Uncertainty in business
When we refer to uncertainty in business we traditionally think about a future on which we have poor or no probabilities at all. For example, what will be the price of the oil within 2 years, more than 50 dollars the gallon or less? Will the Moore law continue? Is there any chance of doing a purely optic computer within ten years sweeping all the electronic companies? Or developing quantum computer? What about the medical progress, for example, telemedicine? Is it necessary to build new hospitals if most of the cares will be made at home? What about surgeons if robots make operations more precisely than they do? These are few examples of some uncertainties that cannot be reduced to a calculus using probabilities, even subjective, because these events will happen or not and it is impossible to guess with what probability. If they happen, business environment will dramatically change, whereas if not, nothing changes. All these examples fall in the realm of non-probabilistic uncertainty. Hereafter, when talking of uncertainty, we mean non-probabilistic uncertainty because we are convinced that disruption and innovation are not probabilistic in nature.
What are the main vectors of innovation nowadays? There are essentially twofold: startups and new economy tycoons taking advantage of a more or less enduring monopoly. The two phenomena are linked because the bosses of the new economy were generally creators of successful startups (unicorns). Let us remark that the older monopolies were generally not as eager to feed innovation as new moguls. Peter Thiel (2014), captured quite well this new situation by stressing the role of the new monopolies in the digital economy.
At this step, we are talking of tech startups run by transformational entrepreneurs which are striving for becoming unicorns These startups, whose number seems to be growing (Guzman & Stern, 2016), increase dramatically the uncertainty in business by two ways: on the one hand, introducing new technologies, products and services, and on the other hand, by the capital accumulation resulting from a temporary monopoly in case of success, because they will often use this money to try to reiterate their success with new ideas. A study by Hathaway and Litan (2014) suggests that that the risk of being toppled by innovation for an existing company is stronger coming from incumbents than from new companies. In other words, the successful new monopolies are more threatening than newcomers. This is consistent with many press releases mentioning Bezos, Musks, Thiel and others’ investments or acquisitions. One of the phenomena of the new economy is a very large dispersion in turnover and financial results. Very few startups are successful but with prodigious results (Dalle, Courtois, & Linassier, 2015).
These successful companies have the money to re-invest in research and support startup creation.
A good measure of the uncertainty is the number of new companies in the ‘Fortune’ ranking each year. It averages to about 20 newcomers each year, but in fact relatively few, less than 1% of the 500 are recently created newcomers, most of them are spin-offs or split parts of older companies. So that, one can guess that the probability for a company to be removed from the list by the competition of a new business issued from a startup or spin-off, is between 0. 5 and 1% each year. This is consistent with the fact that many companies are in the list for decades. We know that the probability threshold above which it is psychologically rational to consider that an event may impact you is about 10-4 and only 10-3 if you have the feeling that you control the situation (Pomerol, 2012, Chapter 9). Considering that it is a psychological trait of managers to be overconfident (March & Shapira, 1987), one can deduce that most of the managers only consider events with more than a 10-3 probability. Thus the probability of being, relatively suddenly, impacted by a new business, that we evaluate between 10-2 and 10-3, ought to enter in the range of consideration of many managers. Hence, uncertainty in business is not very high but, on the one hand, deserves consideration. On the other hand, it happens to be correlated to new technologies, especially information technologies and, in this field, the risk of being impacted occurs at a faster speed now with the introduction of very powerful chips and new algorithms, in particular learning algorithms and AI. For example, in the incubator AGORANOV, one-third of the 40 incubated companies uses AI technology. Some of them will undoubtedly create new businesses and will make older ones rapidly obsolete.
The usual way of reasoning in the risk, is to draw scenarios and then endow event nodes with probabilities. Then, by folding back from the leaves, you can determine the decision path which brings the maximum of expected utility (Raiffa, 1968) and Pomerol (2012) in a practical context.
When there are no probabilities at all or no credible probabilities, it is always possible to think about decision in terms of scenarios. We have emphasised the possibility to use contextual clues in place of probabilities (Pomerol, 1998, 2001). This resulted in the introduction of contextual graphs (Brézillon, Pasquier, & Pomerol, 2002). Contextual graphs were used in several practical situations for decision support. Observing the way people use these graphs we saw that there are essentially two means to deal with uncertainty in a world without probabilities (Pomerol, 1998, 2001): (1) postponing action and (2) having two or more irons in the fire.
Postponing action means that, while you could act, you try to save time by making irreversible decisions the latest you can without damage. Of course, all good management class books teach that: ‘it’s more important to make timely decisions than better decisions too late’, but the truth is that it is wise to make only reversible decisions as long as you can when you totally ignore what will follow on.
This leads to the second aspect of ‘uncertainty management’: to keep several irons in the fire. In the long run it is generally impossible to pursue simultaneously incompatible goals, but in the short run, it is rare that long term goals be immediately exclusive. For example, many companies are clever enough to pursue the improvement of mature products and simultaneously launch or at least prepare competitive new devices. We will see other examples in the next paragraph.
Assume now that you are somebody concerned by the type of uncertainty, we just talked about, that may hit your business. One way to be aware of the threats is to watch what is going on in your field. Thus, it is of foremost importance to watch what the startups are doing. One of the best ways for doing so is to invest in some of them, or at least to make what all the venture capitalists do: thoroughly study and pick information on all the emerging startups bringing new devices or ideas in your domain of activity.
Nevertheless, remember that information or knowledge does not reduce uncertainty perse, but one can hope that a thoroughly scrutinising may allow to spot weak signals of future changes. In your own business, postponing action means that you let the startups, in your domain, try as many novelties as they can, observe the results and choose when it seems that something interesting is ongoing. When a startup has invented a new device or service that works it is the moment of a quick decision, either to imitate the innovation, or to propose 34 J.-C. POMEROL a deal or an acquisition. By doing so, you limit the risk for your own business to be jeopardised. You also limit your investment in research by letting startups test possible markets. It is a kind of passive externalisation of the research because you do not invest at the beginning, just observe.
Thinking in term of active externalisation, for a large company, consists of encouraging innovation either inside or outside. In other words, it amounts to have several irons in the fire. Inside, it is the process of allowing small units to develop new ideas or products, pursuing, for example, researches on patents that are not yet exploited. However, it maybe very expensive and difficult to determine what is worth developing. On the other hand, it is possible to encourage people to quit the company with the possibility to ex to exploit some knowledge they originated inside the company and licencing them. This leads to the creation of spin-offs. A lot of big companies are now involved in such swarming programmes, which is one of the ways to explore the environment in their field and reduce the threats on their own business, because they still keep some links with the spin-offs and can reintegrate the knowledge in case of success. In AGORANOV, we have several such swarming programmes supported by large companies.
Thus, helping startups and facilitating their exploration and testing activities is, if not a way to control innovation, at least a mean to be aware of and be able to react to novelties reducing, ipso facto, uncertainty in your business. It has been observed that large companies more and more externalise research in universities (more than 50% according to H. R. Rawlings, president of the AAU, in a 2014 interview). But since many innovations issued from labs are developed by startups linked to universities, it is alike for these companies to externalise research to the couple universities startups. Paradoxically, the big companies of the new economy are now almost alone to develop, on the one hand, internal research probably because their bosses try to reiterate their initial success. But, on the other hand, they also buy many startups as do also traditional companies. In France, a financial journal (Les Echos) recently printed that ‘Big companies urge to buy startups for promoting divergent thinking’.
This journal gave an impressive list of the largest, French originated, multinational companies that bought startups. Taming uncertainty by not putting his/her eggs in the same basket and keeping many irons in fire is a reality in many wise companies.
Conclusion
The main competition in business, in the past and now, always comes from innovation.
Schumpeter’s ideas are still operating (Schumpeter, 1974), but there is now many differences.
The first one is that a large part of the capital is immaterial. The second is that a part of the rent is partially devoted to non-yet-existing innovations, via capital risk, while a century ago it was directed to the development of recently invented technologies such that electricity, combustion engine, phone, etc.
In the time of Schumpeter, most of the capital was invested in existing industries with huge physicals immobilisations (steal, railway, spinning, etc.…). On the contrary, nowadays a part of the capital is invested in startups, via various investment funds. Yet, more surprising some monopolists, feed innovation like GOOGLE or PAYPAL (see Thiel, 2014). This situation generates a very strong instability and uncertainty for investors and for annuitants. The rentier economy on which was based in the nineteenth-century capitalism is almost dead for this reason among others. Paradoxically, investment in startups increase business uncertainty for others, but it is also an answer to protect your own business, reducing the probability of being push out by keeping an eye on what is going on, according to the old adage that invites to keep two irons in the fire: one in your current business, the other in the startups that could impact it in case of success or even better, jeopardise the business of your competitors.
One can wonder how long this chain reaction can last: unicorns supporting new business development which in return paves the way to new successful companies. As far as there is a positive return for venture capital the system may continue. When the innovations relying on the web and the general networking will slow down, it will be probably be more difficult to make money with successful startups and the investors will go back to more steadily investments. Smaller returns favouring established businesses, we will go back to a strictly Schumpeterian period in which the largest part of the money goes to existing businesses.
It is plausible that this time is already here, according to Guzman and Stern (2016): ‘Even as the number of new ideas and potential for innovation is increasing, there seems to be a reduction in the ability of companies to scale in a meaningful and systematic way’. This implies that the return for investors, especially venture capitalists, is likely to decrease, leading for deceptive times and a decrease of the investment in startups. Unicorns were rare, they will become extremely rare.
Once more Schumpeter was right, there are economics periods whose milestones are the disruptive inventions: steam machine, railways, electricity, combustion engine, telecommunications and computers. Now, we are in the networking era. This epoch is very agile, because innovations rely on immaterial capital and services. This is why the startups are blooming.
The next era will be the DNA age, more technical, more demanding as regards material capital, and returns coming with longer delay. Hence, our guess is that the startup bubble comes to an end. Already, the number of startup creations is decreasing in certain sectors.