What Happens When Global Indexes Start Repricing Nigeria? Three Scenarios for Yields, Liquidity and the Naira

Executive Summary

A benchmark index changes who is obliged to hold a bond, not what the bond is worth. J.P. Morgan's decision to include Nigerian naira government bonds in its new GBI-EM Edge index, at a reported weight of 7.40 percent, therefore matters less as an announcement than as the first link in a chain: benchmark weight, portfolio adjustment, flows, market liquidity, yields, and finally the cost of borrowing and the currency risk the naira still carries. Each link can strengthen or break, and the chain has no fixed outcome.

This article models that chain over a ten-year horizon and sets out three labelled hypothetical scenarios: the weight is absorbed and held, flows arrive and then leave, and liquidity thins before flows arrive. It reads the closest precedent, India's phased entry into J.P. Morgan's flagship index, for what passive-flow estimates overstated, and it separates what has been observed since 14 September 2026 from what is assumption. As of 8 October 2026, no dated series shows Nigerian yields or foreign flows repricing because of the index.

Introduction: The Announcement Is Not the Repricing

The conventional reading of index inclusion is a flow story: a large index adds a market, trackers buy, yields fall. That is directionally plausible and analytically incomplete. It treats the weight as a purchase order, when a weight is only a rule about how much of a market a benchmark-following portfolio should own relative to the benchmark. Whether that rule becomes buying depends on how much money follows the benchmark, how much of the market foreigners can reach, and what else is happening to rates and the currency while the rule is applied.

The question sits downstream of two other arguments the Upside Journal has already made. The investor-confidence analysis of Nigeria's capital market argues that access was never the binding constraint, and the reading of the announcement itself covers what J.P. Morgan decided and why benchmark visibility matters. This article takes both as given and asks only what happens next, in sequence, under stated assumptions.

It is written as a structural projection, not a forecast. Every scenario is a hypothetical built from labelled assumptions, and every observed fact is attributed to the source that reported it. Where the evidence runs out, the text says so.

What Has Been Reported, and What Has Not

The facts available as of 8 October 2026 are narrow, and they are reported rather than read from the index provider's own documentation, which was not located. Nairametrics, citing J.P. Morgan, reported on 14 September that Nigeria carries a 7.40 percent weight in the new GBI-EM Edge index, below the 8 percent country cap, comprising US$17.47 billion of bonds across 16 instruments with an average yield to maturity of 17.1 percent against a benchmark average of 10.39 percent.

A Reuters report carried by Kitco said J.P. Morgan planned to launch the index by the end of September, citing an unnamed source. No effective date was found in the sources opened. The index was described as tracking about US$328 billion across 425 bonds in 26 markets and 24 currencies, and Daba Finance reported Africa at about 44.5 percent of the index, with Egypt and Morocco at the 8 percent cap, Nigeria at 7.4 percent and Kenya at 6.91 percent. Eligible bonds need at least US$250 million outstanding and more than 2.5 years to maturity, as reported.

Two points need to stay separate. First, the US$328 billion is the size of the bond universe the index covers, not the amount of money that tracks it. No figure for assets benchmarked to GBI-EM Edge was found, so no inflow estimate is made here. Second, GBI-EM Edge is a new index. It is not Nigeria's return to the flagship GBI-EM Global Diversified, which Nigeria left in September 2015, so the money that follows each index is a different pool.

What the first weeks show

The observable record since the announcement is thin. On the day of the announcement, 14 September, Nigeria's Debt Management Office offered N1 trillion across a new 16.80 percent September 2036 bond and a reopening of the 15.45 percent June 2038 bond. Subscriptions totalled N1.49 trillion, with marginal rates of 16.79 percent and 16.85 percent, as BusinessDay's report, carried by Aboki Forex, set out. That auction is a single observation priced on the day of the news, and it cannot show a before-and-after effect.

The same report quoted a fixed-income trading head at CardinalStone saying there were no foreign inflows into the local bond market yet and that any effect might be lagged, while a ratings executive at Agusto and Company expected inclusion to lower yields and, with them, funding costs. The official naira rate closed at N1,329.15 per US dollar on 15 September, 0.2 percent weaker on the day. Those are market opinions and one day's price, not a repricing series.

The record therefore supports one statement: as of the sources opened, foreign positioning had not visibly changed. Everything after that statement is a model.

The Causal Chain: Six Links Between a Weight and a Yield

The chain has six links. Each is a mechanism with its own conditions, and each can be tested against data as it arrives. The labelled assumptions below are used throughout the scenarios that follow.

Link one: from benchmark weight to a required position

A benchmark weight is the share of an index a market represents. A fund that tracks the index passively holds each market at roughly its weight. A fund that is benchmarked but actively managed treats the weight as a reference point and may hold more or less. The first link therefore splits investors into three groups: passive trackers that must buy, benchmarked active managers that may now consider Nigerian bonds because they would otherwise run a visible gap to the index, and unbenchmarked investors for whom the index changes nothing.

The economic consequence is that the index creates demand only in proportion to assets that follow it. Index size is not that quantity. Assumption A1 is that the pool of money benchmarked to this specific index is modest at launch relative to the flagship index, because the index is new, and that it grows only if allocators adopt it as a reference.

Link two: from required position to portfolio adjustment

Funds do not buy a market in one day. They adjust in phases, constrained by liquidity, by tracking-error limits and by how much of a market they are allowed to own. India offers the clearest recent illustration. J.P. Morgan announced in September 2023 that Indian bonds would enter its flagship index from 28 June 2024, rising to a 10 percent weight by 31 March 2025 at one percentage point a month, according to a Business Standard analysis. Phasing is a design choice that spreads the purchase, and it also lets the market anticipate it. Assumption A2 is that any Nigerian adjustment is gradual and partly anticipated, so that some pricing occurs before the formal weight is applied.

That anticipation has a cost to anyone reading yields as a clean signal. If the market prices a known event early, the yield change at the effective date understates the total effect, and the effect before it cannot be attributed to the index with confidence.

Link three: from adjustment to realised flows

Passive estimates are upper bounds, not forecasts. Business Standard reported that analysts had projected passive inflows of US$20 billion to US$25 billion into India during the inclusion window, with bullish estimates up to US$30 billion including active repositioning. Foreign purchases of India's fully accessible route bonds reached about 1.09 trillion rupees, roughly US$14 billion, which the same report described as 40 to 45 percent below the passive estimates. Flows were also front-loaded: net inflows of about 92,302 crore rupees arrived between September 2023 and June 2024, before formal inclusion began.

Two lessons transfer to Nigeria without transferring any number. Realised flow can fall well short of a passive estimate, and it can arrive before the date the index names. Assumption A3 is that the timing and size of Nigerian flows depend more on global rate conditions and the naira outlook than on the index mechanics.

Link four: from flows to market liquidity

Flows change liquidity in two directions. Foreign buyers add depth when they trade actively, and they can remove it when they hold to maturity or when they exit together. A market whose price is set by a small number of large holders has depth that looks adequate until those holders move. The liquidity question for Nigeria is whether an index-driven investor base trades the 16 eligible instruments or simply holds them.

The economic consequence is that liquidity decides whether a yield move survives. A yield that falls on inflows and rises sharply on a single exit is not repricing, it is volatility. Assumption A4 is that the 16 eligible bonds concentrate trading in a short list of benchmark-eligible issues, which can widen the gap between index-eligible and other maturities.

Link five: from liquidity to yield

The reported 17.1 percent average yield on Nigeria's index bonds sits against a benchmark average of 10.39 percent. That gap is the reason a benchmark-following investor would look at the market at all, and it is also the measure of what the market is paid to carry: sovereign risk reflected in a B- rating, inflation, and currency depreciation risk. A yield premium narrows only if investors accept less compensation for those risks, which requires evidence that the risks themselves have changed.

The India case shows the same mechanism running in reverse. A market participant quoted by Business Standard attributed outflows in April to June 2025 to a yield spread over US rates that narrowed below 200 basis points. Cumulative outflows exceeded 31,000 crore rupees, about US$3.7 billion by the report's conversion, and the report attributed the selling mainly to profit-taking and rupee volatility. Yield compression drew in capital, and compression below a threshold sent it back out.

Link six: from yield to borrowing cost and currency risk

Government borrowing costs follow yields, but only for the portion of debt that is issued or refinanced at the new level. A lower yield on newly issued bonds does not change the coupon on bonds already outstanding. The fiscal saving from a given yield decline therefore accumulates slowly as old debt rolls over. The currency effect runs through a different channel: foreign holders of naira bonds carry exchange-rate risk, and a flow that arrives to buy bonds must first buy naira.

That risk is not hypothetical. Nairametrics, citing J.P. Morgan, reported naira returns against the dollar of negative 48.7 percent in 2023, negative 41.9 percent in 2024, positive 6.7 percent in 2025 and positive 8.1 percent for 2026. A yield of 17.1 percent has to be read against that record, because a foreign holder's return is the yield minus depreciation. The yield premium and the currency risk are the same fact seen from two sides.

The verdict on the chain: an index can start the sequence, but each later link is set by liquidity, rates and the currency, none of which the index controls.

The Nigerian Precedent: Eligibility Has Been Lost Before

This is not Nigeria's first time inside a J.P. Morgan benchmark. Nigeria was once a member of the flagship GBI-EM index and was removed in September 2015, when, as Intellinews reported on 9 September 2015, Nigerian authorities expressed strong disagreement with the decision to expel the country. The reasons for the removal were not stated in the part of that report that could be read, and no inference about motive is drawn here.

The relevance is structural. Benchmark membership is conditional, and the conditions are set by the index provider's view of accessibility, not by the issuer. A weight that was granted can be withdrawn, and the investors who followed it must then sell. Assumption A5 is that the durability of Nigeria's place in GBI-EM Edge depends on the same practical conditions that govern every other market in it, which makes the market's operating conditions a continuing variable and not a settled one.

That has a direct consequence for the scenarios. Any path that relies on sustained passive holding should carry the risk that the index treatment changes, and the effect of a later exit would be larger the more of the market had been built around the weight. The verdict: membership earned by market conditions is held by market conditions.

Scenarios: Three Hypothetical Paths Over Ten Years

The scenarios below are hypotheticals constructed from the assumptions above. They are not forecasts, none carries a probability, and each is defined by which link in the chain dominates. The first five years are the period in which the evidence would separate them. The second five years describe what each path implies if it persists.

Scenario one (hypothetical): the weight is absorbed and held

In this path, benchmarked investors add Nigerian bonds gradually, and most keep them. The assumptions are that global rate conditions stay permissive, that the naira remains stable enough for foreign holders to keep the carry, and that the market is deep enough to absorb purchases without disorderly moves. Foreign ownership of eligible bonds rises and stabilises, and the investor base becomes more diversified.

To keep scale honest without making a flow forecast, consider ownership shares of the eligible bonds. Applied to the reported US$17.47 billion of index-eligible bonds, foreign ownership of 5 percent, 10 percent and 20 percent would equal about US$0.87 billion, US$1.75 billion and US$3.49 billion. These are derived arithmetic on a reported stock, not estimates of what will occur. For comparison, Business Standard reported that foreign investors held about 6.8 percent of eligible Indian issuance after that process, a market with a different size, rating and currency history.

The consequence is that borrowing costs fall only to the extent that new foreign demand persists through issuance cycles. Savings accrue as debt is refinanced, so the fiscal effect appears over years, not quarters. The path rewards patience and is the least dramatic of the three, which is also why it is the hardest to attribute to the index.

Scenario two (hypothetical): flows arrive and then leave

In this path, early buying compresses yields, the premium over the benchmark narrows, and the same narrowing removes the reason to hold. The assumptions are that global rates rise or the naira weakens, and that a large share of the new investor base is return-sensitive. India's April to June 2025 outflows illustrate the mechanism, not the magnitude.

The consequence is a cycle: the cost of capital falls, then rises, and the market may end with a larger average foreign share than before, and a more volatile one. In this path the index has not failed. It has delivered what it is designed to deliver, which is access for benchmarked capital, and has exposed the market to the conditions of benchmarked capital.

The verdict: where capital arrives because of a spread, it leaves when the spread closes.

Scenario three (hypothetical): liquidity thins before flows arrive

This path starts from a different place. Foreign managers are slow to adjust, the new index attracts a small tracked pool, and dealers who make markets in Nigerian bonds see little additional volume. The assumptions are that the index is adopted slowly, that settlement and repatriation frictions deter active managers, and that the eligibility rules push trading into a short list of bonds, leaving off-index maturities less liquid.

The consequence is that the visible effect of inclusion is on market structure, not on yields. Eligible bonds trade at tighter spreads than ineligible ones, issuance concentrates in eligible maturities, and the benefits of benchmark status go to a narrow part of the curve. The CardinalStone comment that any effect might be lagged is consistent with this path, and also with the first, which is why a single comment cannot separate them.

The verdict: a market can be on the map and still be too thin to trade.

Which Indicators Would Separate the Paths

The scenarios are distinguishable by data that will become available over the next four to eight quarters. The test is not whether yields fall. It is which link is moving first and whether the movement persists once global rates and the naira are accounted for.

What should be tracked, and in what order?

Five series do most of the work. Foreign holdings of Nigerian government bonds, from central-bank and custodian data, show whether positions are being built and held. Yields on index-eligible bonds against comparable off-index maturities show whether the benefit is concentrated in the benchmark list. Auction subscription and bid-to-cover ratios show domestic and foreign depth at the primary stage. Bid-offer spreads and secondary-market turnover show whether liquidity is deepening or merely moving. The official and parallel naira rates, together with external reserves, show whether the currency is absorbing flows without strain.

The BusinessDay report cited above put reserves at US$54.61 billion on 14 September 2026, up from US$41.84 billion a year earlier, which is the best-supported currency fact in the material reviewed.

A repricing claim needs a dated before-and-after series with global rates held in view. A market participant noting that yields fell after an announcement is a statement about timing and not about cause.

The signal that rules out an interpretation

Each scenario fails a specific test. If foreign holdings rise while eligible-bond yields stay flat against off-index maturities and spreads stay wide, scenario three is the better fit. If holdings rise and yields compress, then holdings fall within a year while the premium narrows, scenario two is. If holdings rise and stay while the premium narrows slowly and refinancing costs decline, scenario one is. None of these conclusions is available yet.

The verdict: the useful measure is not the first yield move but whether holdings, liquidity and the currency move in the same direction together.

Where the Index Stops: Boundaries of the Argument

Three limits apply. First, this article does not estimate inflows. The magnitude of money benchmarked to GBI-EM Edge is unknown from the sources opened, and converting index size into an inflow figure would substitute an unobserved number for an observed one. Second, it does not assess whether inclusion guarantees capital. That question belongs to the Contrarian analysis of the index as a gatekeeper, and the answer here is only that eligibility does not set the amount.

Third, the article does not assess Nigeria's fundamentals. Sovereign rating, inflation and fiscal policy affect every link, but arguments about them belong to the confidence analysis. The role of the index is narrow: it changes who is obliged or encouraged to look. A separate, reported classification change at FTSE Russell, which BusinessDay said moved Nigeria to frontier status from 21 September 2026, is a second channel and not part of this chain. The Upside Journal's First HoldCo piece on FTSE frontier inclusion shows how that channel behaves at company level, where concentrated ownership can limit how much index-tracking money is able to hold.

Two further limits matter for the reader. The reported yields and weights are as of the announcement and may be revised when the index goes live. And a ten-year horizon contains events no scenario here models, including a change in global rate regimes, a change in Nigeria's exchange-rate arrangements or the creation of competing benchmarks.

Why This Matters for Treasury Teams and Capital Allocators

For corporate treasurers and industrial borrowers in Nigeria, the sovereign yield curve is the base on which private borrowing is priced. The capital-structure problem in African industrialisation turns largely on how expensive long-dated local currency funding is. A lower, more stable sovereign curve would help, and a curve that lurches with foreign flows would not. The distinction between scenario one and scenario two is therefore a distinction between a falling cost of capital and a more volatile one.

For allocators, the scenarios turn the index into a monitoring problem. A benchmarked manager's decision is shaped by tracking error, liquidity and the currency, in that order of operational importance. For an investor assessing where African capital markets are heading, the sequence of effects matters more than the headline, and the broader thesis in the Upside Journal's account of the African capital builders depends on whether the market that benchmarks open is deep enough to support the institutions that build on it.

For Nigerian policymakers, the lesson from the Indian precedent is that benchmark entry is an opportunity to lengthen the maturity profile of debt, widen the investor base and build secondary-market depth while conditions are favourable. Whether the opportunity is used is a matter of debt management and market structure, not of the index.

What would change this reading?

Four developments would. A published effective date and tracked-asset figure for GBI-EM Edge would replace the unknown in link one. A dated series of foreign holdings would show which scenario is unfolding. A change in naira policy or reserves would alter link six. And a credible demonstration that off-index bonds reprice together with eligible ones would weaken scenario three.

Key Takeaways

• An index weight is a rule about holdings, not a purchase order, so inclusion creates demand only in proportion to the assets that follow the index.
• The reported figures, a 7.40 percent weight, US$17.47 billion across 16 bonds and a 17.1 percent average yield, describe eligibility and carry, not flows.
• India's phased entry shows passive-flow estimates can overstate realised foreign purchases, which were reported 40 to 45 percent below the passive estimates.
• Capital that arrives because of a yield spread can leave when the spread closes, so lower yields and lasting liquidity are different outcomes.
• Government borrowing costs fall only as debt is refinanced, so any fiscal benefit from inclusion appears over years and is hard to attribute to the index.

Conclusion

The chain from index weight to borrowing cost has six links, and the announcement establishes only the first. What follows depends on how much money follows the benchmark, how deep the market is when that money moves, and how the naira behaves while it does.

The three scenarios are not predictions. They are a way to read the data that will arrive, and each can be confirmed or ruled out by the same short list of series. As of 8 October 2026, those series do not yet exist. The discipline this requires is to hold the question open: an index changes who must hold Nigerian bonds, not what they are worth.