And now the hard part begins.
The Fed’s decision to reduce its monthly bond purchases to $75 billion from $85 billion suggests this is not the beginning of the end, but the end of the beginning, or some other cliche (your mileage may vary).
Later today, the Federal Reserve will release its weekly data on its balance sheet, which is expected to tip the scales at more than $4 trillion in reserve bank credit. Given that this is a slowing in growth of the balance sheet, and not a reduction, it’s not stopping there.
So this “taper” means the rate of growth finally slows, as the Fed starts to buy fewer Treasuries and mortgage securities each month.
Ben Bernanke, in his farewell presser (you could see the glee in his unmoving, implacable face), suggested the Fed might reduce purchases by a similar amount at each subsequent meeting. Assuming that goes to plan, this implies a slow reduction of purchases until November, when logically, instead of cutting from $15 billion to $5 billion, it may just do away with it altogether.
Adding what’s left to buy in December, that brings the balance sheet ultimately to about $4.5 trillion. This then needs to eventually be unwound through allowing bills, notes, bonds and mortgage securities to mature or be sold – the ultimate in threading a Monster truck through the eye of a needle.
Taking a closer look at that balance sheet, letting it all mature out (sort of the way the Mets are paying disgruntled ex-outfielder Bobby Bonilla for another million years, and you’re forgiven for not getting that reference) would eventually allow the Fed to exit the markets around the time the earth crashes into the sun. Or, more accurately, in the realm of 2024 to 2028 or so, depending on the situation (assuming no other catastrophic recessions that require a stop or reversal of the tapering process).
It isn’t likely, but for kicks, here’s how the balance sheet looks:
The Fed holds about $2.18 trillion in Treasuries and $1.48 trillion in mortgage-backed securities. Breaking it down by maturity, it’s got about $748 billion in notes maturing between one and five years from now, all of it in Treasuries; $864 billion in five-to-10-year notes, most of which are still Treasuries, and $572 billion in longer-dated governments and the bulk of that $1.48 billion in mortgage notes as well.
Of course, it’s not all going to be sold. At the beginning of 2007, the Fed held about $900 billion in marketable securities, of which only about $150 billion were of a maturity greater than five years. Any effort it makes is going to be slow – Bernanke said that himself in a speech in February 2010.
Draining reserves through reverse repos and other agreements will be a primary way of going about business, and that will help tighten the Fed’s control over the funds rate again.
With regard to the Treasuries and mortgage debt, the Fed is likely to discontinue rolling over that debt for starters – it doesn’t want to simply sell it all and panic the bond market. This version would slowly reduce the amount of outstanding debt, sure, but the rate early on would be so slow as to be non-existent.
With just $750 billion in debt (as currently constructed) maturing in the next five years, by the end of 2018, the Fed balance sheet – assuming an additional $500 billion in purchases and no other offsetting operations (though there will be) – would still be somewhere around $3.8 trillion.
Bernanke, in that same speech, said the Fed “anticipates that its balance sheet will shrink toward more historically normal levels and that most or all of its security holdings will be Treasury securities.”
When would that be? Well, again, the mortgage debt is almost exclusively super-long-dated stuff, so the Fed most certainly at some point will start selling securities – when the “economic recovery is sufficiently advanced.”
Daniel Thornton, economic adviser at the St. Louis Fed, leans toward the idea of selling bonds rather than just paying additional interest on excess reserves – which has a number of other complications.
The problem with the sales, he says, is the Fed could lose money, though most of the bonds were at least bought when bond prices were much lower and yields were higher (the buying that’s been done in the 2012-mid 2013 area is probably the most problematic, particularly the longer-dated issues).
“The likelihood of significant capital losses could also be reduced by selling longer-term securities and simultaneously purchasing short-term securities in advance of reducing the overall size of the Fed’s balance sheet,” he wrote in August. So there’s that.
The Fed’s exit strategy appears to be on a timeline similar to removing U.S. troops from the Korean Peninsula, in other words. Will the markets freak out? Not if they’re bored to death, which seems to be the objective. But hey – whatever works.