Total Portfolio Approach.
Why is TPA more effective than SAA? How to make a successful transition
This article is intended for those familiar with the distinction between SAA and TPA. For those unfamiliar, I’ll provide a link to a detailed description at the end.
Either way, this is a good example of why understanding systems theory is extremely useful when making significant decisions, such as transitioning from SAA to TPA. Your organisational and strategic context should determine how you make decisions. As we go through the answer to the question, we will extract six principles for use when making changes in your organziation
I have included links to all the tools and techniques mentioned in this article.
TLDR; Summary
Q: Why is TPA more effective than SAA?
A: TPA is consistent with the principles of anthro-complexity, while SAA holds us in an outdated systems dynamics paradigm.
Q: Why is this an advantage?
A: TPA inextricably links near term performance and long-term sustainability. (adaptation and evolution).
The TPA is part of a broader paradigm shift from the ‘Systems Dynamics’ that has driven the last 100 years of economic thinking toward ‘Anthro-Complexity’ – An emerging paradigm that focuses on how humans interact within complex systems.
A large group of humans making individual choices regarding a multitude of assets spread across a vast geopolitical landscape can be defined as an "anthro-complex system."
SAA is akin to systems dynamics in that, we strive to optimize our utility by first understanding the flow of energy in the system and then making choices about how we interact with that system. This means that the capacity for prediction underpins investment behavior.
TPA is akin to anthro-complexity in that it recognizes that the sum of the parts is not the same as the whole. When done well, TAP provides a mode of operation that lessens the need for prediction and replaces it with the capacity to adapt.
When faced with uncertainty we have three strategies: Ignore, suppress or adapt.
SAA seeks to ignore or suppress, TPA seeks to suppress at times and adapt at others.
The world is becoming more complex and interrelated. We can no longer afford to ignore or suppress uncertainty.
Making the transition will require new thinking in:
Governance and Strategy
Decision Making
Constraints Mapping
Risk Management
Leadership and Teamwork
AGLX are global leaders in developing the strategy, tools, techniques and methods needed to successfully transition from SAA to TPA.
Why is the Total Portfolio Approach a better strategy than Strategic Asset Allocation?
Much of the discussion around TPA is centered on why it is a better approach than the more traditional SAA. The ‘answer’ has primarily been couched in terms of the self-reported outcomes of the few funds that have already implemented a TPA to some degree.
The advantages of “more joined up”, “dynamic” “, quality decision making” “, goals driven” and “less drag”[1] These are symptoms of a well-implemented TPA, but they don’t provide any information about why a TPA is theoretically better than an SAA. There is an answer, and that answer lies in the intersection of complexity science and biology.
First Principle: Context is King.
Your organizational and environmental context should determine your investment approach.
Systems Dynamics – The paradigm that underpins SAA.
In systems dynamics, we strive to optimize our utility by first understanding the flow of value in the system and then making choices about how we interact with that system to extract value. Systems dynamics assumes that the relationship between the components of the system is linear and knowable. It assumes that every effect has a knowable set of causes, and that the system can be modelled in part and as a whole. Systems dynamics privileges the views of analysts and experts, treating the global financial system as a large machine. We sometimes refer to this as the ‘engineering paradigm,’ and it is reflected in the language we use, such as ‘drivers of value,’ levers for change, buckets of capital,’ ' portfolio construction,’ and ‘engine room of the economy.’
In a systems dynamics world, we can extract value from the system by modelling it, defining rules, goals and boundaries before allocating resources.
If the assumptions in systems dynamics were valid, then the economy would function within a set of known or knowable parameters. We could model these using Gaussian statistical assumptions, as we would seldom see events outside three standard deviations of the mean. It would conform to ergodic assumptions, meaning that the long-run average performance would be a good indicator of the present value. There would be volatility, uncertainty, and, from time to time, shocks, but we would seldom see large crashes or unprecedented events. When we experienced shock, we could quickly update our models and reduce future risk.
The word for this activity is robustness. Mechanical systems are designed to be robust, resisting change without altering their stable configuration. Success depends upon the ability to predict and plan for the changing environment. Failure should be avoided, and recovery is costly. Robust solutions struggle in complex environments. Market shocks and unprecedented events are well known to most of us. This has implications for risk management – a topic for future discussion.
Second Principle: The utility of models and forecasts is inversely proportional to the level of complexity in the system.
Anthro-complexity – The paradigm that (should) underpin TPA.
Anthro-complexity, derived from anthropology and complexity science, views the global management, financial, and economic environment as a complex adaptive human system. Complex systems exist far from equilibrium and exhibit many interesting features, such as emergence, nonlinearity, and self-organization. The causal relationships between the components are not knowable in advance and usually not repeatable. Complex adaptive systems comprise many agents, which are entities capable of acting to change the system, such as individuals, automated systems, and policies. Agent behavior changes the system, and the system changes agent behavior. This gives rise to a constantly dynamic chain of action and reaction at many levels of scale. The behavior of buyers and sellers on a trading platform is a classic example of this complex interaction. Anthropology is the study of human behavior and human systems, patterns of behavior and cultures.
Anthro-complex systems differ from other complex adaptive systems, such as rainforests or ant colonies, because humans possess agency, objectives, and a level of intelligence that enables far more diverse adaptations in far shorter timescales. Our systems evolve more rapidly and exhibit greater complexity than those typically found in nature.
A large group of humans making individual choices regarding a multitude of assets spread across a vast geopolitical landscape can be defined as an "anthro-complex system."
Your organization is a complex adaptive organic system that is part of a more complex adaptive global ecosystem.
Note: There is no intent to describe the organization as a metaphor or simile of a specific living thing. This can lead us to simplistic comparisons such as ‘the brain or heart of the organization’. Your organization is a living thing, and that thing is different from all other living things, just as you are different from all other humans, and humans are different from all other species.
Third Principle: Organizations and markets are ecosystems, not machines.
Organisms don’t survive and evolve because they predict the future with clarity. They survive and evolve by applying a simple heuristic:
“To remain alive, living systems must act in a way that minimizes the difference between their expectations and their reality”.[2]
The correct term for this difference is ‘variational free energy,’ and it can be approximated via Bayesian inference. In colloquial terms, we might call situations where there is a mismatch between expectation and reality ‘surprise’. For example, you can look at a fish and assume by its shape and its gill structure that it ‘expects’ to live in water, extract oxygen from its environment and find food through swimming. Taking a fish out of water creates a mismatch between these expectations and the new reality. The 1987 market crash created a mismatch between the implied ‘expectations’ of the global investment community and reality. Surprise! Examining an investment portfolio allows us to infer the expectations of the individuals who designed it regarding risk and return.
This surprisingly simple idea that organisms ‘minimize the difference between their expectations and their reality’ means that the one ‘strategy’ employed by living organisms allows them to survive in the short term and evolve over the longer term. In your organization, the same strategy should create good returns in the short term and a sustainable investment vehicle over the long term. – The AGLX adaptive strategy process is directly correlated with this idea.
So, how do organisms minimize surprise? There are three broad strategies: Ignore, Suppress or Adapt.
1. Ignore: Take action to avoid the sensory input.
An example of this is a company built on the charismatic promises of its founder. These companies may attract ‘blind’ capital and charge into the market with no regard for sensory information. Companies like these most often hit a surprise via bankruptcy, insolvency, or criminal charges. It is a strategy, but not a good one.
Organizations often ignore surprise when they have a culture where people lack incentives to speak up against established ideas or decisions. It is easier to do what we did yesterday and avoid that cost.
Fourth Principle: You can’t opt out of complexity.
Sticking with an SAA approach in the face of increasing complexity has the same effect as a toddler holding their hands over their eyes to hide from something they don’t like. Blocking it out doesn’t make it go away.
2. Suppress: Take action to modify the environment so that it aligns with our expectations.
Conventional strategic planning is underpinned by the systems dynamic paradigm. The goal of conventional strategic planning is for the organization to become the author of its value creation by optimizing its position in the ecosystem.
An environmental equivalent is the evolution of an ecosystem on an isolated island. The sea provides a barrier to the wider world. The ecosystem becomes optimized for a very specific set of local conditions; there is very little evolutionary pressure. When the world eventually contacts the small island through the introduction of new species, disease and human activity, the level of surprise is often catastrophic for that fragile ecosystem.
SAA is a way to suppress surprise. The Boards asset allocation preferences are usually based on prediction and historic performance. As each asset manager competes for resources, it constrains the organization as a whole from shifting away from this prediction. Time-based reporting can stifle the direct and immediate feedback needed for adaptation to occur. We have suppressed the means for error correction.
Tendencies toward tribalism, social isolation, and tunnel vision are endemic to the human condition. SAA reinforces this tribalism by creating the conditions where it becomes a cornerstone of the organizational culture. Separation of asset classes, risk management and competition for capital creates silos. Privileging the view of a few anointed experts creates a narrow scope for interaction with the wider business and market. The result is a set of local optima, based on a set of predictive assumptions. An isolated archipelago of small islands.
3. Adapt: Update our assumptions and change our actions so that future sensory input is not surprising.
TPA minimizes surprise by providing the mechanism for self-correction. The reference portfolio serves as a model for our assumptions, and the fund's performance against this model provides the necessary feedback to adjust our beliefs and actions, ensuring we are more attuned to the shifting market reality. We move from periodic reporting to continuous monitoring. We have a strong signal for course correction. This action is described as ‘dynamic resource allocation’; the flow of information from assumption to action to feedback provides the mechanism for adaptation.
In the short term, this activity provides adaptive capacity in the form of learning, and in the longer term, this adaptation begins to resemble evolution. Performance and sustainability are inextricably linked. The pattern of action in a well-enacted TPA is like that of complex adaptive systems in nature; however, in the human environment, we overlay deliberate intent, agency to act, and intelligence. The ability to harness change and uncertainty is called resilience. In practice, this means maintaining a capacity for renewal in a dynamic environment. Adaptation provides a buffer that protects the system from failure. We can take actions based on incomplete information, allowing us to learn and change affordably.
These are typically small actions across a broad range of asset classes, designed to interact with the market and identify potential new sources of value. We might refer to this as:
Fifth Principle: Maximize the surface area for luck and learning.
In ecosystems, we see this principle in the foraging behavior of animals, growth patterns in plants and natural curiosity in humans.
While adaptation does reduce surprise, it can never reach a stable equilibrium or be eliminated. Other organizations in the ecosystem are constantly changing the strategic environment through their actions associated with minimizing surprise. Their agency, objectives, and intelligence will be different from yours, creating a continually dynamic and competitive environment. Adaptation is a constant necessity driven by ongoing market pressure. This brings about the following principle:
Sixth Principle: Adapt or die.
When transitioning from SAA, this inertia manifests in areas such as education, governance, leadership, and, of course, portfolio construction.
So, what does this mean when making the shift from SAA to TPA?
Shifting from an SAA to a TPA is not an area where there is best practice or a recipe for success. Your organization should adopt the principles of TPA in their context. What works for one fund may not work as well for you. Be adaptive and find your version of a successful TPA transition.
Governance and Strategy
We use AGLX Adaptive Strategy to provide the enabling infrastructure for this shift. It expresses strategy in terms of a clear direction, main effort and way of working that embodies the anthro-complexity theory in practical, actionable and measurable terms. This allows the board to communicate their intent in a way that enables the organization to move in the right direction, at the right tempo, while maintaining the capacity to adapt. This is embodied not only in a reference portfolio but in clear expectations around organizational performance and culture. We refer to these as a Shared Understanding of Success.
Constraints Mapping
This is the least understood but most valuable aspect of the transition process from SAA to TPA. Once the strategic intent is clear, you need to gain an understanding of the constraints that affect the implementation of a TPA. Your constraints will be uniquely yours, and this is why copying someone else’s journey or looking for best practices is unlikely to be successful. Principle One: Context is King. At AGLX, we see constraints in four broad categories.
Rigid constraints – These include externally imposed boundaries, such as legal and compliance matters, as well as self-imposed boundaries, such as ethical or sustainability policies, that constrain investment decisions.
Governing constraints - Modulate activity by providing expectations regarding outcomes. Experts can use their judgment to make decisions within these constraints. This includes factors such as risk tolerance, goal-based investing, and dynamic asset allocation.
Enabling Constraints – These guide while allowing for distributed decision-making. Principles are a good example of an enabling constraint. Enabling constraints can include things like shared decision-making models, the constraints map itself and the adaptive strategy. Many early-stage or venture funds operate mainly within constraints, as the market is too small and immature to create firm objectives.
Dark Constraints – These are where we can see the effect of the constraint but not the cause. For many organizations making the shift from SAA to TPA, a legacy of tribalism, centralized decision-making, and a ‘beat the market’ culture will create dark constraints. These will impede the shift to TPA.
Planning a shift from SAA to TPA involves understanding and managing constraints. AGLX has pioneered constraints mapping as a management information and decision guidance tool.
Decision Making
Shifting from SAA to TPA necessitates a shift toward decision-making processes that align with how people make decisions in complex situations. Making decisions in the right context creates flow and maximizes the potential for value creation and risk mitigation. Organizations are often hindered by outdated linear decision-making processes, resulting in disjointed and inconsistent decision-making, a lack of accountability, and organizational inertia. AGLX utilize Cynefin[3] and OODA[4] As decision-making frameworks, both are complexity-friendly, unlike other methods such as multi-criteria analysis or balanced scorecards. A shared decision-making process is essential to enable teams to become more collaborative and joined up.
Risk
The TPA should enable a more resilient approach to risk. Resilience is the outcome of two components: adaptive capacity and the ability to activate that capacity in times of need.
‘Robust’ risk methods tend to become fragile in complex situations and often hinder the adaptive response needed to shift away from emerging risks or toward opportunities.
Resilient risk methods acknowledge three domains of risk: Probable, Possible, and Plausible. Each of these is fundamentally different in terms of the relationship between predictability, likelihood and consequence. An integrated approach should acknowledge this difference and provide coherent and authentic actions in each domain. We support this shift with our AGLX Integrated Risk Management framework.
Leadership and Teamwork
Those leading the transition from SAA to TPA understand that the tribalism that helped us as hunter-gatherers now impedes us in an anthropocentric complex world where we must be more connected and joined.
Most organizations invest in leadership development. The AGLX Adaptive Leadership program is designed to equip leaders with the knowledge, skills, and experience necessary to lead in fast-changing and uncertain times. This is particularly important in a TPA environment, as decision-making is distributed to the investment team.
The reliance on SAA has diminished the importance of teamwork. AGLX has developed the "Flow Learning Lab" to provide teams with an interactive experience designed to help them develop resilient capacity in real-world conditions. Organizations that can distribute, foster, and promote anticipatory awareness, adaptive capacity, and empathy across teams will outperform those that view these qualities as individual traits. Your culture is an emergent property of leadership and teamwork.
Shifting from SAA to TPA creates the conditions for sustainable success in the face of increasing global uncertainty and complexity. Making the shift requires doing new things in new ways. AGLX are a globally recognized experts in helping organizations become confident in complexity. We have grounded our theoretical knowledge and applied many of our tools and techniques at the New Zealand Superannuation Fund over the last 6 years.
For more information, please contact:
stevem@aglx.com
[1] https://www.thinkingaheadinstitute.org/news/article/its-a-drag-why-tpa-is-superior-to-saa/
[2] M. Kirchhoff, T. Parr, E. Palacios, K. Friston, and J. Kiverstein, “The Markov blankets of life: autonomy, active inference and the free energy principle,” Journal of The Royal Society Interface, vol. 15, no. 138, Jan. 2018.
[3] https://en.wikipedia.org/wiki/Cynefin_framework
[4] https://en.wikipedia.org/wiki/OODA_loop