The bitter medicine of quantitative easing

Barry Ritholtz wrote an opinion piece on Bloomberg today arguing that it's hard to criticize the Fed's QE programs simply because we don't know what would have happened without them. Since this is not a "controlled" experiment in which we can compare a patient taking experimental medication with the one taking a placebo, there is no way to tell if the therapy had worked. All we know is that the patient has undergone a slow recovery and according to the "doctor" may have been worse off without the "treatment".

"If you are testing a new medication to reduce tumors, you want to see what happened to the group that didn't get the test therapy. Maybe this control group experienced rapid tumor growth. Hence, a result where there is no increase in tumor mass in the group receiving the therapy would be considered a very positive outcome."

This argument was used a number of times in recent years, including for example with the American Recovery and Reinvestment Act of 2009 - the $840 billion "stimulus" bill. There are all sorts of estimates on how many jobs the bill saved/created and how many GDP points were added. Was it effective relative to other job creation programs? We of course will never know because we can't peer into an "alternative universe" where the stimulus bill had not passed.

But maybe we are asking the wrong question. Let's for a moment stay with the medication analogy that Mr. Ritholtz introduced. Experimental medication is usually applied in dire cases when the patient's health is deteriorating and traditional therapies had not worked. The use of the first round of quantitative easing, QE1, was just such a case. It was necessary to stabilize the banking system that was frozen - an extreme problem that called for radical measures. But what about QE3? Mr. Ritholtz argues that with other parts of the federal government dysfunctional, the Fed was simply the only game in town to get the economy moving.

However was the US economy in such a disastrous shape in the summer of 2012 that it called for another extreme intervention? Clearly growth was uneven and the labor markets remained wobbly. Nevertheless a recovery was taking place. A patient who is getting better, albeit slowly, is generally not given an ever larger dose of experimental medication in hopes of miraculously accelerating the recovery.

Rather than Mr. Ritholtz's tumor analogy, let's think about QE as delivering excessive doses of experimental pain killers. Yes the patient may feel better at first, but as we all know, prolonged use could create some nasty side effects. The key side effect of course is addiction - which over a long period of time requires one to administer ever larger doses in order to obtain the same effect. And now you are not just fighting the disorder but also the withdrawal symptoms. That is precisely what is taking place these days (see post). Furthermore, the uncertainty surrounding the QE "withdrawal symptoms" is what had put some of the economic activity on hold, activity that is only now beginning to return (see post).

What's particularly troubling about QE is that even after the "injections" are taken away, the nation's banking system is saddled with the "long-term side effect". The US monetary base is now near $4 trillion, with some $2.5 trillion of it sitting on banks' balance sheets in the form of excess reserves - a situation with no precedent. Removing it would require the Fed to sell its securities holdings - something the central bank is not planning to do. This bloated monetary base is going to be with us for a while even as the Fed's securities purchases end - an "experimental drug" whose long-term effects remain unknown.

SoberLook.com

From our sponsor:

5 years of QE and the distributional effects

As we approach the fifth anniversary of the start of the first quantitative easing program, some are asking the thorny question about the so-called "distributional effects" of these unprecedented programs. Who really benefited since the first QE was launched? There is a great deal of debate on the topic, but here are a couple of facts. Financial asset valuations, particularly in the corporate sector have seen sharp increases. For example the S&P500 index total return (including dividends) has delivered 144% over the 5-year period. Those who had the resources to stay with stock investments were rewarded handsomely.

Source: Ycharts

But what about those who didn't have such an opportunity? For example savers, particularly retirees who had to stay in cash? They were hurt severely by record low interest rates (negative real rates - see post). And those who had neither the savings nor significant stock investments, relied on house price appreciation or growth in wages. The housing recovery has certainly been helpful (for those who kept their homes), but according to the S&P Case-Shiller Home Price Index, US housing is up less than 5% over the past five years. Not much of a "wealth effect" for those without stock portfolios. And when it comes to wage growth, the situation isn't much better. The chart below shows hourly earnings growth of private sector employees.

It therefore shouldn't be a surprise that the three rounds of quantitative easing over the past five years rewarded those who had the wherewithal to hold substantial equity investments. Everyone else on the other hand - which is the majority - was not as fortunate.

Perhaps the best illustration of these distributional effects is in the chart below. It shows the relative performance of luxury goods shares with wealthier clients vs. retail outfits that target the middle class. The benefits of QE are clearly not felt equally by the two groups.

Source: JPMorgan

So as we prepare for the Janet Yellen's ultra-dovish Fed (see story), it's worth thinking about the past five years and the cost of growing distributional effects in the United States. For now there is plenty more cheap money to help those with large stock portfolios.

JPMorgan: - There are debates about whether a 0% cost of money helps anything except financial asset prices ... All we know is that the Fed has a story to tell (“cheap money is good”) and they are sticking to it.

SoberLook.com

From our sponsor:

Fed's expansionary policy is not going achieve projected unemployment levels

CS put together an updated chart of the Fed's unemployment projections (discussed here). The expectations for the "longer run" unemployment rate continue to be unrealistic and are based on credit driven excesses of the pre-crisis era. Someone needs to explain to the FOMC forecasters that a structural change in US employment (discussed here) had raised the level of "natural" unemployment rate. Expansionary policy, no matter how aggressive, is simply not going to produce the desired results.

Source: CS

SoberLook.com

MBS could be part of the Fed's easing program

JPMorgan recently conducted its MBS (mortgage backed securities) investor survey. The key questions were focused on the Fed, as the next policy steps will be particularly critical for the MBS market. Let's look at the responses to two of the questions.

1. Identify the most likely policy course at the next FOMC meeting.

62% believe there will be some form of an easing action. Interestingly this number is lower than Goldman's proprietary model assessment of 75%. But consistent with that earlier post there is 2:1 expectation that this easing will be an extension of Operation Twist rather than some outright balance sheet expansion (QE3).

Source: JPMorgan

2. Assign the odds to the easing action involving MBS purchases, no matter what the program is. Somewhat surprisingly half the participants believe the chances are now over 40%. Only 8% of MBS investors believe the chances are below 20%.

Source: JPMorgan

Of course it could be wishful thinking, but that would indeed be good news for MBS investors, as any indication of such purchases will further improve the value of these securities. However even without any Fed action, MBS paper has done reasonably well as shown by the MBS total return index below.

JPMorgan MBS Index that tracks the total return of the fixed-rate mortgage-backed securities issued by the U.S. agencies FNMA, FHLMC, and GNMA (Bloomberg)

The index is up 1.6% YTD and 4.6% from a year ago - not bad for securities yielding between 25bp and 300bp depending on the maturity. The question of course remains whether such action will help improve US economic growth by lowering mortgage rates even further. The 30-year mortgage rate across the US has already hit another record low last week and people who refinanced at 4% or even lower are once again thinking about refinancing.

30y fixed mortgage rate (Bankrate.com)

Currently there is roughly a 200bp spread between the 30-year mortgage rate and the 10-year treasury. The Fed's goal would be to compress this further by flattening ("twisting") the agency curve (see FNMA curve below) while selling more 2-year notes or even dumping shorter term MBS to keep the balance sheet constant.

FNMA actives curve

The Fed may be able to tighten the spread to treasuries another 25-50bp, but beyond that the central bank would be buying securities with such a low spread/yield that it will be nearly impossible to sell in the future without taking a loss.

Unfortunately the Fed's options are so limited, this may be the most impactful action they can take at this stage. And that explains why a number of the JPMorgan survey participants assign a significant probability to MBS being a part of the next easing program.

SoberLook.com

5 reasons QE3 is now off the table

The third round of quantitative easing (QE3) by the Federal Reserve appears completely unlikely at this stage. Here are the reasons:

1. PPI ex-food/energy, a leading indicator for CPI which is monitored closely by the Fed, has been rising at a decent clip with the latest number at 3% year-over-year.  This will make the FOMC think twice about injecting additional liquidity.

PPI ex food and energy YOY (Bloomberg)

2.  Capacity utilization, though low by historical standards, has been growing.  The latest number is quite strong, particularly excluding utilities which are operating below capacity due to warm weather. In fact manufacturing output increased 0.9% m/m in December.

Capacity utilization relative to 2007 (source: the Fed)

3. Whether by design via "operation twist" or due to flight into treasury markets because of the European concerns, the Fed got their desired result of flattening the treasury curve.  The flattening in the last six months has been quite sharp. 

US treasury curve move in the last 6 months (Bloomberg)

That curve flattening in turn brought down mortgage rates considerably. Incremental outright asset purchases will accomplish little in that regard.

30y mortgage rate (Bankrate.com)

4. It is not clear that quantitative easing has a meterial impact on broad money aggregates.  The chart below shows the M3 aggregate as calculated by Capital Economics (the Fed no longer computes that number).  QE2 was started in November of 2010, when M3 growth was already climbing off its lows.  And in spite of tremendous injection of liquidity into the system, M3 growth continued to lag materially the narrower money aggregates. 

Source: Capital Economics (* M3 is their calculation, not the Fed)

5. QE3 is the last "bullet" of any consequence the Fed currently has in its arsenal.  With short-term rates locked in at zero for two years and mortgage rates at historical lows, the impact on the US economy from additional liquidity will be minimal if not negative. Therefore the only time the Fed would consider using its last bullet is if we experience a global credit crisis such as a sudden default of a major EU financial institution or a large eurozone sovereign.  And the eurozone and the ECB have proven that they will simply not allow for that to happen.

SoberLook.com

The spike in Fed's balance sheet is not QE3

While the world watches the ECB, China, and now North Korea, the Fed's balance sheet has reached a new high last week.

Assets at the Fed

There are two components of this balance sheet increase: a large spike in the Central Bank Liquidity Swap (discussed earlier) and a substantial purchase of mortgage-backed securities (MBS).

Components impacting the increase

This has led to some speculation about the fed's next move, but the rationale here is actually quite simple.  Just as was the case last year, with mortgage rates hitting new record lows, the refinancing activity has stayed elevated, particularly since August.  This caused a decline in the principal amounts of MBS held by the Fed (blue chart above).  Last week's purchase was simply an adjustment to compensate for some of that decline.

There is broad consensus however that at some point it may no longer be just an adjustment.  Dealers think the Fed will soon begin an MBS purchase program that will drive mortgage rates to extremely low levels.

Businessweek: The biggest bond dealers in the U.S. say the Federal Reserve is poised to start a new round of stimulus, injecting more money into the economy by purchasing mortgage securities instead of Treasuries

As a result, the 30 year fixed mortgage could hit 3.5% or even lower.  The last FOMC meeting however provided no hint of such program. The Fed is not in a hurry to launch something like this for two reasons.  First, at this stage it is the only viable tool that will have a material impact on asset prices and they want to reserve it for a potential escalation of the European crisis.  Second, the Fed is not totally immune from political pressure, and such a move will generate a great deal of criticism.  Given the relatively decent economic data coming out of the US, "operation twist" is all we are going to get for now.

SoberLook.com

The ECB loan to IMF - a new form of euro-zone QE

Even though some form of a tighter fiscal integration in the euro-zone is still in the works, here is the latest idea being bounced around in Europe to address the crisis immediately.

A few observations on this approach:

• This structure is supposedly going to get around the rules that make it hard for the ECB to buy paper directly. 
• Bond purchases may include new issue or even focus entirely on new issue paper or a direct emergency loan to governments.
• It is in fact a form of QE for the euro.  The ECB is not expected to sterilize a loan this size.
• The US is about 18% on the hook should the sovereigns fail via it's exposure to the IMF
• Germany is going to have a tough time with the idea even if it gets around the ECB restrictions because of their inflation fears. 
• Sovereign spreads are reacting quite positively to this proposal.  Below is the 5-year Spain bond spread to Germany.

5-year Spain bond spread to Germany (Bloomberg)

Later in the day (as expected) we saw resistance to this proposal from US politicians concerned with the US exposure to IMF.  The US has a veto power at the IMF (Germany has been asking the US for a while to give up the veto right) and some politicians want to use this power to stop IMF from taking large sovereign risks (per structure above).

The Hill: “I’m adamantly against the IMF being involved in this,” [Sen. Tom Coburn (R-Okla)] said. “We’re throwing good money after bad down a hole that I think is not a solvable problem,” he said. “Europe is going to default eventually, so why would you socialize their profligate spending,” he added. Coburn estimates the U.S. could be liable for as much as $176 billion if the IMF shores up Italy and Spain and the European Union collapses.

It's not clear how Coburn arrived at $176 billion, but it's definitely going to resonate with the US voters.  Sovereign spreads widened on the news:

French 5yr spread to Germany (Bloomberg)

SoberLook.com

Bernanke's next steps

Let's take a quick look at options currently available to the Fed and their potential next steps. With the dollar under pressure and bank reserves at historical highs, one would think the Fed is getting uneasy with all the liquidity in the system. Now that the short-term liquidity facilities are winding down, much of the new securities purchases will increase the balance sheet and grow the money supply. Many beleive this will surely lead to inflation.

To address this, the Fed has the following two tools (outside of outright securities sales):

1. Purchase new securities (RMBS, Agency paper, etc.) on repo (sterilized purchases). The Fed would effectively buy the securities and immediately lend them out for some period, taking in cash collateral. This takes these securities out of the market, but does not increase the money supply because the proceeds from these sales would not be available to the dealers (the proceeds become the cash collateral). This could be done not only with new purchases, but with securities already on Fed's balance sheet (about $1.5 trillion worth). To accomplish this on a scale that makes a difference, the Fed needs to set up repo lines with banks and dealers outside of the Primary Dealer group. The primary dealers may not have the capital to absorb such massive amounts of repo transactions on their own.

2. The Fed could also raise rates. But this wouldn't be simply raising the Fed Funds Target Rate. Instead the Fed now has the ability to raise interest rate on the reserves that banks keep with the Fed. That immediately creates a floor on rates because banks have no incentive of lending at levels at or below the reserve rate. Instead they can simply deposit the funds with the Fed on a riskless basis. This tool has been used by other central banks for decades.

The first tool may be set up relatively soon, particularly for new purchases, but it's usage should be fairly modest in the near-term. The rate increases however are months away. Here are two reasons for the Fed's dovish approach:

1. The Fed will not take any rate action until they see improvement in employment. And as we discussed earlier, this may take a while. This is particularly true because many recent jobs (the "bubble jobs") were created on the back of construction spending.

2. The Fed (among numerous measures available to them) watches one key indicator quite closely: the rate of change in "broad" money supply relative to the "narrow" money supply. It's a measure of how effective the liquidity injections have been in stimulating lending. Banks can be loaded with cash, but if they don't lend, the cash is not making it's way into the broader money supply (the banks effectively stay overcapitalized). And that means the broader economy is not benefiting from the liquidity the Fed had provided, which limits it's growth. The chart below shows the relative growth of M1 (narrow measure) and M2 (broader measure). Until M2 picks up significantly, the Fed will do very little in terms of tightening.


source: St. Louis Fed


Inflation is unlikely to pick up until credit is available in the broader economy to allow corporations and individuals to pay higher prices. With broader money supply responding this slowly, significant price and wage increases are unlikely in the near-term.

The possibility of the Fed actually selling securities from it's balance sheet outright is even less likely. Such sales may impact long-term rates, which may have a negative effect on housing and the consumer, and the Fed will categorically not go there. The RMBS securities, the agency paper, and even treasuries they have bought, will stay on Fed's balance sheet for years to come, possibly to maturity.

A signal from the Fed?

At first glance, there wasn't anything unexpected in the FOMC statement today. But reading more into the wording, the following difference in the language between this statement and the one from last month may have significance:

From the FOMC statement August 12:

In these circumstances, the Federal Reserve will employ all available tools to promote economic recovery and to preserve price stability.

From the FOMC statement September 23:

In these circumstances, the Federal Reserve will continue to employ a wide range of tools to promote economic recovery and to preserve price stability.

Many Fed watchers have interpreted this as a signal that the Fed is planning to do something different. Some beleive it is related to the beginning of an end to the liquidity buildup. Yes, the rate will stay at zero, but it is likely that the remaining securities purchases will be "sterelized", meaning some won't be outright buys, but instead may be repo transactions . The Fed may not want significant additional quantitative easing.

The US quantitative easing has just begun

With the FOMC meeting currently under way, it's worth reflecting on the Fed's implementation of the monetary policy this year. Surprisingly, according to the latest Credit Suisse research, there hasn't been much quantitative easing in 2009. But how could that be possible, given the way the Fed has been growing its balance sheet? The chart below shows securities held outright by the Fed. How can this NOT be a form of quantitative easing?



Credit Suisse argues is that the monetary base has not really grown much in 2009, as the following chart shows.


Source: Bloomberg

So where is all the cash going from the Fed's purchases? It has to end up in the banking system and show up in the monetary base. The argument Credit Suisse makes is that all the short-term lending the Fed had put in place last year as an emergency measure has been shrinking in size, effectively offsetting the securities purchases.



Banks are de-leveraging, trying to reduce their borrowing from the Fed. Effectively securities purchases put Fed’s money with the banks, while the banks in turn pay back the Fed on their loans, thus neutralizing the impact. As we discussed before, this has slowed down the pace of the Fed's balance sheet growth significantly.



In addition some of the cash for the Fed's recent lending had come from the SFP program in which the US Treasury has issued treasury bills with the proceeds to be used by the Fed for its emergency lending. That type of program does not add any new cash to the system, because while the Fed injects cash (by lending to banks, etc.), the Treasury takes the liquidity out by selling bills. That in fact was the original purpose for SFP.

But SFP is expected to be wound down shortly. The impact of the reduction in short-term funding facilities will end as the programs come to a close. Therefore there will be nothing more to offset securities purchases going forward. That means that if the Fed continues purchasing paper at it's recent pace, quantitative easing finally will kick in with force.



This will end up ballooning bank reserves and truly “flooding” the system with dollars. The chart below from Credit Suisse shows their projection for reserves (translating into a rapidly rising monetary base):



The Fed is keenly aware of this problem going forward, particularly with the dollar weakness. Once SFP as well as the short-term facilities wind down, every dollar of purchased securities will be a new dollar “printed”. That is why the Fed is now supposedly putting together a new securities reverse repo program with the dealers. The idea is that going forward the Fed will be buying new securities by borrowing money from the dealers rather than “printing” new dollars. The Fed will place its securities with the dealers as collateral when it borrows. In fact the central bank may choose to borrow against the existing securities as well. New security purchases will be putting liquidity into the system, but by borrowing from the dealers, the Fed will be temporarily taking liquidity out.

Eventually the Fed will have to outright sell the trillion plus of securities it holds, taking liquidity out permanently (the reversal of quantitative easing) – a dangerous thing to do in this economy. And the more purchases it makes going forward, the harder on the economy it will be to reverse it. For now however, the reverse repo effort will have to do, by (at least in part) compensating for the wind down of SFP and the short-term emergency lending programs.

Quantitative easing right on schedule

The Fed is gearing up to purchase treasuries all along the yield curve. Looks like quantitative easing across the board in the next two weeks. They picked a period between the Treasury auctions to try to bid up the paper. Let’s see if it works.