Dear Fed: hike rates, don't taper QE of Treasuries

Part 1: How QE really works and the benefits it provides

This is the first of three articles in which I will share my thoughts on how the US Federal Reserve could improve its conduct of monetary policy. In this edition, I will discuss how quantitative easing (QE) works, why it was less effective in Europe and Japan and the benefits that continuing QE provides to the US economy.

The Fed has begun deliberations on when and how to start tapering its asset purchases that they initiated following the COVID-crisis in July 2020. Currently, the Fed is buying $80bn Treasuries a month and $40bn Agency MBS per month. The Fed is debating when to start tapering purchases, how fast to taper, and the composition of tapering (Treasuries vs. Mortgages). This debate misses the forest for the trees. Instead of ending QE of Treasuries, the Fed should be raising interest rates.

The insistence on QE being completed prior to raising rates is based on tradition; the Fed has been influencing the short-term rate for more than 100 years, but only added QE as part of their “unconventional” toolkit after the Global Financial Crisis (GFC). The second reason for the Fed’s insistence on tapering QE before rate hikes is based on an incorrect understanding of the importance of the channels through which QE influences real variables. The Bank of England published a wonderful chart in 2009 highlighting the 3 ways that they believe that QE works. I will work from the bottom to top in terms of examining the potential transmission mechanisms.

Source: Bank of England

The 3 potential channels of QE’s transmission

The first channel I will discuss is the idea that more money in the economy created by QE purchases leads to bank lending (bottom channel in the chart above). The basis for this theory is that when a Central Bank (CB) buys fixed income securities from banks, they are replacing securities on the bank’s balance sheets with cash, which they then have the capability to lend out and boost nominal GDP. While this scenario might be plausible in a world where there is substantial loan demand, we have not been in this environment since the GFC. In a world of low loan demand and bank deleveraging, the impact on QE on bank lending is marginal. Instead, QE simply replaces the security assets on a bank’s balance sheet with cash that gets put back onto the Fed’s balance sheet as excess reserves. This is what happened to the US after the GFC. Despite substantial QE, deleveraging banks had little incentive to grow their loan books.

The second channel on the chart is also known as the Portfolio Balance channel. I am going to spend the most time on this channel as this is what Central Banks incorrectly believe is the most important way that QE works.

The idea behind the Portfolio Balance channel is that by buying Treasuries the CB is reducing the term premia that is a key component of longer-dated bond yields. This raises asset prices and creates a positive wealth effect that stimulates spending through 3 different means. First, by taking out the supply of bonds, the CB creates cash for investors who deploy it into riskier assets like stocks and real estate. Second, a lower discount rate raises the value of longer-dated assets like stocks. Third, the reduced yields help borrowers make their payments and refinance existing debt.

There are some things that are correct about the Portfolio Balance channel. Importantly, it is the primary effect of QE in crisis situations. For example, if the Fed had not increased their balance sheet by $3 trillion during March 2020 there is no question that yields at that time would have been higher. This was a liquidity crisis and the Fed buying Treasuries and MBS from dealers who would not hold them unquestionably reduced yields and increased asset prices in the very short-term. Another way the Portfolio Balance channel works is in terms of credit or credit-like securities like mortgage-backed securities (MBS). Mortgage rates would be a lot higher if the Fed was not involved in the market; similarly, the effect of the European Central Bank’s (ECB) QE on corporate spreads in Europe and on peripheral sovereign spreads is beyond doubt. Finally, the reduced yield on credit or credit-like securities (MBS) does allow borrowers to refinance into lower rates, which can prevent system-wide deleveraging which is disinflationary.

However, the inconvenient truth for the proponents of the Portfolio Balance Channel is that Treasury yields typically rise after QE instead of fall as would be predicted by the theory - see chart below. How could this possibly happen if the primary way that QE works is by reducing government bond yields?

10y US Treasury Yields

Source: Bloomberg

I believe the third channel (top of the chart) is the most important mechanism through which QE has real effects on the economy in that QE directly increases inflation expectations. Market participants see the increase in money supply and believe the real purchasing power of their dollars will be lower in the future, which is another way of saying they expect higher inflation. This perception usually does manifest itself in a weaker currency, but this may not happen, particularly if other CBs are also engaging in substantial QE that is typical in a global economic slowdown. Thus, the most obvious effect of QE is higher inflation expectations within a country. The expectations channel is better able to explain the seeming contradiction between QE taking out the supply of government bonds and yet the fact that yields rise.

Higher inflation expectations can lead to higher bond yields directly, but also indirectly through an asset allocation framework. If you think that inflation expectations are going to be higher you shift your portfolio away from fixed-rate instruments like bonds and instead buy assets that offer some form of inflation protection like stocks. I do not mean to say that the Fed’s treasury purchases have no impact on yields, but even after more than a decade of “unconventional” monetary policy, the Fed still only owns about a third of Treasuries that mature >5 years. The reality is that an extremely important global asset like US treasury yields is not simply influenced by Fed demand, but expectations for inflation, real growth, productivity, demographics, trade flows, etc., and in general the expectations component of QE overwhelms any reduction in yields from lower net supply (outside of crisis environments).

If the main way that QE works is through the expectations channel rather than through the portfolio balance channel, it has a profound impact on the way that the Fed should view the present situation. One argument commonly given for ending QE is we do not need to worry about a drastic rise in yields given the starting point for yields is so low. I agree, ending QE is not likely to cause yields to rise drastically, but the risk is the opposite happens i.e. as QE ends, yields could fall as inflation expectations drop as well. If the Fed truly wants credibility on their inflation target than maintaining QE is a much easier way to obtain it.

The biggest pushback to the idea that QE directly influences inflation expectations is the case of Japan and Europe. Although, I would say at the extremes the argument still holds true. For example, imagine if the Ministry of Finance of Japan issued a 565trn Yen (nominal GDP of Japan) 0% perpetual bond and the Bank of Japan said they would print money to buy the whole note. This is a permanent increase in the money supply because it is perpetual (never needs to be repaid) and carries a 0% coupon. Suppose the Ministry of Finance went on a spending binge or put the money in Japanese citizen’s bank accounts. Is there any doubt that the purchasing power of the Yen would fall dramatically, and the country would experience inflation?

QE after a Balance Sheet Recession and the problem with Japan/Europe

There are two ways that money supply can increase, and each has a different impact on the economy. The most powerful is the bank lending channel. In a fractional reserve banking system, banks do not need to back their deposits with 100% cash and instead can lend out most of this capital. As a result, when citizens put cash into the bank as deposits, the money supply multiplies as bank lending increases and spending finds its way back to the financial system as more deposits. Bank lending is the most powerful way that money supply can increase as it tends to support economic activity directly. In the US, this was the primary mechanism of money creation prior to the GFC.

Now let’s imagine a world where banks are not seeing loan demand, need to recapitalize and regulators push to improve the liquidity of their assets. In this world, the asset side for banks is changing to be much more heavily focused on securities. As loans are repaid, this capital is not deployed into new loans but instead used to purchase securities. If the CB does QE it still increases money supply, but it is simply replacing the security assets on a bank’s balance sheet with cash that ends up getting deposited back on the central bank’s balance sheet as excess reserves.

It is for this reason that QE did not increase inflation expectations in Japan for many decades. After the real estate crash in Japan in 1989, many of the country’s banks were technically insolvent, but given the bankruptcy laws within the country and support from the government, they did not file. However, despite not declaring bankruptcy, bank managers were clearly not in the mood to extend credit to new businesses via loans. Instead, they hoarded securities particularly Japanese government bonds (JGBs). When the BOJ initiated QE they simply replaced the securities with cash on the bank’s balance sheet and the banks then hoarded this cash with little effect on the real economy.

This behavior is typical after a particularly insidious type of recession called a Balance Sheet Recession (BSR). A BSR occurs after a dramatic fall in asset prices after many years of debt growth. In Japan, the BSR affected the corporate sector; in the US it affected the household and financial sectors after the GFC. An entity might be technically insolvent through its balance sheet (assets might be worth less than debt) but may be able to survive as it continues to use cash flow to pay interest expense and rollover debt. In a BSR, low interest rates have the ability to reduce the interest burden, but generally do not stimulate investment or further expansion because debt capacity has been reached. At the same time the fiscal multiplier is very low, because transfers are used to retire debt rather than for investment or consumption. ?

I do not mean to be flippant and suggest that QE is the only thing that matters for inflation expectations. In fact, after a BSR the fiscal side is as important as monetary policy in keeping inflation expectations elevated as the government is the only sector willing to expand debt and keep the overall system from deleveraging. Total credit (which is extremely difficult to measure given shadow banking) is much more important than even money supply in determining economic outcomes. The chart below shows non-financial debt + household debt + government debt as a % of US nominal GDP. As you can see, debt levels increased after the GFC, which is a stark difference between what the US experienced after the Great Depression compared to Japan after its housing crash in the late 1980s. In fact, one needs to view inflation after the GFC in context. Many commentators mock the idea that inflation is a monetary phenomenon and use the post-GFC experience to argue their case i.e. money supply increased dramatically without substantial inflation. I would argue the opposite, it is astounding and a credit to monetary and fiscal policy that we did not have outright deflation after the GFC when home prices fell 26% nationally, over 1mm homes were in foreclosure and the financial system was basically bankrupt.

US Total Debt to GDP

Source: Bloomberg

There are another two reasons that the US was able to maintain higher inflation after the GFC than both Japan and Europe. We now take it for granted that European rates should be substantially lower than the US, but this was not the case as recently as Q4 2011 when 10y German bunds actually yielded more than the 10y US Treasuries. From there, however, German bund yields sank dramatically relative to the US. What explains this relative performance is that unlike the US, the European economy engaged in fiscal deleveraging while they were still recovering from a BSR. As concerns over peripheral debt loads grew throughout 2011, bond yields on many of these countries spiked. As a result, European policymakers started to clamp down on peripheral fiscal spending, and overall credit in the system declined causing inflation expectations and nominal bond yields to collapse. The US also started to rein in fiscal spending with the Budget Sequester Act of 2011. However, the reason that total debt in the system did not decline is because the non-financial corporate sector used the low nominal yield environment to increase their debt levels dramatically. This prevented the overall system from deleveraging in a way that was absent in Europe.

The third reason that both European and Japanese inflation expectations and realized inflation were lower than the US is because of the negative policy rates they instituted. Both Europe and Japan have bank-dominated lending channels and negative front-end policy rates are particularly damaging for the banking system who were already struggling given their holdings of peripheral debt and the downturn in the housing cycle. Negative rates hurt the bank’s earnings power and prevented them from extending lending, which as I discussed earlier is the more powerful form of money supply increase. The irony is that the US would actually benefit from a negative interest rate policy in recessions. One of the reasons the Fed was inclined to raise the policy rate last cycle was because they wanted to create policy space to cut rates to 0% in the event of a recession. They saw the experience that Europe and Japan had with negative rates and viewed a 0% Fed Fund Rates as their lower bound. Unlike Europe, the US has a capital-markets dominated lending channel (bond and equity issuance) with banks having a much smaller role in credit creation particularly after the GFC. While negative rates hurt banks, it does facilitate borrowing via the capital markets as companies use low rates to relever their balance sheets.

QE’s likely effect on the US dollar

A major advantage with maintaining QE is that it will likely keep the dollar weak. The US dollar is the most important financial asset in the world, not only because it is the world’s reserve currency, but also because there is a vast Eurodollar system dependent on the US maintaining global liquidity. The lack of dollar liquidity was one of the root causes of the market reaction to COVID. The Fed had orchestrated relatively tight monetary policy heading into COVID with its balance sheet at $4.2 trillion, the same level it had been in March 2014. When COVID initially struck and global investors sought US dollars, they needed to liquidate their assets to raise cash. Foreigners own about 61% of GDP worth of US assets, and this selling is one of the reasons that COVID turned into a liquidity crisis that necessitated the Fed reopening swap lines to foreign central banks. While I do not think that it is the responsibility of the Fed to foster global growth, it is na?ve to think the US can remain immune to a slower global economy. The last cycle (GFC to COVID) was marked by a strong US dollar, weak global growth and also weak US growth.

There are other advantages to a weak dollar policy. Flush with superfluous fiscal spending, American consumers wallets are brimming, and the US is currently running a $85bn/month goods trade deficit, close to its largest ever. The only other time in American history that we ran similarly large current account deficits was in the mid-2000s. It was an environment where strong consumer spending masked underlying fragilities. Current account deficits (a country’s purchases in excess of its income) are not necessarily a bad thing if the spending is focused on investment that will enable the country to raise its productive capacity and eventually pay off its creditors. If the current account deficit is excess consumption, it amounts to simply pulling forward future consumption and lowering long-term growth and productivity. Also, when you run a current account deficit you need to finance this externally either short-term (portfolio flows) or long-term (Foreign direct investment), in the case of the US this has been the former, which is risky and creates fragility in US asset prices as was seen in March of 2020.

Fixing the US current account is not easy. In a world where trade was completely free and open, there would not be large current account deficits (US) or surpluses (Germany, China). If a country consumed more than it produced than it would experience higher inflation, which in turn would weaken its currency, making imports more expensive, reducing demand for them and balancing the current account deficit. Unfortunately, we live in a world where the countries with significant current account surpluses are not rectifying internal imbalances by fostering enough domestic consumption. In the case of Germany, the country subsidizes its export industry via the underpayment of labor at the same time the country benefits from being in the Euro area with a currency that is much weaker than if Germany was still on the Deutsche Mark.

Before the GFC, China had an artificially cheap currency that allowed it to run historic current account surpluses in an attempt to rapidly move up the value chain (agricultural to industry) and in the process accumulated a record amount of US financial assets. This savings glut suppression of long-term US interest rates is arguably one of the main causes of the US housing boom and the GFC thereafter. While China’s currency may no longer be undervalued, they are still subsidizing industry relative to consumption by restricting citizens from holding assets abroad and locking them into the Chinese financial system where returns offered are well below the level of inflation.

Without other countries willing to rectify problems with their domestic economies, it seems very reasonable for the Federal Reserve to target a weaker US dollar to help US industry. There is no law that says as a country gets richer, more of its GDP needs to come from the service sector. Germany is a testament to how this is not true, and it is easier to generate productivity in manufacturing relative to services. While you can get some productivity from services the pace is much slower; think of the difficulty in increasing productivity for a waiter relative to a worker on an assembly line looking over a process and making continuous improvements.

The Fed would surely push back on a role that involves currency management as they leave this to the Treasury department. This is not consistent with other countries (ECB commonly talks about the value of the EUR), the Fed’s history (Fed policy in the 1920s was based on helping the UK return to the gold standard) or their dual mandate (the currency is one of the most important factors in inflation and global growth). Besides, the Fed controls the supply of US dollars, the Treasury’s power to control the price of USD through words and moral suasion pales in comparison.

Conclusion

The QE of Treasuries provides key advantages that can help the US economy by maintaining inflation expectations and potentially weakening the US dollar. However, while I believe that the Fed should keep QE of Treasuries, it should end its purchases of MBS for reasons we will discuss in further notes. Most importantly, the Fed should be raising interest rates immediately. In the next essay, I will discuss the mistakes the Fed has made with its interest rate policy over the past 30 years and how that has made the economy structurally weaker.

Important Disclosures

For professional, institutional and accredited investors only. The views expressed are those of the author and are subject to change. Other teams may hold different views and make different investment decisions. While any third-party data used is considered reliable, its accuracy is not guaranteed. Forward-looking statements should not be considered as guarantees or predictions of future events. The value of your investment may become worth more or less than at the time of original investment. Past results are not a reliable indicator of future results. Commentary provided should not be viewed as a current or past recommendation and is not intended to constitute investment advice or an offer to buy or sell securities. Wellington assumes no duty to update any information in this material in the event that such information changes.