The market is abuzz with speculation about "if "and "when" the Fed will launch an extra wave of asset purchase. As headline news improve, the debate revolves around "if". As they deteriorate, the issue becomes "when".
Very little consideration is given to the content of those potential asset purchases, as the assumption is that QE3 will look just like QE1 and 2.
In QE1, the Fed purchased Mortgage Backed and Agency securities as well as Treasuries while in QE2 it concentrated on Treasuries. Between November 2008 (QE1) and mid-2011 (QE2), the FOMC added up a grand total of $2.350Bn to the size of the Fed balance sheet.
QE2 was complemented with the current "Maturity Extension Program," whereby the Fed sells short-dated securities and buys long-dated ones, in an effort to lower the yield on government bonds. The goal is to cut long-term rates, boost investment and encourage risk taking by fighting risk aversion: the idea is that if the Fed buys Treasuries, lower long term rates will percolate elsewhere in corporate bonds and stock prices.
Should the current effort not yield the expected results, there is a need to think again about the nature of the tool: If Quantitative Easing has not worked as well as expected up until now, is it because its volume has been too small and therefore we need more of it, or is it because it is too blunt an instrument to achieve the objectives of a higher level of investment in the economy? The total amount of QE ($2.350Bn) is equivalent to over 15% of GDP and the size of the Fed balance sheet now amounts to close to 20% of GDP.
This should raise questions as to the appropriateness of a policy aimed at more of the same. In parallel, as the Fed buys assets, it "prints" the money used to fund those purchases. The rationale behind increasing the size of the Fed's balance sheet is, among other reasons, to inject liquidity in the economy, to generate "lendable funds" to finance growth.
A good chunk of those funds now sits on the Fed liability side, invested at 0.25% in idle excess reserves. They currently amount to $1.450Bn as of May 16th, so almost 10% of GDP after a peak of $1.618Bn in July last year.
This begs the question as to what would more of the same achieve. In our view, the current policy's shortcoming is in the "easing power" of the assets purchased by the Fed: whereas its purchases of mortgage bonds succeeded in impacting the relevant interest rate (namely mortgage rates), its purchases of treasuries only partly succeeded in their aim at increasing risk taking.
There are two reasons why this is in fact not working very well:
First, there is a lack of coordination inside the official sector, where the right hand acts without regard to what the left hand is doing. Indeed, as the Fed buys more long-term bonds and boosts the demand for Treasury securities, the US Treasury lengthens the average maturity of its debt, boosting the supply of the same securities: the Treasury increased the average maturity of outstanding debt during QE1 and QE2 from 47 months in March 2009 to almost 59 months in June 2011. According to some estimates, this pushed the 10-year bond yield up by 27 basis points during QE1 and 14 basis points during QE2. So, had the Treasury not taken advantage of the Fed policy to lengthen the average maturity of its debts, the 10-year yield would have been lower at the end of QE2.
Second, long-term treasury rates are now in a world of their own and become to some extent disconnected from the rest of the economy. Therefore, the Fed is now "jumping off a pancake," cutting rates which are already very low and spending trillions in raising the price of an already overvalued asset, namely long-term bonds. It thereby runs the risk of contributing to a bubble with all attendant risks. In a 2004 article co-authored with Vince Reinhart, Chairman Bernanke himself was expressing this view: "Even if large purchases of, say, a long-dated Treasury security were able to affect the yield on that security, the policy may not have significant economic effects if the targeted security became "disconnected" from the rest of the term structure and from private rates, such as mortgage rates."
So where do we stand?
- The Fed carries $2.3Tn of assets already
- It is "jumping off a pancake," compressing yields that are already very low.
- There's little evidence that risk taking is encouraged as evidenced by the stock market spooked by the European drama.
The conclusion is obvious: there is no point in more of the same. QE in its previous form is not really working, at least not as fast as expected.
So what next?
The conundrum is to jump-start the economy back to life by fighting excessive risk aversion. This implies two things: consumers need to feel confident enough to spend. Entrepreneurs need to access capital cheaply enough to make projects viable in an uncertain environment.
By asking the question that way, the answer becomes obvious: the stock market is key in both those aspects.
Therefore there is a case for the Fed to start buying stocks. By raising the value of stock portfolios, a rising stock market impacts consumers through the well-documented wealth effect. By raising P/E ratios, it lowers the cost of capital that counts, namely the one funding private sector investments rather than government deficits.
For those reasons, rather than inflating a bond bubble, our view is that the Fed needs to impact the assets that are directly relevant to is objective, namely stocks.
This is the most appropriate way of achieving what has been its objective from the start: wealth effect considerations are at the heart of Fed policy and of QE in particular. The tentative thinking was that it could be achieved indirectly through the bond market. That avenue has been exhausted as long-term rates near the absolute lows and in fact penalize savers, namely fixed income investors and pension funds, inflicting heavy collateral damage which hurts the Fed's objective as aging Americans in particular depend on bond yields for consumption.
As obvious as it is, the answer undoubtedly will meet some objections. Let us review them:
Objection 1: "It is not appropriate for a central bank to buy stocks. It is taking on market risk and favoring some stocks over others."
The issue of risk is a real one even if it has somewhat lost its sting since the Fed balance sheet carries "Maiden Lane" type SPVs. The response to this risk issue is that "exigent circumstances" mandate an unusual move as they have mandated similar moves in 2008.
As to the appropriateness of buying stocks for a central bank, the Bank of Japan has created the precedent in similarly difficult conditions. To avoid favoring stocks, the Fed should simply buy indexed funds (SPY or broader ones such as the Russell total market index).
Section 13 (3) of the Federal Reserve Act allows the Board, "in unusual and exigent circumstances," to lend to any corporation or even individual. It has been used for the Maiden Lane LLC to salvage part of Bear Stearns' assets.
Across the pond, the Bank of England purchased £325Bn ($500bn) without buying this directly but by lending the equivalent amount to a subsidiary called "Bank of England Asset Purchase Facility Fund Limited."
The Fed could therefore lend money to an entity tasked with the stock purchases.
Objection 2: "The Fed is not legally authorized to buy stocks."
As mentioned, the Fed did lend to Special Purpose Vehicles holding risky assets (Maiden Lane) under the authority of section 13(3) of the Federal Reserve Act. It could therefore proceed similarly for stock holdings.
Should lawyers feel that the precise wording of section 13(3) prevents its use to finance a subsidiary or a special purpose vehicle, the Treasury could buy the stocks and pledge them to the Fed for financing. This would work just fine.
Objection 3: "Buying stocks during an election year would favor the incumbent President."
The politics would indeed likely weigh against such a move before election time. Similarly, the mood in the FOMC appears so far unripe for bold innovation, barring significant deterioration of the economy.
But should current modest growth stall in the coming months, we feel this option will look appealing to both the Fed and to whichever administration comes out of the November elections.
So, should this come to pass, what are the investment implications? Well, the current downturn in stock prices might present good opportunities of progressively increasing exposure to stocks through broad based instruments as we near election time: SPY, VTI, IWV. Given the tentative nature of the Fed move, this increased stock exposure should be implemented cautiously and progressively.
Correspondingly, the current bond bubble might correct somehow should QE use other tools than bond buying. For the less risk adverse, the progressive and limited build-up of a short position in TLT might appear attractive.
Disclosure: I am long BRK.A.