After rebounding sharply from mid-July to end-August amid a broader market rally and staff-level agreement with the IMF, Sri Lanka’s eurobonds have plummeted back to the mid-July lows on the back of rising risk aversion towards EM and the realisation that Sri Lanka’s debt restructuring could be a long and messy process. Indeed, the SRILAN 7.55 03/28/2030s closed at US$25.3 on 4 October, just above the mid-July low of US$24.1.
We last conducted a recovery analysis in March and updated the analysis in July to account for higher exit yields and a deteriorating macro outlook, calculating a recovery value of US$33-40 for the ‘30s at a 12-14% exit yield in our more bullish scenario and US$24-29 in our more bearish scenario (which we acknowledged, due to worsening economic conditions since our March analysis, was likely a more appropriate 'base case').
We have since gained some clarity on Sri Lanka’s reform plans from its interim budget, the press release from the IMF’s staff-level agreement and the recent investor presentation by the Central Bank of Sri Lanka (CBSL) and Ministry of Finance. While details are still scant, we know that Sri Lanka is targeting a 2.3% of GDP primary surplus by 2025, a relatively hefty 6.3pp consolidation compared with the projected 4% of GDP deficit in 2022 and 8pp from the 5.7% deficit achieved in 2021.
For context, an adjustment of this magnitude would be in the top decile of all IMF-supported programs for low-income countries since 1990 (although Sri Lanka is a middle-income country). Further, a 2.3% of GDP primary surplus compares to an average deficit of 1% of GDP from 2010-19 and is above the maximum 0.6% surplus that Sri Lanka has achieved since 1990. As such, the adjustment appears very ambitious even with low-hanging fruit on the revenue front, and will be difficult to deliver.
The government also laid out its restructuring timeline. It hopes to obtain “financing assurances” from its official creditor committee by mid-November and to launch parallel discussions with the private creditor committee to make way for Board approval of its IMF programme by mid-December and the completion of creditor negotiations and implementation of the restructuring agreements by mid-2023, at the latest.
This, too, seems very ambitious. The official creditor committee has yet to be formed, a process that took 6.5 months after a staff-level agreement with the IMF was reached in Zambia, plus another 2.5 months before financing assurances were granted and Board approval was secured. This suggests mid-2023 could be a more realistic target for financing assurances and Board approval, although there may be more urgency from official creditors.
As in Zambia, the key creditor will be China, which holds c50% of bilateral debt and c10% of public external debt in Sri Lanka (versus c75% and c40%, respectively in Zambia). India also comprises 12% of bilateral debt and 4% of public external debt, meaning non-Paris Club creditors comprise two-thirds of the bilateral debt stock in Sri Lanka. This could delay the process if either creditor refuses to come to the table or transparently share information, with China and India effectively dictating the pace of progress.
Beyond the headline macro targets and restructuring timeline, there is little clarity on how Sri Lanka’s IMF-backed reform and restructuring programme will be designed. And, unfortunately, the assumptions underlying the IMF’s debt sustainability analysis (DSA) will not be made public until Board approval is obtained, and can only be shared with the financial advisors of the private creditor committee on a non-disclosure basis until that point is reached.
Given the high degree of uncertainty associated with Sri Lanka’s restructuring (see related reading), we refrain from a holistic update to our DSA and recovery analysis from March. Instead, we take a different approach by outlining a 'breakeven' scenario for the Sri Lanka ‘30s and assessing whether the actual restructuring is likely to be more or less favourable for bondholders, allowing a more agile response to the recent volatility. We augment this with lessons drawn from Zambia’s ongoing restructuring.
We start with a relatively punitive restructuring proposal to be conservative, pencilling in a 50% nominal haircut, five-year maturity extension and 50% coupon reduction. Assuming the restructuring is completed in mid-2023 as planned and interest is capitalised in the meantime, the ‘30s would break even at a 15.8% exit yield, which implies significant upside (as both the terms of the restructuring and exit yield are very conservative).
In Zambia, the IMF/World Bank have called for debt relief on net present value (NPV) terms of 35-45% at a 5% discount rate. As a “Market Access” country, Sri Lanka is subject to a different DSA framework than “Low Income” countries like Zambia. That said, both countries have a similar debt stock (a projected 123% of GDP in Zambia in 2022 versus 122% in Sri Lanka in mid-2022 and a projected 137% by year-end) and composition (external debt is projected to be 57% of the stock in Zambia in 2022 compared with 58% in Sri Lanka in mid-2022), meaning Zambia’s case could still serve as a rough guide.
The illustrative restructuring scenario above corresponds to NPV relief of 46-58%, depending on the definition used (58% at a 12% discount rate for the value of both the new and existing bonds, 52% at a 5% discount rate for both the new and existing bonds, and 46% at a 5% discount rate for the new bond and face value for the existing bond). Even under the more conservative 46% calculation, this exceeds the 35-45% requested in Zambia, a country that is likely viewed as having a lower debt carrying capacity than Sri Lanka and therefore likely subject to more stringent sustainability thresholds.
Using this quick and dirty methodology, Sri Lanka’s eurobonds have likely dropped below recovery value after the recent sell-off. To be clear, the restructuring process is likely to be long and messy. Even if timely financing assurances are provided by key bilateral creditors like China and India, there will likely be lengthy debate between Sri Lanka and its private creditors on how to treat SLDBs and domestic debt, potentially dragging the process out even further (with the decision to exclude domestic debt from the restructuring Zambia setting an interesting precedent that will likely be hotly debated in Sri Lanka).
That said, we think bonds have now fallen enough to incentivise investors to hold on for what is likely to be a bumpy ride (for which they will presumably be compensated via accrued interest). As such, we upgrade our recommendation from Hold to Buy across the curve (excluding the ‘22s, which have gone back to trading at a c10-20% premium after temporarily closing the gap - we think that is excessive notwithstanding the pending legal action, which could slightly bolster recovery values but whose prospects are uncertain and which the Sri Lankan government has launched a motion to dismiss).
We will consider reviewing our recommendation if the ‘30s begin to rise toward the upper end of our projected US$24-40 recovery value range at 12-14% exit yields or if the premium on the ‘30s again collapses below 5%.
Sri Lanka: Upgrade ‘22s to Hold on collapsed premium, August 2022
Sri Lanka: IMF talks mark beginning of long road, April 2022
Sri Lanka announced external debt restructuring, April 2022
Sri Lanka finally turns to the IMF, March 2022