Zambia is seeking to restructure its foreign debt, according to Bloomberg reports on Tuesday, after issuing a request for proposal for advisors. It becomes the second country over the last week – that we know of – to seek some kind of debt treatment or relief on its bonded debt, in relation to the impact of Covid-19, following Ecuador's suspension of some upcoming coupons last week. Details are sparse however, and we only really have the Bloomberg report to go on. We outline some initial thoughts here.
Zambia's announcement may not be much of a surprise. Even before Covid-19, the writing has been on the wall for some time. It may be that the economic impact of the coronavirus pandemic has finally pushed Zambia over the edge, given its pre-existing vulnerabilities, and the muddle-through strategy finally reached its limit. Indeed, the final nudge may simply be that the government has virtually run out of reserves – gross international reserves fell to US$1.28bn at end January, according to central bank data.
However, the timing looks fortuitous. We asked the question last week whether some sovereign borrowers would follow Ecuador's lead and look opportunistically to take advantage of current circumstances to seek some kind of relief on bond payments, especially now under the cover of the joint IMF/World Bank Call to Action to the G20 to suspend debt payments owed to official bilateral creditors from certain poor countries (for which Zambia would be eligible). That Zambia has seemingly rushed into this announcement ahead of formal G20 agreement on such official sector debt relief may seem curious (although it may have been planning the RFP for some time, we don't know) – it might have wanted to test the waters with official bilateral creditors first. Indeed, while this decision is in some sense long overdue (market participants had been expecting a bond restructuring for a few years now), the delay in action until now may have worked in the authorities' favour if they are able to capitalise on current conditions to obtain an even more generous treatment than they would have previously been given.
Moreover, with elections approaching next year, a debt restructuring at this juncture – along with multilateral financial support – may even help the incumbent President Lungu who may have previously feared the implication of such a decision on his government's economic credibility and his support base, and if executed quickly, could allow time for gains to be shown to the public. Conversely, any IMF-imposed conditionality, that this President has hitherto resisted, could undermine his popularity (although at this stage, the nature of any IMF involvement, if it all, is unclear).
Some questions and issues at this early stage:
1. It is not clear what Zambia is looking for. The report states that Zambia is seeking a "reorganisation" of its foreign debt. Bondholders may take some comfort from the word reorganisation, which may imply a mild treatment, but we think it could cover a number of different outcomes, ranging from simple maturity extension, coupon holidays, to principal haircuts. That said, we think a strong and well-coordinated bondholder group, along with the bonds' collective action clauses (CACs), provide some protection for bondholders. The report suggested the government is seeking a collaborative approach.
2. The perimeter of the restructuring. The article refers to foreign debt, and the intention seems to be that it will cover its entirety – multilateral, bilateral, banks, bonds, export credit agencies, and others. Official ministry of finance figures show public external debt was US$11.2bn at end-2019. We think eurobonds and China may account for at least half of this. Eurobonds amount to US$3bn, so about 27% of the total. We think China accounts for another 25% (at least). It is not clear whether eurobonds are the main (or really sole) target – given their size, annual debt interest payments (about US$237mn), and most importantly the looming US$750mn maturity of the 2022s, they probably should be a focus – or whether a more comprehensive treatment is envisaged in practice.
3. Role of China. We think bondholders – and the IMF/IFIs – would find it difficult to accept a solution that did not include China, given its share of the debt and (likely) debt service. But whether China will participate, and on what terms, is unclear. Zambia's previous efforts to seek bilateral restructuring with China, which were announced two years ago (February 2018), haven't appeared to have gone anywhere. Nor may China want to cede to the degree of transparency that a comprehensive treatment would entail, especially if a restructuring took place in the context of an IMF programme. There is already a precedent of China participating in a wider restructuring in the context of an IMF programme, in the case of Republic of Congo, but Zambia may be a much more high profile case. But similarly, we doubt a China-only restructuring, without IMF involvement, will be enough to restore sustainability.
4. Could the authorities even decide after negotiations with bondholders to exclude the bonds? This seems doubtful to us, especially given the proximity of the 2022 maturity (it may have been viewed as an option by bondholders a few years ago, but not now, in our view). Omitting the bonds would put more of the onus on China, although one might assume it could even accept a generous bilateral agreement as the price for avoiding the transparency that a comprehensive agreement would mean. Yet if the IMF are involved, we doubt a China-only restructuring would be acceptable to the Fund (it may not provide enough liquidity relief), while a China deal that didn't involve the IMF (or bondholders) would leave residual questions from bondholders over the authorities' policy commitment that could mean secondary market yields fail to decline to restore market access in a reasonable time. The game dynamic between the IMF, China, and bondholders will be important. We think bondholders would want an IMF programme; failure for the IMF approve to one, for whatever reason, could endanger the chances of a bond restructuring.
5. Role of the IMF. Despite many words and false dawns over recent years about the authorities' intention to seek an IMF programme, including three aborted attempts at securing one since 2014, the government has consistently failed to do enough to convince the Fund (and market participants) that they are serious – which essentially comes down to fiscal commitment and borrowing restraint, and which led the IMF to suspend programme discussions in Autumn 2017. Given the poor track record, we suspect bondholders would rather the IMF be part of the solution to impose external policy discipline and oversight (and the IMF in turn demand that bondholders are too). Whether the authorities will now submit to an IMF programme is still unclear in our view – one reading is that the authorities' hand has now been forced by the current global economic crisis and that maybe the IMF has suggested a debt restructuring would be a prior action to a new programme. Another potential hurdle (and trigger for this decision, now) could be the approaching presidential election, due August 2021, whereby the window for approval of an IMF programme will shorten as the election approaches – leave it much longer, and the opportunity may have gone for several months. That said, under the IMF/WB's Call for Action, the IMF may be (or may have to be) more pragmatic in its response to Zambia, when recent relations have been somewhat frosty. The IMF already assessed Zambia's public debt to be at high risk of debt distress, and called for a 2% primary surplus over the medium-term (commitment basis) in its adjustment scenario – see its 2019 Article IV from August 2019. The implied fiscal effort may now need to be greater given current circumstances, which may therefore require burden-sharing if it is not seen as feasible or desirable, although bondholders will argue they should not need to suffer a harsh treatment in response to what should hopefully prove to be a temporary shock (Covid-19) and to make up for the government's previous policy laxity. The government may also seek financing from the IMF's emergency facilities to help it fight the impact of Covid-19.
6. Domestic debt may be excluded. While the report mentions foreign debt, it is possible that either domestic debt is excluded or will be treated separately. Domestic public debt was about US$5.7bn equivalent at end-2019 according to MOF figures (about one-third of public debt).
7. What is total public debt? Figures from Zambia's ministry of finance put total public debt at cUS$16.9bn at end-2019 on our estimates (about 70.5% of GDP). However, it is higher according to IMF methodology. We estimate public debt was projected at about US$20.4bn in 2019 on IMF figures, according to the 2019 Article IV. The IMF includes arrears (about 9% of GDP) and external guarantees (4% of GDP), although the IMF figure may be an overestimate to the extent that new borrowing over the remainder of the calendar year was less than they had projected. The IMF projected public debt at 91.6% of GDP in 2019, with public external debt of 60% of GDP and public domestic debt of 32% of GDP. The fall in the kwacha this year (down 22% against the US dollar in Q1 according to Haver) would likely raise the debt/GDP burden even further, other things unchanged. Of course, this will have implications for any future DSA from the Fund.
8. Timing. The report states that banks have been asked for proposals, which suggests the process is very much at an early stage. A US$42.5mn coupon on the 2024s is due on 20 April. It remains to be seen if the authorities will pay it. After this, there is a bit of space until July.
Implications for bondholders
It is too early to say for sure what this means for bondholders. One important question will be whether Zambia's situation is seen as a liquidity problem rather than a solvency one; that is, some short-term easing in the debt service burden along with stronger policies will provide time to restore macrostability without the need for debt reduction. That was certainly the view from many bondholders before Covid-19. Indeed, stronger policies, an IMF programme, and a bilateral restructuring agreement with China may have obviated the need for any kind of bond treatment at all, or at least, if one was necessary, only a mild one.
We think the former liquidity diagnosis is still appropriate, but arguments may be had either way. Debt service on bonds (interest of US$237mn and a looming US$1bn maturity which has little prospect of being refinanced at reasonable rates) is heavy relative to reserves (official reserves of only US$1.3bn, and an adverse outlook for export earnings, especially copper prices, which might pressure the balance of payments and suggest limited chances to allow reserve accumulation). But stronger fiscal policies (cutting expenditure, including capex as well as current spending and transfers, and widening the revenue base, while exerting greater control over borrowing and improving public financial management) should also put debt on a more sustainable trajectory. And while public debt/GDP is high, at 90% (on IMF figures), that doesn't necessarily imply debt is unsustainable under stronger fiscal policies which also support confidence in the currency. That said, an argument could still be made that the weak medium-term growth outlook (with real GDP growth of around 3% as per the IMF's adjustment scenario, rather than the high average growth rate seen historically; over 7% in the ten years to 2014) may still imply an unrealistic fiscal effort to stabilise the public debt ratio at these high levels, let alone reduce it.
We assume for now that this is a liquidity problem, which might suggest a combination of maturity extension and/or coupon reduction, grace period. Indeed a temporary coupon holiday for each of the three bonds, say over the rest of 2020 (with unpaid coupons capitalised or agreement on a new repayment plan for unpaid coupons, a la Cote d'Ivoire 2012) could be the best outcome for bondholders (the PV loss from deferring one coupon would be relatively minor in the big scheme of things). However, this wouldn't address the 2022 maturity, so we think this would have to be accompanied by a maturity extension of the 2022s at least – but it is not clear to us whether the 2022s can be isolated specifically in this way without evoking calls for equitable treatment to extend a restructuring to all the bonds and making it a more complicated, and longer, exercise.
We focus on the 2022 bond to keep things tractable. The problem with a simple five-year maturity extension of this bond, however, is that it would mean the new principal falls due in 2027 when the existing 2027s also mature. However, the 2027s amortise in the last three years of its life in three equal payments, so this mitigates against producing a lumpy debt service schedule. We assume a four-year extension here to represent a possible "best case" (we also assume bullet; if it was a longer extension, we could also amortise it to spread the debt service burden). There is no principal haircut. We also assume – as an extreme case – that one coupon (September 2020) is written off (there is no strong rationale for this, other than perhaps being consistent with the G20 Call to Action) otherwise the coupon is unchanged at 5.375% over the longer life of the bond, resuming in March 2021. This might be the most basic solution, and any worse terms – such as principal haircut or lower or no coupons for longer – would imply lower recovery values, excluding the possibility of interest capitalisation or higher coupons further out to compensate for PV losses from maturity extension.
Our estimated recovery values for our simple illustrative re-profiling of the 2022s under a four-year maturity extension and one forgone coupon are shown in Table 1, with recovery values (PVs) expressed per unit of existing principle for value today. They imply a 20-25% PV reduction over a range of more normal exit yields (12-15% in the current market context), when comparing the PV of the new bond with the PV of the existing bond discounted at the same rate, although recovery values – in the 60s – are still higher than prevailing secondary market prices (indicated around US$41 mid, as of cob 31 March, and implying a yield of c50%; indicative prices across all the bonds fell to cUS$31 mid, as of cob 1 April). Recovery values would be lower still if the bond was exchanged into a longer maturity (eg an eight-year, 2030, to avoid the repayment schedule over 2024-2027), upon principal reduction, and/or if more coupons are suspended (eg those over 2021), although in such circumstances we might expect bondholders to seek compensating measures to recover lost PV such as higher or step-up coupons further out, or capitalisation of missed coupons.
We also provide two alternative illustrative scenarios below – a nominal haircut and a longer maturity extension:
First, re-profiling of the 2022s as above, a four-year maturity extension and one forgone coupon, with a 20% principal haircut (Table 2). This implies a 35-40% PV reduction over the 12-15% range for exit yields, with estimated recovery values in the 48-55 range.
Second, an eight-year extension (to 2030), amortising in the last three years, with three equal annual payments, no principal haircut, and two forgone coupons (September 2020, March 2021), resuming unchanged coupons thereafter over the extended life of the bond (5.375% from September 2021). Our estimated recovery values are shown in Table 3. The implied PV reduction rises to about 30-40% over the 12-15% range for exit yields, with estimated recovery values in the 50s.
We assign a Hold to all of Zambia's US$ bonds at current prices (we had a long-held Sell on Zambia). We see upside potential in a post-Covid-19 global recovery and if bondholders can agree more favourable restructuring terms, but it is too early to call this now. There are also downside risks from the length and depth of the current crisis, the resulting macro-economic impact, globally and on Zambia, the impact on Zambia's debt sustainability, policy commitment and implementation, and what this might imply for potential treatments and recovery rates, especially if this calls for principal haircuts.
|PV of new bond**|
|Implied PV reduction|
*Price/yield combinations for the existing 5.375% 2022 bond under current terms based on Tellimer Research calculator. Figures may differ from those derived from Bloomberg YAS. **Four year maturity extension (2026), bullet, unchanged coupons, but September 2020 coupon is forgone. No principal haircut. ***Indicative mid-price as of 31 March 2020.
PV of new bond**
Implied PV reduction
*Price/yield combinations for the existing 5.375% 2022 bond under current terms based on Tellimer Research calculator. Figures may differ from those derived from Bloomberg YAS. ** Four-year maturity extension (2026), with 20% principal haircut, bullet, unchanged coupons, but September 2020 coupon is forgone. ***Indicative mid-price as of 31 March 2020.
|PV of new bond**|
|Implied PV reduction|
*Price/yield combinations for the existing 5.375% 2022 bond under current terms based on Tellimer Research calculator. Figures may differ from those derived from Bloomberg YAS. ** Eight-year maturity extension (2030), amortising, unchanged coupons, but September 2020 and March 2021 coupons are forgone. No principal haircut. ***Indicative mid-price as of 31 March 2020.