At ACT’s Cash Management Conference 2016 last month, Frances Hinden, Vice President of Treasury Operations,  talk on ‘Global treasury management at Shell’ provided vital lessons and tips for all sizes of MNC. 

The scale and size of Shell’s global cash management is, as Hinden put it, “Unusual.,” see figure below.

Shell global cash management 

Source & Copyright©2016 - Shell International Ltd

However, Hinden believes that their problems and challenges are not unusual, they are basically the same for all multi-national corporations. 

Lessons from Shell’s centralisation journey 

Shell have a centralised model and focus for all their cash management using central accounts, central market agreements, etc. Everything is run by one team and bank relationships are owned by a team of four Global Cash Managers in London.

Key events and lessons in Shell’s centralisation journey: 

• moved ‘everything’ into a centralised treasury operation organisation in 2009
• able to demand appropriate facilities from their banks (but at cost-effective price) to match the importance of having Shell as a customer
• the culture in the three cash management focused share service centres - in Glasgow, Krakow and Chennai - is very different due to the variations in profile: age, male/female, level and areas of expertise, and style. So when considering moving an operation/process to a SSC, Shell now base the decision on what fits best with the culture in the SSC team. Since the first Shell SSC was set up in 2006, they have ‘moved things around’ to improve the match of the culture with the capabilities of each SSC. They plan to continue to do this as they understand what best fits with each SSC 
• classically, Shell started their SSC journey by documenting current process (in bank reconciliations) and then moved the templated process into the SSC produced some small savings. The real savings were only achieved when they decided to have one standard process for the whole of Shell regardless of country, language or bank PLUS continuous process improvement combined with concentrating all ‘bank recs’ in one team. Since then the number of items outstanding has decreased dramatically
• using a single standard process approach for all functions in their SSCs has enabled Shell, for example, to reduce the level of payment fraud to a minimal level because ALL payments have to go through the standard process. Anyone, from CEO down, wanting to make a payment have to go through the payment process. No exceptions are allowed, Hinden, “If the payment doesn’t go through the process, it doesn’t happen.” 
• setting up a opening bank account standard process is proving more difficult. A standard process has been set up for their two principal banks in which a new bank account can be opened in 24 hours, but bank account opening process in banks in emerging countries is ‘more of a challenge’
• Shell now call their SSC teams Finance Operations to avoid having the SSC mentality of just following processes. The SSC teams are now focused on “what are we trying to achieve together” with their Shell colleagues. They now feel part of the same team. Also Hinden believes that there is an advantage of all being part of Shell who are all on the same KPIs and not worrying about service centre agreements with external organisations.

Cash flow forecasting systems and lessons

Hinden described Shell’s cash flow forecasting processes and the lessons learnt from their experience:

• some 18 months ago Shell had a team extracting and re-typing the data from over 30 ERPs in Shell to provide short-term forecasts - ranging from ‘on-the-day’ to 25 days. This has been replaced by a new cash flow forecasting system in which the cash flow data is sucked automatically out of all the A/P and A/R systems into an SAP HANA database (a new advanced system, see). A model in SAP then uses this data to produce the short term forecasts. 
• lesson - cut out the unnecessary activities and reports: the savings in FTEs from the new system have been small so far, even though some 50% of the cash flow forecasting process has been automated. Investigation as to why there were so little FTE savings showed that many of the staff were doing other jobs, e.g. answering the phone to colleagues asking, “Can you tell me what my balance is on the current account.” Shell stopped much of this unnecessary work by:
  • asking for every report, for every request to treasury, “If not working in treasury, why do you need to know this?” Or asking, “Do you really need this information, what do you do with it?” The answer in 80% of cases was, “Actually, I don’t need it.” The 20% who do need the  forecast are mainly in JVs and/or in regulated markets where detailed knowledge of the amount of cash is essential 
  • cutting medium term forecasts reports that were not useful, e.g. end of month forecasts
  • only retaining reports that have value, e.g. rolling and end of year reports
• lesson - accept the 80:20 rule: cash flow forecasting in the OECD /free flow big countries and in companies 100% Shell owned is easy and takes little time and resources. While the small unregulated countries - where the currency needs to kept offshore and is not freely convertable - and/or in non-wholly owned ventures takes most of the cash flow forecasting time and resources. (Hinden expects that in 8 months time when 95% of cash flow forecasting process has been automated, they will have only managed to save 20% of the FTE time spent on cash flow forecasting.) 
• lesson - don’t try to achieve unnecessary levels of cash flow forecasting accuracy
• lesson - focus on the cash flow forecasting that you need, not on what you would like
• lesson - cash flow forecasting is all about people. The systems are just supporting the cash flow forecasting process. It is the people that matter in cash flow forecasting.

---

CTMfile take: The volume of Shell’s cash management is ‘unusual’, but what impressess is the clarity of thinking about operations and the progressive refinement of their processes. There are important lessons for other corporates, particularly on cash flow forecasting.