The Future of CDC

Written evidence submitted by Dr Sarah Bracking, University of Manchester

I am writing to submit a short Memorandum in respect of the future of the CDC, within the published terms of the current IDC enquiry. I am a Senior Lecturer in Politics and Development at the University of Manchester and specialise in the political economy of African development with particular reference to the role of public development finance institutions (DFIs) of which CDC is the UK national example. I have written a book on the subject called Money and Power: Great Predators in the Political Economy of Development (Pluto Press, 2009). More specifically I have recently completed a project, in which I was the principle investigator, commissioned by the Norwegian equivalent to DfID, Norad, about the future of Norfund, the Norwegian equivalent to CDC. The study was a comparative one, and also featured other European DFIs, including detailed material on Norfund, Swedfund and CDC. The focus was particularly about the consequences of the domicile arrangements for private equity funds contributed by DFIs in the context of the government restrictions in place in Norway and Sweden on the use of secrecy jurisdictions/tax havens by public DFIs.

This document was used by Norad to inform a report to the Ministry of Foreign Affairs in Norway. These can be obtained by contacting

postmottak@norad.no<mailto:postmottak@norad.no>. The documents make clear, there are a number of serious research gaps which pertain to an assessment of the developmental impact of private sector development interventions. It is in two separate documents:

- Future Directions for Norwegian Development Finance", dated 14th October 2010,

ref.no: 0902364-55". (ps.48)

- Supplementary documentation to the report "Future Directions for Norwegian

Development Finance", dated 14th October 2010, ref. no: 0902364-54. (ps. 168)

These documents are the property of Norad according to the contract between Norad and the University of Manchester.

Given the complexity of the issues around the developmental impact of different types of delivery mechanism of capital to the private sector (private equity funds, direct investments, guarantees and so forth); their differential developmental impact by sector (infrastructure, microfinance, industry and so forth); by choice of partner; and by choice of domicile; I would expect the Committees work to be challenging.

My comments below, in relation to this inquiry, concern:

- the effectiveness of CDC compared with other similar institutions;

. the reforms proposed by the Secretary of State for International Development on 12 October 2010 and the feasibility of achieving desired results given the CDC’s current resources, including staffing;

- the extent to which the proposed reforms will be sufficient to refocus CDC’s efforts, especially with respect to poverty reduction;

I also raise issues around tax justice and domicile (the place where CDC investments are legally registered). In general I believe that there are critical problems in the current means of evaluation, management and direction given to CDC in terms of delivering a development benefit. The argument below is that the development effectiveness of CDC as currently operating is limited, although it cannot be accurately measured given the paucity of public data and research. The reforms proposed are to be welcomed, but geographic targets should be replaced by more specific criteria on poverty reducing investments, while the issue of domicile of companies and funds should also be addressed. The proposals are too general at present to be significantly addressed, but increased use of alternatives to the current exclusive use of equity funds is to be welcomed. These could include direct investments, loans, the provision of risk instruments, investment guarantees, mezzanine financing and commercial and mutuals investments. Increased use of these instruments would mean an increased coordination with ECGD would be required. Issues of tax justice must be considered alongside these changes, and restrictions to secrecy jurisdiction use put in place.

Executive summary

1. The use of the fund of funds model critically impairs CDC’s management of developmental impact, political risk, reputational damage, consolidated and counterparty risk in the underlying investee companies, due diligence monitoring and the oversight of investment partners. The model relies on leverage and influence, but the positive effect of augmenting private investment flows is largely unproven and must be offset against the model’s disadvantages in terms of reducing the fiscal base in developing countries, increasing business opacity and for privileging the interest of investors unduly in relation to other stakeholders.

2. The overwhelming use of secrecy jurisdictions for fund of funds investments (secrecy jurisdictions are also termed ‘tax havens’ or ‘offshore financial service centres’) exacerbates most if not all of these investment governance problems.

3. The overwhelming use of secrecy jurisdiction domicile for CDC investments reduces the rightful and morally just fiscal resources available for developing country governments, which would otherwise be available to them had the funds and companies in which CDC invests been domiciled in their countries of actual operation. This impairs the development of education, health and public services upon which development critically depends.

4. It is hard to accurately calculate a developmental impact which includes tax losses and offsets these against figures for employment created and tax paid, which are the indicators most often used by DFIs in order to promote the benefit of their work (see EDFI, 2009). Public announcements by DFIs of developmental benefit generally rely on a thin evidence base and weak methodology. They are not subject to systematic publically available evaluation.

5. The current impact methodology in use by DFIs and the European Development Finance Institutions (EDFI), their representative body are inadequate. Most centrally there has been an erroneous, but widespread use of aggregated figures for employment created and tax paid in companies in which DFIs have contributed investment capital, rather than a pro rata calculation relating to their actual proportional ownership within the enterprises.

6. There is a dearth of relevant research within the DFIs or within the wider research community which can accurately assess the developmental impact of different forms, types and scales of private sector development intervention. This is due to the uncritical assumptions that have been held by DFI managements that any investment must be good for growth, and that growth is roughly commensurate with development, and also, in turn, with poverty reduction. Each link in this associational chain is problematic in practice.

7. The current instructions issued to CDC by DfID concerning the deployment of funds are only roughly related to actual outcomes, partly because of these flaws in logical association, and partly because of the globalised and multi-layered complexity of sovereign domicile in corporate structures and in the funds and companies in which DFIs invest. For example, the requirement to put certain percentages of invested funds into geographic areas such as sub-Saharan Africa, or ‘the poorest countries’ are impossible to accurately measure, even by CDC itself, since it places the vast majority of its funds in key secrecy jurisdictions under investment contracts. It is the Fund Managers who agree to meet these DfID criteria by proxy. However, they again are not required to disclose the degree of on-lending of DFI-originated funds to underlying portfolio companies which are similarly routed to secrecy jurisdiction based holding companies, parent companies and trust companies. The level of reporting of Funds to DFIs is not exhaustive in this regard since they do not routinely practice country by country accounting. It is not always clear where the final destination of funds invested actually is from these reports. In short, it would be a challenge for the participating parties – the Fund Managers and their DFI contributors - to know themselves what actual scale of funds are truly delivered into the national geographies to which they are mandated.

8. There is a perverse effect of mandating investment to the poorest countries in anticipation of an enhanced poverty reduction effect, in that the local capacity to absorb such funds tends to be highly concentrated in enclaved, extractive and export-oriented sectors owned by a small proportion of the population. The political economy effect of this is to increase inequality and frustrate efforts to make political and economic elites accountable, particularly since many of these transactions occur ‘offshore’. Much more empirical evidence should be generated on the types of investment that DFIs support, the sectors in which they invest and the ability of these to catalyse broad based and poor-poor growth. The mandating of funds to the ‘poorest countries’ does not at present, and as currently operationalised, have a provable beneficial developmental effect, and could quite easily be contributing to these adverse social and political consequences which may in fact hinder growth.

9. DFIs align themselves to the investment models and prerogatives of private sector actors in order to augment the total amount of capital available for investment, by pooling their resources with these partners in Funds, normally domiciled in a secrecy jurisdiction. However, there is currently no convincing evidence that total capital flows to Africa are expanded as a consequence of this investment model. But there is evidence of this structure being harmful in many other respects, including in a reduced fiscal base, in reduced corporate accountability, in patterns of ‘round-tripping’ investment, and in the privileging of the ‘international investor’ and Fund Manager in comparison to other local stakeholders which include nationally domiciled shareholders, workers and affected communities. The issue of tax justice should preclude any further use of this ‘offshore’ intermediated model.

10. Given the above, and the problems with risk assessment given in more detail below, the use of public funds in this manner by CDC must be of critical concern to the public interest and UK tax payer. The governance of CDC must be reformed in order to increase its public accountability, the quality of its accounting and transparency of reporting of its accounts, and its monitoring and assessment of risk and development impact. The influence that it believes it has as a contributing investor to private funds must be proven in terms of its ability to actually induce reforms and best practice in those firms as a condition of its participation, and not as something aspired to and promised by a third party Fund Manager. An increased use of direct investments is to be preferred.

The submitter

11. I have been researching and writing on development finance for close to 20 years, and have cognate interests in the political economy of development, political corruption, democratisation and poverty reduction. I am currently employed as a Senior Lecturer for the International Development programme at the University of Manchester, UK.

Factual Information

12. The CDC provided a list of its domiciles by number of funds invested in each place, but not by value to a recent University of Manchester research team (see Bracking et al, 2010a, Bracking, 2010b). The majority are domiciled in secrecy jurisdictions. The CDC currently invests solely using a ‘fund of fund’ model, investing in private equity funds in an intermediated model. From the data presented by CDC in 2010, it is clear that the intermediated model relies heavily on OFCs, with 71.67 per cent of all its funds and subsidiaries - by number not value - domiciled in either London, Mauritius or Cayman, which deductively act as three central business hubs. Using the IMF list of tax havens for 2007, it is possible to classify the domicile of its funds (excluding subsidiaries) in 2010, which adds up to 144 in tax havens, or 80 per cent of all the CDC’s investments by number (Bracking et al, 2010, para 1.7, page 10).

13. The CDC has a ‘fund of funds’ structure and about half its portfolio by value is routed using domiciles out of London. Actis, which includes the legacy investments, is managed from London and probably accounts for the majority of, if not all, London-based domiciles. CDC has 59 fund managers overall, one each for Actis and Aureos, and a further 57 with whom it sub-contracts. Actis employs 49 per cent of CDC’s overall capital – USD 3,361 million. It attracts to its funds a further USD 80 million from other DFIs, and USD 2,789 million from commercial investors. Aureos, a joint venture with Norfund aimed at SMEs, invests a further six per cent of CDC capital. CDC’s 57 other managers direct 45 per cent of its capital (Bracking et al, 2010, para 1.8, page 10).

14. There is an intimate connection between one particular type of investment model – the intermediated private equity investment fund – and domicile in a secrecy jurisdiction. The lesser association of direct investments with this type of domicile is clear from the domicile patterns of Norfund and Swedfund, but is still prevalent for larger direct investments (see Bracking et al, 2010a, 2010b).

15. There is some evidence from CDC that using sub-contracted in this way does augment private investors’ participation. Data from the CDC Development Report 2008, suggests that the leverage rate of different types of fund is indeed better for funds that are sub-contracted, rather than in funds and subsidiaries in the ‘legacy’ portfolio. Thus, Actis has USD 3,361 million invested, with USD 80 million from other DFIs (making USD 3,441 million), which manages to attract USD 2,789 million from the commercial sector. Expressed as a ratio, this is approximately 1: 0.81. Conversely, the 57 new fund managers between 2004 and 2008 had USD 7,137 million invested by CDC, with an additional USD 1,433 million from other DFIs (making USD 8,570 million), which together attracted a further USD 19,718 million from commercial investors, giving an equivalent ratio of approximately 1: 2.3 between the ‘public’ and private sectors, or nearly three times as much per unit of invested DFI capital.

16. CDC is unique among DFIs to employ all its capital through private sector Fund Managers, albeit that the fund with nearly 50 per cent of its capital is managed from London, in the new private fund company Actis, jointly invested by DfID and private owners (who used to be CDC public employees). Ergon explains that the objective of this arrangement was to achieve a ‘step change in CDC’s economic impact and catalytic role’ (quote assigned to DfID, 2010: 14), in that CDC resources are less than one per cent of total international private equity to developing countries, so to have a bigger impact it was thought that they should invest though private equity Fund Managers to influence commercial investors in the same companies available to them (see Ergon, 2010: 14). This model thus relies on leverage, influence and perhaps more critically, following the pack, rather than directly investing in stand-alone projects. However, it is hard to attribute to CDC alone the responsibility for the behaviour of these investors, since the counterfactual case – that they would have done it in any case – is unclear. It is certainly possible that this investment model attracts investors who would otherwise have invested ‘onshore’ in any case, displacing local savings and investment. It is not clear that this is additional capital in terms of the investment market as a whole.

17. There is little evidence that using sub-contracted funds in this way produces any developmental impact that couldn’t be attributed to a like volume of private investment per se. The additionality is not proven such that the deployment of CDC funds in other structures must be seen as desirable given the adverse tax effects.

18. DFIs and investors view secrecy jurisdictions as protecting them from expropriation of their investments, and as guaranteeing them an effective disputes mechanism for conflicts between co-investors. Leaving secrecy jurisdictions is generally associated with an increase in investment risk, in particular in the categories of political risk, expropriation risk and exchange rate risk. However, the risk ‘protection’ offered by funds could concentrate rather than spread risk, in a similar fashion to the manner in which off-balance sheet items from financial institutions, opaque structured investment vehicles and structured investment products, such as offshore SPVs and CDOs concentrated risk, triggering the current global financial crisis (see Government Commission Report (NOU, 2009)).

19. Also cross investments and repeated ownerships in the underlying portfolio can also concentrate risk, such that the ‘consolidated risk’ – which aggregates that of the underlying companies could be extremely toxic. European DFIS are only beginning to account for this, although to the author’s knowledge CDC may not. Relatedly, the Government Commission in Norway correctly concluded that tax havens enhance counterparty risk (NOU, 2009). Moreover, the public is the risk carrier of last resort, through the respective central banks of DFI owner countries. Thus, it would be prudent to remove certain business types and organisational forms from the risk-bearing chain – such as secrecy domiciled Funds and international business companies (IBCs) - since these contribute only opacity, and work only in the interests of capital holders at the expense of capital receivers. International companies which are not obliged to keep proper accounts are unsuitable as business counterparties, since the transaction risks are too great. Thus, the depiction of secrecy jurisdictions as a means of managing some types of risk to the investor – expropriation, payments – must be offset against their contribution to the inflation of other types of risk – counterparty risk, consolidated risk and reputational risk. Withdrawing from secrecy jurisdictions will have the likely result of enabling a more systematic means of risk accounting, as well as a reduced exposure to non-transparent risks (see Bracking et al 2010a, para 2.2.2, page 19).

20. The issue of management of risk. In Bracking et al (2010, 19-20) it states that:

" It is clear from the financial crisis of 2007-09 that when the risk management agencies were under the employ of the organisations releasing the financial products, they were bound to have a conflict of interest, which would lead to error. The system here has a similar flaw, wherein those defining and benefiting from the fund product, are also calculating the investment risk, or assigning it to third parties. That the DFIs are significantly cross-invested adds to the image of safety, but actually – particularly with the type of investments that DFIs do best in, which by their nature tend to be high risk – this herd effect gives another false veneer of safety. DFIs could be subject to significant reputational damage, should this risk concentration fail them. The current arms-length due diligence and risk management of intermediated funds is thin. It is efficient and cuts down administrative costs in order to plough more money back into ‘development’, but that will prove a short-sighted strategy if a political or reputational scandal emerges. DFIs should calculate counterparty, consolidated, reputational, social and political risk, as well as financial and investment risk themselves, or assign this role to an impartial third party, such as an office within the home Public Accounts Department. A European development finance regulator should be able to conduct ad hoc tests of procedures in member states."

This holds equally true as a recommendation for this review.

21. Due diligence. CDC does not carry out due diligence on all their co-investors. The argument that they check all major shareholders but the very small ones is an argument incorporating much hazard: a ‘small’ investor in a large equity fund is often a very ‘big fish’ once he or she is back in their domestic context, capable of wielding much power and control over local markets, communities and workers. Thus the moral hazard is that such people are empowered in relation to others, with no apparent checks on their business practice or legal record. The due diligence of small investors is left to contracted Fund Managers.

22. DFIs sometimes point to the reductions in administrative costs attribute to the use of funds. However, fees for fund managers are born by the fund as a whole, deducted before the fund is closed. These fees are taken from the pool of investable capital, and are contributed in part by tax avoidance in the country of effective residence of the underlying portfolio company. In other words the cost structures of different types of investments are not commensurate as some costs which would be accounted for in a public organisation are subsumed into the fee structure of a fund. Further research is required on relative cost structures. Suffice to say that an increased use of direct investments by CDC would require more in-house investment managers, the cost of which could be born in part from the reduced indirect cost of fees and expenses in sub-contracted funds.

23. DFIs also assert that the use of secrecy jurisdictions reduces the transactional costs of administering investments, and thus allows for a better flow of funds to the end user. However, Bracking et al (2010a, 21-22) argue that:

"Increased transactional costs due to withdrawal from secrecy jurisdictions in terms of red tape, bureaucracy, and in altering the culture of facilitation payments can be expected in the short and medium term. Overcoming such high transactional costs could be part of the core developmental remit of a DFI by supporting domestic finance markets, enhancing institutional quality, which would eventually lead to better developmental outcomes. Operating guidelines could be changed to reflect the increased role DFIs would have in the reform of soft and hard market infrastructures and operating cultures. In order to ameliorate the increased costs associated with the active search for investment opportunities (a role previously assigned to the Fund Manager), DFIs could consider advertising available funds through a calls to tender type system, directed at the private sector in target countries, with clear investment criteria that potential firms and funds must demonstrate that they can meet."

Because investments in developing countries are considered hard to implement in a direct fashion is not a good enough reason to stop trying, since the market environment improvements that a DFI presence can catalyse have multiplier effects for other investors. This is part of the developmental benefit that can be expected from PSD intervention.

24. Active ownership by DFIs in more embedded national structures can make a real difference to institutional strengthening, regulatory improvements, and measures of market and risk perception, in order to catalyse capital market growth. Institutional strengthening and enhanced institutional quality is positively correlated to further growth in investment flows, savings and investment in a virtuous cycle. Research from Kaufmann et al (2005) and others around the ‘governance and institutions matter’ theme, suggests that local investment and savings growth is just as, if not more important, to sustainable economic growth and development in the long-term than short-term injections from ‘outside’ (also see Addison, 2010; Chang, 2002; Di John, 2010). Recent development success stories generally start with a strong, developmentally oriented state, not a private sector which is designed to bypass it (See Chang, 2002).

25. Richard Murphy has recently estimated the tax loss to developing countries from the DFI use of tax havens for the EDFI portfolio as a whole, using data produced by the DFIs, at in excess of EUR 430 million1, a year on average over the last five years. Actual corporate taxes paid on the EDFI portfolio might be as low as EUR 270 million2 per annum in direct conflict with the EDFI claim that their investments generate approximately EUR two billion3 of tax revenues per annum for developing countries (Murphy, 2010).

26. In Bracking et al (2010a) we estimated for 29 companies in which Norfund invested, the tax loss (or gain) for developing countries for 2008, as compared to what the same companies would have been liable to pay had they been domiciled in the country of actual operations. For these, the amount of tax underpaid, or tax losses for developing countries because of domicile in secrecy jurisdictions has been calculated for 2008 at over (gross) USD 14.6 million. This is tax that would have been paid, at the rates prevailing in the countries of their actual operations, for these 29 companies if they had been domiciled in the same jurisdictions as their operations, rather than in a secrecy jurisdiction. [An explanation for how these figures are derived is found in the accompanying report Supplementary Documentation for the report "Future Directions for Norwegian Development Finance", Bracking (2010b)].

Recommendations for action

27. DfID should make provisions in law, policy and practice to ensure that:

· CDC does not invest in private equity funds in secrecy jurisdictions as determined by the existence of 1) differential tax treatment for non-nationals and 2) company law which allows beneficial owners to remain secret and company accounts to remain private. CDC invests in funds in which at least 10 per cent of the onlending to the underlying portfolio companies occurs in the legal jurisdiction of the domicile of the fund. This is in order to enhance the developmental impact and accountability of Funds to developing countries’ regulatory authorities and the transparency of corporate structures.

· CDC should not make direct investments to companies where the parent is in a secrecy jurisdiction and the funds are routed through it. There is no positive developmental function that can be attributed to a firm having an offshore parent, but many negative consequences thereof.

· DfID should redesign the directions given to CDC for the deployment of funds so that they can be realistically and accurately measured in terms of territorial domicile, rather than the deceptive data which is currently published by CDC about the geographical spread of its investments.

· DfID should redesign investment management criteria to ensure that CDC activities maximise developmental impact and poverty reduction. This will require use of a new development impact matrix which should include criteria concerning direct poverty reduction effects, downstream and upstream business generated, employment, environmental impact, governance effects, economic displacement effects and modelling of the type of growth effects to be expected by sector and type of enterprise. In other words, a systematic means of monitoring developmental impact should be employed by CDC.

· If CDC maintains the use of private equity funds its participation should be conditional on their adherence to international accounting, environmental and human rights standards and to fair remuneration.

· HM Government should support moves to establish a European regulator of European DFIs to carry out independent checks of DFI operational procedures.

Summary

28. The development effectiveness of CDC as currently operating is limited, although it cannot be accurately measured given the paucity of public data and research. The reforms proposed are to be welcomed, but geographic targets should be replaced by more specific criteria on poverty reducing investments, while the issue of domicile of companies and funds should also be addressed. The proposals are too general at present to be significantly addressed, but increased use of alternatives to the current exclusive use of equity funds is to be welcomed. These could include direct investments, loans, the provision of risk instruments, investment guarantees, mezzanine financing and commercial and mutual investments. Increased use of these instruments would mean an increased coordination with ECGD would be required. Issues of tax justice must be addressed alongside these changes and restrictions to secrecy jurisdiction use put in place.

17 November 2010

References

Addison, T (2010), "Revenue mobilization for poverty reduction: What we know, what we need to know", in Lawson D, Hulme D, Matin I and Moore K (eds.) What Works for the Poorest? Poverty reduction programmes for the world’s extreme poor, Practical Action Publishing Ltd, Bourton on Dunsmore, ps. 253-262

Bracking S, Hulme D, Lawson D, Sen K and Wickramasinghe, D (2010a) "Future Directions for Norwegian Development Finance", dated 14th October 2010, ref.no: 0902364-55". (ps.48) Bracking S (ed.) "Supplementary documentation to the report "Future Directions for Norwegian Development Finance", dated 14th October 2010, ref. no: 0902364-54. (ps. 168)

Chang, H-J (2002), Kicking Away the Ladder: Development Strategy in Historical Perspective: Policies and Institutions for Economic Development in Historical Perspective (Anthem Studies in Development and Globalization), Anthem Press

Di John J (2010), "The political economy of taxation and resource mobilization in sub-Saharan Africa", in Padayachee V (ed.) The Political Economy of Africa, Routledge, London

EDFI Secretariat (2009) 2008 Comparative Analysis of EDFI members, EDFI ASBL, Brussels

Ergon Associates Ltd (2010), Decent work and development finance: Background paper for Decent Work and Labour Standards Forum, March 2010, sponsored by DfID, mimeo

Kaufmann, D, Kraay A, and Mastruzzi M (2005), "Governance Matters IV: Governance Indicators for 1996-2004" World Bank, May

Murphy R (2010), Investments for Development: Derailed to Tax Havens. A report on the use of tax havens by Development Finance Institutions. Prepared for IBIS, mimeo

Norwegian Official Report (NOU), Government Commission, Norway, (2009) Tax Havens and development: Status, analyses and measures, Official Norwegian Reports, no 19, Government of Norway, Oslo