After the housing bust, the Fed decided to help homeowners by keeping mortgage rates low. To implement this policy, it embarked on a $40 billion per month purchase of mortgage-backed securities (MBS). This drove the prices of MBS higher, resulting in record low mortgage rates. However, in June of 2013, Bernanke announced that the bond buying would be reduced. This caused the MBS market to sell off, but in September, the Fed decided to hold off on scaling back purchases, resulting in a rebound in bond prices. The MBS buying program was finally terminated in October, 2014.
Before I delve into the analysis, I will quickly review the characteristics of mortgage-backed securities. Many years ago, most mortgages were issued by your local bank, which serviced this mortgage over the life of the loan. Then someone had the idea that banks could sell these mortgages to another company, who could bundle these loans into packages that could be sold to the public. This process was called securitization. The mortgages were sliced and diced into "shares" of the mortgage pool, with each share receiving a portion of the interest and principal as the underlying mortgages were repaid. These shares were referred to as mortgage-backed securities. The majority of MBS are pooled together by government-sponsored enterprises (NYSE:GSE) such as Ginnie Mae, Freddie Mac, and Fannie Mae and are called agency MBS. If the MBS is sponsored by a private company, it is called a non-agency MBS.
Mortgage-backed securities are like bonds since they have a coupon or interest rate, but they are different in that the principal is repaid incrementally. Since mortgages may be paid off early, mortgage securities are tracked in terms of their "average life" rather than a fixed maturity date.
There are many types of MBS. The simplest is called a pass-through participation certificate, where the principal and interest paid by the home buyer passes through to the MBS holder. Other, more complex structures are collateralized mortgage obligations (CMOs) and collateralized debt obligations (CDOs). These more complex structures usually allow investors to buy specific slices or tranches of the mortgages to satisfy specific investment objectives. These subprime tranches were one of the major drivers of the 2008 financial meltdown.
To determine value, you would need to model the probabilities of prepayments and defaults, which is complex, especially for retail investors. Luckily, rating agencies -- such as Standard & Poor's and Moody's -- came to the rescue and rated these MBSs, similar to the way they rated bonds.
MBS are important to the economy. They allow the banks to reduce risks by offloading mortgages to buyers, typically an investment bank. This frees up funds in your local banks and permits them to offer more loans to borrowers. The MBS market now exceeds more than 7 trillion dollars and makes up about 12% of Barclay's Global Bond Index.
To asses this asset class, I analyzed the risk and reward associated with MBS Closed End Funds (CEFs) over the past few years to determine the "best" investment. There are many ways to define "best." Some investors may use total return as a metric, but as a retiree, risk is as important to me as return. Therefore, I define "best" as the fund that provides the most reward for a given level of risk and I measure risk by the volatility. Please note that I am not advocating that this is the way everyone should define "best"; I am just saying that this is the definition that works for me.
There are 9 CEFs that invest primarily in MBS. However, many of these are small and illiquid. To reduce the trade space, I selected candidates based on the following criteria:
- At least a five year history (MBS CEFs are relatively new so only a few have histories greater than five years)
- Market cap greater than $100 million
- Average trading volume greater than 50,000 shares per day.
Only the following three CEFs satisfied all of these conditions:
- BlackRock Income Trust (NYSE:BKT). This CEF trades at a 7.4% discount (which is smaller than its 5 year average discount of 8.8%) and has a distribution rate of 5.8%, funded primarily by income with only a small Return of Capital (NYSE:ROC) component. The portfolio holds 131 securities, most of which are agency pass-through (72%) and Agency CMO (21%). The fund utilizes 28% leverage and has an expense ratio of 1%. Over the past 5 years, the fund has been profitable except for 2013, when it broke even in terms of Net Asset Value (NYSE:NAV) but lost 5.6% in terms of price.
- Western Asset Mortgage Defined Opportunity (NYSE:DMO). This CEF trades at a premium of 4.6% which is well above the 5 year average discount of 1.7%. Over the past 5 years the fund has traded at both a discount and a premium, with the premium reaching as high as 8% and the discount reaching 8%. The fund sold at a discount for most of 2015 but this turned into a premium in late November. It holds 324 securities, spread primarily between non-agency residential mortgages (51%) and asset-backed bonds (28%). About 99% of the holdings are below investment grade. The fund utilizes 31% leverage and has a 2.2% expense ratio. The distribution is 10.6% and is funded by income and capital gains, with no ROC. The fund lost only 0.4% in price in 2011 and has been profitable for every year since.
- Nuveen Mortgage Opportunity Term (NYSE:JLS). This CEF currently sells at a discount of 9.5%, which is a larger discount than the 5 year average discount of 5.3%. The distribution is 6.7%, which includes a small amount of ROC (about 8% of the total distribution). The fund has 204 holdings, primarily in non-agency residential mortgages and commercial mortgages. This fund was launched on Nov. 25, 2009, and has a limited term of 10 years, at which point all assets will be distributed to the shareholders of record. About 92% of the holdings are below investment grade (with 60% below B). The fund utilizes 27% leverage and has an expense ratio of 2.2%. The fund lost 12% of price in 2011 (7% in NAV) and also loss 5% in price in 2013 (but NAV gained 8%).
For reference I also included the following Exchange Traded Funds (ETF) in the analysis.
- iShares Barclay MBS Bond Fund (NYSEARCA:MBB). This ETF tracks the Barclay Capital U.S. MBS Index, which reflects the performance of investment-grade mortgage backed pass-through securities issued by GSEs. This fund yields a 2.1% and has an expense ratio of 0.29%. This will be used to compare MBS CEFs with MBS ETFs.
- iShares Barclays 20+ Year Treasury Bond (NYSEARCA:TLT). This ETF tracks the performance of long term Treasury bonds and yields 2.6%. It has an expense ratio of 0.15%. This will be used to compare the MBS funds to long term Treasury funds.
- SPDR S&P 500 ETF (NYSEARCA:SPY). This ETF tracks the S&P 500 index. This is not a good comparison metric for MBS but it will show how MBS compares with the general stock market in terms of return and volatility.
To analyze the risks and return of MBS funds, I used data over the past five years (1/7/2011 to 1/7/2016). The results are shown in Figure 1, which plots the rate of return in excess of the risk-free rate of return (called Excess Mu on the charts) against the historical volatility. To make comparisons easier, I assumed zero as the risk-free rate.
Figure 1. Risk Versus. Reward Over Past Five Years
As is evident from the figure, MBS funds had a wide range of returns and volatilities. DMO had the greatest return but also the largest volatility among the CEFs. Was the DMO return commensurate with the increased risk? To answer this question, I calculated the Sharpe Ratio.
The Sharpe Ratio is a metric developed by Nobel laureate William Sharpe that measures risk-adjusted performance. It is calculated as the ratio of the excess return over the volatility. This reward-to-risk ratio (assuming that risk is measured by volatility) is a good way to compare peers to assess if higher returns are due to superior investment performance or from taking additional risk. In Figure 1, I plotted a red line that represents the Sharpe Ratio associated with MBB. If an asset is above the line, it has a higher Sharpe Ratio than MBB. Conversely, if an asset is below the line, the reward-to-risk is worse than MBB. Similarly, the blue line represents the Sharpe Ratio for TLT.
Some interesting observations are apparent from Figure 1.
- Among all the funds analyzed, MBB had by far the lowest volatility but also the lowest return. Taking these together, MBB had an excellent risk-adjusted return that was better than all the other assets except DMO.
- DMO had both the highest absolute return coupled with the best risk-adjusted return. Thus, DMO was the clear winner over the 5 year period, with a risk-adjusted performance that even beat SPY.
- DMO had a volatility slightly lower than SPY and TLT.
- All the MBS funds had lower volatility than SPY and TLT.
- Both BKT and JLS had risk-adjusted performance worse than MBB. These CEFs had a risk-adjusted performance similar to TLT.
In addition to risk-adjusted returns, I also want to assess if adding MBS funds to an equity-focused portfolio would provide diversification. To be "diversified," you need to choose assets such that when some assets are down, others are up. In mathematical terms, you want to select assets that are uncorrelated (or at least not highly correlated) with each other. Therefore, I calculated the pair-wise correlations associated with these MBS funds. The results are shown in Figure 2.
Figure 2. Correlation Matrix Over Past Five Years
As you can see, none of the assets were highly correlated with each other or with Treasuries. MBB had the highest correlation with SPY with Treasuries but that was only about 68%. So, overall, mortgage-backed securities provided excellent diversification, both to Treasury and equity portfolios.
I next wanted to see what would happen to the MBS CEF performance if I reduced the look-back period. I re-ran my analysis with a 3 year look-back period and the results are shown in Figure 3.
Figure 3. Risk Versus. Reward Over Past Three Years
The results were similar except that SPY was in a bull market and outperformed most of the funds with the exception of DMO. DMO still provided outstanding performance on both and absolute and risk-adjusted basis. MBB outperformed the BKT and JLS. BKT had about the same risk-adjusted performance as TLY but JLS lagged.
The final analysis reduced the look-back period to 12 months (1/7/2015 to 1/7/2016). The results are shown in Figure 4. SPY, with the terrible start to 2016, has not fared well over the past 12 months. Neither has TLT, but the MBS still continued to shine. All the MBS funds has positive returns and DMO was still the best of the lot.
Figure 4. Risk Versus. Reward Over Past Twelve Months
In summary, MBS funds have performed well over the past 5 years. Clearly the best performer has been DMO, so it is not surprising that it is selling at a premium. However, I do not like buying at a premium and if history repeats, DMO will sell at a discount sometime later this year. Thus, my strategy would be to keep an eye on DMO and buy as soon as the premium disappears. But make no mistake, MBS funds have relatively high volatility (about the same as the stock market) and if you decide to add them to your portfolio, they will need to be managed carefully.
Disclosure: I/we have no positions in any stocks mentioned, but may initiate a long position in DMO over the next 72 hours.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.