Factor Investing in Fixed Income (EP.138)

Episode Summary

Executive Summary: In this episode, hosts Benjamin Felix and Cameron Passmore explore the complexities of fixed income investing, arguing that bonds offer time-varying expected returns through term and credit premiums that can be captured by adjusting portfolio exposure based on yield curve shape and credit spreads. They contrast this with Bitcoin mania, discuss return predictability in financial planning, and share community updates. The key insight is that bond returns are more forecastable than stock returns, allowing for dynamic strategies that may enhance returns without increasing risk, challenging the view that bonds have no expected return in a low-yield environment.

Main Topics: Fixed Income Factors and Dynamic Strategies (Priority: 5/5): Ben explains that bonds have two main risk factors (term and credit) with time-varying expected returns, unlike stocks. The yield curve shapes and credit spreads can predict near-term bond returns, enabling strategies that vary maturity and credit exposure to potentially enhance returns by 30 basis points annually with lower volatility. Bitcoin ETFs and Crypto Mania (Priority: 4/5): Canada launched the first North American Bitcoin ETFs (Purpose BTCC, Evolve EBIT), attracting significant assets. The hosts discuss Bitcoin's cult-like following, environmental impact (mining energy usage comparable to Argentina), and caution against using it as a retirement strategy, especially near retirement. Return Predictability in Financial Planning (Priority: 5/5): Ben presents evidence that equity returns are predictable over long horizons based on valuation measures like earnings yield. Using constant expected returns in Monte Carlo simulations can mislead planners; a model allowing mean reversion improves success rate estimates by up to 30% for long retirement periods. Community Engagement and Merchandise (Priority: 2/5): Updates include: new mugs with corrected guest list, free socks with orders, over 2,200 community members, CE credit poll for Canadian advisors, Goodreads and Peloton group growth (20+ and 30+ members respectively). Book and Media Recommendations (Priority: 2/5): Ben recommends Jennifer Risher's 'We Need to Talk: A Memoir About Wealth' for insights into wealthy families' challenges. Cameron started Satya Nadella's 'Hit Refresh'. TV recommendation: 'Your Honor' with Bryan Cranston (now ended). Bad Advice Critique: Bitcoin for Retirement (Priority: 3/5): A listener submitted an article suggesting near-retirees invest in crypto to 'double income' in 3-5 years. Ben and Cameron condemn this as irresponsible, noting Bitcoin's 70% historical drawdown and dangerous timing near retirement.

Key Arguments: Fixed income has two workhorse risk factors—term and credit—which provide higher expected returns when compensated by current market conditions (steep yield curves, wide credit spreads). Bond returns are more forecastable than stock returns because cash flows are fixed; changes in yields primarily reflect changes in risk premiums rather than default expectations. Over 90% of credit spread variation is driven by expected credit premiums. Cap-weighted bond indexes tend to increase exposure when premiums are low (e.g., more credit when spreads narrow, more long-term bonds when curve flattens), making them structurally disadvantaged. Dynamic strategies using simple index-based rules (varying maturity and credit exposure based on yield curve shape and spreads) can add 12-17 basis points annually with lower volatility, per Dimensional's 1974-2015 simulation. Equity returns exhibit predictability over long horizons: high valuations (low earnings yield) imply lower future returns. Historical average returns are negatively correlated with future returns, while implied equity risk premiums are positively correlated. Using constant expected return assumptions in Monte Carlo models is inferior to incorporating time-varying expected returns with mean reversion, especially for long retirement horizons (up to 30% higher success rates in preliminary modelling).

Data Points: Purpose Bitcoin ETF first-day inflows: $165 million - Assets under management on launch day Purpose Bitcoin ETF second-day inflows: Over $200 million - Additional assets by end of second trading day Bitcoin mining energy usage ranking: Between Norway and Argentina - If Bitcoin were a country, its energy consumption would rank in top 30 globally Credit spread variation explained by expected credit premiums: Over 90% - From 2014 paper on 40 years of US investment-grade corporate bond data Excess return of variable maturity strategy: 12 basis points annualized - Dimensional simulation using indexes, 1974-2015, compared to static 50/50 credit/government intermediate benchmark Excess return of variable credit strategy: 17 basis points annualized - Same simulation period, varying credit exposure based on spread levels Combined variable term and credit strategy excess return: 30 basis points annualized - Simultaneously varying both dimensions; lower standard deviation than static benchmark Historical Schiller CAPE mean: ~16 - US market average since 1871, compared to current level of ~32+ Improvement in retirement success rate with time-varying expected returns model: Up to 30% - Preliminary PWL model results for long retirement periods (e.g., 60 years) Bitcoin price decline (2018-2019): 70% - Historical drawdown mentioned as context for risk

Pivotal Quotes: "With bonds, the short-term forecasts... With Schiller PE, 10 years, you're going to explain about 40% of the differences... With bonds, it starts to become more like 40% for one-year forecasts." — Benjamin Felix: Explaining the key difference between bond and stock return predictability—bond returns are forecastable over much shorter horizons. "If you're using historical returns as an expected return, as returns go up, realized returns... future expected returns may be getting lower. But if you're using historic returns as your expectation for the future, as they go up, you're also assuming higher returns in the future. It can be a little bit deadly." — Benjamin Felix: Highlighting the dangerous feedback loop of using historical averages during bull markets, which leads to over-optimistic planning assumptions. "This isn't just saying you know that this might be a thing that sticks around and maybe keeps pace with inflation. This is saying that you're going to double your income... three to five years before you retire. Yeah, that's something." — Cameron Passmore: Critique of bad advice article promoting Bitcoin as a near-retirement 'double your income' strategy.

Implications: DIY investors should not blindly hold aggregate bond ETFs—dynamic strategies based on yield curve shape and credit spreads may improve risk-adjusted returns. Financial planning assumptions must incorporate time-varying expected returns to avoid over-optimism or excessive conservatism. Bitcoin as a retirement near-term strategy is highly dangerous; advisors should educate clients on sequence-of-returns risk and proper asset allocation.