Exchanging Nickels for Dimes?

Indeed, I think one of the biggest lessons to come out of this episode will be that thinking of monetary policy in terms of a interest rates is confused.

what concerns me about such an operation is that it increases the interest rate exposure of the central bank by a factor of 15 (or so depending on the average change of the portfolio's maturity). unless they hedge this (I am not sure that the capital markets could easily absorb such an amount of concentrated interest rate risk)any small change in the rates of interest (due to some exogenous shock lets say)would totally wipe out the central bank causing the taxpayer to make them whole. this could amount to a large loss. If rates rise from 2% per annum to 4% there could be a total loss of some 20% of the outstanding portfolio (i.e. $80 billion)and that is a conservative estimate.

Could the Fed become insolvent? I mean, I suppose they have liabilities in an accounting sense, but if a bunch of banks want their reserves, doesn't the Fed just print them up (or have the Treasury do it)?

They way I see it (and I was taught), is that a central bank is just like any other institution with a balance sheet. When the central bank prints money it is in essence issuing a zero coupon infinite duration bond on behalf of the sovereign. Should the central bank engage in investments that turn out to be worth less then originally thought, then there will be a mismatch between those liabilities and the assets that back them. This causes those liabilities to be discounted by the market as the output lost on the investment has to come from somewhere, it cant just disappear. this is a direct transfer from the holders of the central bank's liabilities to the investors who shorted the central bank's investments. Just like pre-bankruptcy GM couldn't issue debt to restore it to solvency (as the market was already discounting the wealth destruction that took place in the price of its debt and equity) a central bank cannot issue more currency to restore it to solvency. This will just causes more inflation then the discount that the market will already place on the high powered money(this usually happens in the FX market first before appearing on CPI). The central bank can either reduce the amount of its liabilities (by retiring money), or if monetary policy needs to be expansionary the treasury can issue (tax collecting) debt to the central bank(the bank of england has an official arrangement like this with HM Treasury over their QE induced purchase of guilts) . As to the commercial banks, as the wealth loss gets translated into the market for the central bank's liabilities,these banks have a large incentive to reduce their demand for reserves as to avoid any depreciation of their own capital. this is often why developing countries force commercial banks to hold certain amounts of liabilities, so that any mismanagement does not lead to large runs on the currency.

I don't get what Hamilton is driving at. Is he suggesting a form of "carry trade" for the Fed is actually a good thing? What is the point? To hike short rates and lower long rates? Its all just zero sum in the financial system. It doesn't create quantitative easing as bills need to be rolled. Its strikes me as unecessary government manipulation. I have an idea. Why not sell 2 trillion of bonds and buy 40 trillion of S&P options on margin?

I meant futures, but what the heck, they could buy some options too.

Hamilton is talking about once was called "Operation Twist" by the JFK crowd that sought to tilt the Tresury yield curve by swapping the quantities of debt along the curve. The Treasury could do this on its own; it doesn't need the Fed.

The expectations theory says no effect; the preferred habitat theory says some effect.

But none of this matters. QE is not about swapping debt. It is about the Fed buying, say, $1 trillion of long Treasuries with freshly created cash, currency or bank reserves. That could have a profound effect on bond yields and ultimately on aggregate demand. (Hello, Prof. Sumner.)

I see no downside to more QE right now. To say that more QE will cause inflation is to say that it will raise aggregate demand. But if the Fed is "pushing on a string," it can't raise aggregate demand. Both can't be true, unless you believe the aggregate supply curve is vertical at a 9.5% unemployment rate