Treasury Today Country Profiles in association with Citi

Centralisation – should companies revisit the treasury model?

Have approaches to organising treasury changed at all in the past few years? Once, centralisation was a key mantra of treasury. But advances in technology, coupled with new counterparty and other more invisible risks that companies face today, mean that a completely central model may not be necessary or even beneficial. If you are considering reorganising your treasury or setting up new structures, here are some key points to consider.

The debate between the proponents of regional treasury and those promoting centralised control from headquarters is by no means new. The arguments change from time to time but they can be summarised in two simple questions. Should the head office function be to gather together all the liquidity to maximise returns and manage global risk? Or should regional offices be in daily control to minimise transaction costs and local risk exposure?

Multinationals centralise their operations for reasons of convenience and cost. Above all, central control – based on comprehensive knowledge of where the cash balances are – means that the financial reporting lines – from treasury to the CFO and board of directors and, ultimately, to investors – are clear.

This model applies best to those companies which operate in countries which are broadly similar in terms of banking and regulation. Millions of dollars scattered in local branches of a global bank can be consolidated at home, giving the treasurer immediate visibility. A multinational headquartered in any of the OECD countries can be run with a handful of core functional currencies (dollar, euro, yen and sterling) all of which are freely convertible against each other.

Adding one or two other currencies for geographical reasons (US companies might use the Canadian dollar and Mexican peso, for example) or for industry sectors (Australian dollar or South African rand for commodity-based businesses, for example) is not too difficult. Customer and supplier credit, bank lines and cash management can therefore be handled relatively simply even if the sheer volumes may dictate some complexity.

The practicalities of centralisation

But centralisation is much more difficult in practice, particularly for companies that operate in countries which do not all share the same ways of doing things, or in industry sectors that have differing practices. There may be restrictions on repatriation or convertibility of local currencies or regulatory impediments. A list of some of these barriers to centralisation would include: tax treatment (ranging from withholding tax on interest, stamp duty and other charges and also double taxation treaties which can be incomplete, bilaterally unfavourable or simply burdensome), transaction reporting to central banks, legal agreements, contractual arrangements and local best practice.

The management guru’s saying of ‘Think global, act local’ might be appropriate guidance for manufacturing processes and marketing, where a car or computer can be produced on a global platform and tailored to local needs. But treasury management is not suited to the disciplines of mass production. The financial supply chain is not based on a gradual accretion of value but an intricate sequence of debits and credits, each of which contains a financial opportunity and a related risk.

A risky business

The recent financial crisis has also highlighted how the risk components of treasury management have become clearer in their linkages, and consequently more dangerous. The traditional need for visibility has been joined by the need to search for what might be invisible. As Donald Rumsfeld, the US Defence Secretary, put it so memorably: “As we know, there are known knowns; there are things we know, we know. We also know there are known unknowns; that is to say we know there are some things we do not know. But there are also unknown unknowns – the ones we don’t know we don’t know.” Although he was ridiculed for these apparently confusing words, they do hold a fundamental truth.

If you look back just two years ago, American International Group (AIG) was the world’s biggest insurance company with a triple-A rating from the agencies. Greece was an investment-grade credit in the Eurozone with its national accounts audited to the (political) satisfaction of the European Union. AIG, now effectively nationalised by the US government, was laid low by its incursion into financial guarantee insurance. AIG’s implosion was, ultimately, manageable.

The US government stood behind it (at present a $180 billion bailout although the final cost may well be less) but its real businesses were strong enough to attract suitors. But it could have had catastrophic consequences given that the insurer’s product lines impact everything from corporate cash flow security to retiree employment benefits and company contributions.

The Greece debacle was also a bolt out of the blue. Although the European Union had threatened fines against the country for its flouting of the Growth and Stability Pact’s ceilings on the public sector deficit and total debt, Greece continued to borrow at euro interest rates and investors continued to buy. Only a massive last minute intervention from the combined weight of the Eurozone’s Central Bank (backed by credit lines from the Bank of England, the US Federal Reserve, the Bank of Canada and the Swiss National Bank), non-euro EU members and the IMF, stopped a disaster.

The treasurer’s role in getting centralisation right

These two single events illustrate better than any other how treasurers need to be aware of risks that go well beyond the day to day business. Treasurers have to perform an uncomfortable balancing act between highly technical processes and seeing the bigger picture as it evolves. Because treasury is facing all types of new and invisible risk, treasurers are once again looking at their structure and approach to see which model works best to identify and stop threats to the business. Is it full centralisation, in theory to be able to see all risks? Will local management produce better control and understanding of local risks? Or is a hybrid model of centralisation combined with localisation, the best of both worlds? Can it really work in practice?

A few guidelines can help identify which model or combination will work best for treasury. First, cash is king, no matter where it is. If you have subsidiaries which are fairly self-standing, there is no advantage to centralising control.

Accounts payable and receivable under control, a reasonable cash balance in a safe local bank (or group of banks) and credit lines to cope with variations in the local demands for working capital will take care of themselves because there can be the right balance of stock. “Inventory management is key,” says one treasurer in the distribution supply business. Efficient stock management reduces funding costs (and, therefore, interest expense) and avoids the need to discount or additional marketing costs. Business management and treasury management can work hand in hand at the local level.

The local approach – pros and cons

In the local model, bank relationships are also simple and the technology or physical processes are equally straightforward. The latter point may be contentious when you consider the huge amounts of paperwork which remain the norm in many non-OECD countries. Paper is slow and time-consuming but it can only be managed locally. A central treasurer working on the basis of weeks-old documentation would be better off with guesswork. The local treasurer is in the best place to know. “Centralisation is the best thing,” says one treasurer, “but there are certain functions that need a physical presence.” In many countries, the cheque or bill, which requires local handling, remains the most secure and legally enforceable means of payment.

In the same circumstances, credit risk is easily managed. Local sales teams know their customers and managers know their suppliers. Following best practice will keep credit insurers happy (and premiums low) with relatively straightforward claims and collections procedures. In established businesses, the cash flow forecasting should also be quite simple. Seasonal variations can be built into modelling. It is boring but reliable.

But local management can become complacent. Danger signs which might be apparent to an outsider can be ignored if they just creep up. That was the Greek problem. Unlike Enron and Lehman – where the critical faults were hidden inside – there was plenty of public data available to make analysis possible about Greece’s position. Cultural variations can also be too defensive, making local managers unwilling to admit to head office that things are moving out of control.

So keeping overall cash visibility with balancing mechanisms is important even when the autonomous units are doing well enough by themselves. The treasurer needs to have a global perspective so that local exposures can be identified and, where possible, mitigated in advance of contagion.

Secondly, centralisation can not only lead to reduced transaction costs, but it may also produce savings in maximising interest income while minimising interest expense. Cash balances can be swept within and across currency zones so foreign exchange is the most obvious area for centralised treasury. Interest rate exposures are also, effectively, a function of foreign exchange so centralisation provides immediate visibility and control.

What about the banks?

Global banks are also part of the argument for centralisation but also equally a reason not to centralise some processes, particularly banking. On the one hand, they are leaders in technology-intensive processes and products and offer a consistent service across countries. Crucially they have access to the international SWIFT messaging network and corporate treasurers now can also gain access to the system as members or through service bureaus. A handful of very sophisticated global banks have also built portal capability which allows you to view balances and aggregate information whether their accounts or another bank’s.

The global banks can also provide greater flexibility. A single relationship with a centralised treasury means that they can also see the whole of a company’s operations and extend credit on a global basis without worrying too much about industry or country limits if they are confident that the corporate’s systems can provide a sufficient degree of comfort. This possibility has become more important for regional conglomerates since the credit crunch began in 2008 where markets performed very differently in different regions.

The counterargument of course is counterparty risk. While centralisation may give a company the ability to pretty much work with one bank only globally, in practice that model is dead. No company is willing to have all its cash management ‘eggs’ in one bank’s ‘basket’.

Best practice

The third argument for a degree of centralisation is spreading best practice around a multinational group. While certain practices must continue to be dictated by local needs, principles can be led from the centre. However, cautions one treasurer, this must process must be driven by example. “There is a risk of losing what is really going on if you force everything into a template.” Best practice can also have a wider impact because no company can be entirely independent.

It must have credibility in the eyes of the outside world: shareholders, regulators, analysts, ratings agencies, investors and other stakeholders. Unlike local relationship bankers, these outsiders do not have the time or skills to investigate every local operation of a multinational. They rely on centralised information in order to form opinions. Internally consistent reporting practices based on central principles inspire confidence even if they have to be expressed slightly differently, in local or international standards.

So the regional and global models both have their proponents. But perhaps the biggest barrier companies face in revisiting their treasury structures, relationships and practices is that, in reality, companies have just grown, without a managed growth plan. Almost every example of centrally-managed organic growth is supplemented by acquisition and often divestment.

Bringing together different legacy systems may simply cause things to grind to a halt so often treasury is left to muddle along and just get used to it. Treasurers do not set the corporate strategy but have to make sense of what they get.

Checklist of change:

What can I centralise? Approaches are changing:

Funding and investing

Many companies in the past have centralised their funding to treasury HQ whether through intercompany lending or giving HQ full responsibility for bank lines or public issues. Because of the credit crunch, this has changed for many companies who now look to local markets where they operate for both bank loans and public markets. It is worth looking at local financing opportunities in your key markets where some banks, awash with liquidity, may be loaning at better prices. The same applies for investing short-term funds, although beware of counterparty issues.


Visibility has become an absolute treasury essential. If you can centralise information then you can still have processes and people to deal with risk and opportunity on the ground. Centralised information is key to sound management and preventing problems. Any application or system that can assist in centralising information is worth having. This is a good time to review initiatives like SWIFT; bank portals (that have come on in leaps in and bounds in terms of centralising information); investment portals and risk software, and of course new releases in TMS that have placed information and usability of that information to the fore.

FX-hedging and other risks

Most risk mitigation processes should be handled centrally, but in some instance, hedging can be best done locally, for example with certain commodities hedging or local insurance. Processes and policies should be centralised, a lesson learned well after the crisis, particularly in the FX hedging arena.

Cash management

Review your cash management structures: does it still make sense to pay for certain services where there are too few gains because of the current funding and investing environment? Can funds be better employed locally than pooled back to the treasury centre?


Centralisation of payment processes almost always lead to savings. Depending on the local markets in which you do business, a regional shared service centre might be best placed. Related credit risk is best handled locally, although procedures and policies may be centralised


The trend to centralise to one global bank has died after the financial crisis with almost all companies reverted to working with more banks. Spreading around the business is not only good for counterparty risk but for giving business to those from whom you may be seeking funds. This trend also allows treasury to choose the most appropriate bank for a country or region rather than compromising on service because of a global relationship.