Tunisian bond prices have continued to slip as investor optimism over a possible IMF programme has dwindled, economic and political pressures have increased, and global financial conditions have tightened, pushing the US$-denominated TUNIS 5 ¾ 01/30/2025 bond from just above US$80 at the beginning of the year to US$56.6 at the time of writing (albeit still above the low of US$52.5 reached at the height of the EM sell-off in May).
Bonds have now fallen below our target price and base case recovery value of cUS$65 calculated in December. However, Tunisia’s political and economic outlook has also worsened since then (increasing the probability of default) and rising global rates have pushed up exit yields (lowering recovery values).
As such, we conduct an updated debt sustainability analysis (DSA) and recovery analysis to determine if the sell-off is sufficient to upgrade our recommendation from Sell to Hold.
We find that bonds are now roughly in line with our updated recovery value of US$50-54 and target price of US$54-57 with relatively symmetrical risks, and upgrade our recommendation from Sell to Hold on the TUNIS 5 ¾ 01/30/25s at US$56.6 (31.1% YTM) at the time of writing on 14 June on Bloomberg.
Reform and IMF prospects remain dim
Rhetoric about IMF programme prospects from the government has been positive in recent weeks, with Finance Minister Boughdiri saying last week that he expects to agree with the IMF on resuming negotiations in the next few days. In preparation, Boughdiri said that Tunisia will begin to gradually reduce food and energy subsidies in 2023 and will cut the public wage bill by 5% over the next three years. The wage and subsidy cuts will be offset by cash transfers to lower-income Tunisians. The government will also roll out plans to restructure Tunisia’s struggling state-owned enterprises (SOEs), but Employment Minister Nasreddine Nsibi clarified that “no SOE will disappear because of this plan".
These are likely to be the key prior actions for any potential IMF programme, and represent a walk back from recent statements that Tunisia would not agree to subsidy or wage cuts. Food, energy and transport subsidies are budgeted to increase from TND6bn (4.6% of GDP) in 2021 to TND7.3bn (5.2% of GDP) in 2022, but this was based on an average Brent oil price of US$75/barrel, with each US$1/b increase above this level leading to TND137mn of extra spending, according to central bank (BCT) Governor Marouane El Abbassi. If Brent averages cUS$110/b on the year, as implied by the futures market, this would increase the fuel subsidy bill by TND4.8bn (3.5% of GDP).
Indeed, El Abbassi says that the budget deficit is now projected to rise to 9.7% of GDP this year versus the initial 6.7% target, due mainly to higher subsidy spending. This compares with a 7.5% of GDP deficit in 2021, which came in below the budgeted 8.3% deficit and the 9.4% deficit recorded in 2020. Meanwhile, spending on wages made up 57% of government spending in Q1 22 and is budgeted to stay flat at 15.5% of GDP in 2022, among the highest in the world and nearly double the non-oil producing EM median of 8.7% of GDP in 2020 (IMF estimate). Against this backdrop, the need to cut wage and subsidy spending in Tunisia is clear.
However, the powerful public sector union, UGTT, has remained steadfast in its condemnation of any plans to reduce to the role of the state, including wage and subsidy cuts and privatisation, and is drawing up its own reform plan to present to the IMF. A recent Afrobarometer poll shows that the vast majority of Tunisians agree with the UGTT’s stance on subsidies, with 86% reporting that they want to keep subsidies in place.
Meanwhile, President Saied’s political capital is largely tied up in his constitutional referendum, now officially scheduled for 25 July. A "national dialogue” is ongoing but is by all appearances simply meant to rubber-stamp Saied’s proposals, and is thus being boycotted by the UGTT and most opposition parties. The UGTT is planning a national strike on 16 June after the government failed to negotiate with the union on the planned constitutional and economic reforms.
Saied has also continued to tighten his grip on the country, swearing in a new and friendly Independent High Authority for Elections (ISIE) ahead of the referendum, personally appointing new governors for 13 of Tunisia’s 24 provinces, and firing 57 judges (prompting an indefinite nationwide strike of judges). Saied is now effectively running the country via one-man executive rule, having dissolved parliament and taken control of the judiciary.
Against this backdrop, it is difficult to imagine the government being able to commit credibly to reforms and implement the politically difficult prior actions needed to secure an IMF programme. The UGTT is ultimately big and powerful enough to bring the country to a halt if Saied does not take its economic demands into consideration, which are fundamentally incompatible with the IMF’s likely reform requirements, and the constitutional referendum will keep Saied on a collision course with the UGTT and opposition and will continue to absorb most of his bandwidth and political capital.
Debt sustainability analysis
Central government debt in Tunisia rose only slightly from 77.8% of GDP in 2020 to 79.2% in 2021 per government estimates, less than the 85.6% initially projected in the 2021 budget. However, the trajectory moving forward is less favourable, given the expected widening of the deficit in 2022 and worsening of the real growth-interest rate differential. Further, Tunisia’s contingent liabilities are significant, with official data on Tunisia’s 30 largest SOEs showing a debt stock of 40% of GDP at the end of 2019 (of which 15% is guaranteed by the government) and circular arrears complicating matters even further (with the government owing 9% of GDP to SOEs and SOEs owing 6% of GDP to the government). These estimates are obviously quite stale, and have likely increased since.
We forecast a sharp increase in central government debt to 90% of GDP this year (c85% of GDP on a PV basis) and 97% by 2027 if the government can deliver 6pp of fiscal consolidation over this period (or 4pp relative to 2021), an ambitious target that would see the primary deficit decline from a projected 6.6% of GDP this year (or 4.7% of GDP in 2021) to 0.6% of GDP in 2027. In this scenario, the stock of public and external debt relative to GDP would remain above the IMF’s sustainability thresholds for low-income countries with a “strong” debt carrying capacity, but external debt to exports and external debt service to exports and revenue would fall within the relevant thresholds by 2027 (thresholds are presented for illustrative purposes only as Tunisia is a lower middle income country).
However, a one standard deviation shock (based on 2010-19 data) to our baseline growth and primary balance projections and a 200bps rise in the interest rate on new domestic and external debt would push central government debt to 120% of GDP by 2027. Adding a 30% exchange rate shock relative to the end-2023 forecast would boost debt to 135% of GDP by 2027, showing a high degree of sensitivity to exchange rate depreciation. In this scenario, public debt would clearly be unsustainable.
And even in the baseline forecast, which assumes a heavy 6pp fiscal consolidation and uses relatively favourable macro assumptions, it is not clear that public debt can be put on a sustainable trajectory with reforms alone.
As such, even if Tunisia secures an IMF programme, there is a risk that its debt stock will be deemed unsustainable and that the IMF will be unable to provide financing without a debt restructuring. But because super senior multilateral debt makes up c50% of Tunisia’s external debt stock, nominal haircuts on the rest of the external debt have very little impact on the end debt level if they are not also applied to domestic debt. Even in our base case with significant fiscal consolidation, a 45% nominal haircut on all bilateral and private external debt would be required to bring public debt below 85% of GDP over the forecast horizon, as assumed in the IMF’s February 2021 reform scenario. This is well above our original base case restructuring of 20% haircuts from last August and in line with our original worst case of 45%.
With the combined shock, an even more punitive nominal haircut of 70% on both domestic and external would be required to bring debt to 85% of GDP. Overall, we revise the nominal haircut assumption in our base case from 20% to 25% (which would stabilise public debt at 90% of GDP by 2027) and from 45% to 50% in our worst case (which would stabilise public debt at 85% of GDP). While public debt remains elevated in both scenarios even without the impact of potential shocks, this is sufficient to bring Tunisia’s external public debt indicators within IMF thresholds. When combined with a three-year maturity extension and coupon step-up of 1.5% to 6% in annual 1.5% increments, this reduces threshold breaches during the middle years as well (we assume a five-year maturity extension in the worst case).
In addition to a traditional DSA, we can also examine a potential restructuring from the balance of payments perspective. Reserves appear to be healthy in Tunisia, with net reserves roughly flat on the year at US$8.05bn (four months of imports). However, we have previously argued that the underlying external position is worse than it looks, with nearly the entire reserve increase seen from 2018-20 driven by short-term borrowing in the form of trade credits. The stock of trade credits reached US$5.5bn by the end of 2020, of which US$3.9bn was owed by Tunisia’s struggling SOEs, with regular anecdotes of shortages of key goods since last year suggesting that the SOEs may be having trouble rolling them over.
Although this data is now extremely stale (it should be updated in the BCT’s Annual Report this month), it shows Tunisia’s reserve buildup and subsequent stability has largely been a function of short-term borrowing. Tunisia’s net international investment position (NIIP) reached an eye-watering -158% of GDP in 2020, surpassed only by Mongolia and Jamaica in our sample of 43 emerging and frontier markets, of which -96% is “other investment” (ie external borrowing and trade credits). If Tunisia remains locked out of the market, it will be dependent on continued bilateral support to continue funding its large external funding needs, leaving it extremely vulnerable to the tightening global financial conditions.
Further, Tunisia’s current account deficit is projected by the IMF to widen to 10.1% of GDP in 2022 from 6.2% last year, driven largely by higher global food and fuel prices. Indeed, over the first four months of the year, it has already risen to 2.7% of full-year GDP versus 1.7% in the same period of 2021, with a US$2.2bn widening of the trade deficit on the back of a 30% yoy rise in imports and 25% rise in exports. Adding the projected US$4.6bn current account deficit and public external amortizations of US$2.8bn (per World Bank IDS estimates) results in gross external financing needs of US7.45bn (almost 17% of GDP) in 2022, c5pp above the 2010-19 average.
Stale data on Tunisia’s IIP and external amortisation profile makes it difficult to quantify the amount of relief that could be required to maintain reserves at sustainable levels in the event of a hard stop in external financing inflows but, without significant fiscal consolidation and large financial support from the IMF and other multilateral and bilateral partners, a balance of payments crisis is inevitable in Tunisia. The proximate trigger would likely be difficulties rolling over Tunisia’s trade credits, which will either trigger a large rise in import shortages (likely resulting in severe political instability) or a sharp drawdown of reserves (likely culminating in a debt crisis).
We conducted a similar exercise with Pakistan at the beginning of the month, and concluded that Pakistan does not have major solvency issues but is facing a severe liquidity challenge. Tunisia is in the opposite predicament, in that its liquidity constraints are not yet binding but it is not clear that its debt stock can be put on a sustainable path. Pakistan’s debt is projected to decline from just over 70% of GDP to less than 60% over the next five years and is projected to stabilise even in the combined macro shock, with the primary deficit in FY 21/22 just wide of the debt-stabilising level (2% of GDP).
Tunisia’s debt, on the other hand, is projected to rise from just under 80% of GDP to almost 100% over the next five years and to explode to 135% of GDP in the combined macro shock, with its primary deficit in 2022 projected to be c5.5pp wide of the debt-stabilising level (1.2% of GDP). Thus, even with an ambitious fiscal consolidation, Tunisia’s debt will be at a high risk of distress.
Therefore, while Pakistan is more likely than Tunisia to default in the next 6-12 months due to comparatively severe liquidity constraints (we think default is inevitable without the resumption of its IMF programme but can be avoided if it is put back on track), Tunisia’s probability of default is higher over a longer horizon (as very large and difficult reforms will be required just to stabilise its debt burden).
Further, Tunisia’s recovery values are likely to be lower than Pakistan’s (cUS$60 base case) due to the need for larger nominal haircuts, partially offset by less severe frontloaded cashflow relief. Conversely, recovery values are likely to be higher than the recently defaulted Sri Lanka (cUS$40 base case), which has more severe constraints in terms of both solvency and liquidity. This should result in a base case recovery value somewhere between the two (see below for our detailed calculations).
We update our scenario analysis, last updated in December, to reflect a greater probability of default (from 70% to 75%), higher exit yields (from 10% to 12-14%), and slightly more onerous restructuring scenarios (see above). We have three scenarios in total:
Scenario 1: Tunisia avoids a default through difficult reforms, which will likely require an IMF programme;
Scenario 2: Base case restructuring of 25% nominal haircuts, three-year maturity extension, and a coupon step-up from 1.5% to 6% in annual 1.5% increments; and
Scenario 3: Worst-case restructuring of 50% nominal haircuts and five-year maturity extension, and the same coupon step-up as scenario 2.
Although Tunisia likely needs an IMF programme to backstop its reform agenda and unlock funding in scenario 1, there is also a risk that the Fund does not think debt is sustainable and thus needs to be restructured to unlock IMF funding. If Tunisia proactively secures an IMF programme, it is equally likely that this results in scenario 1 (no default) or scenario 2 (credible reforms/restructuring) unfolding.
Conversely, if Tunisia defaults without proactively securing an IMF programme (eg 'hard default'), it could either result in scenario 2 (if it nonetheless is able to formulate a credible IMF-backed reform/restructuring plan after the fact) or scenario 3 (if economic conditions deteriorate sharply before the restructuring is finalised and/or reforms are not credible).
In the restructuring scenarios, we assume that Tunisia muddles through for the next 12 months before defaulting, since we think its liquidity constraints are not yet binding but that restructuring is inevitable without major reforms. Although we use this simplifying assumption for illustrative purposes, the timeline is highly uncertain and will depend on Tunisia’s ability to attract ad hoc bilateral and multilateral financing to paper over the cracks.
The table below provides the recovery value in each scenario and weights by probability to arrive at a target price for the TUNIS 5 ¾ 01/30/25s:
The upside scenario sees bond prices rising to US$86 based on a 12% yield if Tunisia is able to secure an IMF programme and avoid a restructuring. In our base case restructuring, the recovery value is between US$50-54 based on exit yields of 12-14%, and, in the worst-case restructuring, it falls to US$32-35 at 12-14% exit yields. Weighting by probability and holding the yield constant at 12% in scenario 1, we arrive at a target price of cUS$54 for the TUNIS 5 ¾ 01/30/25s at a 14% exit yield and US$57 at a 12% exit yield in scenarios 2 and 3, which is roughly in line with the current price of US$56.6.
For illustrative purposes, we also show the recovery value based on various exit yield and nominal haircut assumptions (holding the three-year maturity extension and coupon step-up constant). We find that, using our base case restructuring assumptions, the current price is consistent with an exit yield of 11%. Conversely, at a 14% exit yield, bonds are pricing nominal haircuts of just 12.5% with a three-year maturity extension and coupon step-up.
Bonds are approaching recovery value, but there is therefore still some downside risk and bonds are likely to drop further if the economic and/or political situation deteriorates further and markets begin to position for a greater likelihood of default (probably not to the high 30s as we see in Sri Lanka, but likely somewhere in the mid to high 40s).
That said, while we see little reason to be optimistic that an IMF programme is imminent or that the reform outlook will improve given the increasingly fraught political environment, bonds are now trading roughly in line with our base case recovery value at 12-14% exit yields, with potential upside if global risk sentiment improves and Tunisia’s yields drop in line with broader market improvements.
Overall, pricing risks are roughly symmetrical and the ongoing sell-off has bonds trading in line with recovery value. As such, we upgrade our recommendation on the TUNIS 5 ¾ 01/30/25s to Hold from Sell at US$56.6 (31.1% YTM) at the time of writing on 14 June on Bloomberg, based on a revised base case recovery value of US$50-54 and target price of US$55 (based on a US$54-57 range).
Tunisia inches closer to default, March 2022
Tunisia: IMF optimism may be premature, February 2022
Tunisia: Why the external position is worse than it looks, December 2021
Tunisia: Downgrade to Sell with reforms and IMF on backburner, September 2021