Minutes from the Federal Open Market Committee’s (FOMC) March meeting confirmed that the Federal Reserve (Fed) has been actively discussing the potential reduction of their balance sheet. Some Fed officials have mentioned the possibility of selling securities outright, but outright sales could cause a disruption in the market, which the Fed explicitly is trying to avoid. For this reason, most of the discussion on reducing balance sheet size has instead been around the idea of ending the Fed’s current policy of reinvesting the proceeds of maturing bonds. This method would not be completely market neutral, given that it would remove a large buyer from the market sometime later this year, but would likely be less disruptive than outright sales. Current Fed holdings (as of April 6, 2017) total $4.5 trillion, with the majority of the holdings composed of Treasury bonds, $2.5 trillion, and mortgage-backed securities (commonly known as MBS), where holdings total nearly $1.8 trillion [Figure 1].
Impact of Fed policy on mortgage-backed securities market
We discussed the potential impact of balance sheet normalization in this week’s Weekly Economic Commentary, but wanted to shift the focus to the potential impact on MBS as well. The Fed originally started purchasing mortgage-backed securities following the financial crisis in order to push prices higher and yields lower in support of the battered sector. Over time, and several rounds of quantitative easing (QE), the Fed purchased nearly 20% of the outstanding MBS market ($1.8 trillion of $8.9 trillion outstanding at the end of the fourth quarter of 2016 per SIFMA data). Total issuance in the MBS market has averaged a little over $1.8 trillion per year over the last 5 years, so if the Fed were to sell $1.8 trillion into the market at once, there would likely be a major impact, causing prices to fall due to a massive increase in supply. Since the Fed does not want to cause a market impact of this type it is unlikely to take this route.
The more likely way for the Fed to shrink its balance sheet would be to no longer reinvest principal when bonds mature. For a normal bond, interest payments are received over the life of the loan, and the full amount of principal is paid back at maturity. However, mortgage-backed securities are different than most high-quality bonds in that when income payments are received, they include both principal and interest (just as most homeowners pay both principal and interest each month when making their mortgage payments). Additional principal payments can also come when homeowners pay off their existing mortgages whether by refinancing due to lower rates, or due to selling their homes and purchasing new ones. This means that even though the majority of the Fed’s MBS holdings don’t technically mature for more than 10 years, the impact to markets could be felt sooner, though still at a much slower rate than if the Fed were to sell securities outright.
Overall the fundamental impact of stopping reinvestment of mortgage-backed securities proceeds may be relatively small in the near term. However, we don’t discount the potential for markets to react emotionally ahead of time, which could put pressure on MBS if news of actual steps toward normalization began.
Yields still attractive relative to interest rate risk
One of the key reasons we continue to believe MBS are attractive relative to other high-quality bonds is that they offer the potential for more yield per unit of duration (interest rate risk) than comparable quality offerings. Duration has increased from when we originally upgraded our view of the sector in September 2016, but yields have risen as well, so MBS still offer more price protection in a rising rate environment, even if the difference isn’t as stark as it once was.
A general rule of thumb is that if rates rise by 1%, an investor can expect the price of their bond holdings to decrease at a level consistent with its duration in percentage terms. For example, if a bond has a duration of 5 years, it could be expected to drop approximately 5% if rates rose by 1%. There are of course complications in the math that don’t make it this clean in reality, but this generalization can give investors a broad idea of how their bond holdings may perform.
Investors like to be compensated for risk and in the bond market this means receiving more yield (paying a lower price), if risk levels are higher. To measure the yield per unit of interest rate risk, we can divide the yield by duration. A higher number indicates that a bondholder is receiving more compensation per unit of risk. Using this measure, Figure 2 shows that MBS currently offers 0.13% more yield per unit of duration than higher yielding, but longer duration corporate bonds, 0.27% more than Treasuries, and 0.15% more than the Bloomberg Barclays Aggregate Bond Index.
The Fed’s discussion of plans to end reinvestment of principal payments from maturing bonds is on balance a negative development for the bond market, though the fact that the Fed doesn’t want to create market volatility means that the impact may be less severe than some may fear. For the MBS market, the current structure of the Fed’s balance sheet, with more than 99% of Fed’s nearly $1.8 trillion in MBS holdings maturing in more than 10 years, likely means that any fundamental impact to the MBS market would be gradual. We do caution that the forward-looking nature of markets means mortgage-backed securities performance could start to falter if markets begins to believe that the Fed is ready to move. However, Treasuries and, to a smaller extent, the broader bond market share this risk, and MBS still offer the potential for additional yield per unit of interest rate risk, leading us to believe that MBS continue to offer relative value.
Article by LPL Financial