Planning with uncertainty

As the world has become more interconnected both politically, economically, and socially, the future has become more uncertain. Political, economic, and social events in one part of the world, increasingly affect other parts of the world. Some examples include the fall of the Berlin wall in 1989, the collapse of American financial markets and the Great Recession - precipitated by the subprime mortgage crisis - in 2008, and in recent years, the militarisation of the South China Sea, the Covid-19 pandemic, and uncertainty about climate impacts and global climate policies. 

As uncertainty has increased, technological development and humankind’s ability to use technology to shape and influence the environment has also increased. This has led to a desire to invest more often in very large projects (both government and non-government) involving large sums of money and often spanning many years and even decades. Typically, the investment is irreversible – once committed, the costs are sunk – the decision cannot be unmade.

The real options approach has emerged over the last forty years as both a framework and a toolkit to support better strategy in the presence of these large uncertainties. What might appear to be the best decision to meet what is the most likely (or central case) future, may have little or no value if a different future eventuates to which that decision isn’t suited. The more uncertain the future, the more likely this is to be the case.

The real options approach recognises this problem by encouraging balanced analysis across all potential futures, with the strategy capable of adapting to be suited to most, if not all futures. This may mean that the design is not optimal for any of the futures that might eventuate; rather it is optimal for the potential futures as a set. The real options approach achieves this through:

• Designing as much flexibility into the investment as is economically feasible, such that it can be adapted cost effectively to different futures.
• Delaying irreversible commitment to elements of the investment where feasible and using the period of delay to invest in information to improve understanding of the likely future. This might include investing a small portion of the total investment cost in research and development, which can then be used to improve the design of the investment to suit the set of potential futures.
• Making smaller upfront investments to establish a range of options but delaying the commitment of the full investment until greater certainty about the future exists. For example, designing and planning an infrastructure project may be only a few per cent of the total cost but can be a large part of the time to implement. Spending a little more to design multiple options or added flexibility can create a range of options. Once the future is clearer, the best option to suit the unfolding future can be implemented in a much shorter time frame.
• Favouring approaches that offer inherent flexibility to adapt to new information – including smaller, modular additions to capability through time, use of approaches that lend themselves to being diverted to alternative uses and often favouring ICT approaches over large engineering approaches.

Change in the submarine strategy

The cancellation of the submarine contract with France and move to a yet undesigned nuclear-powered alternative from the US and UK is widely seen as strengthening Australia’s ties to the US and the UK ¹. But it has arguably done harm, that may persist for some years, to Australia’s relationship with France and the wider EU, and to an already strained relationship with China. Implications for relationships with closer neighbours such as Malaysia and Indonesia are unclear.

The rationale behind this decision appears to involve both an assessment that Australia’s security environment has changed since the decision to acquire the French submarines was made and the emergence of Australian access to a specialised nuclear drive from the US. What was considered an appropriate solution in 2016 appears no longer to be considered most suitable. In making the announcement, Australia’s prime minister referred to Australia as playing its part in promoting a free and open Indo-Pacific region including in the South China Sea ². The key difference between 2016 and 2021 driving the change in strategy is the continuing rise of China, its militarisation of the South China Sea and the sense that it is taking a more hawkish outlook in the region.

Willingness to change strategy in the light of new information is essential if risk and uncertainty are to be managed well. However, changes as dramatic as these demand sound processes to weigh the competing merits – and to guide the evolution of the strategy, to limit unnecessary costs and risks, and to provide access to upside opportunities likely to emerge from the uncertain environment.

Considering the submarine strategy as a set of real options

In the above context, Australia’s decision to change its submarine strategy may be best assessed within a real options framework. The submarine strategy appears to involve adaptation to changing information about access to leading-edge nuclear power technology and changing assessments of the geopolitical risks to Australia’s north.

Experience, including across a range of major strategy work undertaken by ACIL Allen, has highlighted that serious bias can creep into strategy planning processes and associated business case development. This bias is often evident in decision-makers’ preference for certainty and a desire to move forward with investments or commitment to strategy.

This certainty and clarity about future direction has appeal, economically and politically – but not if it creates the near certainty of locking into being systematically wrong. Decision-makers have been slower to fully recognise the value of building strategic flexibility through the acquisition and maintenance of options in the capability development process ³. The real options approach involves placing the management of uncertainty at the centre of the investment or strategic objective. Uncertainty is a core part of the question – not just a constraint on how well we can answer the question.

Clearly, nuclear power plants could add greatly to the capabilities of a submarine fleet to undertake certain tasks – especially ones requiring avoidance of detection, over long periods of time and at great distances. This technology affords flexibility with potentially high value in some circumstances.

The change in strategy replaces a process developed over several years for delivering a conventional submarine capacity, with a new process that is expected to take many more years and at a higher cost to then deliver a fleet two-thirds the previous size, albeit with some enhanced capabilities. While stated to be built in South Australia, it will probably involve lower levels of local involvement in the build (especially the propulsion system) and may have less associated development of onshore support capabilities. All these have implications for the options to be secured, and their value to Australia, through time.

The submarine decision will delay Australia’s access to its next generation of submarines, at least by several years, with this likely to trigger heightened costs and/or a lower value of existing submarine capability ahead of commissioning of the new submarines. In real options terms, these differences in cost and capability during the interim phase are highly relevant, alongside a comparison of the eventual capability and the value of that capability. Two elements of this potential loss of value because of delayed delivery are:

• The special appeal appears linked to current perceptions of risks in the Indo-Pacific and more specifically, the South China Sea – more so than the demands for deployment on tasks closer to home. What if the existing perceived threats have evolved to a point where existing tensions are resolved or no longer relevant before the submarines arrive?
• Over the next 30 or more years, some combination of smart space, air and surface tracking and deployed autonomous sensor devices or vessels might render a large nuclear submarine highly detectable ⁴. Might evolving technologies favour a shift in approach towards a larger number of smaller, harder to detect and possibly uncrewed vessels – paralleling the development of air fleets in recent years?

The previous strategy of acquiring a greater number of conventionally powered submarines to be delivered from the 2030s was potentially better suited, or at least more cost effectively suited, to meeting Australia’s more proximate defence needs beyond the 2030s. It also would have avoided the Chinese Government seeing the strategy as directed at containing China. However, the conventionally powered submarine strategy should also be seen as having been a high-risk investment in light of the uncertainties discussed above.

The process of switching to the new strategy appears to have created a new set of risks, and new costs. Rather than assessing decision making through a comparison of the capabilities of two different submarine designs, it should be assessed through a comparison of two different design and capability deployment processes, over several decades. This includes different access to capability through time and different value of that capability through much of that time, all of which is dependent on how external uncertainties play out.

Using a real options approach would focus on the differences between the two strategies, and the options and option value each might deliver through time. Notably, re-entering the basic design phase delivers new tasking flexibility with potentially high value. But it would also have focused on the range of Australia’s options whose value has been degraded because of the shift:

• delay in availability
• damage to the relationships with France
• possible reactions from China
• potentially heightened risk of being drawn into a major regional engagement
• likely increased cost, limiting options to do other things with extra taxpayer money.

There also remain important questions in relation to the management of non-routine work (emergency failure etc.,) or work where the delays in sending a vessel to the US or UK for repair could entail costly delays in access to vital capability. Questions also arise as to the potential for either the French or the new submarines to rapidly lose option value in the future if new submarine detection capabilities emerge and are deployed, for example in the South China Sea.

Where to from here?

A new direction for Australia’s submarine strategy appears to have been set, with bipartisan support. It is now critical the options that have been created with the reopening of the submarine design process are not wasted.

ACIL Allen has, many times, applied a real options framework to planning major long-lived infrastructure ⁵. Typically, this favours strategy with greater flexibility and lower levels of irreversible commitment to very high costs. Real options reasoning and probing will often support approaches that had in the past been ruled out as uncompetitive, based on biased assessment methods that undervalue the way that the creation and selective use of options can deliver both reduced downside risks and greater net value.

Options reasoning applied to defence fleet planning should support building greater flexibility into platform decisions. This means less focus on optimising the design for current perceptions of threat and current technologies, and instead planning for and executing a steadily evolving fleet design that exploits flexibility built into the strategy from the start. This includes identifying plausible directions in which technologies may evolve (own/ally and hostile) and factoring into planning the need to build resilience against, or flexibility to exploit such developments.

Of course, there is little here that has not been discussed elsewhere and, I assume, factored into planning ahead of the recent announcement of a switch. Military planners know a lot about planning with uncertainty, though not necessarily in relation to modern, options-based principles for comparing the economics of strategy alternatives in the presence of high levels of external uncertainty.

I will close with a few related questions that could warrant closer consideration:

• Would the switch in strategy still be justified were we to know that the current geopolitical tensions would be resolved within the next 25 years?
• Would the switch in strategy still be justified were we to know that technology advances over the next 30 years would make a large submarine easily detectable?
• What are the implications, in terms of capability value, in shrinking the planned fleet size from 12 to 8, especially in the light of more distant tasking?
• There have been reports ⁶ of the US Navy ordering Orca drone submarines from Boeing ⁷. Similarly, there are reports of advanced work by the Chinese on drone submarines ⁸ with suggested deployment from the 2020s. Might Australia’s longer-term need for submarine capability in its immediate region be better served by these emerging technologies?
• Is there scope for sustaining an adequate existing submarine capability for long enough to seriously assess these issues?

Of course, sometimes it is justified to commit to large and rigid infrastructure projects even though emerging technologies may in time deliver a better solution – that may be true here. In some areas, long lead times are needed, even though uncertainties are large, and some risk cannot be avoided. On the other hand, seeking to optimise strategy within currently proven technological and geopolitical possibilities can be very costly – this will almost certainly sacrifice the delivery of high value and cost-effective flexibility and options.

Experts on the ground understand the threats, opportunities, and uncertainties. They need to be supported by the right processes and tools for evolving and justifying policies and building capability. Too often, they have been discouraged from this by assessment protocols that embed serious bias. In the case of submarine strategy, such bias could prove very costly.

References

^[1]  Indeed, the AUKUS defence arrangement that has accompanied the announced strategy switch is likely to have been one of the major reasons for considering the switch. AUKUS delivers new options and new risks.

^[2]  ABC News Broadcast (16 Sep 2021) published here (see statement at around 35 minutes).

^[3]  Real options as investment tools were formalised in the late 1970s, and even then seen as addressing serious bias problems in the way that traditional cost-benefit assessments were being undertaken. Much of the interest came out of finance streams, with a strong focus on the tools for valuing options. These tools can be powerful, but their complexity often proves a strong disincentive to use. In fact, it is a mistake to view real options as being just a toolkit, when it is in fact an extremely powerful way of constructing the investment problem and its objective, and helping shape strategy design. This different way of looking at how the question are asked, and how this perspective can better guide strategy design and comparison of competing strategies, is what is meant here by the real options framework. Robust comparisons are often possible even where accurate valuation is not. Reverting to a seriously biased alternative, in the form of traditional investment planning tools, without probing differences in flexibility value, is not an adequate response.

^[4]  Compare the once mighty battleship, rendered obsolete in the space of a few years across the early 1940s, as aircraft carriers with nimble fleets of fighters and bombers proved far more potent. and GasMark.

^[5]  ACIL Allen has applied the real options framework to strategy planning and management across a wide range of fields, including: urban water planning in the context of climate and rainfall uncertainties; urban transport system planning in the context of technology uncertainties; R&D planning and assessment, ranging from single projects to entire strategies, such as reviews of the whole of CSIRO; strategy in relation to large-scale research infrastructure investment, including Australian strategy in relation to the Square Kilometre Array radio telescope; management of policy evolution in the context of rapidly changing information – such as in relation to climate change; and management of a range of evolving fleets of military capability. In many cases, the assessments have highlighted the very different pictures, and indicated strategies, that flow from more traditional assessment tools vs the real options approach.

^[6]  “Orca XLUUV: Boeing’s whale of an unmanned sub”, published online here

^[7]  With claimed capabilities for mine countermeasures, anti-submarine warfare, anti-surface warfare, electronic warfare and strike missions.

^[8]  “China reveals secret programme of unmanned drone submarines dating back to 1990s” published online, here.