The Fed's ability to print money (seigniorage) is valuable, worth about $25B per year, history suggests. But it's far from unlimited in its intrinsic value, probably not much beyond that historic figure. In particular, since for each dollar the fed prints to buy an asset it records a liability, the effect on the Fed's net worth or solvency is zero.

So the first reason the Fed is not like an S&L, bank, or hedge fund is that the no S&L, bank or hedge fund would be allowed to run 10, 20 and 30 Year Treasuries at a leverage ratio of 55 to 1. The (counterparty) market wouldn't allow it.

Second, the Fed is not akin to a S&L or hedge fund simply because all of the Fed's liabilities (currency and reserves) are implicitly backed by the U.S. Treasury and Congress. Put more plainly, taxpayers own the Fed's losses.

Given the difficult interest rate environment going forward, at a minimum we should assume that the $25 billion in seigniorage income the Fed's become accustomed to will not be forthcoming. Furthermore, we should clearly not expect the $88 billion in profits sent by the Fed to the Treasury last year to repeat itself. The budget impact will be felt immediately, as even currently, constituents bemoan the sequester (lower government spending), rising taxes (higher government revenues), and higher healthcare insurance premiums and penalties (deemed taxes as well).

Lastly, with a modest 2% rise in interest rates (across the curve as the Professor McKinnon suggests), a mark to market, OR hold to maturity, loss of $800 billion will be felt by taxpayers.

Certainly, Fed watchers, and policy wonks can debate the effects of QE on employment, and the distortions to markets. However, I for one, side with Fed Governor Jeremy Stein regarding the need for a definitive exit plan, and think I understand where Bernanke last May was coming from when he first spoke of an exit. The $3.5 trillion, and growing, balance sheet is something which sooner, rather than later, we will have to exit from...