How the Fund Management Industry Miscalculates Risk
The following is a guest post by Jaskaran Singh, a risk management professional with over 30 years experience in the finance industry.
The majority of the Fund Management Industry identifies the geographic exposure of an entity by headquarter location, place of primary listing or place of incorporation. In a globalized world where a company receives revenue streams from around the world this obscures country risk. This calls into question the reliability of industry asset allocation and diversification strategies and the prediction capability of conventional portfolio risk modelling techniques.
Fund Managers are primarily in the business of managing risk and it is essential that they understand the risks they face in managing assets within their chosen portfolios. They are, therefore, duty bound to identify, measure, manage and control the risks in their portfolios and construct their portfolios in the full knowledge of the identifiable risks and the likely returns they expect to achieve. Using the twin concepts of asset allocation and diversification as their building blocks, they then construct portfolios aimed at maximising return for any given level of risk or to reduce the risk for any given level of return. They achieve this by allocating capital to different types of assets. Thus, diversification through asset allocation in different and, hopefully, uncorrelated asset classes and sub-classes, investment styles, size, sectors and geographies, is key to managing investment risks.
None of the above is controversial and there is general agreement and acceptance amongst portfolio managers of this reality. Why then does the majority of the Fund Management Industry and their data providers who serve them, choose not to measure geographic exposure correctly?
In an increasingly global world (notwithstanding the events of 2016) companies are doing and will continue to do business across the globe. Materials, goods and services are often produced and sourced in one region of the world, processed in another and sold in a third. All of these actions incur risks which are broadly covered under the over-arching concept of country risk (of which geographical exposure is the foundation). The primary listing, the place of incorporation or headquarters of the company provide no mitigation or ability to control or minimise the risks associated with doing business, and, therefore, reliance upon them, especially in an increasingly global world, is inappropriate and improper.
When companies choose to invest internationally, they have to make a careful assessment of a host of risks: political, economic, financial, legal, tax, security, reputation, financial transparency, money laundering, terrorist funding, public transparency and accountability, corruption etc. Based upon their analysis and understanding of these factors, they take investment decisions. Thus, over the past 30 years, based upon assessment of risks and cost of doing business, the world has witnessed a growing trend of global expansion by businesses outside of their home territories. This has helped open new markets for their products as well as enabled them to establish manufacturing and servicing hubs in countries which offer lower costs of production and attractive infrastructure and technological advantages when compared to their home markets.
So, if the companies in which the Funds Management Industry invests have changed, why is the Industry (and its data providers) still stuck in the past and using the old fashioned method of measuring geographical exposure?
Is this because the Industry can’t be bothered?
Or is it because the Industry considers the geographic and country risks in their portfolios to be insignificant?
Or is this perhaps based upon the false premise that what the investing public isn’t aware of does not need to be explained? But then, why are institutional investors not raising this as an issue? Why have the regulators not picked up on this gap?
Or is it that the industry believes it is too difficult to do and so (apart from a select few exceptions like Capital Group) prefer to bury their collective heads in the sand?
I raised this issue with Morningstar in October 2016 but, other than thanking me for my analysis, they have neither accepted nor rejected my argument that the industry needs to rethink how they go about identifying geographical exposure.
Whatever the reasoning of the Fund Management Industry (and its data providers) for their failure to change, what is apparent to me is that theirs is an abject failure to recognise the risks in the portfolios they manage and get paid for. The Industry has a duty of being transparent and is required to disclose sufficient information to its investors to enable them to make investment decisions. In turn, it has sufficient regulatory backing to demand the same level of transparency from the companies that they invest in. Regulators also mandate that decision-making by the Fund Managers is supported by robust risk management policies, systems and processes with clear linkage into the asset valuation process. So risk management processes used should enable the Fund Manager to monitor and measure the risks of the positions they take and the contributions of the assets to the overall risk. This requires risk limits and systems that enable risks to be accurately measured on the basis of sound and reliable data.
When the Fund Management Industry has no realistic idea of the geographic exposure in their portfolios, the risks they do actually identify and the measurement of these risks cannot be considered reliable, as they ignore the operating environment in which companies operate. This, in turn, calls into question the reliability of their asset allocation and diversification strategies and the prediction capability of their modelling of portfolio risks. If the risks cannot be predicted, how can they be reduced, mitigated or controlled?
Identifying the environment in which the target companies operate is a key building block of any risk assessment before an investment decision is taken. The geographical spread of these operations is, therefore, a key early enquiry that needs to be made. If this is not done, or is only limited to the target company’s domicile base, the risks and rewards of the international strategy of the target company and the Fund Manager remain unknown, as is the impact upon the share price and potential value of the target.
Thus, using the domicile or the primary listing of the target company to identify its geographical exposure is fundamentally incorrect as it turns a blind eye to the sovereign and country risk exposures that businesses are exposed to in trading internationally.
Governments in the developed world do not provide any implicit or even explicit mechanism to guarantee the solvency of corporations. The only exceptions to this are major banks of systemic importance and certain companies in the crucial defence sector. Moreover, country risk is not merely an issue for financial institutions lending to, and corporations trading in or with, emerging market countries. Nor is it simply a question about the foreign exchange holdings of the host country. It, if ignored, has serious repercussions for the investor, as it can seriously impact on share price and not just in emerging markets as BP PLC learnt at huge expense in relation to the Macondo spill some years back (Management lost sight of the highly litigative nature of American society, compared to other nationalities. This became toxic for BP PLC when it mixed with politics and environmental risks.). The fund management industry, therefore, ignores country risk at its peril.
If further proof is required of the importance of assessing country risk, I draw the attention of readers to the innumerable times the Oil & Gas Industry has faced losses through the nationalisation, confiscation or expropriation of their assets. Mexico (1938), Iran (1953), Sri Lanka (1962), India (1973), Bolivia (2006), Venezuela (2007 & 2010) are just some of the many examples. Some countries such as Russia, Kazakhstan, Equatorial Guinea, Bolivia, Venezuela and Ecuador have, in recent years, all sharply increased tax rates and royalties on oil & gas assets. Exxon, Conoco Phillips, BP, Royal Dutch Shell, Total and many other big players have all been the victims of these actions.The Oil & Gas sector is not the only sector that has felt the impact. The banking & financial sector too has seen its fair share: India (1969), Mexico (1982) and Venezuela (2009-10) are just a few examples. In fact, Venezuela under Chavez nationalised a wide range of industries in sectors ranging from Agriculture, Oil & Gas, Finance, Glass containers, Gold, Steel, Power, Telecommunications, Tourism and Transport.
How should geographic exposure be identified?
The first step for assessing country risk is the need to accurately measure the geographical exposure. The correct way to identify the geographical exposures should necessarily be based on one or more of the following methods:
- Domicile of the Sources of Revenue: Revenues being less
capable of being massaged than net profits and operating profits are, therefore, more reliable.
- Domicile of the Revenue Producing Assets.
- Domicile of the legal system upon which reliance is placed to recover revenues and assets in the event of bankruptcy.
- Domicile of the legal system upon which reliance is placed when damage to licensing, royalties and patents are seen as the principal risk to the success of the business.
- Domicile in which key personnel and systems are based and supported when the risk of loss of key personnel and systems is identified as the key risk to the success of the business.
The Risk Manager and the Risk Committee of the Fund Manager must, based on careful investigation and consideration, decide which of the above listed options is best suited to identifying and measuring the geographical exposure of the target company to be incorporated into the investment portfolio. There is no single best option and the decision must be taken by identifying the key ingredients that ensure the success of the target company. However, if a single choice must be made, then, in my opinion, the one based on economic exposure (i.e. revenues) is perhaps the best of the above listed options because companies are already providing this information in some form within their annual report and accounts.
How much do the results of measuring geographical exposure by the above mentioned methods differ from those produced by the industry’s approach?
This is best shown by way of some examples. As I am a UK based small investor, let me show you some examples from within my own portfolio. This necessarily focuses on individual companies and not an ETF or a mutual fund.
I. Arm Holdings PLC
Arm Holdings PLC is incorporated in the UK and its principal listing (prior to it being acquired by Softbank) was also in the UK. This led Morningstar to categorize and count Arm Holdings for purposes of geographic exposure as a 100% UK based company.
However, when one examines the sources of revenues of Arm Holdings PLC (taking year end 2015 data), the following picture emerges:
|Europe ex Eurozone||2.76%|
This data clearly shows that an investor buying shares in Arm Holdings PLC is not being exposed to the UK but has a broad international exposure. This data was available to Morningstar and displayed in the stock report PDF it published but still it chose not to apply this information to the geographic piece. While the Revenues analysis provides a useful tool for assessing the geographic spread of Arm Holdings’ business, the nature of the company lends itself to at least two other ways for assessing the geographic exposure of the company:
i. If its employees and systems are viewed by the Risk Manager and Risk Committee to be key to its success then an analysis of the domicile of the 3,975 strong workforce (at the end of 2015) could provide guidance. The information contained in the following table has been extracted from the 2015 Annual Report & Accounts:
|Rest of Europe||628||15.80%|
Here, the data needs further granularity so that the number of people domiciled in the Eurozone, Europe ex Euro Zone, Emerging Europe, Developed Asia and Developing Asia (data by country would be better but even data by region is useful) can be more clearly identified. Fund Managers and institutional investors should not find it too difficult to ask and get this information. Even data providers like Morningstar are well placed to do so.
Notwithstanding the gaps in the table above, it does clearly show that on the basis of the strength of the workforce, an investor buying stock of Arm Holdings will have just under 40% exposure to the UK.
ii. There is yet another choice available to the Risk Manager & Risk committee to investigate: As Arm Holdings’ revenues come principally from licensing and royalties and they have a large number of patents in place, the need to protect these may be considered critical to its continued success. If so, then the fact that all these are protected under English law, the Risk Manager and Risk Committee may choose to identify the geographic exposure as entirely UK.
It hardly needs to be stressed here that there is an active choice that has to be made and based on an analysis and understanding of the risks. This,in turn, should be reflected in the stock report.
II. Experian PLC
Experian PLC has its headquarters in Dublin, Ireland, but its principal stock listing is in the UK. Hence Morningstar and the Fund Management Industry categorizes Experian PLC as 100% UK exposure. However, if we analyse its 2016 revenues, the following picture emerges:
|UK & Ireland||21.4%|
It shows that categorising Experian PLC as 100% UK exposure is misleading and clearly wrong. The USA is,by far, the biggest market for the company. The troubles of the Brazilian economy have resulted in it slipping below the UK & Ireland contributions in this year.
The business of Experian PLC lends itself to three others ways of determining its geographical spread. These are:
i. If we look at where Experian’s non-current assets are domiciled, their latest annual report as at 31 March 2016 shows the following distribution:
|North America||US$ 3,494m||58.4%|
Here too the message that comes through is that Experian PLC’s fate is not dependant upon the UK. It is clearly more reliant on North America while Latin America (primarily Brazil) due to its growth potential is also more critical to its success.
ii. Experian states that its expertise lies in data, analytics and technologies. It uses this expertise to give customers the power to assess, predict and to plan to achieve their goals. It also counts its brand as a key strength. Given this information,the Risk Manager and Risk Committee may wish to determine the domicile of its employees. Taken from their 2016 Annual Report, the picture that emerges is:
|UK & Ireland||3,569||21.6%|
This, again, highlights the fact that Experian PLC cannot be considered to be wholly geographically based in the UK. There is nothing to stop the Risk Committee to choose a subset of the whole employee count to arrive at their answer. Their reasons must be supported by demonstrable facts and reasoning.
iii. The amount of capital deployed to each region of the world also offers us another means of determining the geographic mix of Experian PLC. The 2016 Annual Report gives the following information for our benefit:
|UK & Ireland||13.9%|
|EMEA/ Asia Pacific||7.5%|
Here, again, it is clear that North America and Latin America are key to the success of the company and that the UK & Ireland, despite being the domicile of the company, is third in the pecking order as far as capital employed in the business is concerned.
The choice of method to be used for determining the correct geographic exposure must be decided by the Risk Committee. This decision should be explained and form part of the final investment decision.
Both the examples of Arm Holdings PLC and Experian PLC show that the results of the alternative methods of identifying geographic exposure are very different to the results achieved by relying on country of primary listing or domicile of a company’s headquarters. It also shows that plain reliance on country of primary listing or domicile of a company’s headquarters is not a useful criteria for identifying the geographical mix of the business and that the alternative means available are better at determining the right answer. It is disappointing that neither the Fund Management Industry nor the data providers like Morningstar make an effort to get this right. As the data is incorrect, how can investors rely upon it while taking decisions to invest? This raises the question whether Fund Managers are failing to meet their undertakings to the Regulators who expect them to provide accurate and reliable information in a transparent and accountable fashion to the end investor.
If the geographic exposure is wrong, as I have shown above, it is highly unlikely that the Fund Management Industry has a proper handle on what country risks their portfolios contain. This means that while doing their due diligence, the Fund Managers are unable to ask the right questions that should be asked to determine and understand the risks. It also means that once the investments have been made, there is no active process to manage the country risks that are out there in the portfolio.
Am I alone in raising my voice on this issue?
Happily, mine is not the lone voice on this issue. One voice is from within the Industry itself – the highly respected fund management giant, Capital Group. It makes the point that a company’s domicile provides little information about its potential success or its future share price and believes that far more meaningful is a company’s economic exposure, with revenue being the best available measure.
It has also pointed out that over two-thirds of companies that are constituents of the MSCI All Country World Index provide country-by-country revenue data while the remaining provide regional breakdowns. However, they do admit that notwithstanding enhanced reporting requirements introduced by accounting standards, much work needs to be done to get the broader corporate world to voluntarily provide data in usable form. Here, I would like to suggest that the Fund Management Industry (and its data providers) needs to fully acknowledge this issue and use its position and influence to mould behaviour of companies so that they provide this information.
The EDHEC – Risk Institute1 too has questioned the usefulness of analysing geographic equities exposure based on the stocks place of listing, incorporation or headquarters. Like the Capital Group, they too prefer to identify geographic exposure by disaggregating sales (i.e. using economic exposure i.e. revenues) by their geography. They studied the geographical exposure by sales of the constituents of several developed market indices (e.g. S&P 500) over a 10-year period: 2003-2012, and found these to have a significant exposure to non-domestic regions and by the end of the 10-year period this exposure had increased significantly2. The same finding was reinforced when the constituents of the S&P 500 and STOXX Europe 600 were weighted by percentage of sales in foreign markets3. They further found that the emerging market exposure of these indices was already material in 2003 but almost doubled by 2013.4
MSCI has also recognised the benefit of identifying geographic exposures5 through the use of economic exposure and produced some indices based on economic exposures for use of investors who wish to “tilt” their portfolios accordingly. However, they take the view that the original market-cap is the pure index and, therefore, they do not advocate a wholesale shift to indices based on economic exposure. I have an issue with this. Market capitalisation is a reflection of the economic value of the business as perceived by the market and incorporates all information that is known. Thus, any positive or negative impact of activity outside of the home market would feed through into the market capitalisation of a company that has cross-border activity, while one that has purely domestic activity will not be impacted. If that is accepted to be true, then wouldn’t the indices using economic exposures to identify the correct geography of their portfolios be the real “pure” index instead of one based on domicile or primary listing? In other words, I believe MSCI should actually ask investors to shift to economic exposure based indices. This will no doubt mean that they will have to swallow a bitter pill while accepting their original indices were flawed; I do believe this is part and parcel of being a highly professional organisation and would not do any harm to their reputation. It may, on the other hand, enhance it.
Once the geographic exposures of the company are correctly identified, the Fund Manager must ensure he understands the likely impact on his investment from the manifold factors that constitute country risk. This would normally include periodic review with company to identify the risks and understand the steps taken by the company to mitigate the risks. Purchasing political risk insurance, hedging currency risk, and settling legal disputes through international arbitration are just some of the available options to the company. Where the company has a local presence in an emerging market, working with local partners, hiring local staff, providing security detail to protect to fixed assets and key foreign employees, borrowing local funds from a local bank are just some of the options available to the company to safeguard its interests and for the Fund Manager to remain informed about.
The Fund Manager, armed with knowledge of the cross-border exposures contained in his investment portfolio, is better placed to take corrective action when adverse events have the potential to hurt its value. He may be able to buy a “protective-put” on stocks or a subset of the portfolio of stocks to prevent potential loss in capital invested.
The Fund Management Industry and its data providers are, by ignoring the real geographic mix of their portfolios, failing to recognize the real risks in the portfolios they manage and get paid for. It is time for the industry, its data providers, trade bodies and regulators to heed this wake-up call for change.
1. [Accounting for Geographic Exposure in Performance and Risk Reporting of Equity Portfolios.]↩
2. [S&P 500 had exposure of 19% to ex-Americas region in 2003 and increased to 27% in 2012.]↩
3. [S&P 500 % of sales rose from 29.9% in 2004 to 38.7% in 2013. STOXX 600, 41% to 53.3%.]↩
4. [STOXX 600 exposure increased from 10.7% to 22.7% in the period.]↩
5. [Economic Exposure in Global Investing: Tilting Portfolios based on Macroeconomic Views.]↩
Recommended: Charting the Federal Reserve's Assets - 1915 to 2012