Obtain the data you need to make the most informed decisions by accessing our extensive portfolio of information, analytics, and expertise. Sign in to the product or service center of your choice.
Global companies are invested in a diverse variety of markets -
increasingly in the developing world. According to UNCTAD, in 2018
58% of FDI went to developing economies. While that year saw a drop
of 19% annually into developed economies, developing economies saw
an increase of 3% annually. Total FDI into developing economies has
increased nearly 50% in the past decade.
However, rapidly developing environments are particularly
difficult to predict and are also often less understood by
investors coming from the developed world. Companies investing
outside their "home base" have to take into account a complex
spectrum of country risks. At the extreme end, violent conflict or
expropriation can rapidly make an operation less viable and
certainly less profitable. Less extreme events can also have a
significant impact on investment return - contracts being
frustrated by local counterparts, demand contraction and changes in
taxation, regulation, currencies and costs. By anticipating
potential impacts of this kind, a company can assess if the
expected returns on an investment are sufficient to compensate for
country risk in locations where there is significant
uncertainty.
When evaluating potential investments, companies need to
quantify these country risks in financial terms in order to more
profitably evaluate, compare, and select investment opportunities.
Current tools like sovereign risk indicators reflect only a very
limited view of commercially relevant risks and don't account for
the significant variability of risks by sector. In fact, these
tools often over- or undervalue country risks significantly, by
ignoring the on-the-ground realities in exchange for market
sentiment.
By looking at the specific risks that impact projects on the
ground, we can get a more precise measure of the cash flow impact
of risks.
Peru
In Peru, sovereign risk significantly undervalues the risk.
Though Peru is not a particularly risky country, there are
government stability, terrorism, and protest risks that can have a
significant impact on the operating environment in Peru, especially
for certain sectors and locations. For example, government
stability risks are likely to be high in the next year despite
incumbent president Martin Vizcarra on 5 June having survived a
congressional vote of confidence over his proposed judicial system
and campaign financing reforms. Indeed, if Congress fails to
approve these reforms, the president could call again another
congressional vote of confidence; if he loses, he can dissolve
Congress and call parliamentary elections to break the impasse.
In addition, although incidents of social conflict have declined
from a 12-month high of 202 in September 2018, efforts to restart
stalled mining developments in Arequipa and Piura, as well as
demands of compensation in Apurimac, are likely to provoke
anti-mining demonstrations in the next year.
As for terrorism risks, the Sendero Luminoso (SL: Shining Path)
insurgent group has declined in strength and scope since its peak
in the late 1980s and early 1990s, and no longer poses a security
threat beyond the remote coca-growing areas where it operates in
the central-southern Valley of the Apurímac, Ene, and Mantaro
rivers (VRAEM). However, within the VRAEM, SL poses a significant
risk. SL cooperates closely with drug-trafficking groups and its
presence in the VRAEM gives rise to ongoing risks of extortion,
sabotage, property damage, and kidnapping for firms operating in or
around VRAEM. Natural gas, construction firms, and military
helicopters face a higher risk of attack within the VRAEM.
Mozambique
Though it is a riskier environment in general, in Mozambique
sovereign risk indicators significantly overvalue the risk,
particularly with regard to key sectors such as Liquefied Natural
Gas (LNG). Mozambique suffered a credit rating downgrade and
economic downturn after the International Monetary Fund (IMF) in
2016 discovered USD1.4 billion in 'hidden loans' awarded to state
entities. The IMF subsequently suspended assistance and a new
program has not yet been forthcoming.
However, the fiscal and regulatory environment for LNG
investments has been relatively favorable and predictable for
foreign investors, unlike in the case of regional neighbors such as
Tanzania, where a political agenda focused on resource nationalism
has significantly disrupted finalize of a Host Government Agreement
for new LNG projects. In Mozambique, by contrast, Anadarko on 19
June 2019 took a final investment decision for a USD25 billion
onshore LNG project, demonstrating the divergence between sovereign
and country risk indicators when applied on a sector-by-sector
basis.
Quantified results
By taking into account the level of risk of these important
factors, we find a significant difference in the Country Risk
Premium as opposed to the Sovereign Debt indicator, and very
different premiums based on the sector. Mozambique, though a
riskier operating environment overall, is much less risky than the
sovereign debt indicator would suggest; while Peru is a less risky
operating environment in general, but riskier in certain sectors
such as mining.
By quantifying these on the ground risks, and mapping them to
sectors of interest, we can achieve a much more precise measurement
of the impact of country risk on project or sector cash flow. This
means better investments. Digging into the drivers of that impact
make more targeted mitigation possible, to further lower the risk
of investment. As all that investment flows to risky markets, this
kind of understanding is crucial for having the best strategy.