Ticker Symbols

TGLMX (I Share) TGMNX (N Share)

Investment Advisor

TCW Investment Management Co

I Want Information

Mortgage Centric and Financial Cycle Aware

Dec 16, 2019

TCW Total Return Bond Fund

  • AUM

    $5.83 billion

  • Inception Date

    Jun 17, 1993

  • Portfolio Holdings

     

  • Portfolio Turnover

     

Q: How has the fund evolved since inception?

A: The TCW Total Return Bond Fund is an intermediate-term fund that has focused on securitized mortgages since its inception in the early 1990s. It is somewhat unique because it has never had exposures to traditional corporate securities—and we have no intention of changing that mandate.

Part of the reasoning behind this mandate comes from our analysis of how securitized mortgages perform versus other types of investments. Over the last 25 to 30 years, both corporate bonds and agency mortgages beat a duration-matched portfolio of Treasuries. On average, the corporate index is significantly more volatile, while the mortgage index has produced slightly lower excess returns. However, when adjusted for risk, we found that the excess returns of mortgages looked particularly attractive over the long term.

About 10 years ago, we began to include other types of securitized assets. The fund specifically invests at least 50% of assets in agency and non-agency mortgage-backed securities as well as government securities like money markets and Treasuries. It can have very high exposures relative to its benchmark, the Bloomberg Barclays U.S. Aggregate Bond Index. For example, today the fund has over 50% of its assets in agency mortgage-backed securities alone, versus the benchmark’s exposure of 27%.

Currently, the fund has approximately $5.8 billion of assets under management. Firmwide, TCW has over $80 billion in securitized products under management.

Q: What are the underlying principles of your investment philosophy?

A: We are alpha-focused while remaining keenly aware of the cycle as well as deeply knowledgeable about the benchmark. Fundamental valuation is of utmost importance to us, and we believe prices are generally more volatile than fundamental value, providing alpha-generating opportunities. The fixed-income market tends to be mean reverting with cycles of various lengths affecting interest rates, sectors and individual securities.

So, rather than trying to time the market for interest rates, we instead attempt to assess where we are in the financial cycle and use dollar-cost averaging for positions and sectors. Although diversification in the fund is also essential, we won’t invest in securities that we believe are fundamentally rich or fully valued simply for the sake of diversification.

Another core principle is our belief that mortgage-centric assets have specific characteristics that make alpha generation easier to achieve. Cash flows within this asset class have a wide range of alternative outcomes. Modeling these complex cash flows and their potential returns requires expertise, technology and a good deal of infrastructure, which is where our proprietary technology and tools adds real value.

Our alpha expectations are not constant over the cycle. Generally, we believe it is easier to achieve early in the cycle, and we will become more defensive later in the cycle.

Q: Would you describe your investment process?

A: Investing in a mortgage-centric fund is somewhat different than investing in a traditional multi-sector fund. With securitized products, each security is completely unique with highly individualized potential cash flows. Securities could be backed by just a few loans or by thousands, and they can have different underlying borrower profiles, loan characteristics, and cash flow waterfalls including various structures for credit enhancement or an agency guarantee.

In this fund, both security analysis and portfolio management are team-managed, though with some specialization within mortgage subgroups. Investments are based on a framework positioned relative to the index. Because we are mortgage centric, our exposures to agency mortgage-backed securities tend to be overweight versus the index, but within that overweight we invest in the sectors, coupons, and issuers with the most relative value.

Our investment process is bottom-up with asset selection focused where we believe alpha will be generated given the current positioning in the financial cycle. It is also stochastic, meaning that when we think about the interest rate process and moves in credit spreads, we look at a wide range of potential outcomes.

Position sizing is determined by a host of factors. If an asset is in the index, we look at how much exposure we already have relative to internal targets based on relative value. Liquidity is another consideration. The majority of the fund owns highly liquid agency mortgage securities as well as Treasuries, for example—while a small portion of the fund’s assets can be liquid in good times and less so at others, represent excellent fundamental value. Our goal is to make sure the portfolio is well-diversified with a specific focus on less-liquid assets.

Q: What is unique about evaluating a mortgage security compared to a corporate security?

A: Unique to mortgage security analysis is that mortgages tend to be negatively convex, so when rates are particularly volatile, they tend to underperform non-callable bonds. Falling rates typically lead to an increase in prepayments and shorter average lives for mortgage securities; the converse occurs when rates rise. That’s the nature of selling volatility. The good news is that investors are rewarded for selling volatility through receiving a yield spread achieved over government securities with similar credit quality.

Borrowers have the right to prepay a mortgage any time, and often do so for a host of reasons, so our job as investors is to thoroughly understand how a pool of borrowers might prepay mortgages under various rate environments, as well as under various economic scenarios. For example, when home prices go up rapidly, people have a much easier time prepaying and refinancing their mortgages because they have more equity in their properties. Should the opposite happen, as it did during the great financial crisis, we expect to see an increase in defaults, but a reduction in voluntary prepayment of home loans.

Along with understanding prepayments, we must also look at what’s happening with interest rates and adjust our models if needed. For example, during most of my 35-year career in mortgage investing, the market assumed that a sustained negative interest rate environment wasn’t actually possible. Yet today, central bankers around the world have made negative interest rates a reality, so we adjusted our models for stochastic interest rate processes accordingly.

Q: How does your research process work?

A: Pools of mortgages can often look a lot alike on the surface—they could have the same issuer, the same 3.5% coupon and the same 30-year maturity in common. Our proprietary tools allow us to discern value among similar-looking investments by providing us with real-time information, analyzing pools of mortgages on a loan level basis.  Because each type of securitized asset is so unique, we developed separate models for agency mortgage securities, commercial mortgage-backed securities and non-agency residential mortgage-backed securities. Regardless of asset type, our thought process remains the same: we pick securities that we believe have cheap fundamental value relative to the market.

Currently, there are 100 million securitized single-family loans in the U.S. Of these, over 60 million are in agency pools, either Fannie Mae, Freddie Mac, or Ginnie Mae. The remaining are non-agency securitizations. We gather monthly information on all these pools, then we look within at individual securities—and there are hundreds of thousands of individual CUSIPs. By simply entering a CUSIP, our portfolio managers can access details about the loans backing a pool.

Because understanding how pools might prepay given various interest rate and economic scenarios is crucial to determining whether an agency MBS is a good value, we analyze many factors to determine what borrowers will do when rates, home prices, or other economic indicators move. For instance, we might look at borrower’s FICO score, geographic location, the diversity in different regions of the country, or a pool’s average home equity as determined by current loan-to-value (LTV), or the average size of its loans at origination.

However, what’s most important to valuation isn’t often found in the averages, but rather in the distribution of a variety of a loan’s characteristics. Say the borrowers’ average FICO score is 700, but in looking at the pool, we see a barbell chart. Its tail represents the few loans from borrowers having high credit ratings, but a significant majority of borrowers have far lower ratings. Tails like this tell us a lot about how that pool may prepay under different rate environments, as those loans maybe most sensitive to moves in changing economic environments.

Loan balances, too, provide information about prepayments. A loan for $75,000 will prepay much less efficiently than a $500,000 loan. Not only would we expect on average the borrower of the smaller loan to be have less sophistication, but more importantly, the loan’s fixed costs like title and appraisals costs can be much higher as a percentage of the balance. A loan’s originator and servicer can also help us evaluate how it might prepay. Our technology tracks and ranks them over time, so we know which servicers are aggressive in refinancing and which are less so.

Q: How do you deal with default situations wherein the real title owner is unknown?

A: This was a significant issue early in the recovery. When the market fell precipitously, home prices dropped dramatically, borrowers defaulted at very high rates, and many title issues arose. In fact, in a lot of cases, the originators were sued. Since then, these lawsuits have largely been settled, with cash flows going back to borrowers as well as to some of the security holders—and we received additional profit from some settlements.

We know which securities still have potential lawsuits pending, the ones in which recovery is anticipated, and how much those are likely to be in terms of points on the loan. The faulty titling issues have been resolved, so what’s left are the borrowers who have been making payments for over 10 years now. These legacy non-agency mortgages originated in the height of the housing market and have positively self-selected borrowers at this point. At its peak, the securitized market for this legacy non-agency mortgage market was about $2 trillion; today it is approximately $450 billion.

Q: How is your team organized? Who has the final say in the decision-making process when selecting a security?

A: The fund is team managed with a generalist/specialist structure. Four generalist portfolio managers act like a board of directors and guide the specialists who head each sector. The generalists review all strategies and portfolios, determine risk budgets and sector allocations, then dictate these to the sector specialists.

It is up to as the sector specialists to make decisions about individual CUSIPs and meet targets provided by the generalists. We believe that the most knowledgeable person in a particular area should be the one who makes individual security selections. However, that individual receives guidance related to the overall risk from the generalist, as well as guidance regarding where value can be found from the sector specialists.

Q: Can you illustrate your portfolio construction process?

A: The key is the consistent implementation of our guiding principles: being benchmark aware, alpha focused, and dollar cost averaging. In part, construction is dictated by fundamental valuations. When a particular security or sector looks cheap, we will buy and overweight. If it gets cheaper, we will continue to dollar-cost the average.

Obviously, it’s extremely important to understand what’s available in the marketplace, at what price and where similar securities have been trading. There are thousands of pools of agency mortgages offered each day across Wall Street. We collect this inventory daily from dealers and use proprietary tools to bring together relevant data and make it accessible by the team. We also buy bonds in what is called Bid Wanted in Competition (BWIC), which is basically an auction. BWICs are also tracked in real-time and trade information is kept historically by our system.

Typically, the fund has hundreds of positions; mortgage securities tend to prepay over time rather than all of a sudden as would be the case with a corporate maturity. Our bottom 100 positions can be quite small, while the top 20 to 50 are relatively much larger.

Q: How do you define and manage risk? What kind of risk do you primarily focus on?

A: Volatility and alpha go hand in hand. They are not constant, and neither are our expectations for them. We view alpha as a risk-adjusted measurement that helps us understand the potential volatility in portfolio and security values while we try to achieve a measure of excess returns over a full cycle.

As one way to measure   relative portfolio performance and alpha produced, we look at the standard deviation of excess returns over time.   Late in the financial cycle, we would anticipate less alpha and less volatility as our positioning in portfolios would become more defensive.

Q: What lessons did you learn from the financial crisis of 2008–2009?

A: The most significant change that has taken place in fixed income investing post the Great Financial Crisis (GFC) is greater involvement by government, specifically by the global central banks. Quantitative easing since the GFC—the magnitude of which has never been seen historically—has been a material factor in the economy’s long, but slow recovery. Ultimately, the cycle has been lengthened tremendously, but at the expense of slower growth, an enormous amount of new global debt and a greater exacerbation of the wealth gap.

Although equity prices have risen in tandem with central bank balance sheets, they have increased much faster than incomes. Asset prices have become decoupled from reality, increasing much more quickly than GDP. Fed policies have kept rates artificially low, lifting price-to-earnings multiples in equity markets and yet profits have stagnated. Low rates have stimulated a boom in global debt and we have seen a slowdown in corporate revenue, which strongly indicate that we are late in the financial cycle. We’ve watched yield curves flatten and invert; an astounding 27% of bonds around the world now are trading at negative interest rates. I believe this suggests we are pushing into policy that makes little economic sense and is unlikely to provide the results that central bankers’ desire.

The reality is that recessions are not optional. However, when a recession is delayed by the issuance of massive government and corporate debt, there is far greater risk that it will be much more severe.

Annual Return 2019 2018 2017 2016 2015 2014 2013 2012 2011
TGLMX 7.27 0.80 3.44 1.53 1.08 5.80 1.66 13.41 4.13