Rethinking 'Investment Beliefs'.....My 50-Year Journey

Print the PDF

“Belief is trust, faith, or confidence in someone or something.”

“Belief is an attitude that something is the case, or that some proposition about the world is true.”

“A person can base belief upon certainties, probabilities, or matters of faith.”According to GOOGLE

 

Investment Decisions, Investment Beliefs, and Fiduciary Duty

It is a truism to say that behind every thoughtful investment decision lie thoughtful investment beliefs.  Paraphrasing the GOOGLE definitions, these beliefs reflect trust, faith, or confidence that some propositions about the world of investing are true, and hence should be reflected in how investment decisions are made. These beliefs may be grounded in certainties, probabilities, or simply reflect matters of faith.

At the same time, the exercise of ‘fiduciary duty’ in pension organizations is a commitment to act in the best interests of the plan’s stakeholders. In an investment context, this requires ensuring that the organization’s investment processes are based on a set of investment beliefs that reflect the organization’s best thinking on how the world of investing works. So logically, every pension organization has a fiduciary duty to clearly set out its investment beliefs in its Statement of Investment Policies and Goals.

Page 4 of this Letter sets out the investments beliefs that work for me today after 50 years of thought, research, and practical experience. Between here and page 4, allow me to retrace the 50-year journey that led to the investment beliefs I hold today.         

The Start of the Journey

After the Royal Military College of Canada and a brief military career in the 1960s, I emersed myself in  post-graduate economics, first at Western University and then McGill. Realizing full time academia was not the right career path for me, I interviewed for a new position Sun Life Financial had just created. The job was to understand Modern Portfolio Theory (MPT) and assess its relevance to investing Sun Life’s own  assets, and those of its clients. I got the job, which required doing two things: 1. Read the MPT literature and meet its academic creators, and 2. Understand institutional investing as it was practiced in the late 1960s/early 1970s.

It did not take long to discover that academic MPT and ‘real word’ institutional investing at that time were parallel universes with zero overlap. MPT assumed known return distributions, covariances, and risk      tolerances. Further, if all market participants have access to all available information, and process that  information identically, markets are price-efficient and lead to the Efficient Markets Hypothesis (EMH). In contrast, ‘real world’ institutional investors believed that they had a competitive advantage in attaining and using information that would produce excess returns (i.e., ’alpha’) for them and their clients.  

So whose investment beliefs to believe….the academics’ or the investment professionals’? I spent the 1970s addressing this question. How? By asking investment professionals and their investment models to predict the relative performance of the stocks they were following. Could they generate a positive correlation between their predictions and actual subsequent outcomes? The general answer was ‘yes they could’, but with the caveat that the correlations were low (i.e., 0.2…not 0.8).ⁱ Were such low levels of predictive ability sufficient to generate positive portfolio ‘alphas’ net of transaction costs? In a 1979 article titled “Can Active Management Add Value?” in the Financial Analysts Journal (FAJ), Jim Farrell and I provided a ‘yes’ answer to that question, but with the strong caveat that was only the case if the low quality of the predictions was explicitly recognized in portfolio management processes.

In fact, the ‘low prediction quality’ reality was generally not recognized in how portfolios were managed in the 1970s. Instead, the combination of too much turnover and too high fees produced mostly negative rather than positive ‘alphas’ in that decade. This set the stage for the triumph of the academics and their EMH: the introduction of low-cost, low-turnover passive management through index funds.      

Enter Drucker and Keynes

That 1979 FAJ article marked my exit out of applied portfolio theory into the broader space of the design and management of retirement income systems. An important trigger was Peter Drucker’s 1976 book “The Unseen Revolution” where he argued that workers would end up owning the means of production not through violent revolution as Marx had argued, but through their pension plans. Who would manage these pension plans? Drucker visualized high-quality pension organizations governed under fiduciary  umbrellas requiring them to act on the sole best interest of plan participants. Fast-forwarding to today for a moment, these kinds of organizations are no longer just figments of Drucker’s imagination. They now exist in the real world.ⁱⁱ

What about the investment beliefs of these Drucker pension organizations? Should they simply accept the EMH and manage their retirement savings pools passively? Or should they rethink the meaning of ‘active management’? Personally, I chose the ‘rethink’ option, guided by the great economist John Maynard Keynes. While I had read parts of his 1936 opus “The General Theory of Employment, Interest and Money” during my graduate studies years, I had somehow missed Chapter 12 titled “The State of Long-Term Expectation”. Belatedly reading it in the 1980s, I realized there was much more to investment beliefs than either the EMH or those that drove professional active management in the 1970s.

Keynes did not write “The State of Long-Term Expectation” as an academic, but as the manager of Cambridge University/King’s College Endowment Fund with 15 years of practical experience under his belt. His three key insights were:

• Investment professionals seem to be continuously engaged in ‘beauty contests’, trying to outwit each other in predicting which stocks investors will find most attractive a few quarters hence.
• Real investing is based on understanding how well (or poorly) businesses allocate capital to be sustainably profitable over the long-term, and to focus the portfolio on companies that score well on this basis.
• Investment committees seem to prefer being ‘beauty contest’ conformists…..”preferring to fail conventionally to succeeding unconventionally”.

And how well did Keynes do with this unconventional long-term approach to investing? Cambridge University’s David Chambers and Elroy Dimson estimate Keynes’ ‘alpha’ to be an astounding 8%/yr over the 1921-1946 period. Keynes is not alone in producing extraordinary results using unconventional long-term approaches to investing. Our May 2017 Letter titled “Fostering ‘Long-Termism’ In Investing” cites multiple studies of long-term unconventional investment approaches that have reached the same conclusion. These findings, along with the accelerating institutional shift towards sustainable finance, are now redefining ‘active management’ and lengthening investment time-horizons. At the same time, this shift is increasing collaborative efforts between investment institutions (e.g., through PRI, FCLT, ICPM, etc.). For more on the shift to sustainable finance, see our October 2020 Letter titled “Sustainable Investing – A Path to a New Horizon”.

What About Asset Mix?       

In 1987 I wrote a sequel FAJ article titled “Pension Fund Asset Allocation: In Defense of a 60/40 Equity/Debt Asset Mix”. It was a response to the emerging LDI movement to fully immunize DB Plan balance sheets with 100% Bond portfolios. The article argued for a more affordable ‘going-concern’ rather than a solvency mindset in investing retirement savings by emphasizing equities. In that same spirit, I wrote a Letter in 2020 titled “Future Proofing Pensions: Integrating the Wisdom of Peter Drucker and John Maynard Keynes”. It argued that from a ‘going-concern’ perspective, the traditional 60/40 asset mix policy had become inefficient for two reasons. First, the 40% Debt piece no longer offered a positive real return, and second, the robust application of Keynesian monetary/fiscal policies after WWII had effectively eliminated a 1930s-type global economic depression as a plausible scenario. This second factor materially reduces capital impairment risk for well-capitalized businesses with sustainable business models, and hence also for their investors.

What do these two asset mix pieces, written 33 years apart, tell us about investment beliefs? Four things:

• Asset pricing matters. What was a sensible ‘going-concern 60/40 investment policy’ in 1987 was no longer so sensible in 2020.
• If asset pricing matters, then asset pricing models matter. They should be explainable and grounded in reality (e.g., I have profitably used Myron Gordon’s R=Y+G Model for decades).
• Investment risk is time-horizon sensitive. What may be risky to investors with short horizons (e.g., those with material solvency risk) may not be risky for investors with ‘going-concern’ long horizons (e.g., multi-generational pension plans).
• If time-horizon matters, then metrics/tools that help imagine/visualize ‘risk’ over those horizons also matter. Examples might be asset volatility metrics for short horizons and plausible capital market scenarios/narratives for long horizons.ⁱⁱⁱ   

An example of these investment beliefs in action follows.

Investment Beliefs in Action

Through this publication, we have been setting out long-horizon investment narratives with integrated asset pricing implications for a long time. The most recent effort was in the September 2018 Letter titled “Short-Term vs. Long-Term Investment Risk: Really Understanding the Difference”. Its key messages were:

• We have been living in a ‘Mature Capitalism’ era since 2010, the 5th investment era since the end of World War II. The referenced Letter named the four prior eras ‘Pax Americana I’ (1950-1970), ‘Scary Seventies’ (1971-1980), ‘Pax Americana II’ (1981-1999), and ‘Double Bubble Blues’ (2000-2009).
• As predicted by the Gordon Model, equities materially outperformed bonds in the two 20-year Pax Americana eras, marginally outperformed in the 10-year ‘Scary Seventies’ era, and materially underperformed in the 10-year ‘Double Bubble Blues’ era. Over the full 60yr (1950-2009) period, the S&P500 had a real return of 7.0%/yr vs. 2.3%/yr on a 10yr T-Bond  portfolio, a realized risk premium of 4.7%/yr.
• At the start of the ‘Mature Capitalism’ era (i.e., 2010), the standard Gordon Model prediction showed the S&P500 had only a modest prospective risk premium over 10yr T-Bonds using a history-based real G of 1.5%. However, this assessment was skewed by two factors: 1. The increasing weight of non-dividend-paying growth stocks in the Index, and 2. The increasing use of net share buy-backs rather than dividend payments to return cash to shareholders. Both factors would boost G at the expense of Y in the Gordon Model. In fact, the S&P500 Y at the start of the 2010-2020 period was 2.6%, well below a history-based average 4%, while G was 5.5%/yr, well above a history-based average 1.5%. Y+G together add up to a Gordon Model return of 8.1%. The actual 2010-2020 S&P500 real return was 11.9%, as Y over the period fell from 2.6% to an even lower 1.5%.^iv The 10yr T-Bond portfolio total real return was 2.0% over this period.
• All this sets up the challenge of telling possible stories of where the ‘Mature Capitalism’ era may go from here…..a task to be taken up in a future Letter.

This Letter ends by setting out five investment beliefs that reflect what I believe can be learned from the past, and which are relevant for investing retirement savings in the decades ahead.  

Investment Beliefs for the 21st Century

Consider these five beliefs:

• Time-horizon is a critical factor in framing investment risk and return prospects.
• Active management is justified as investment returns are modestly predictive over long horizons.
• Active ownership and collaboration strategies reduce agency costs and improve global environmental and social outcomes.
• Successful active management and active ownership strategies require requisite skills, talent, and modelling capabilities….as well as effective governance and efficient organization design.
• Risk-reducing diversification and cost-reducing organizational structures are both ‘free lunches’ for investment organizations with effective governance and efficient organization   design.

I believe these five investment beliefs are both defensible and relevant. Do you?

Keith Ambachtsheer

Endnotes:

1. My modest contribution to the investment literature at that time was to name these correlations ‘Information Coefficients’ or ICs for short.
2. For example, see the recent JPM article by Ambachtsheer “The Canadian Pension Model: Past, Present, and Future”.
3. See the Kay-King 2020 book “Radical Uncertainty” for more on this.These findings are updates from the ones reported in the 2018 Letter. The decline in Y over the 2010-2020 period from 2.6% to 1.5% explains the difference between the Gordon Model Y+G return of 8.1% and the actual real S&P500 return of 11.9%. 

KPA Advisory Services is pleased to share this edition of The Ambachtsheer Letter with all readers; if you wish to become a KPA Advisory Client/gain access to ALL Letters, please see the Services page on our website.

The information herein has been obtained from sources which we believe to be reliable, but do not guarantee its accuracy or completeness.