FolioBeyond Fixed Income Commentary For November 2020
Performance Summary
FolioBeyond’s algorithmic Fixed Income strategy returned +2.42% and +1.25% in its dynamic and static volatility versions, respectively, for November. While each version is designed to fit the needs of different investors, both strategies use the same multi-factor optimization framework. The dynamic volatility version targets the trailing 1-year volatility of the commonly followed Bloomberg Barclays U.S. Aggregate Bond Index (“AGG”), which leads to higher risk targeting in more volatile market environments. The static volatility version targets constant annual volatility of 3.5%, which generally corresponds to the long-term average volatility of the AGG.
The Treasury yield curve flattened slightly in November, with the 10-year yield declining by 4 basis points and the 2-year yield rising by 2 basis points. Credit exposures were the main drivers of our strategy returns as short duration High Yield Corporate Credit and Bank Loan exposure were the main contributors to positive returns. Long duration Treasuries and intermediate US Agencies also added to the performance.
Source: FolioBeyond’s returns are from SMAs on Interactive Brokers (from January 1, 2019, for Static Volatility and from November 1, 2020, for Dynamic Volatility) and back-tested simulated results prior to that. AGG and Multisector Bond Category returns are from Morningstar.
* FolioBeyond Dynamic and Static Volatility returns are nets of underlying ETF fees and 30 bp assumed management fee. Although the information herein is believed to be reliable, FolioBeyond makes no representation or warranty as to its accuracy, and information and opinions reflected herein are subject to change at any time without notice. The past performance information presented herein is not a guarantee of future results.
Highlight: Empirical Duration and Correlation Effects
While theoretical duration measures are easy to quantify, realized or empirical durations are more variable for credit sensitive bonds. A simple cash flow duration measure on credit bonds can be associated with a set of underlying assumptions that include unchanged yield spreads to Treasury benchmarks and unchanged default expectations. For small interest rate moves within a stable macro environment, the above assumptions would generally apply. However, interest rate moves of any significance are typically accompanied by a changing macro environment, changing yield spreads, and possibly revised default expectations. Consequently, investors often observe that empirical durations can deviate significantly from model durations.
As an illustration, in recent weeks, a modest selloff in bonds (rise in Treasury yields) has been associated with potentially good economic news (e.g. positive vaccine testing results), lower default expectations leading to credit yield spread narrowing versus Treasuries, and cash movement out of bonds into stocks. Therefore, empirical durations on certain credit bonds have been shorter than the model duration measures would suggest. On certain days, some credit bonds have even exhibited negative duration (credit bond prices going up as Treasury prices declined).
The above paradigm may hold in the near-term if the market does not experience a major, sustained selloff in Treasuries. For larger moves in Treasuries, however, the impact of cash flow duration will come back into play as well as possible extension risk for bonds with callable features. This may lead to empirical duration of credit sensitive bonds being longer than cash flow duration measures. For MBS bonds, the negative convexity effects will be even more pronounced as prepayment expectations will slow dramatically if interest rates go up significantly.
In addition to properly understanding the effects of varying empirical durations, correlations can provide both positive and negative portfolio effects. A long-duration Treasury exposure is likely to be a good offset against credit exposures that underperform in a bond market rally. Floating rate Bank Loans will not move in tandem with long-duration High Yield Municipals in a yield curve steepening selloff. Therefore, while portfolio diversification should be one of the main goals within a broad investment strategy, it is critical to capture the appropriate risk variables and correlation effects in producing a properly diversified portfolio mix.
FolioBeyond’s algorithms captured the effects of subsector performance in different recent environments, including the effects of correlation and risk levels that are based on empirical data adjusted for forward-looking implied volatility levels. These aspects are all incorporated within a multi-factor optimization framework that also includes relative value relationships, short-term and long-term momentum effects, and stress testing. Furthermore, the model optimization is run daily with up-to-date market data and analytics, which is critically important especially in regime shifting environments. In order to avoid frequent transactions, however, the rebalancing threshold is set to a band which still allows it to capture the major moves in the markets.
Please contact us to explore how our advanced, low-cost portfolio optimization solutions can be utilized to fit into your overall investment goal.
