Local Knowledge: Five Reasons Not to Centralise Treasury - Part I
Given all the column inches devoted to Treasury centralisation over the last decade, it may be surprising to find an article on why you wouldn’t centralise Treasury. The reasons to centralise have been well, and repeatedly, articulated by others and will be taken for as read here. This article however attempts to summarise some of the issues that a multi-national corporate needs to consider when weighing up the options. For the record, the author remains, on balance, convinced of the benefits of Treasury centralisation, whilst having little faith in ‘one size fits all’ solutions. The issue is finding the best fit solution for your organisation, given the specific structures and constraints within which it operates. So don’t consider this list definitive, but use it to stimulate your thought processes about the best central/local mix for your organisation.
The modern multinational is complex entity and its relations with external entities (suppliers, customers, shareholders, financiers) are complex in their turn. When considering centralisation of any activity, the key question has to be “will the central team have an adequate grasp of the complexities of this business or market to support it remotely”. If you are not sure that the central team can afford or achieve adequate depth of understanding, maybe you should do it locally.
Part of the complexity that needs to be assessed is related to local custom and practice. In the Anglo-American model, we are used to think about laws and codes of practice etc, but this is not universally true. In many countries where we do business, “custom and practice” is a more significant factor. We therefore need to ask whether we have a deep enough understanding at the centre of local custom and practice before we undertake to support a process centrally.
Another element of complexity relates to managing financial risks. In a number of industries there are significant variables arising directly from the nature of the business (e.g. tender to contract risk, particularly in the heavy capital goods industries). It is unlikely in many cases that a central function will have, or can develop in the short term, the depth of understanding of these business variables possessed by someone actually in the business front line; indeed there are several examples of major multinationals with central Treasuries, who have kept the identification of financial risk (as opposed to its management) at business level.
This is an extension of the “It takes a local….” point above, but extended to the environment in which Treasury works. The issue is local knowledge and awareness. Part of the Treasury task is to be aware of the environment in which we work and how it is changing. If the local environment is both complex and rapidly evolving (the People’s Republic of China might be an example for many corporates) then how well can a central, team, possibly far remote from the environment, stay in touch with the evolution? How aware can they be aware of threats and opportunities? If there is an opportunity to negotiate a better deal, for instance with a regulator, can this be more effectively done by a local or a remote team. The point doesn’t just apply to perceived exotic locations; how many European based Treasuries can accurately assess the impact on their US businesses of the Cheque 21 initiative?
A consensus seems to be forming around the view that forecasting commercial cash flows is one of the toughest nuts that Treasury has to crack. This is accompanied in many cases by a view that commercial flows are best forecasted by those who are close to the businesses that generate the cash flows. There is a case to be made for centralised cash flow forecasting based on statistical analysis, but in most corporations pragmatism rules and cash flow forecasting is left as a local function, even by centralised Treasuries.
In summary then; there are good reasons to centralise, and good reasons not to. Some elements are ripe for centralisation; mostly these are operational and repetitive tasks, capable of a high degree of automation. Functions requiring a high degree of technical knowledge and awareness of the external environment are far riskier to centralise. The best fit for your corporation will be determined by the nature of your business, the geographical locations of your businesses and depth in which you need to support those businesses. In practice, few Treasuries are fully centralised or decentralised; the art lies in deciding where on the spectrum you should be; how you get there in practice is the subject of a completely different article.