We present our top picks for the second half of the year – an update to our top picks for 2022, published in December (which consisted of Buys on Cote d'Ivoire, Mozambique, Pakistan and Uganda local, and a Sell on Tunisia).
Our revised Top 5 comprises Buys on hard currency bonds in Ghana, Mozambique, Nigeria, Pakistan and Uzbekistan. Compared to our December piece, this represents quite an overhaul – which we feel is understandable given market conditions – with three new additions (Ghana, Nigeria, and Uzbekistan), while Mozambique and Pakistan are unchanged.
The year so far
What a scorcher, and that's the US inflation print last week (9.1% yoy in June) – its highest for forty years – not the first ever 40c temperature recorded in the UK this week.
Soaring US (and global inflation) has been the main investment theme this year, along with Russia's invasion of Ukraine in February. Rising global inflation already reflected post-Covid policy stimulus, supply chain disruptions, labour market shortages, and capacity constraints, although the further leg up this year is due to higher food and energy prices following Russia's invasion of Ukraine. The impact of higher oil and commodity prices – which many EM would typically benefit from – has however largely been offset by tighter global financing conditions as central banks raise interest rates aggressively to tame inflation, which is at multi-decade highs for many countries.
But with inflation not having been this high for a generation, and after the huge changes in the structure of the economy and policymaking over the intervening period, this could pose a new challenge to policymakers.
A more aggressive interest rate path in the US, as the Fed tries to get ahead of the curve, has rattled markets and raised EM financing costs while the US dollar has strengthened (DXY is up 10% this year). The outlook for the US Fed Funds policy rate has shifted from a consensus of three 25bps hikes by the end of 2022 last December, and a Fed Funds rate of 2.1% by 2024, to a terminal rate of around 3.5% (from 1.50-1.75% currently) – which would take it to its highest rate in 15 years when it reached 5.25% over 2006-07 (the post-GFC peak was 2.5% in 2019). That could present quite a shock to the external funding conditions of many Frontier issuers, many of whom have only been in the market for a decade or so, and few will have experienced conditions like these.
As a result, US 10-year bond yields have risen 150bps this year to 3% (touching c3.5% in mid-June), although the consensus in December was for the ten-year to rise by only c50bps by end 2022 to 2.0%, and to 2.3% by end 2023.
But curiously, before the Fed has even got on top of inflation, market attention has now shifted from inflation to recession, although we don't expect rate relief anytime soon. The probability of a US recession, according to Bloomberg data, is at its highest since the 2008 GFC (excluding the exceptional Covid period). However, despite all the talk of US recession, few are actually forecasting it at this stage while the implications for EM bonds of a US (and global) recession are not clear cut in our view.
Meanwhile, weaker Chinese growth, due in part to the authorities' zero-Covid policy, but also China's own real estate crisis, means it cannot pick up the baton to support global growth as it may have before.
Consequently, most risk assets – including EM hard currency bonds – have had a torrid time this year in the face of the twin external shocks of war and rates.
The total return on the Bloomberg EM Sovereign USD index is -21.8% year to date (as of cob 15 July), after a terrible June (-6.3%) – the worst month this year, and the worst since March 2020 – with the yield having risen by 331bps this year to 8.15% and the spread up by 169bps to 506bps. Indeed, after a period when the spread remained fairly well contained – much to our surprise, given the global headwinds – (and so rising EM yields were really just a function of the rise in the risk-free rate), the spread has risen by more than 100bps since 6 June. The spread (the EM country risk premium) is now, with the exception of the outset of the Covid pandemic, at a post-GFC high. The spread is now 140bps above the pre-Covid (2019) average and 225bps wide of its post-Covid low.
As a result, EM hard currency sovereign bonds are on track for the worst annual performance in over twenty years (if not longer), and worse than during the GFC.
Within the overall index, HY has now outperformed IG on the downside, with a total return of -23.1% on the Bloomberg HY subcomponent compared with -20.5% on the IG portion.
The HY yield has risen 522bps year to date to 12.6%, with the HY spread rising 355bps to 951bps. Meanwhile, the IG yield has risen 204bps to 4.63%, with the HY spread rising 46bps to 153bps.
There have been few hiding places, with all the country constituents in negative territory. Croatia has however been the best performer (-2.5%), perhaps buoyed by its 1 January 2023 expected euro membership. This is followed by Suriname, albeit in default (-4.3%), Poland (-4.9%), Vietnam (-5.7%) and Iraq (-5.8%). Other oil/gas exporters (Trinidad and Kuwait) have also outperformed (-6.0%).
At the other end of the scale, Russia (now in default) and its client state Belarus (flirting with default) are the worst performers, down 100% and 75% respectively in the index – and both have been excluded from the index due to sanctions. These are followed by Ukraine (-75%), amid expectations of a default (perhaps in September), Sri Lanka (-53%) which defaulted earlier this year, and El Salvador (-48%), amid expectations of a default (perhaps in January). Ecuador (-39%), Lebanon (-38.5%), Argentina (-38%), Zambia (-37%) and Pakistan (-36%) are in the chasing pack, followed by a raft of SSA issuers.
Indeed, the EM credit rout is such that nearly one-third (30%) of the Bloomberg index is in distress (defined as a spread of 1000bps or more), while we count 28 countries with yields over 10% on their sovereign dollar bonds. At these levels, many markets – not just frontiers but smaller EM too – have lost market access, which will make meeting financing needs more challenging, and/or pose a risk to those with more immediate refinancing needs (eg upcoming maturities).
Our analysis flags several countries as being particularly vulnerable to the tighter global financial conditions witnessed this year based on large gross external financing needs and lack of market access, although to be clear, some already had pre-existing concerns. These include Egypt, El Salvador, Ghana, Kenya, Pakistan and Tunisia, while of course Ethiopia is pursuing a common framework treatment (which may or may not include its bond) and Sri Lanka has already defaulted.
Lack of market access and high external funding needs could force more countries to the IMF (and other multilateral lenders), following Ghana's recent lead, to bilateral partners, or to consider more creative financing structures (like the Bahamas), or they will be forced to tighten policies, or to resort to potentially inflationary domestic financing, sharp exchange rate adjustments, capital controls or – in the limit – default.
This fate has already befallen Sri Lanka – the first sovereign default this year – although it was followed by Russia (which Moscow disputes). They could soon be joined by Belarus and Ukraine (however, the latter three are due to a specific set of circumstances).
In this regard, the IMF MD, Kristalina Georgieva, expressed concern about a wave of EM sovereign distress in her address to the G20's FMCBG on 16 July. That said, those seemingly most at risk tend to be smaller markets – and well flagged – so mitigating the threat of a wider systemic crisis, in our view, although investors may be concerned about domino effects and uncertainty over the international policy response. In addition, multiple simultaneous debt defaults could test the sovereign debt architecture.
Meanwhile, EM hard currency sovereign bond issuance has been subdued, with issuance in the first half of the year amounting to US$58.4bn, about 44% below the first six months of last year, and portfolio debt flows into EM have eased up considerably. Inflows in the first half of this year amounted to US$29bn (about US$5bn a month), according to the IIF, although ex-China inflows amount to US$40bn (US$6.7bn a month). This contrasts with ex-China inflows of US$209bn in the whole of 2021 (about US$17.4bn a month). Still, at least they are net inflows (there has, surprisingly, been only one month of net outflows this year, in April, although the data is subject to revision).
In addition, EM has been also been weighed down by a stronger US dollar (the Bloomberg EM FX basket has fallen by 9% ytd vs the US dollar). Weaker EM currencies could feed domestic inflation while a stronger dollar increases the local currency cost of servicing foreign currency debt. However, most of the dollar's strength has been against DM currencies, and so EM local currency has displayed some resilience (the Bloomberg EM local currency index is down 9.6% YTD in total return terms). This was driven by pre-emptive hiking of EM central banks, but could begin to reverse as rising DM rates erodes differential with many EM already having done the heavy lifting. Also with many EM now showing similar or even lower rates for LCY debt than US$ debt, this will obviously undermine the attractiveness of EM domestic debt, while investors may also shift to better liquidity in international bonds. For instance, YTW is 4.36% for Bloomberg EM local currency index versus 8.05% for Bloomberg EM hard currency index, although strictly the indices are not strictly comparable.
Yet, it is still unclear if we have reached the bottom, given the backdrop, despite the market falls so far. That said, with many EM bonds now so distressed, we think the sell-off has unlocked pockets of value, especially in higher-yielding frontier markets and smaller EM, although further weakness could be in store in the near term.
Our Top 5 picks
We summarise our top picks below.
1. Buy Ghana 8.125% 2032
We elevate Ghana to our Top-5 with a Buy on GHANA '32s at a price of US$47.6 (yield 21.8%) as of cob 18 July on Bloomberg (mid price basis). This follows our upgrade to Buy from Hold on 9 June after prices fell in line with even our more pessimistic estimates of recovery values, which – for illustrative purposes – assumed a combined 50% coupon haircut and a 25% principal haircut, thereby opening up some value.
While mainly a value play, given weak fundamentals, our positive view has been reinforced by the government’s surprise announcement on 1 July to seek an IMF programme – much earlier than we had expected. This may help to reinforce policy discipline and stave off default, which investors had increasingly come round to thinking was all but inevitable without Fund support. Indeed, the bonds jumped 6pts initially on the news, although they have since given up those gains, and more, and are now below the level they were before the announcement.
The request for IMF help does however raise some interesting questions in terms of programme design, which may be weighing on investor sentiment after the initial euphoria wore off. First, the size (and composition) of the required fiscal adjustment, and whether it is feasible, especially in light of continuing revenue underperformance and the IMF’s own fiscal realism tool in last year’s AIV DSA that cast doubt on the ability to achieve consolidation. Second, from a bondholders’ perspective, whether any new IMF programme will seek burden sharing (PSI), given debt sustainability appears to be on a knife-edge, pushing Ghana into Common Framework territory. Public debt is a high but not scary 80% of GDP.
However, Ghana has a domestic debt problem too, which also feeds into its high interest burden. Interest is over 7% of GDP and two of thirds of this is interest on domestic debt. Hence PSI on the eurobonds may provide only limited results. There is also some breathing space on the eurobonds in that Ghana’s first sizeable bond maturities are not until 2025/26.
The next milestone will be Ghana’s delayed mid-year budget review, which the government confirmed on Monday will now take place on 25 July. However, after over a year of fiscal ineptitude, we struggle to see what it can usefully say after barely three weeks of talks with the Fund.
2. Buy Mozambique 5% 2031
We retain Mozambique in our Top-5, with a Buy on the MOZAM 5% 2031 at a yield of 15.9% (price US$66.6) as of cob 18 July on Bloomberg (mid price basis). While we still like Mozambique, having it as one of our top picks for 2022, we recognise the bonds have fallen by c20pts this year, although most of this decline has occurred since June. Until then, Mozambique had been one of the best performers in the Bloomberg index – justifying our pick. It has now suffered a total return of -19.2% ytd, roughly in line with the index.
We reiterated our positive view on Mozambique following the IMF/WB Spring Meetings in April. This came after the IMF announced staff-level agreement on a new programme in March – a US$456mn (150% of quota) three-year ECF was subsequently approved on 9 May – and our discussions at the Spring Meetings served to underline our positive view on the authorities’ strong ownership of the intended programme and its reform commitment.
Our positive view on the bonds is based on the expected onset of LNG production (first gas is due in H2 from the ENI FLNG project) and strong ability to pay, given low debt service on the bonds (and little supply risk in the near term), together with the government's prudent approach to macro-management. Interest on the bonds is only US$45mn per annum until 2023 (although it then increases to US$81mn when the coupon steps up from 5% to 9% in 2024), so Mozambique shouldn’t be defaulting on this in the near term, and risk further jeopardising its LNG projects, while reserves (gross basis) were US$3.5bn at end-2021. The lower public debt burden is also a positive (projected by the IMF at 101% of GDP this year, although we still await the publication of the staff report for more detail). The IMF programme, which improves the outlook for donor funding (and should ensure Mozambique is fully funded over the programme period), is the icing on the cake.
Nonetheless, there are risks, primarily the fragile security situation (mainly in the north), possible social pressures (including arising from higher food and fuel prices), risks from higher inflation (10.8% yoy in June) and currency devaluation (with concomitant risks to the debt burden given some four-fifths is external), the timing of LNG production, governance and capacity, and climate vulnerabilities.
3. Buy Uzbekistan 3.9% 2031
We elevate Uzbekistan to our Top-5 with a Buy on UZBEK '31s at a yield of 9.0% (z-spread 617bps) as of cob 18 July on Bloomberg (US$68.6 on a mid price basis). This follows the Buy we assigned to the bond on 22 April (when the yield was 5.9% and with a z-spread of 317bps). However, we recognise the bonds have fallen by 17pts since then (spread 300bps wider), with a total return of -20.1% ytd, roughly in line with the Bloomberg index.
Still, we think fundamentals remain strong and valuations are even more compelling now. Our view on the bonds followed the IMF’s 2022 Article IV mission in April which painted a reasonably positive picture despite the impact of the war in Ukraine (the board review was concluded in June and the staff report was published). Growth this year was revised down (from about 6% pre-war to 3-4%), primarily due to heavy dependence on remittances from Russia, and inflation (around 12%) was seen as high but stable. However, we think policy space (low public debt, strong liquidity buffers) may help to mitigate the economic impact, and is supportive of ability to pay. With four dollar bonds now, amounting to just US$2.2bn in size (only 3% of GDP), total interest payments are still only US$96mn a year, albeit with the next maturity, US$500mn, in 2024.
Uzbekistan remains at low risk of debt distress, according to the IMF assessment, with public debt projected at 38% of GDP this year, albeit some 10ppts higher than pre-Covid levels. Moreover, a high level of international reserves (over twelve months’ import cover) and long maturities further mitigate the risk of debt distress. Indeed, on our calculations, international reserves (including gold), are c50% of GDP. This makes Uzbekistan a net public external creditor.
Investors may however be concerned about the rising debt burden (public debt is deemed sustainable but the trajectory is upwards), high external financing needs, the risk that Uzbekistan gets drawn into Moscow's sphere of influence, as Putin seeks to re-create the Soviet bloc, domestic and regional unrest, and concerns over governance and transparency.
That said, we observe that with the Uzbekistan sovereign dollar curve pretty flat, investors are not getting paid for duration. The yield on the February ‘24s at 8.5% (that’s 8.5% for essentially 18m paper) compares to 9% on the ‘31s. Hence, the ‘24s could be an interesting short-duration trade for defensive investors (with a price of US$94.5). Still, the ‘31s are also interesting at a low cash price of US$69 (incidentally, that is virtually the same price as Mozambique but Uzbekistan has much stronger credit metrics); for a country with debt/GDP of 40% and reserves equal to 50% of GDP, Uzbekistan should not be defaulting.
4. Buy Nigeria 7.875% 2032
Nigerian credit has sold off sharply this year, with the Bloomberg Nigeria Sovereign spread rising by 495bps (-27.6% total return) versus 157bps (-21.5% total return) for the Bloomberg EM Sovereign Aggregate. The NGERIA 7 ⅞ 02/16/32s now yield a whopping 14.56% (1,161bps z-spread) and have a cash price of just US$66 as of cob on 18 July, implying a high degree of default risk.
This has occurred despite a sharp rise in oil prices, which comprises c5% of GDP, c45% of revenue and c85% of exports. However, chronic mismanagement of Nigeria’s exchange rate regime has limited the improvement of Nigeria’s current account and weighed severely on its financial account, with gross reserves actually declining to below US$39.5bn, from US$40.5bn at the beginning of the year, despite the oil price environment and the naira trading at a 50% premium on the parallel market.
Further, refusal to remove Nigeria’s costly petrol subsidy means that every US$10 increase in oil prices increases the budgetary cost of subsidies by 0.5% of GDP, according to our estimates. As such, a rise in oil prices is a mixed blessing, helping to improve Nigeria’s external position but weighing on its budget. Further, a sharp decline in production (from 1.88mbpd in 2019 to 1.4mbpd in the first half of the year and just 1.2mbpd in June, well below the 1.77mbpd June OPEC quota) has limited the overall benefit from higher prices, so its underperformance relative to other oil producers is justified.
However, the extent of the sell-off is surprising, with bonds now pricing in a high risk of distress despite a public debt stock of just 35.3% of GDP at the end of 2021 (or 39.3% of GDP if external debt is valued at the parallel exchange rate). That said, public debt is projected to rise to 44% of GDP by 2027, according to IMF estimates, and debt service absorbed an eye-popping 96% of federally retained revenue in 2021 (or 47% of total government revenue), pointing to severe liquidity problems (although a negligible primary deficit, which reached 0.3% of GDP in 2021, should limit fiscal financing needs).
Nigeria’s external liquidity position, however, is less concerning, with just US$1.6bn of external amortisations due annually from 2022-26, per World Bank IDS data (and less than US$2bn of eurobond amortisations due from 2022-25 before the US$5bn maturity hump in 2026). Adding an estimated current account deficit of c1% of GDP (US$6.2bn) over this period, and Nigeria’s gross external financing needs are negligible compared with its US$39.5bn reserve stock (although a sharp decline in oil prices would leave Nigeria vulnerable, with reserves just barely treading water with Brent above US$100/bbl).
So, while we remain quite negative on Nigeria’s fundamental story (see here for a detailed analysis), Nigeria’s risk of debt distress is quite low in the coming years, at least based on consensus oil price projections (indeed, Nigeria scores highly on both our debt sustainability and external liquidity indexes). Further, with elections looming in February 2023, there is potentially some upside risk if a more reform-minded government emerges (although it is not clear if either candidate has clear economic policy preferences, and neither is likely to have the will or ability to tackle Nigeria’s big structural problems).
With the NGERIA 7 ⅞ 02/16/32s already pricing a high risk of distress (prices reflect a 10% principal haircut at a 14% exit yield, a 50% coupon cut at 10%, and a combination of the two at 8%, for illustrative purposes), this should, in theory, put a floor on further price declines (although the ongoing market rout means bonds could nonetheless drop temporarily below that floor until global financial conditions start to loosen). As such, we reiterate the Buy recommendation we initiated last month, with the ongoing sell-off creating a compelling value proposition for Nigerian eurobonds given limited near-term default risks (notwithstanding a weak long-term fundamental story).
Note: This preceding text is an updated excerpt from our full recommendation published on 21 June. See here for our full analysis and investment thesis.
5. Buy Pakistan 7.375% 2031
Pakistan’s eurobonds, one of our top picks for 2022, have been among the worst performers this year as political uncertainty and reform fatigue have repeatedly derailed its IMF-backed reform programme and rising commodity prices, large gross external financing needs, and tight global financial conditions have weighed on reserves. Through 18 July, the Pakistan Bloomberg Sovereign Index had widened by 1,318bps (-37% total return) versus 157bps (-21.5%) for the Bloomberg EM Sovereign Aggregate.
We have repeatedly said that default is inevitable without the resumption of Pakistan’s stalled IMF programme, with a massive gross external financing requirement of US$41bn in FY23 versus liquid FX reserves of just US$10.3bn (1.5 months of imports). While a staff-level agreement was reached with the IMF last week, which should unlock further financing and temporarily stave off the risk of a balance of payments (BOP) crisis, bonds have continued to plummet since.
Indeed, the PKSTAN 8 ¼ 04/15/2024s dropped by c12pts yesterday and the PKSTAN 7 ⅜ 04/08/2031s by c6pts, reaching prices of just cUS$56.2 and cUS$47.5, respectively. The PKR also depreciated by 5% to a record 221/US$, bringing ytd depreciation to 20% (which should rein in some of Pakistan's external pressures if the PKR is allowed to continue serving as a shock absorber, with a limited associated impact on debt sustainability given domestic debt making up 2/3 of the total stock).
The weakness was driven by a by-election victory by recently ousted PM Imran Khan’s PTI over the weekend, prompting renewed calls for an early election (due by August 2023 at the latest) and associated concerns that the reform programme could fall back off track. This also prompted Fitch to downgrade Pakistan's credit outlook to negative yesterday on concerns of rising external vulnerabilities and risks to reform implementation.
There is certainly some merit to this caution given Pakistan’s weak track record with the Fund and a political backdrop that will make reform implementation difficult, with the PTI’s by-election victory seen by many as a referendum against PM Sharif’s reform efforts. This raises the risk that the ruling PML-N will either backtrack on its promises to bolster its chances at the polls, or that former PM Khan will reverse the recent reforms if he wins the eventual elections (which seems likely).
If this happens before Pakistan’s external position has improved sufficiently and/or external financing conditions have loosened, the risk of BOP crisis and associated default will again rise sharply. That said, with the PKSTAN 7 ⅜ 04/08/2031s now trading well below both our base case recovery value of US$59 and our bear case of US$53 based on our June analysis and 12% exit yields, we think the risk of default is fully priced in.
While the upside is capped by political uncertainty and a weaker long-term reform outlook, if spreads on the PKSTAN 7 ⅜ 04/08/2031s drop to 1,000bps over time – double the c500bps spread that prevailed in early 2021 – this would imply 50% of upside. Further, downside is capped by our estimated recovery value even if Pakistan defaults (which, admittedly, is likely if they do not stick to their IMF targets).
The 15.5% current yield also provides further downside protection even if Pakistan defaults, but not right away, with the potential for bondholders to collect some hefty coupons if Pakistan makes a couple of payments first (the US$1bn sukuk due on 5 December still trades at US$86, suggesting that even the pessimistic market thinks there is some runway, although there could also be technical factors at play).
As such, despite an elevated risk of distress, we think Pakistan’s eurobonds are compelling from a value perspective, especially given the ongoing price drop and recent steps towards restoring its IMF programme, which the market has not priced in at all. We affirm our Buy recommendation on the PKSTAN 7 ⅜ 04/08/2031s at US$47.5 (20.6% YTM) as of cob on 19 July on Bloomberg.