...the Federal Reserve has purchased large amounts of longer-term securities in recent years. The Fed's resulting securities portfolio has generated substantial income but may incur financial losses when market interest rates rise.
He goes on to explain the size of the problem:
....the Fed’s recent securities purchases have caused its balance sheet to grow enormously. Just before the financial crisis, the Fed’s largest financial asset was about $0.8 trillion in Treasury securities, and its chief liability was a similar amount of currency outstanding in the form of Federal Reserve notes. The Fed now holds about $2.4 trillion in Treasury and federal agency securities. These assets are roughly balanced by a similar amount of currency and bank reserves, which can be thought of as the electronic equivalent of currency...Furthermore, besides producing a larger balance sheet, the Fed’s purchases have shifted the composition of the Fed’s securities portfolio toward longer-maturity securities. Indeed, the duration of the Fed’s portfolio—which is roughly a measure of average maturity—rose from between two and three years before the financial crisis to between four and five years now. The longer duration of the Fed’s portfolio implies that its market value is more sensitive to changes in interest rates. The combination of a larger securities portfolio with a longer duration implies that the Fed is taking on more interest rate risk than usual.Here are the mechanics of the problem:
In understanding the Fed’s interest rate risk, it is useful to separate the effects of rising short-term interest rates from the effects of rising long-term interest rates. In general, when short-term interest rates rise, the manager of a portfolio financed by short-term liabilities faces increasing interest expenses. Similarly, when short-term interest rates rise, the Fed will pay a higher interest rate on bank reserves, which increases the funding cost of its securities portfolio. In contrast, the Fed’s interest income that is generated from its holdings of fixed-coupon longer-maturity securities will be essentially unaffected. Thus, rising short-term interest rates will squeeze the Fed’s net interest income...In 2010, the Fed earned $82.9 billion in interest income, which is equal to an average coupon yield of around 4% applied to a $2 trillion portfolio of longer-term securities. The interest expense for funding these assets last year was only $3.1 billion, which is equal to the Fed’s reserves rate of 0.25% applied to more than $1 trillion in bank reserves. If short-term interest rates were to rise, the Fed’s net interest income would fall as interest expenses rose and its fixed-income earnings changed little. Importantly though, currency, which now represents about 40% of Fed liabilities, has a zero funding cost. So, short-term interest rates would have to rise rapidly to quite high levels—in the neighborhood of 7%—for the Fed’s interest expenses to surpass its interest income. Such an outcome appears very unlikely.
So here's what is interesting, although Rudebusch says that 7% short term rates are unlikely, the Fed just adjusted its accounting procedures to prepare for such an outcome. Here's Rudebusch explaining how the accounting change would work, without advising that this is a new accounting method just introduced by the Fed:
In such unlikely circumstances, the Fed’s capital base would be maintained by letting remittances to the Treasury fall to zero. In the most extreme case, future remittances would also be reduced (and recorded as a change in deferred credit), but the Fed’s capital base and financial position still would remain completely secure.
A few things should be noted about Rudebusch's discussion. The entire problem of the Fed going cash flow negative only comes about because of Bernanke's introduction of a new "tool", the paying of interest in excess reserves. Prior to the introduction of this "tool", using the old methods of monetary policy, the Fed could never have been in a position to go cash flow negative.
Secondly, this new "tool" creates an additional inflationary bias to the Fed. One would think that short-term rates would be climbing during a price-inflationary period, either because markets would be demanding an inflation premium or the Fed was tightening rates to battle price inflation. But if the short-term rates stay high for long enough, the Fed could find itself in the position of liquidating assets to pay for the climbing short-term obligations (Thus putting even higher upward pressure on rates) or having to print money to buy short-term rate securities to push rates lower (Creating even more price inflation).
Obviously, this scenario, if played out, puts the Fed in a box. And the box is the result of the mad scientist-instincts of Fed Chairman Bernanke, who has been using the American economy as his laboratory. Rudebusch is quite right in saying that short-term rates are not near 7%. But, I am far from sure that he is correct as to whether two years out short-term rates won't be at 7% or higher, at such point the Fed will have to deal with one of the consequences of Bernanke's mad tool creation---significant inflation (by printing more money) or deflation (by printing more money) appear to be the only solutions at such time, other than the Fed going bankrupt. The final option being one that some of us would not mind seeing, with the ultimate result being that the Fed is ended to put it out of its misery.