Examining Bond Portfolio Trends: Second Quarter 2020
Beginning in May 2012 we started tracking portfolio trends of our bond accounting customers here at the Correspondent Division of CenterState Bank. At present, we account for over 128 client portfolios with a combined book value of $11.4 billion, or $89 million on average per portfolio. Twelve months earlier the average portfolio size was $65 million. For the past several years there was little to no growth in average portfolio size due to solid loan demand. Thus, the increase noted in the first half of 2020 speaks to softening loan demand as the pandemic hit hard in the second quarter. That trend will likely continue in the months ahead as we grapple with the economic aftershocks caused by the virus.
The pandemic continued to dominate market action in the second quarter with the most volatile moves occurred in early April and June with yields spiking but other than those two periods the quarter was one long sideways trade with yields severely range bound. Investors knew April would be awful from an economic perspective but the yield lows of March 9th held during the second quarter. The 10-year traded as low as 0.30% on that day before closing at 0.54%. While yields flirted with that closing low time and time again, it was never bested until early August. With the Fed at the zero-lower bound on fed funds, and the longer-end facing halting reopenings and fading stimulus, the low yield levels may yet be challenged in the third quarter.
Illustrating the range bound nature of the Treasury market in the second quarter, two-year yields fell 6bps ending June at 0.15%, right on top of fed funds. Meanwhile, 10-year yields fell only 1 bp to 0.66%. That outperformance by shorter-term Treasuries bull-steepened the 2yr-10yr Treasury spread from 42bps as the quarter began to 47bps at quarter-end. With the economic bounce already struggling, yields are likely to be at these ultra-low levels for awhile. We continue to believe it will be a fitful and slow recovery that is likely to keep the fed funds rate anchored at zero and longer-end Treasuries sporting 0-handle yields through this year and most likely into early 2021.
With that backdrop, lets look at how portfolio allocations and performance have changed over the past year.
Let’s begin by revisiting allocations a year ago, as shown in the pie chart to the immediate right. The MBS/CMO sector comprised 54.57% of portfolios, municipal allocations stood at 22.70%, Agency/Treasury investments were at 18.77% and the “Other” category (CDs, corporates and other floaters) was 3.96%.
Fast forward one year to June 30, 2020. The MBS/CMO sector comprised 57.06% of portfolios for a 2.5% increase during the past year as portfolio managers reinvested returning principal, and more, finding value in the sector as yield spreads to Treasuries generally widened over the last twelve months. Municipal allocations increased from 22.70% to 26.11% (23.00% tax-free, 3.1% taxable) as the shock of lower tax equivalent yields from the 2018 tax cuts were slowly replaced with a realization that muni’s still represent some of the best yields for a typical bank portfolio. Agency/Treasury investments dipped during the year decreasing from 18.77% to 12.65% at quarter-end as some called and matured bonds were reinvested elsewhere (most likely MBS and muni) and not back into the sector. The “Other” category increased slightly from 3.96% to 4.18% with corporate bonds constituting nearly half the category while CDs totaled 1.3% as lower short-duration yields reduced the popularity of the group with portfolio managers.
Now let’s look at portfolio performance trends. The graph below tracks average portfolio tax equivalent book yield, duration, unrealized gain/(loss) as a percent of book value, as well as 10-year Treasury yields and average portfolio size over the last two years.
As 2019 began, Treasury yields were dropping following the market volatility that characterized the fourth quarter of 2018. That drop in market yields finally started to show in portfolios as book yields peaked in the second quarter 2019 at 2.75% and finished December at 2.71%. Yields edged lower again in the first quarter dropping 3bps to 2.68% but they really fell under the weight of the Fed’s lower for longer policy with the book yield ending the second quarter at 2.39%, down 29bps during the quarter. As shown, longer-term Treasury yields (green line) collapsed during the first quarter starting at 1.92% and finishing the quarter at 0.67%. That collapse in Treasury yields will continue to pressure portfolio yields lower as we move through 2020.
Looking at portfolio duration (orange line), the graph shows 2019 started with durations at 3.28 years and ended the year at that same level. As Treasury yields moved decidedly lower during the first half of 2020 durations eventually shortened with the second quarter ending at 2.80 years. We expect the drop in yields to an ultra-low range will continue to pressure durations lower as prepayments and call expectations react accordingly.
2019 began with portfolios at an average unrealized loss of –1.86% of book value. With the Fed shifting to rate cuts in mid-year 2019, unrealized gains accelerated from 0.16% as of June 2019 to 1.22% at year-end 2019. In the first half of 2020 unrealized gains rose again ending June at 2.67% of book value. Gains peaked at 3.00 in May but we suspect a run to that level again as rates have dipped again in July and August.
Finally, new investments in the second quarter of 2020 focused on the MBS/CMO sector with purchases equaling 64% of the $2.3 billion in total par (61% in MBS and 3% in CMOs). The 64% MBS/CMO allocation compares to a legacy total of 57%. Of the purchases, 58% were in 30yr pools with 20yr pools at 29%. The municipal sector followed with 19% of total purchases (17% tax-free and 2% taxable). The 19% muni purchases compared to 26% legacy muni allocations. 11% of purchases were in Treasury/Agency which compares closely to a legacy allocation of 13%. The average tax equivalent book yield for the second quarter purchases was 1.56% with an average effective duration of 2.93 years. We will update this data again in October to track how allocations and performance characteristics are trending in 2020.