Here I have the actual mix of holdings, and the total (red) matched against the ten year yield. But the general rule still holds, whatever mix of securities held by the Fed, the ten year yield jumps as the rate of purchases increase.
Remember, in the reserve model of the economy, a five year reserve is held against the need of a firm to buy a five year machine at a random opportune moment. The firm sees the need for an additional machine, it can go to the bank and show that it has been holding 15% of the purchase price against a loan for the purchase. These liquidity reserves are fixed, and well managed by the firm. The most common depreciation schedule is likely the ten year, that is why I watch it and it seems to be the knee of the curve.
When the fed takes out the supply of treasuries, from 1 to ten, or any large interval in between, then there are fewer of the liquid reserve instruments available. To an approximation, the economy has to adjust, replacing the reserve instruments removed with a mixture of cash and ten year notes. Hence the
As a side note. When the Fed is buying the instruments Congress wants to sell, then the Fed is mismatched, in term structure, to the private sector. Its earning from treasuries held drops more than the seigniorage to Congress increases, it is a net loss to the Fed. Hence, as it supports the debt ridden Congress more, its return on holding drops, as we have seen.
So what is this tightening, loosening thing? Depends on what the target is. When the Congressional term structure is mist matched to the private sector, then the Fed has to choose. Match Congress in the near term so Congress can make the budget and tighten on the private sector or visa versa? The Fed can tighten on the private sector and make money for Congress, but maybe not in time for Congress to make its budget. But frankly, I see the financial industry as always being even in the game, able to adjust to whatever distortion the Fed creates. Congress and the Fed are stuck, in equilibrium they both go broke at the same time.
Here is the effective interest paid by Congress, and these are net returns from the Fed:
2008 32/800 = .04
2009 47/500 = .094
2010 80/800= .1
2011 75/1600 = .05
2012 88/1600 = .055
2013 78/2100 = .037
When Congress needs more seigniorage, the Fed returns drop and the effective interest drops. It looks to me that they will be simultaneously broke some time in the middle of the year. If Congress can hold on then we can delay the moment until our next regularly scheduled recession, in early 2015.
No comments:
Post a Comment