Letter number 104 of May 2017

  • NEW

News : Strategy in the digital age

by Pierre Dussauge

The emergence of the “digital economy” and the success of companies such as Google, Amazon, Facebook, Air BnB and others such as Uber, Blablacar or LinkedIn, whose business models would appear to be very different from those of “old economy” companies, has resulted in claims that strategy, as we used to see it not all that long ago, was ready for the scrap heap. Totally new approaches will have to be invented to understand the new reality and take relevant strategic decisions in this new context! Although it is true that an analysis of many of the entrepreneurial successes over recent years does present a challenge to traditional strategy models, it would be jumping the gun to say that nothing is as it was before and that our traditional approaches have lost all relevance. 

First of all, it should be noted that the large majority of digital success stories have been achieved by companies with a very special type of set up.  They’re called platform companies. Such companies play no role in the production of goods or services by inserting themselves into a production chain that runs from production of raw materials to final user, but play an intermediary role between groups of different users who need each other. Without the platform, these groups would find it very difficult to make contact with each other or they would interact with each other a lot less efficiently. Uber puts drivers in contact with potential passengers who, without the eponymous application, would be unable to meet up, would be faced with major uncertainty with regard to fares and both drivers and passengers might well fear for their safety. Uber facilitates contact, “regulates” fares charged and creates, through its reciprocal rating system, the trust that is indispensable for the smooth running of the transaction. Air BnB or Blablacar, in other sectors, do much the same thing. However, offering a platform for putting diverse users in contact with each other is a lot different from playing a role in the construction of value added on the production chain for goods or services. Consequently, approaches such as Michael Porter’s famous “5 Forces” model, designed for analysing power relationships within a sector, is totally inappropriate for any relevant analysis of the platform business.

These platform businesses have very specific features which alter the terms of competition among rival platforms.  They are often impacted by what we call “network externalities”, i.e.  the attractiveness of the platform for users increases in line with an increase in the number of users. The greater the number of Uber drivers, the more attractive the service is to passengers, and the greater the number of Uber passengers, the more profitable the service is to drivers, which creates a virtuous circle in which success begets success. One might be tempted to think that nothing is as it was before and that competition in all sectors has become an endangered species with the emergence of one dominant player. In any event, traditional strategy models may be starting to look old-fashioned as they encourage the creation of greater value for customers at lower costs than competitors, with the idea that a slightly higher level of efficiency on at least one of these two parameters will enable a company to compete successfully against its rivals on the market.

Hold it right there! It is undeniable that we need to adapt our approaches and ways of thinking to the specific features of business platforms.  But in doing so, the fundamental concepts of strategy remain relevant. The virtuous circle arises because more users, on the one hand, and drivers, on the other, increase the value of the service offered by Uber. This is boosted by size effects which result in Uber’s unit costs dropping with the rise in the number of users. And if Uber’s managers had thought in these terms – the terms of traditional strategic reasoning only – perhaps they would have realised that “network externalities” for their type of activity are extremely geographically delineated. If I live in Berlin, London or Madrid, the number of Uber users and drivers in Paris will only have a very minor impact on the perceived value of their service for me. Similarly, the volume of activity in one city will only impact very marginally on the unit costs of the service in other cities. This simple and very mundane strategic analysis could have saved Uber the trouble of barking up the wrong tree in China. This type of analysis also explains why Blablacar is dashing to set up shop in as many countries as possible. Its management seems to have understood that “network externalities” are operating in their favour, but country by country. So they need to set up shop and create a substantial presence before anyone else does.

Yes, there are some traditional strategy tools that have perhaps become less relevant for analysing the strategy of digital companies. But the underlying concepts and approaches remain entirely valid. In Porter’s “5 Forces”, it is less the forces themselves that are important than the idea that competition conditions relating to a given activity must be carefully examined in order to decide whether it will be a good strategy or not.

Moreover, it should be noted that platform companies are not something new. Auction houses such as Christie’s and Sotheby’s are platforms that put buyers and sellers of artworks in contact with each other. Similarly, estate agents put people who want to sell houses in touch with people who want to buy them. What has changed is that digital technologies have led to the multiplication of the number of activities for which a platform company is able to considerably improve the value of the service for different groups of users seeking a more optimal way of getting in contact with each other. Ebay for example, provides access to many more objects for sale than traditional auction  houses, so sellers are able to reach a large number of buyers throughout the world, not only those who have chosen to turn up at the sale. But the explosion of the number of platform companies that we have witnessed over recent years cannot go on indefinitely. Sooner or later, the number of activities suitable for digital intermediation will start to fall off and it will be increasingly difficult to create new digital services of this type.

All in all, the digital eruption should encourage us to extend the scope of strategic thinking, to understand in which contexts the different approaches apply, and to be more aware of the limitations of the tools being proposed. More fundamentally, what these considerations suggest is that the tools and models are more specific to economic, sectorial, technical, historical or even geographical contexts than the underlying perspectives and concepts, which do have a more universal and timeless nature.

Pierre Dussauge is Professor of Strategy at HEC Paris and the creator of the fully digital strategy training course, strategy@HEC Paris, which enables learners, over 6 months and going at their own pace, to review or acquire the basics of strategy, a subject that has vital links with finance. For more on this programme, click here.

Statistics : Average maturity of government debt

The US government is thinking about issuing ultra-long debt dated debt (40, 50 or even 100 years) which would be a première in the USA where the longest debt issued so far is with a  30-year maturity.

Long ago, the UK went in that direction, which explains the maturity of its government debt roughly double that of other major countries. It is helped by the way pensions are financed in the UK: funded pension schemes and not pay-as-you-go schemes.

Research : Investments: the CEO cycle

With Simon Gueguen, teacher-researcher at the University of Paris-Dauphine

There are several factors that influence the level of investment of companies: business cycle, financial constraints, degree of regulatory stability, political uncertainty, etc. In a world where the agency problem did not exist, CEOs would always choose value-creating projects taking constraints into account. The article that we present this month [1] shows that there is in fact a “CEO cycle”, i.e. that investment choices vary in line with the time during which the company’s CEO has been in office.

The study, which covers 3,000 US firms between 1992 and 2009, looks both at investments (internal investments such as the acquisition of fixed assets and investments in external growth) and at disposals. The main conclusion is as follows: early on in a CEO’s term of office, disposals are frequent and then they gradually fall off. The opposite is true for investments which are relatively infrequent early on although they increase over time.

The effect is significant. The average investment rate measured by Pan et al[2] is 9.4%. A standard deviation increase in the number of years since the appointment of the CEO, i.e. four years, results in an increase of two percentage points in the investment rate (half a point per year). This is the same order of magnitude as for effects already identified in academic literature (business cycle, financial constraints, etc.).

This result can be interpreted in several ways. For example, the explanation could be based on the fact that changes in CEO most often happen during periods of low investment. However, Pan et al verify that the CEO cycle appears regardless of the reason for the change, ill health, death or retirement of the predecessor, change notwithstanding a good performance, etc. Another explanation could be the learning effect. If the CEO’s ability to identify good projects increases over time, it is only rational to invest towards the end of his/her mandate. Here again, Pan et all exclude this possibility because on the contrary, they observe a deterioration in the quality of investments chosen over time[3].

Finally, the explanation advocated is that of a CEO cycle linked to the gradual increase in agency problems. For a long time, we have known that CEOs have a tendency to overinvest (for reasons relating to compensation or prestige): we refer to this phenomenon as empire building[4]. By obtaining a certain amount of control over the board of directors over time, CEOs can gradually increase the amount of investments, by choosing them on the basis of criteria that are specific to themselves and that do not necessarily result in value creation. Then, when a new CEO is appointed, he/she gets rid of some of the assets resulting from these investments, and a new cycle begins.

Pan et al suggest a rather radical solution for avoiding overinvestment: frequent changes in the management teams of listed companies. This conclusion remains open to discussion given that the effects of the stability of the management teams in terms of all of the company’s policies (choice of investment and others) are multiple and are yet to be studied. But these results once again highlight the specific role of managers in the performance of companies, in a world where the preferences of the different agents involved are not always aligned.


[1] Y.Pan, T.Y.Wang and M.S.Weisbach (2016), CEO Investment cycles, Review of Financial Studies, vol.29(11), pages 2955 to 2999.

[2] Sum of acquisitions of fixed assets and external acquisitions related to the assets’ book value

[3] To do so, they measure market reaction to the announcement of these investments

[4] This idea is presented by Nobel Prize winner Oliver Williamson (1964),The Economics of discretionary behavior: managerial objectives in a theory of the firm, Prentice-Hall


Q&A : What is the cost of capital for a bank?

The cost of capital for a bank is its cost of equity, not the weighted average of its cost of equity and of its cost of debt. The cost of equity is measured in the same way as the cost of equity for a non financial company using the CAPM and its equity beta. 

 This cost of equity is then compared to the return on equity to check whether the bank, over the course of several years, created or destroyed value.

 One might ask why there is a difference between financial and non-financial groups on this point?

 The main reason is that a bank is fundamentally nothing but an intermediary contrary to a non-financial firm. This is illustrated by 2 facts: it does not price its products (loans) quoting a price (€x), but a spread: Euribor + x%, Euribor being its marginal cost of funding (the price of its raw materials). Its turnover is not the sum of all interest received on loans. It is the sum of all spreads received on loans. 

 The concepts of fixed assets and working capital which are key for a non-financial firm, and are the basis for computing return on capital employed (ROCE), are useless for a bank: customers pay what they owe in due time without payment delays as the bank holds their accounts: it simply take the money on the due date! Inventories are inexistent, suppliers of money are paid their interest if any without payment delays. Tangible assets are very small compared to loans granted by a bank.

 If ROCE is not relevant, cost of capital in its widest acceptance is not either. Hence, you are left with ROE and cost of equity.

New : Comments that we have recently posted on the Vernimmen Facebook page

Regularly on the Vernimmen.com Facebook page[1] we publish comments on financial news that we deem to be of interest. Here are some of the comments published over the last month.

And LVMH becomes the top French market capitalization with €121 billion (between Total at $119bn and Sanofi at €116bn).

The Arnault family bought the Boussac-Saint Frère textile group in 1984 on the brink of bankruptcy. The latter had among other assets Dior Couture, Dior perfumes having been sold years ago in Moët-Hennessy to pay down debts. At the end of the 1980s, the Arnault family had set up a cascade structure to finance its acquisition of control of the LVMH-Moët-Hennessy - Louis Vuitton group, which was then plagued by dissensions between the Vuitton family and its CEO, Alain Chevalier (former CEO of Moët-Hennessy). The Arnault group was made of a stack of holding companies, each holding a stake in the lower level holding and an industrial asset likely to attract investors in addition to the mere holding of control of the lower-ranking holding company.

Over time and with the successful development of the LVMH group, the Arnault family has gradually bought out all the minority interests in the LVMH control line and outdated the intermediary holdings that have become useless (Financière Agache, Au Bon Marché) to the exception of Christian Dior. Christian Dior is a holding company that owns Dior Couture and controls LVMH via a 41% shareholding and 57% of the voting rights. On Monday, it was worth €41bn. The Arnault family owns 74% of this listed holding company.

On April, 25th was announced a deal that could allow the Arnault family to own 100% of Christian Dior and therefore 46.8% of LVMH, given a direct stake of 4.6%, against today economic interest of 36.2 % 

Rather than a merger that would have diluted the control of the Arnault family on LVMH, it is planned:

1 / the purchase by LVMH of Christian Dior Couture for €6.5bn, ie 15.6 times its EBITDA;
2 / the purchase of the minority shareholders of the holding company Christian Dior, either in cash or by remitting 8% of the share capital of Hermès still held by the Arnault group (which is worth € 4 billion based on a market capitalization of €49bn for Hermès).


For the Arnault family, the gross investment is €12.2bn, and the net investment is €1.7bn, net of Hermès shares (€4bn) and Christian Dior Couture's sale price in the holding, corresponding to 2 years of its share of the LVMH dividend.

For LVMH, in addition to the satisfaction of bringing together the two main components of Christian Dior, net indebtedness will rise from 0.4 x EBITDA to 1.1 x, which is a non-subject as illustrated also the percentage of net debts on the value of equity: 8%. . . It should be noted that the growth in LVMH's EPS induced by this transaction is irrelevant since it results from a cost of debt lower than the inverse of the multiple of operating income 4.15% (reverse of 24 times). For further details on this important point, see Chapter 26 of Vernimmen.

The multiple paid for Christian Dior Couture is logically a transaction multiple higher than LVMH's own multiple (15 times the operating result, which is not a transaction multiple but a trading multiple). It is justified by Christian Dior Couture higher growth rate than that of LVMH: + 24% per year since 2011 compared with + 6% per year for LVMH. The gap is strong, although it is likely that it will not last in this magnitude.

For the shareholder of Christian Dior, the transaction results in a sale on the basis of the net revalued assets, which shows the fairness of this offer. Unfortunately, not all majority shareholders have this behavior. . . It will be noted, and this is not lack of salt, that the shareholders of this holding company who were there for Christian Dior and Louis Vuitton (for simplicity), are offered Hermes shares to leave!

Hermès shares that they should sell to acquire LVMH shares if we believe the price trends when the deal was announced: - 4.5% for Hermes and + 3.9% for LVMH.

It remains to be explained why this transaction, expected for 25 years according to analysts, takes place now. In fact, when a group is doing better than good and has no other investment opportunities, it buys its minority interests. What Bernard Arnault implicitly confirmed yesterday: "The best assets (of the sector) are not for sale" he answered when asked about the future external growth of LVMH.