Closer to Shore

Market Review

The Fund (Class A Shares) was up 2.53% in August, outperforming its benchmark by 2.43%. Performance was driven significantly by Argentina and Sri Lanka, with further contributions from Uruguay (in local currency), India (Vedanta), Mexico (in local and hard currency), Suriname and China (corporates). Importantly, the Fund did very well again by not having Brazilian exposure—Brazil’s ongoing weakness was responsible for taking 25 bps off of the performance of the Fund’s benchmark. We’d note that the Fund’s local currency exposures continue to contribute to performance in months when overall emerging markets (EM) local currency performance has been lackluster. That is the case in August, with the J.P. Morgan Government Bond Index-Emerging Markets (GBI-EM) down 33 bps in August, but Mexico, Uruguay and Indonesia were all positive and top local currency performers for the Fund. This is consistent, in our view, with the Fund having uncorrelated positions that have either defensive or re-rating characteristics.

We see three interesting themes characterizing EM debt right now: a) the relatively greater opportunities in hard currency EM debt over local currency EM debt; b) the low margin for error in many EM fixed-income markets; and c) the even greater need for selectivity with many markets approaching pre-COVID-19 highs.

EM local currency has challenges, currently. We wrote more extensively on this last month. The summary explanation is that local currency yields are low in real terms. Policy rates are also especially low, with only three (major) countries—Russia, Mexico and Indonesia—having positive real policy rates. This signals the common sense view that nobody wants their currencies stronger, whatever “USD death” theory has to say. As we’ve noted before, monetary experimentation in the developed markets (DM) has not benefited EM local currency performance. EM currencies have tended to act as shock absorbers protecting the economic health of the EM countries, therefore requiring greater selectivity.

EM hard currency has supports, currently. The U.S. Federal Reserve (Fed) is now buying U.S. corporate investment grade and high yield bonds—we think that has created a ceiling on those spreads, essentially. That leaves EM debt as the only market-priced sovereign or corporate debt around, to be, admittedly, hyperbolic. Hard currency debt has done very well in the era of DM monetary experimentation, compared to local currency. This is because the key dynamic remains—low interest rates. This clearly anchors hard currency debt (depending on credit quality), while having intermittent effects on local currency debt whose interest rates are already low and whose currencies have underperformed in the quantitative easing (QE) era.

Finally, there’s low margin for error in the many EM countries with very low interest rates (local or hard), particularly with many markets at pre-COVID-19 highs. In a risk-on scenario, shouldn’t the U.S. curve steepen? Won’t low-spread, high-duration hard currency bonds suffer in that scenario? Won’t low-yield, high-duration bonds in local currency suffer in that scenario? In a risk-off scenario, shouldn’t credit spreads, especially long duration, suffer in that scenario? Shouldn’t EM local currency markets also sell off in that scenario? If you’re down 4%-8% on a long-duration exposure (let’s “back-of-envelope” assume either a 50 bps or 100 bps rise in spreads or yields, very few exposures have the carry to make it back in under a few years. Thus, our low duration, combined with positions with re-rating and defensive characteristics.

The Fund made significant changes in August, continuing the profit taking that characterized July—Angola was reduced and Jamaica and the Dominican Republic were closed. These were our three highest VaR positions. These investments worked very well, very quickly, but they are also now, by definition, more correlated with global risk given their much higher prices. VaR was a key lens for bonds that had substantial rallies, as was upside/downside, spread/yield and spread/yield vs. duration. The three positions that were reduced or closed were Jamaica (long duration), Angola and the Dominican Republic (long duration). The Fund used these proceeds to establish exposure to lower-duration South Africa and Eskom (the state-owned utility). The thinking on Eskom is similar to our thinking on our, so far, successful Pemex view. The view is not a positive one about Eskom on a stand-alone basis, but rather simply that the over 800 bps-handle spread on Eskom is converging to the roughly 400 bps-handle sovereign spread, as the sovereign increases credit support. We are also in the process of reducing positions in Argentina, as discussed in our various publications.

We end August with carry of 5.9%, duration of 4.0 and approximately 25% in local currency. Our largest exposures are Argentina (in hard currency), Indonesia (local), Mexico (MXN-hedged Pemex local) and Uruguay (local). Cash of 7.5% is only a reflection of the final day of the month happening to capture a lot of profit taking (described below), with resources not yet deployed.

Exposure Types and Significant Changes

The changes to our top positions are summarized below. Our largest positions in August: Argentina, Indonesia, Mexico, Uruguay and China.

• We increased our hard currency sovereign and quasi-sovereign exposure in South Africa. There are many macroeconomic concerns about the economy at the moment, but the central bank’s policy—especially its commitment not to waste the international reserves on foreign exchange (FX) interventions—is not one of them. We believe South Africa’s external debt is well managed and there appears to be little risk of default. In terms of our investment process, this improved the country’s technical and policy test scores. As regards adding quasi-sovereign exposure, the spread with the sovereign is quite wide, and we do not think this is fully justified under the circumstances. In terms of our investment process, this resulted in the improved technical test score.
• We also increased our hard currency corporate exposures in India, Israel, and Panama. The Indian corporate bond—which we bought at issue—has a very attractive valuation (initial allocation Bucket 1) in our view. The bond is shorter term and secured, with a favorable yield-to-maturity. In addition, we added the unsecured bonds, because we believe these would benefit from a transaction that is being financed by this new secured issue. Panama’s corporate exposure was a brand new deal, which was very attractively valued (a good technical test score). The company is an investment-grade utility, this new issue came cheap to already extant bonds, as well as to the sovereign (approximately 175 bps wide to the sovereign at issue) with some additional protections offered by security/collateral. Israel’s corporate bond was pricing some 60 bps wide of its largest shareholder, owner/operator of an already successful business in Israel, making this initial allocation Bucket 1 in our process (an improved technical test score). We derived additional comfort from the largest shareholder’s operating history in the region, the company’s solid, long-term contracts at favorable prices and expectations of free cash flow generation now that all major capex has been done. Bonds also came with collateral which we valued in excess of the total amount of debt outstanding post deal.
• Finally, our hard currency sovereign exposures in Sri Lanka and Argentina, as well as hard currency quasi-sovereign and corporate exposures in Argentina, also went up. These increases reflected positive price changes. In Argentina, the main driver was the government’s finalizing the sovereign debt restructuring deal with foreign creditors on August 28. In Sri Lanka, the market continues to internalize the overwhelmingly positive outcome of the recent presidential elections, which improves the country’s changes of getting an IMF deal.
• We reduced our hard currency sovereign exposure in Angola and Romania. There were two main reasons for trimming Angola’s exposure. First, the country’s bonds staged a massive rally after the initial COVID-19-related selloff, so valuations started to look a bit stretched. Second, the market got concerned about a delay in the IMF deal. We do not think that these concerns were fully justified, because the country wanted to upsize its IMF program, but these concerns affected the market sentiment. In terms of our investment process, these developments worsened Angola’s technical test score. As regards Romania, the reduction was technically a switch from the sovereign bond to higher-yielding and better-valued local bonds, which should get extra support from the E.U.’s COVID-19 emergency funds that can improve the country’s growth outlook and boost inflows to the region. We continue to keep an eye on the situation with pension increases, but for now this risk seems contained. In terms of our investment process, this improved the country’s technical test score.
• We also decided to fully exit from our hard currency sovereign and corporate exposure in Belarus. We started to reduce the position in July in the run up to the presidential elections, as we thought the market was underestimating the risk of protracted political impasse, betting on the “business as usual” scenario instead. The subsequent developments showed that our assessment was correct. The elections turned more troubled than expected. There is still no acceptable solution in sight and each day brings new reports about massive demonstrations and police brutality. The incumbent switched from anti-Russia provocations to pleas for help and this path might lead to sanctions. In terms of our investment process, this significantly worsened the country’s policy/politics test score.
• Finally, we reduced our hard currency sovereign exposures in Jamaica and the Dominican Republic. The bonds’ technical scores were getting more and more stretched after a strong rally. The fact that both countries had high VAR and relatively low spread/yield ratios (compared to other positions in the portfolio) also contributed to the worsening technical test scores. The economic test scores for both countries also suffered on the back of concerns about the second wave of the COVID-19 virus and its impact on tourism revenue.

Fund Performance

The VanEck Emerging Markets Bond Fund (Class A shares excluding sales charge) gained 2.53% in August compared to a gain of 0.10 % for the 50/50 J.P. Morgan Government Bond Index-Emerging Markets Global Diversified (GBI-EM) local currency and the J.P. Morgan Emerging Markets Bond Index (EMBI) hard-currency index.

Turning to the market’s performance, GBI-EM’s biggest winners were Indonesia, Mexico, and South Africa. Its biggest losers were Brazil, Turkey and Hungary. The EMBI’s biggest winners were Sri Lanka, Turkey and Egypt. Its losers were Saudi Arabia, Philippines and Qatar.

^†Monthly returns are not annualized.

Expenses: Class A: Gross 2.69%; Net 1.26%. Expenses are capped contractually until 05/01/21 at 1.25% for Class A. Caps exclude acquired fund fees and expenses, interest expense, trading expenses, dividends and interest payments on securities sold short, taxes and extraordinary expenses. Please note that, generally, unconstrained bond funds may have higher fees than core bond funds due to the specialized nature of their strategies.

The tables above present past performance which is no guarantee of future results and which may be lower or higher than current performance. Returns reflect temporary contractual fee waivers and/or expense reimbursements. Had the Fund incurred all expenses and fees, investment returns would have been reduced. Investment returns and Fund share values will fluctuate so that investors’ shares, when redeemed, may be worth more or less than their original cost. Fund returns assume that dividends and capital gains distributions have been reinvested in the Fund at Net Asset Value (NAV). An index’s performance is not illustrative of the Fund’s performance. Certain indices may take into account withholding taxes. Index returns assume that dividends of the index constituents in the index have been reinvested.