Back Matter

Back Matter

Author(s):
Cheikh Gueye, Javier Arze del Granado, Rodrigo Garcia-Verdu, Mumtaz Hussain, B. Jang, Sebastian Weber, and Juan Corrales
Published Date:
March 2014
    Share
    • ShareShare
    Show Summary Details
    Annex: Econometric Analysis of Drivers of Portfolio Flows to Sub-Saharan African Frontier Markets

    The regressions reported in Table A.1 are run on an annual panel data covering the period 1991–2012 for selected emerging and frontier markets. The dependent variable is defined as the ratio of portfolio liabilities (nonresident purchases of domestic assets net of sales) to GDP. The use of lagged regressors helps to minimize potential of endogeneity. The regression includes variables aimed to capture push factors such as the U.S. Treasury bill interest rates, and global risk (VIX), as well as several pull factors such as the inflation level, fiscal balance, gross public debt, output growth, change in reserves, and the stock market development. A stepwise approach was employed to explore the significance of alternative pull variables (many of them are likely collinear). Random effects were used to allow the inclusion of a dummy for sub-Saharan African frontier markets. Running the regressions for sub-Saharan African frontier markets alone produced no statistically significant relationships, likely because of poor data quality and the small size of the sample.

    Table A.1.Determinants of Portfolio Investment Flows(Gross, in percent of GDP)
    Emerging and Frontier MarketsFrontier MarketsSub-Saharan Africa
    [1][2][3][4][5][6][7][8][9][10][11]
    U.S. 10-year Treasury–0.438**–0.388*–0.394*–0.251–0.310–0.320–0.315–0.420–0.0540.0160.018
    bond yield(–2.867)(–2.064)(–2.060)(–1.003)(–1.424)(–1.000)(–0.984)(–0.820)(–0.519)(0.118)(0.144)
    VIX (risk aversion)–0.055**–0.062**–0.063**–0.059**–0.065**(–0.081*–0.082*–0.095*–0.015–0.008–0.007
    (–3.621)(–3.572)(–3.547)(–2.916)(–2.690)(–2.552)(–2.547)(–2.208)(–1.559)(–0.980)(–0.920)
    Domestic interest rate0.2620.4460.4591.970–1.943+–2.154–2.292–1.3960.1102.500+2.500+
    (in U.S. dollar terms)(0.254)(0.344)(0.355)(1.618)(–1.706)(–1.236)(–1.334)(–0.579)(0.110)(1.671)(1.677)
    Growth relative to G7–0.0060.0260.0320.0310.0200.0480.0670.072–0.031–0.066*–0.067*
    average growth(–0.101)(0.385)(0.460)(0.414)(0.246)(0.506)(0.690)(0.536)(–1.455)(–2.002)(–1.990)
    Inflation rate (t-1)–0.020–0.020–0.003–0.009–0.0080.020–0.007**–0.007**
    (–1.299)(–1.346)(–0.192)(–0.484)(–0.427)(0.904)(–3.040)(–3.020)
    Total debt-to-GDP ratio–0.010+–0.010+–0.019**–0.007–0.006–0.010–0.009–0.009
    (t-1)(–1.739)(–1.732)(–3.070)(–0.969)(–0.814)(–0.884)(–1.627)(–1.630)
    Fiscal-balance-to-GDP–7.328+–7.355+–4.649–3.986–3.419–0.9700.9590.976
    ratio (t-1)(–1.818)(–1.821)(–1.346)(–1.083)(–0.816)(–0.383)(1.271)(1.242)
    Change in foreign0.0010.005**0.0610.0100.009
    reserves (t-1)(0.887)(3.826)(1.542)(0.296)(0.074)
    Stock market0.679**0.135
    development(3.187)(0.549)
    SSA’s frontier markets0.588+0.2720.271
    (dummy)(1.715)(1.400)(1.300)
    Constant3.594**3.910**3.934**3.650**3.153**3.848**3.781**4.295*0.4700.878*0.878*
    (5.009)(5.093)(5.060)(3.487)(3.375)(3.160)(3.129)(2.268)(0.758)(1.960)(1.965)
    R-squared within0.0540.0580.0580.0340.0540.0660.0720.0630.0040.0080.008
    R-squared between0.0030.2120.2270.5520.0290.1990.1740.4240.0370.3140.310
    Number of countries4340403723202017363636
    Observations496436436346264212212148419380380
    Sources: IMF staff calculations based on IMF’s World Economic Outlook and International Financial Statistics databases, World Bank Financial Development and Structure database. The sample period covers 2001–12 years.Note: Dependent variable is ratio of portfolio investment liabilities to GDP. t-statistics reported in parentheses and **, *, and + indicate significance at the 1, 5, and 10 percent levels, respectively. Estimates obtained using random effect estimation. Domestic interest rates are adjusted for exchange rate changes. Growth relative to G7 growth is defined as the difference between the one-year-ahead projected growth minus average growth of G7 countries.

    Another set of regressions focusing on sub-Saharan African frontier markets was run using quarterly data and disaggregating portfolio flows in equity and bonds flows (Table A.2). The dependent variables are based on EPFR data, defined for each country as the flows into equity and bond funds of the country, divided by the stock at the beginning of the quarter. A crisis dummy is equal to 1 from 2008:Q4 onward. The estimation is limited by the fact that most pull factors are not available on a quarterly basis. Quarterly output growth was only available for three countries in the sample. Ghana and Nigeria were selected as a subsample because these countries have experienced larger portfolio inflows.

    Table A.2.Determinants of Equity and Bond Flows in Sub-Saharan Africa
    PortfolioEquitiesBonds
    Sub-Saharan AfricaNigeria and GhanaSub-Saharan AfricaNigeria and GhanaSub-Saharan AfricaNigeria and Ghana
    VIX−***−***−***−***−***−***−***−***−***−***−***
    Crisis dummy+***+***++***+***
    Output growth+
    Constant+***+***+***+***+***+***+***+***+***+***+***
    R-squared0.220.270.290.340.020.020.230.300.170.200.29
    Countries5522111122522
    Observations126126535347847892471606660
    Source: IMF staff calculations.Notes: (+,−) indicate the sign of the coefficient, ***, **, and * indicate significance at the 1, 5, and 10 percent levels, respectively.
    References

      Beck, Thorsten, Samuel MunzeleMaimbo, IssaFaye, and ThourayaTriki,2011, Financing Africa (Through the Crisis and Beyond). (Washington: The International Bank for Reconstruction and Development/The World Bank).

      International Monetary Fund, 2011a, Regional Economic Outlook: Sub-Saharan Africa (Washington, April).

      International Monetary Fund, 2011b, World Economic Outlook (Washington, September).

      International Monetary Fund, 2012, “2012 Nigeria Article IV Consultation Staff Report” (Washington).

      International Monetary Fund, 2013a, “2013 Ghana Article IV Consultation Staff Report” (Washington).

      International Monetary Fund, 2013b, “Key Aspects of Macroprudential Policy,” Board Paper (Washington, June).

      International Monetary Fund, 2013c, “The Liberalization and Management of Capital Flows: An Institutional View,Board Paper (Washington, June).

    In this paper, a relatively wide definition of frontier markets for sub-Saharan Africa is adopted. Criteria used to select countries include recent growth dynamics and prospects, financial market development, general institutional conditions and evolution, and political conditions and perspectives. Although some of the countries are not included in investment bank indices, there has been sufficient foreign investor interest over the past five to ten years to warrant their consideration here. The list includes Ghana, Kenya, Mauritius, Mozambique, Nigeria, Senegal, Tanzania, Uganda, and Zambia (IMF, 2011a).

    IMF (2011a) examined capital flow developments in sub-Saharan Africa up to 2009.

    A survey of the issues raised in the Article IV staff reports on the quality of data include (i) large net errors and omissions for the published balance of payments; (ii) data collection on transactions in nonresident securities is still a challenge; (iii) unreliable current/capital transfer split for foreign aid and no detailed data on the costs of embassies abroad; (iv) poor coverage of reinvested earnings; (v) data on outstanding debt stocks and principal payments are inconsistent; (vi) financial account is incomplete, as it does not record substantial transactions in assets; and (vii) flows and stocks of gross international reserves and net foreign assets position often require substantial adjustments.

    For Senegal, IMF staff believes the official numbers to be unrealistically large.

    With a few exceptions, the real exchange rates in frontier markets have remained aligned with fundamentals despite the surge in capital inflows. Until end-2012, only Ghana seems to have an overvalued exchange rate while no other appreciation pressures are recorded for the rest of sub-Saharan African frontier markets economies. Moreover, real exchange rates weakened recently in most of the frontier markets countries.

    In the 2012 Article IV consultation (IMF, 2012), the Nigerian authorities indicated that they were mindful of the risk of capital flow reversals and thus were closely monitoring the surge in inflows, especially by keeping track of the maturity distribution of the securities that were being purchased.

    For example, the temptation to use restrictions on bank deposits in low-capacity settings as tools to control capital outflows would not be advisable, as they can severely disrupt economic activity, confidence in the financial system, and prospects for financial deepening.

    At times, CFMs and macroprudential measures may overlap. Policy tools could be seen as both a CFM and a macroprudential policy. For example, a restriction on banks’ foreign borrowing through a levy on bank foreign exchange inflows or required reserve on banks’ foreign exchange liabilities would aim to limit capital inflows or slow domestic credit. In any way, these policies should be seen as supporting effective supervision and complementing appropriate sound macroeconomic policies (IMF, 2013b).

      You are not logged in and do not have access to this content. Please login or, to subscribe to IMF eLibrary, please click here

      Other Resources Citing This Publication