Guest Post: Why The Fed's Buy-And-Hold (No Sales) Exit Is Not Feasible

Submitted by Martin Sibileau of A View From The Trenches blog,

“…The exit strategy of the Fed is not a stock, but a flow problem. Just like expansionary monetary policy must address fiscal policy, contractionary monetary policy cannot ignore fiscal deficits…”

In the past months and right after implementing Quantitative Easing Unlimited Edition, the Fed began surfacing the idea that an exit strategy is at the door. With the latest releases of weak activity data worldwide, the idea was put back in the closet. However, a few analysts have already discussed the implications of the smoothest of all exit strategies: An exit without asset sales; a buy & hold exit. I have no doubt that as soon as allowed, the idea will resurface again.

Underlying all official discussions is the notion that an exit strategy is a “stock”, rather than a flow problem, that the Fed can make decisions independently of the fiscal situation of the US and that international coordination can be ignored. This is logically inconsistent and today’s letter will address these inconsistencies. Let’s see…

Monetary expansions are treated as a flow process

Conventional PhD wisdom on monetary things tells us that government deficits represent net credits to the system via reserves, as well as to non-government deposits at banks.


 April 28 2013 1

When it comes to bond purchases by the Fed, such wisdom implies that the US Treasury is assisting markets with liquidity. This is not new. As a student, I once heard that “governments must run deficits, so that markets can have a benchmark rate”. My professor meant that “thanks” to fiscal deficits, bonds are issued and markets can proceed with the price discovering process. Today of course, we don’t even have that luxury, courtesy of Quantitative Easing (Not happy with the lesson, I asked Dr. Julio H. Olivera his thoughts on this statement. He chuckled (although Dr. Olivera never really chuckled) and recalled a similar exchange with  John Hicks. According to Dr. Olivera, when Hicks was faced with the same proposition, he replied: “The merchant makes the market”. Unfortunately, I cannot prove this exchange, but thought I would share it with you).

But monetary contractions are treated as a stock problem

Why do I bring this up? Because if deficits are welcome by the PhD standard when it comes to monetary expansions, surpluses should not be ignored, when dealing with monetary exit strategies. It’s only fair…Yet, in the past months there has been a timid incursion into the upcoming debate on exit strategies available to the Fed, but without a single comment on fiscal policies. By now, I have become used to typing CTRL+F “fiscal” (i.e. find “fiscal”), whenever I come across any research note on potential exit strategies. If nothing comes up, it looks suspicious to me.

Once such example was Bank of America’s April 10th note titled “The consequences of a “no sales” Fed exit strategy”, from the Global Economics Rates & FX team. This paper has not a single sentence or thought on the fiscal situation and Treasury issuance forecasts of the United States (the word “fiscal” only shows up once). They are not alone. How can mainstream economics afford to ignore the fiscal side of the problem when facing an exit strategy? They simply treat it as a “stock”, rather than a flow problem.

Terms of the “stock” perspective

As a stock problem, mainstream economists look at a “no sales” exit strategy by the Fed, in these terms:

1.-Not to sell means to hold, while principal and interest payments are reinvested.

2.-The target of a 6.5% unemployment rate is reached and there are signs of a firm recovery underway

3.-Losses on their US Treasuries portfolio are manageable, particularly since the Fed announced its accounting policy change on January 6th 2011, where capital losses may be treated as negative liabilities (Truly, you can’t make this stuff up).  Even putting this fiction away, mainstream analysis is comfortable with a negative impact on the asset side of the Fed’s balance sheet. To assess that impact, reference is made in terms of 10-yr equivalent duration exposure held outside of the Fed. Growth of 10-yr equivalents is expected to stabilize. As I mentioned in the last letter on the Bank of Japan, I side with Shuichi Ohsaki and Shogo Fujita, from Bank of America’s Pac Rim Rates Research team, who argue that volatility in the Japanese bond market could be diminished if the BOJ announced a schedule for buying operations, with the amounts that would be purchased in each maturity sector. In other words, the market does not look at the stock of government debt as a block of exposure that is sizable in equivalent duration terms.

4.-Reserves management, via interest on reserves, can be used to send short-term signals to the market.


In the next sections, I will seek to demonstrate that it is a huge mistake to ignore the fiscal side of this dynamic picture, and that a smooth, no sales exit strategy is fiction. Moreover, I will show that this is a flow, rather than a stock problem. Before I proceed, let me offer you this interesting exchange between Stanley Druckenmiller and Kevin Warsh, which took place on March 5th (Druckenmiller’s intervention starts on minute 5:41)


The flow perspective of the “no sales” exit strategy

To simplify the exposition, let’s look at the cash flow situation of the US government. Like any of us, the government has to collect taxes and pay for expenses. For this particular discussion, it will not matter if the same are ordinary, extraordinary, operating, capital expenditures etc. All I want to do here is to separate this collection of taxes net of expenses –which I will call Primary Cash flow- from the cash flow that has to be used to service debt obligations. In other words, like any of us, the US government will have, after collecting taxes and paying expenses, a primary cash flow with which to service debt obligations:

April 28 2013 2

If the Primary Cash flow (PCF) is not enough to service the debt, unlike us, the government can issue more debt (at least the US government; at least for now). Additionally, the government can liquidate assets. Therefore:

April 28 2013 3

Let’s now look at the demand for the gross issuance and simplify it, saying that the same is purchased either by the Fed, by the rest of the central banks in the world, and by the rest of the world (ROW, i.e. anyone else in this planet who is not a central bank, either in the public or private sector). Under these terms:

April 28 2013 4

Let’s assume that the government sells no assets. If the Fed stopped purchasing US sovereign debt but did not sell any holdings and kept reinvesting the interest and principal payments it received, re-arranging the terms, we obtain:

April 28 2013 5

Let’s further call a Net Demand of one of the agents (i.e. central banks, rest of the world) the difference between its purchases and the collected interest and debt repayments. We can then say that under a “no sales” exit strategy of the Fed and without asset sales, the primary cash flow of the US government equals the sum of the net demands of the central banks and the rest of the world. This is valid at one point in time as well as when we consider the comparative statics of the issue (the term “D” below denotes temporal change in a variable, between t and t+1):

April 28 2013 6

Having arrived to the identity above (the above notations are identities, not equations), let’s look at the context under which the “no sales” strategy would take place. It is a context of a firm recovery, as the Fed has told us and we have every grounds to believe that for this reason, interest rates would tend to rise, as capital moves out of fixed income and credit, into equities. This means that the Net Demand of US Treasuries by the Rest of the World will likely be negative (i.e. Drucknemiller’s observation) or zero, at best:

April 28 2013 7

Let’s take the optimistic view that the Net Demand of the Rest of the World is zero (Clearly, Mr. Druckenmiller does not share this view…and he has every reason not to be). This means that if neither the Fed nor the Rest of the World add US Treasuries to their balance sheets, the primary cash flow of the US government has to be addressed by the Net Demand of central banks, exclusively.

We can think of three different scenarios for the primary cash flow of the US government: A scenario of surpluses (PCF >0), deficits (PCF <0) or balance (PCF = 0).

If the primary cash flow is negative

This is the toughest scenario. It implies that the negative primary cash flow of the US government will be financed by the central banks of the rest of the world. The question here is: Why would these central banks keep accumulating US Treasuries when the Fed itself does not? From here, it is very clear to me that in the presence of continuing fiscal deficits, regardless of where the unemployment rate is, the Fed has no alternative but to continue monetizing the deficits.

But let’s examine this case further. Let’s suppose that by some miraculous intervention, the central banks of the rest of the world would in fact resolve to continue purchasing US sovereign debt, even if the Fed itself wouldn’t. How would this process take place?

There are two ways. Either the currency zones these central banks operate in generate balance of trade surpluses or their respective nations incur into fiscal deficits.

In my last letter, I explained how the latter way worked in Japan under Shirakawa. With regards to the former, to expect a sustainable recovery in the United States (which is the a priori condition for an exit) within a context of fiscal deficits, increasing sovereign debt and balance of trade deficits is a contradiction. Yet some mainstream economists see this as something very feasible, whereby the Debt/GDP ratio falls because the denominator rises faster than the numerator. If this is true, then I am completely wrong and I have nothing else to say. If you believe in the sustainability of this context, please accept my apologies for having taken your time. If you don’t, please proceed to the next scenario analysis.

If the primary cash flow is positive

If the primary cash flow was positive, the Net Demand of the rest of the central banks would be negative.  This would imply a strong and positive savings rate in the United States. The problem is to figure out how the United States can get to achieve a savings rate strong enough to get to this point, in a context of negative to zero interest rates, where nobody has any incentive to save and where the same Fed wants to boost consumption. I asked about this problem (i.e. how the savings rate will improve) to a very well-known economist who gave a presentation this past Wednesday April 24th, at the Oakville Community Foundation. His answer was that the stronger savings rate would come from the public sector. But this explanation seems to me a tautology (i.e. The US government will be cash flow positive because it will save)

The real question in the face of this problem is “What will push the US government and the US to save, when all its deficits are monetized and interest rates are negative?” This is not a new question. In fact, it occupied the mind of Jacques Rueff for decades. Perhaps the first time M. Rueff made public this concern was during an exchange with no other than the same John Maynard Keynes in 1929, during a conference at the Assembly of the League of Nations, in Geneva. M. Rueff suggested that there was indeed an adjustment mechanism for the balance of trade and Keynes asked how such an adjustment could be brought about.

Rueff explained that inflation is nothing else but the creation of purchasing power in a country without a counterpart increase in production. For that reason, it is only possible to run balance of trade deficits indefinitely –like the US has done over the 20th and 21st centuries- if there is inflation. The opposite should also be true: In the absence of inflation, there would be a balance of trade surplus, until all debts are paid (as in this scenario, where the Net Demand of the rest of the central banks is negative).

In summary, to effect a negative Net Demand of the rest of the central banks in US Treasuries, the purchasing power of Americans should be decreased. But how will the United States ever achieve such a state of affairs, when the Fed targets a 6.5% unemployment rate precisely by inflating the purchasing power of Americans? If the Fed is successful, the opposite will have occurred and the nominal purchasing power of Americans will have increased. Therefore, a positive primary cash flow is not possible, as long as the Fed continues boosting asset prices.

How did Keynes react to this view? We have only the testimony of Jacques Rueff on this, which I reproduce below:

“…Et Keynes, qui marchait de long en large –c’était sur la scène d’un théâtre- s’est arrêté brusquement et a dit: “Tiens, mais, cela c’est une idée intéressante, il faudra que j’y réflechisse.”

Je dis cela à mon ami Largentaye, parce que c’est très important pour l’historie de la pensée keynesienne. Cela prouve qu’en 1929 la théorie de la dépense global n’était pas encore au point dans son (i.e. Keynes’) esprit et que c’est plus tard, dans l’ouvrage que M. de Largentaye a traduit, qu’elle s’est élaborée, d’abord dans le Traité sur la monnaie et, ensuite, dans la Théorie générale. Et cela indique, d’ailleurs, le caractère mouvant de sa pensée; ce n’est pas une critique que je lui adresse, c’est plutôt un éloge; c’était un des esprits les plus actifs qui fût…” (J. Rueff, Le système monétaire international”, presentation given at the Conseil Economique et Social, May 18th, 1965).

Finally, if the nominal purchasing power of Americans will not be decreased by the Fed, the real purchasing power will have to fall, with the devaluation of the US dollar. This is a logical conclusion. In a context of global monetary easing, this can only be achieved against gold and…. why not, commodities in general.

If the primary cash flow is zero

This is a simple theoretical conjecture, just like the existence of general equilibrium in the fractionary reserve system and shadow banking we live in. To discuss it is an intellectual exercise of dubious utility.


In this discussion, I sought to show that:

-The exit strategy of the Fed is not a stock, but a flow problem.

-Just like expansionary monetary policy must address fiscal policy, contractionary monetary policy cannot ignore fiscal deficits.

-The fiscal issue PRECEDES the monetary issue. Without first addressing fiscal policy, it is irrelevant whether or not a labour market objective is achieved (i.e. unemployment rate of 6.5%).

-Any analysis of a potential exit by the Fed that dismisses fiscal deficits and focuses on the management of the balance sheet of the Fed only is surreal. It is not enough to claim that buy & hold is better than selling.

-In the case of the Fed, international coordination is required for an exit strategy to succeed.

Bonus: Was Mr. Druckenmiller correct?

As you may have noticed, I was optimistic and assumed that the Net Demand of US Treasuries by central banks would increase (i.e. international coordination) and that the Net Demand of the Rest of the World would remain unchanged.

What I believe Mr. Druckenmiller had in mind is a more realistic picture, where the Net Demand of the central banks would remain unchanged, while that of the Rest of the World becomes increasingly negative. In this context, with the US government continuing to run negative primary cash flows and the Fed shifting from quantitative easing to a buy & hold stance, the supply of US Treasuries would increase and interest rates would rise exponentially. Mr. Druckenmiller was correct.