Did You Forecast Asia? Scenarios In Portfolio And Risk Management

Ross Dawson
Originally published in The Australian Corporate Treasurer

Did you forecast and respond effectively to the ongoing impact of the Asian crisis? The debate continues on whether or not the crisis was predictable, however the reality is that it was not effectively predicted. One of the major reasons is the strong bias in financial markets to making single-point forecasts. By their nature these cannot encompass anything except what is perceived as the most likely outcome, and thus blind us to the unexpected rather than help us to prepare for it.

The greater the degree of uncertainty and unpredictability, the greater the value of using multiple scenarios rather than forecasts. Financial market forecasts have at best a poor track record, and almost never anticipate market discontinuities, which are what we most need to be forewarned about. Using forecasts in setting portfolio strategies results in optimised performance only if they are correct, and sometimes severe underperformance otherwise.

One of the most important problems with basing strategies on single-point forecasts is that it strongly reinforces the natural human tendency to look for information which supports your position or views, and to filter out contradictory evidence. While quantitative models and frameworks are playing an increasingly important role in portfolio management, these are subject to ‘model risk’, which is largely the risk that the people who have implemented the models have not perceived all the relevant parameters. Ultimately it is essential for senior treasury staff, and arguably also board members, to understand the broadest possible scope of market outcomes, their impact on the companyТs financial position, and appropriate responses.

Corporate treasuries, due to their limited resources, usually have to depend largely on their relationship banks for economic and market research. The difficulty is that, not only does each one present and promote a single perspective on the markets, but these views often contradict each other. (Or if they do achieve consensus, itТs even more worrying!) How can you make sense of these differing perspectives, integrate these into your own view of the markets, and from that construct a robust asset and liability management strategy? As importantly, how can you implement effective risk management in increasingly uncertain market conditions?

Scenario planning is a well-established methodology which is used extensively in corporate strategic planning by organisations such as Shell, Motorola, BHP, National Australia Bank and St.George Bank, as a tool for better understanding and anticipating changes in their business environment. Adapted to the financial markets, scenario planning provides a structured approach to developing scenarios which are specifically designed to provide a basis for setting portfolio strategies and in risk management.

Scenario planning is by its nature a facilitated process. This means taking a group of key executives through a structured process which draws on the diversity of their perspectives and viewpoints. Traditional investment committee or treasury management meetings are often dominated by personalities and opinions, resulting in strategies formed from a limited range of input.

To be effective, scenarios must start from qualitative differences – examining the underlying factors which impact financial markets, which can include political, social and technological dimensions as well as domestic and international economic factors. The scenarios can subsequently be quantified into market levels for use in asset allocation models or risk management frameworks if desired, however their value depends on them reflecting different qualitative trends.

Focusing on underlying issues results in internally consistent and thus plausible scenarios, identification of early warning signals which can be readily monitored, and the consideration of the full range of factors which could result in changes in market conditions. Simply considering the impact of different market outcomes without examining their qualitative underpinning provides no insight into whether these scenarios are likely or even possible, and certainly no indication of signals which will help identify these trends.

It is also very important not to develop simply most likely, best-case and worst-case scenarios, which is the most common first step beyond single-point forecasts. The nature of these forecasts is that the focus remains on the single most likely scenario. In addition, plotting the scenarios on a single dimension Ц from СgoodТ to СbadТ Ц clearly fails to explore the full scope of possible variations, and indeed doesnТt suggest looking for optimal responses whatever the outcome. Scenarios are only of real value if they are multi-dimensional, in that they explore several possible dimensions of the critical uncertainties ultimately affecting treasury decisions.

There are many possible approaches to scenario planning; which approach is most effective will depend on factors including the nature of the decision to be made and the group involved. However in most cases the process will include the following elements:

Determine the strategic decisions to be made

The strategic decisions at hand could include an asset or liability portfolio strategy, currency and interest rate hedging strategies, or the timing and nature of funding. In all circumstances the overall objectives and parameters of the organisation and treasury should be considered, including risk tolerance, anticipated cash flows, and time horizons.

Distinguish the underlying trends and the critical uncertainties

In considering the strategic decision, what are the underlying trends which have a high degree of certainty? And what are the dimensions in the environment which both have a major impact on the decision, and are highly unpredictable? This process requires looking beyond the usual issues, and is in itself a valuable exercise in perceiving organisational vulnerabilities and opportunities which may have been overlooked.

Select and develop scenarios

From the underlying trends and critical uncertainties, a small number of scenarios emerge. These are selected to be distinct, complementary and relevant to the decision at hand; to be useful each must represent a plausible future. The evolution of each scenario is then plotted in detail.

Develop contingency strategies and identify early warning signals

For each scenario, an overview of an optimised strategy is developed. This constitutes a contingency strategy for this scenario; elements of this will subsequently be integrated into the overall strategy. In addition, for each scenario early warning signals are identified; these are events or triggers which signal that the key elements of that scenario are likely to occur.

Develop robust strategy across scenarios

Using as building blocks the key components of the scenario-specific strategies, and other elements as appropriate, an overall strategy is developed which is resilient across the range of scenarios. This can be done in a variety of ways, including qualitatively in establishing a framework for strategy, and quantitatively by quantifying the scenarios and using optimisation models.

Apply scenarios to ongoing implementation of strategy

Once the overall strategy has been set, this can be readily adapted over time to changed market conditions, using the contingency strategies and strategy elements identified earlier. It is not necessary to redevelop scenarios frequently; these can be modified as market conditions unfold.

One of the key benefits of using scenarios in financial markets is that it provides a ready framework for filtering and making sense of information overload and contradictory signals. Market information and signals can be understood in terms of the scenarios previously developed, rather than simply supporting or contradicting a single market outlook. As such, using scenarios provides heightened sensitivity to market signals and turning points, and enhanced responsiveness to changed market conditions, which is a vital part of effective risk management.

Applying scenario planning specifically to risk management is a similar process to that outlined above. In this case scenarios are developed which deliberately test the sensitivities of your organisation’s portfolio and treasury function. Traditional stress testing is done using historic market shocks, which are neither likely to recur, nor may be relevant for your portfolio. One of the key advantages of scenario-based stress testing is that it can be applied to both market and non-market risk, including credit, regulatory, and operational risk.

The scenarios thus developed can then be integrated into existing quantitative risk management frameworks, with contingency strategies and proactive measures put in place. As with other applications of scenario techniques, much of the benefit is in treasury and senior management developing a broader perception of risks and eventualities.

Since scenarios are particularly relevant and useful as uncertainty – and thus timeframes – increase, it can be appropriate to integrate funding decisions with the examination of long-term investment decisions. For example, the effective life and payback of a major infrastructure investment is often 10 years or more; over this timeframe forecasts are of limited value and scenario planning is a very important tool in making considered investment decisions.

Scenario planning has proven its value in corporate strategic planning, and adapted to the treasury function has extensive applications in portfolio and risk management, including FX hedging strategies and funding decisions, especially in risk-averse environments.