How GIC incorporates uncertainty into portfolio construction

Limitations of traditional portfolio methods such as mean variance optimisation (MVO) and the 60-40 model drive new approaches at the sovereign wealth fund.
How GIC incorporates uncertainty into portfolio construction

Faced with an inflationary environment, uncertainty over interest rate hikes and an unending pandemic, GIC is exploring new scenario- and simulation-based approaches that incorporate uncertainty into portfolio construction.

The Singapore sovereign wealth fund detailed these approaches in a paper published last November in collaboration with BlackRock after recognising the limitations of traditional portfolio methods such as mean variance optimisation, as well as other methodologies such as risk parity or “heuristic-type portfolios, such as 60-40 or 80-20 – each of them has their own shortcomings,” Grace Qiu, senior vice president of economics and investment strategy at GIC told AsianInvestor.

Grace Qiu, GIC

“Mean variance’s main issue is that it overly relies on one single mean forecast of return expectation, while risk parity assumes certain correlation structures between equity and bonds, which, looking at a rising inflation and rising interest rate environment, may or may not hold going forward,” Qiu said.

Investors, however, continue to use these methods because they are well-known, simple enough and have widely available inputs, she said.

“Take the 60-40 portfolio, for example. They worked extremely well from a return and diversification perspective for the past 20 years. Also, being very simple and transparent, it is attractive from a governance perspective,” she explained.

“That said, moving forward with a challenging investment environment and continued advancement in asset allocation means that more sophisticated and more robust frameworks will become less intimidating and more widely available to the broader investor universe,” she said.

Another example is BlackRock’s use of capital market assumptions (CMA), which offers a framework for return expectations across asset classes over various time horizons.

Ben Powell, BlackRock

“These continue to evolve,” said Ben Powell, managing director and chief investment strategist for Asia Pacific at the BlackRock Investment Institute. For instance, “before the adjustment we made last year, the CMAs were climate agnostic, and ultimately we didn’t think that was right – climate risk is real and needs to be incorporated into how we think about asset class returns.”

As a result, BlackRock began to incorporate the journey of transition to net-zero and how it creates opportunities and risks for different asset classes in its CMAs from February 2021. Other considerations include “more direct linkages to macro scenarios, and blending in elements of a scenario-based approach in allowing for multiple scenarios while using simulation techniques”.

“This incorporation of uncertainty, while different to the incorporation of sustainability, is part of the same journey to continue to try to make our capital market assumption framework as robust, pragmatic, useful and as helpful as possible for people that are using it,” Powell said.


GIC and BlackRock published a paper last November that detailed the scenario- and simulation-based approaches that they were exploring.

Essentially, both share an underlying philosophy that there is no optimal portfolio for all outcomes and a level of flexibility is necessary to cater to the investors’ individual risk appetite.

The scenario-based approach involves the consideration of alternate scenarios and using probability-weighted scenario outcomes to construct a portfolio that is tailored to maximise diversification. The aim is to minimise the opportunity cost of different macro scenarios.

Scenario-based optimisation example. Source: GIC and BlackRock

The simulation-based approach on the other hand looks at worst-case economic and market outcomes of a base case scenario and aims to maximise portfolio returns.

This is done by simulating multiple pathways for asset returns and choosing the worst set of return estimates based on the investor’s aversion to uncertainty. Naturally, this involves stress testing based on specific historical, market-driven or alternative scenarios, which means blending both simulation and scenario-based methods.

Simulation-based example vs MVO. Source: GIC and BlackRock

These approaches have applications in unexpected scenarios such as Covid or other volatile environments, Powell explained.

“Just in the last few weeks, we've seen inflation moving to much higher levels than many were anticipating, and market volatility we haven't seen some period of time, hawkish central bank policies, and policy rates hikes that everyone is expecting,” he said.

“All of these factors have driven some tumult in the market. So investors should brace themselves for more volatility and surprises. And that might persist in the post-Covid or living-with-Covid world. It seems like this is going to be a world where there's volatility and there'll be various dynamic market factors that can change very quickly.”


Another example of application is the integration of the climate crisis into the portfolio construction process.

“Mostly all investors believe climate change is happening, but the ultimate destination in terms of temperature pathways, or how much physical risk damage that's going to occur, I think there's still quite a lot of uncertainty in the eventual outcome,” she said.

“Given the super long horizon that it takes for the physical risk to play out, and also wide uncertainty in government policies in how technology is going to help us to adapt into a hotter world, it is very hard to determine how this secular trend would actually impact the strategic asset allocation at the current juncture. That's why the scenario-based approach is quite an effective way to integrate climate change, taking into account the many plausible paths going forward,” she explained.

Ding Li, GIC

For instance, the different scenarios could be whether the world undergoes a 1.5-degree Celsius increase as is the aim of the Paris Agreement, or a higher 2-degree rise. Or even within the scenario of a 1.5-degree increase, a scenario-based approach looks at the different paths that could lead to that end goal.

“Essentially, it's the transition risk along the way, which we also actively monitor and hope to be able to incorporate,” Qiu said.

“As an asset owner, especially for GIC, we review all the possible return streams in the market to construct a robust long-term portfolio in order to achieve our investment return objective,” Ding Li, also a senior vice president of economics and investment strategy at GIC told AsianInvestor.

“We carefully review all the economic drivers of the asset return and better understand the factors that can impact the long-term return outcome, for example, the investment process, manager dispersion on the market environment, and just like Ben highlighted, the recent inflationary, high volatility, and rate hike environment,” he said.

“All these kinds of considerations will help us to have a full assessment of the uncertainty, then we can incorporate this kind of assessment of that expected return and uncertainty scenarios together to design a more resilient portfolio solution over the long run.”

Ultimately, this approach, while taking into consideration different investor needs and requirements, blends the philosophical principle of building portfolio resilience, BlackRock Investment Institute’s Powell said.

“In terms of a philosophical approach, we think that is the right framing towards a more strategic resilient blended balance, diversified portfolio, and perhaps a move towards a more long term strategic approach to that portfolio construction, we think it can be good for everyone, while acknowledging that different investors have different needs, different beliefs, different requirements and so forth.”

¬ Haymarket Media Limited. All rights reserved.