Strategy

40 / Rotman Magazine Spring 2007

The vast majority of today’s business decisions are made with incomplete information and in the face of an uncer- tain future. Indeed, with the exception of a reasonable expectation that the sun will come up tomorrow, few things in human affairs can be predicted with certainty. The term ‘risk management’ has come to serve as a set phrase in the financial services industry, referring to a formal process for evaluating and controlling a company’s total exposure to loss together with estimates of the probability of loss. However, this term deserves to be reclaimed to a broader usage, as all companies have to deal with risk.

While risk can never be completely avoided, well-trained man- agers have many ways to take uncertainty into account in their strategy planning. Successful long-term planning includes a combi- nation of the following six techniques.

1. Risk Identification In most business situations it is remarkably easy to identify the risks that a firm is facing by simply asking, “What could go wrong?”

For example, if a product depends on components, the inability of a supplier to ship parts when needed is a risk that can be readily anticipated. Where a firm is attempting to develop a new technol- ogy, its research and development efforts will be successful in achieving desired performance – or not.

This is not to say that all risks can be foreseen. Former U.S. Defense Secretary Donald Rumsfeld famously observed that there are both ‘known unknowns’ and ‘unknown unknowns.’ In a military situation, he meant that an army might know that its adversary had troops over the horizon, but not exactly how many (a ‘known unknown’); and there might be some uncertainty as to whether the enemy had certain types of weapons (again, a known unknown). However, his dictum was meant to alert planners that, for all the variables that they could list and worry about, there would always be some others that could not have been anticipated (the ‘unknown unknowns’).

It is hard to describe an ‘unknown unknown’ in business, simply because it is just that – unknown. However, an example would be

A Primer on the Management of Risk and Uncertainty

By David Robinson

Effective risk management depends on a clear understanding of what a firm can control and what factors are beyond its control. Using a combination of the techniques described here can go a long way to effective strategic planning in the face of risk.

Strategy

Near Misses

Plan

ROT045

This document is authorized for use only by Familia Herring in Strategy at Strayer University, 2018.

Rotman Magazine Spring 2007 / 41

something like the emergence of a completely new technology, such as the advent of transistors in electronics. The efficiency and minia- turization offered by these solid-state devices could not have been predicted from the existing development work on thermionic valves.

How can business leaders plan for the unknown unknown? At the extreme, they could calculate the cost of project abandonment due to unforeseen circumstances. It is usually possible to imagine the effects of an unknown unknown, even when a specific cause cannot be anticipated. For example, prior to 2001, the managers of the New York Stock Exchange would have had little reason to include terrorist attacks in their strategic planning. However, they could reasonably plan for ‘something’ (an unknown unknown) that would make it impossible for them to use their own trading floor: burst water-pipes, bomb threats and catastrophic fires in neighbor- ing buildings were reasonably foreseeable as rare-but-possible events that could interrupt access to their building.

One useful source of ideas for things that can go wrong is ‘near misses.’ For example, in the loss of the Columbia space shuttle in 2003, the vehicle broke apart upon re-entry, a problem that had never been seen before. However, the cause of this disaster was damage to a heat-shielding tile that in turn was the result of a strike by fuel tank insulation that had broken away during the launch. The

Control

loss of insulation had been observed before, and was a matter of ongoing concern for NASA engineers.

In the business realm, the hint of a near miss comes when a suc- cessful tactical maneuver is accompanied by exclamations of, “phew – we were lucky!” For example, suppose a firm runs an overly-gener- ous mail-in coupon rebate program, but is saved from crippling expense by the lucky coincidence of a major storm that kept most shoppers out of the malls during the time of the promotion. This is a clue that in future, the firm could face a predictable loss if rebate programs are not more carefully planned and controlled.

2. Risk Assessment There are several ways to take risks into account and develop ways to manage them. The first step is to consider when trend shifts and adverse events might occur, and then to make an assessment of the severity of the impact of these future states. Once risks have been identified, the impact of each can be assessed and assigned to one of three categories: noticeable, moderate and severe.

‘Noticeable’ events are things such as the market entry of a small competitor with an offering that appeals to a market niche; the competitor’s eventual entry has been predicted, and as long as the competitor stays in its narrow market, there is a measurable- but-small loss in sales that requires no particular response. Events that have moderate impact might include the failure of a supplier in a market with many generic alternatives, or the entry of a competi- tor with equivalent technology. A catastrophic impact would be something that makes the current strategy infeasible, such as the

Unknown Unknowns

This document is authorized for use only by Familia Herring in Strategy at Strayer University, 2018.

42 / Rotman Magazine Spring 2007

overturning of a patent or the emergence of a severe side effect to a drug – as happened to Merck with the Vioxx painkiller.

When these dimensions have been added to the list of risks, they can be organized into a matrix that gives general guidance for how to manage them, as shown in Figure 1. Consider first the col- umn of ‘immediate-impact events.’ Few firms have to face a catastrophic event with little warning. Terrorist attacks and the outbreak of civil war certainly count in this category – but thank- fully, these are rare.

Where there is a change in the environment that has an imme- diate impact that is moderate, the prudent managerial course of action is to quickly shift additional resources to the current strate- gy. For example, when a major competitor enters a firm’s principal market, a substantial investment in promotional spending (and a new advertising approach emphasizing either a defensible core position or direct comparison) is likely warranted.

In the case of an immediate event that is assessed to have a merely ‘noticeable’ impact, firms should continue with the strategy already in place and make only moderate adjustments. For example, suppose a niche-player enters a firm’s established market. The firm would continue the same level of promotional spending, but might retrain salespeople with new talking points to specifically address the competition. The key to success here is in correctly assessing an impact as ‘noticeable,’ but not over-reacting. Temporary promo- tional pricing by close competitors is an example of such a situation that should not require a complete rethinking of strategy.

In the intermediate time horizon, where the risk to the firm is not imminent, the key success factor is to anticipate the risks to cur- rent strategy and then identify and monitor the relevant metrics. For example, it is often anticipated that a firm’s customers will migrate from one technology to another at some unknown point in the future. Prudent managers will handle this by surveying customer interest and monitoring the pricing of competing technology. For example, for many years manufacturers of cathode ray tube comput-

er monitors were well aware that flat panel displays would eventual- ly take over their business. However, early flat panels were very expensive. There was no way to predict whether the change-over would be sudden (as happened with adoption of home use of the Internet) or would occur only as consumers replaced their comput- ers (about once every four years). As long as careful observation is in place and managers are not complacent, a firm can take some time to develop alternate strategies.

Where the intermediate future includes some possibly-cata- strophic events, managers should test strategies against the cost of project abandonment. For example, in the tourism and hospitality industry, avian influenza or other contagious diseases might close down all intercontinental travel for a period of time. The prudent course in the face of such a risk is to avoid over-investment that would ‘bet the company’ on one project.

When threats are far distant, the best approach is to have a comprehensive view of strategy planning and to fully explore all the options available. This is especially true in the middle case, where perceived adverse events are likely to have a moderate impact on the company’s operations. Where the impact will be catastrophic, the firm should have an exit strategy in mind. Often this will simply be the abandonment of a specific market, with investment-to-date treated as a sunk cost. For far-distant threats that are likely to have only a minimal impact on the firm, identification of the relevant metrics – such as market share or consumer sentiment – and regu- lar monitoring are all that is required.

3. Risk Tolerance and Diversification All firms and their management teams have a certain tolerance for risk, usually tied to the company’s age, with young companies often willing to bet their entire net worth on a single strategy, while more mature firms may become positively risk averse. Between these two extremes, many firms actively manage their exposure to risk through a ‘portfolio approach.’

Investors are familiar with the portfolio approach to investing, whereby they tolerate owning risky instruments by only putting a portion of their net worth into each one. A well-balanced portfolio is one with investments in many different industries, so that an adverse event that affects one investment will not impact the others. Firms use the same portfolio approach for projects, especially those associ- ated with new product development. A simple product modification to adapt a product to an existing market is low risk, whereas entering a new industry with no developed market is very high risk.

Under a portfolio approach, a firm would not completely avoid high-risk projects, but would limit the number of such projects that it undertakes simultaneously, balancing them with low-risk/moder- ate-reward projects. For example, a firm with too many high-risk projects might license out some patents to other firms to avoid the expenses and risks of new product introduction for some lines of business, while concentrating on the in-house launch of other ideas.

4. Risk Transfer When risk can be clearly identified, one way to eliminate it is to transfer it to another entity. For example, almost all homeowners transfer the risk of loss of their house by fire to an insurance com-

CATASTROPHIC Abandon Manage with Plan an current strategy this in mind – exit strategy and develop a don’t new one over-invest

MODERATE Shift Pre-plan an Research significant appropriate strategy resources to reaction options counter

NOTICEABLE Modest change Plan a Develop in resource possible metrics allocation resource shift and monitor

NOW SOON FAR DISTANT

Approaches to risk management based on immediacy and impact

Strategy

Effect

Figure 1

This document is authorized for use only by Familia Herring in Strategy at Strayer University, 2018.

Rotman Magazine Spring 2007 / 43

pany. Similarly, commodity producers can transfer the risk of price declines by trading in the futures markets to lock in a price and commodity users can avoid price shocks by purchasing future contracts. Southwest Airlines used this technique to avoid a rise in jet fuel prices in 2005. Firms whose strategies are impacted by foreign currency exchange variations can reduce their risk by hedging transactions.

Changes in ownership structure can also reduce risk. For exam- ple, a company that sells its headquarters building to a real-estate investor and then leases back the building transfers the risk of future value of the building to the investor. Taking on a joint ven- ture partner can reduce a firm’s capital exposure to a risky new market. For example LG.Philips LCD is a a joint venture between South Korea’s LG Electronics Inc. and Royal Philips Electronics NV of the Netherlands. It spreads the risk of loss from the massive investment needed to set up a flat-panel TV manufacturing facility, when consumer adoption of the new technology was unknown.

So why don’t firms simply transfer away all their risk? First, there is no insurance or hedging that will protect against what is called ‘ordinary business risk,’ such as the risk that consumers won’t like a new product. Second, all insurance comes at a price. When Southwest hedged its fuel costs, JetBlue, an airline with a similar business model, guessed that fuel prices would decline and decided not to purchase future contracts. The firm that engages in sale-and- lease back gives up the possibility of a capital return on the asset, and a firm that engages in a joint venture gives up part of future profits. In sum, risk transfer is never without cost.

5. Scenario Analysis In his seminal text Competitive Strategy, Michael Porter wrote: “The device of scenarios is a particularly useful tool in emerging industries. Scenarios are discrete, internally-consistent views of how the world will look in the future.”

Scenario planning can be contrasted with forecasting, where careful analysis of historic data is used to offer a precise estimate of a parameter in the future, such as the specific level of demand for a firm’s product. Unfortunately, precise numeric estimates often give a false sense of certainty of what the future will be like. For exam- ple, the worldwide demand for domestic-use DVD recorders in 2005 was estimated at 47 million units, but in fact only about three million were sold.

The difficulty of extending trends into the future can also be illustrated with the example of the market opportunity that has opened up due to the increasing number of Americans who are obese. Furniture manufacturers and hospital equipment suppliers have adapted designs that accommodate people of high body weight. But will the trend continue? From 1980 to 2006, the num- ber of obese Americans rose from 23 to 60 million people: but does that mean that we can expect an additional 37 million obese cus- tomers in the next 26 years, or will the trend reverse itself due to cultural shifts and new treatments for obesity?

The process of constructing scenarios is fairly simple. First, identify the relevant variables, then identify whether they are dis- crete or continuous. Next, gather the variables together to make specific ‘scenarios.’ Remember that Porter said that scenarios must

be discrete and internally-consistent. That means that one scenario cannot overlap another, and ideally all possible futures can be described within the list of scenarios. The requirement for internal consistency may eliminate some scenarios entirely. For example, if the economy is in deep recession, it’s unlikely that consumer spending will be high.

Coming up with a manageable, discreet set of scenarios is not as difficult as it sounds, because of likely links between variables. Consider the situation faced by film maker Kodak in the early 1990s. Effective digital camera technology had just been announced, and the uncertain future contained these possibilities:

(a) Digital cameras remain expensive and are only adopted in professional markets; (b) over a long period of time, digital cameras are adopted by professionals and a small number of ‘serious amateurs,’ and the price of components remains high; (c) digital cameras move to the mainstream over several decades (similar to the adoption of automatic transmission in U.S. cars from 1940 to 1960) and prices of components gradually fall; (d) ‘Grandma goes digital’: fast and broad adoption of the digi- tal format accompanied by a sharp decrease in use of film cameras and a lowering in component costs for digital cameras.

Note that while the cost of components for cameras was uncer- tain 15 years ago, it was reasonably predictable that with mass-market adoption would come manufacturing inefficiencies resulting in lower prices. As a result the possibility: “Everyone wants one and the prices go sky high,” is not in the set of likely scenarios.

Once the scenarios have been identified and labeled, it is pru- dent to make an estimate of the likelihood of each one. For discrete variables, when there is no credible external information, each out- come is equally likely. For example, the chance that a competitor will or will not decide to offer a product in a particular category should be estimated at 50 per cent. However, if the competitor has signaled intentions (“We find Asian markets particularly attractive for future growth”), then the probability of market entry could be estimated higher, say 80 per cent. For continuous variables, such as growth of the economy, planners can use intuition or the average

1. List the critical variables 2. Constrain continuous variables into low, medium and high values 3. Combine variables and eliminate implausible groupings 4. Give easy-to-remember names to each scenario 5. Estimate the likelihood of each scenario 6. Develop a concise but exhaustive list of possible strategies 7. Assess the likely payoff for each strategy against each scenario 8. Develop an algorithm to choose such as minimum acceptable pay-

off, maximum tolerable loss 9. Select a strategy and re-test it using regret analysis

Steps in scenario planning Figure 2

This document is authorized for use only by Familia Herring in Strategy at Strayer University, 2018.

44 / Rotman Magazine Spring 2007

value of published experts’ opinions (individual expert’s opinions are often wrong, and ‘consensus’ of experts who listen to each other’s forecasts are not particularly reliable.)

Fortunately, if scenario planning is done correctly, estimating which scenario is most likely is the least important part of the process. Rather, managers must try to choose a strategy that will be beneficial no matter which scenario plays out.

6. Regret Analysis This essentially entails a ‘re-check’ of the attractiveness of the cho- sen strategy. When a viable strategy has emerged, it should be con- sidered a ‘candidate’ for the firm’s planning, and managers should conduct one last survey of all the options that have been eliminat- ed and reassess them to evaluate ‘regret.’ Regret analysis is a form of protection against biases that tend to creep in when planners begin to believe with certainty the likelihood of one scenario.

As an example, imagine a company that is considering selling software in the China. A major unknown about the future is whether the Chinese government will strictly enforce intellectual-property protection. A reasonable extrapolation from historic experience is that despite top government commitment to enforcement, there is still rampant piracy of computer software (more than 96 per cent of programs in use have been pirated). A thorough scenario analysis suggests that no strategy for market entry (high price, high price and economy versions, economy version only) can lead to any reasonable payoff for the foreseeable future.

However, what does regret analysis tell us? One scenario, though extremely unlikely, is that the future includes a fail-safe method to prevent the copying of software. Competitors who entered early have achieved brand recognition and our firm’s belat- ed entry would be shut out of an emerging profitable market. The level of regret would be high. After the regret analysis, the firm takes another look at the market entry strategies, and chooses the one with the best chance of establishing a recognized brand, albeit at little immediate incremental profit.

In Closing Effective decision making depends on managers clearly under- standing what their firm can control and what factors are beyond its control. The systematic identification and assessment of risk, a frank discussion of risk tolerance, and the use of a variety of the techniques described here can go a long way to achieving effective strategic planning in the face of risk.

David Robinson teaches Marketing at the Haas School of Business, University of California, Berkeley, and has also taught at the Stanford Graduate School of Business. He writes a weekly business column for The San Francisco Chronicle. A longer version of this article is available on his Web site at

http://faculty.haas.berkeley.edu/robinson

R E U N I O N 2 O O 7

This document is authorized for use only by Familia Herring in Strategy at Strayer University, 2018.