Structural Duality in Emerging Markets: HS-to-Sector Mapping and Partner-Role Diagnostics for the Kyrgyz Republic’s Trade (2019–2024)

1. Introduction

Small open economies are especially exposed to external trade shocks because imports and exports directly affect growth, fiscal and external balances [1,2].

Recent literature and policy reports emphasize that supply-chain re-routing and sanctions-related adjustments can materially change trade geography and recorded partner shares [3,4].

Export concentration in a single commodity increases vulnerability by tying foreign-exchange earnings to a narrow channel [8,9].

Official external-sector publications for the Kyrgyz Republic also point to high import concentration from non-EAEU suppliers, with China as a dominant source [5,6].

However, headline totals do not explain which partner–sector links generate the deficit and which links generate the surplus [7].

Research question. Which partner–sector relationships form the Kyrgyz Republic’s two-pole trade structure in 2019–2024, and how concentrated is the “surplus donor” channel relative to the “deficit engine”?

What is new in this paper. We combine (1) a deterministic HS-to-sector mapping, (2) interpretable diagnostics (net trade, import coverage, market share), and (3) a role-based partner labeling (Export Driver / Import Dependency / Mixed Trade) to produce a compact, reproducible duality profile.

Contributions.

• A transparent HS-chapter aggregation into macro-sectors enabling consistent partner comparisons.
• A partner-role diagnostic that makes “deficit engines” and “surplus donors” visible in one figure set.
• Evidence that the surplus side is narrowly concentrated in gold exports to the UK and Switzerland, while the deficit side is dominated by China-linked manufactured imports.

2. Related Work (Short)

Export diversification and concentration are widely studied as drivers of macro vulnerability and growth patterns [8,9].

Trade statistics in small economies can be distorted by re-exports and mirror-statistics gaps, which motivates triangulation and careful partner attribution [3,14].

3. Data and Methods

3.1. Data Acquisition (KaggleHub) and Validation Sources

The base dataset was imported programmatically via kagglehub:

path = kagglehub.dataset_download("aitenir/kyrgyz-republic-export-and-import-2019-2024")

The dataset contains partner-country trade by year with fields: country_name, product_code, product_name, import_som, import_dol, export_som, export_dol, year.

We interpret product_code as an HS-chapter-style identifier and build macro-sectors through a deterministic chapter-range mapping (Section 3.2).

For external context and plausibility checks we refer to NSC/NBKR publications and UN Comtrade/WITS summaries [5,6,7].

Code Availability: https://colab.research.google.com/drive/1ga8ULMZ6kxRulTBIuo896YrwlcUA5Fcn?usp=sharing.

3.2. Feature Engineering and Sector Mapping

We add the following derived columns:

• Sector: macro-sector label based on product_code ranges (HS chapter groups).
• Turnover: $Turnover=Impor{t}_{dol}+Expor{t}_{dol}$.
• Net_Trade: $Net\_Trade=Expor{t}_{dol}-Impor{t}_{dol}$.
• Market_Share: partner-sector turnover share in total turnover.
• Trade_Role: categorical role based on export share in turnover:

  -

  Export Driver if $Export/Turnover>0.8$,

  -

  Import Dependency if $Export/Turnover<0.2$,

  -

  Mixed Trade otherwise.

This mapping enables interpretable partner comparisons and makes donor/engine asymmetry measurable.

3.3. Normalization and ML-Ready Pipeline (Optional Layer)

Trade values are heavy-tailed, so we use Yeo–Johnson power transformation before clustering [10,11]. Where clustering is applied, we use scikit-learn K-Means and interpret clusters as partner categories [12,13].

4. Results: Visual Evidence of Structural Duality

4.1. How to Read the Figures (Visual Legend)

• Color shows net trade sign: red/orange = deficit ($M>X$), blue = surplus ($X>M$).
• Size shows weight: larger shapes/bubbles = larger turnover.
• Efficiency Matrix diagonal: above ${45}^{\circ }$ means $X>M$ (profit zone), below means $X<M$ (cost zone).
• Sunburst roles: Import Dependency (deficit-dominant), Export Driver (surplus-dominant), Mixed Trade (balanced).

4.2. Duality Map (Four Coordinated Plots)

The next four plots provide coordinated evidence: (a) where deficit mass sits (treemap), (b) how roles split by sector and partner (sunburst), (c) which observations lie above vs below the diagonal (efficiency), and (d) which partners donate surplus vs generate deficit (net balance ranking).

Figure 1. Two poles in close-up. Deficit is dominated by China-linked manufactured imports; surplus is dominated by UK/Switzerland via gold exports.

Figure 1. Two poles in close-up. Deficit is dominated by China-linked manufactured imports; surplus is dominated by UK/Switzerland via gold exports.

Figure 2. Structural duality in Kyrgyz trade (2019–2024), Part II. Efficiency geometry and partner contributions to deficit/surplus.

Figure 2. Structural duality in Kyrgyz trade (2019–2024), Part II. Efficiency geometry and partner contributions to deficit/surplus.

4.3. Partner Categories: Turnover vs Net Balance

Figure 3 shows that the aggregate trade balance is shaped by a few outliers. Turnover (log scale) measures partner importance, while the vertical axis shows whether the partner behaves as a deficit source or a surplus donor.

4.4. Main Empirical Message (One Idea)

Across all views, one pattern is stable: the Kyrgyz Republic’s surplus side is concentrated in Gold & Precious Stones exports, and the largest positive net trade is generated by the United Kingdom and Switzerland. In parallel, the deficit side is concentrated in manufactured imports dominated by China. This is the operational definition of structural duality: one concentrated surplus-donor channel partially offsets one concentrated deficit engine.

5. Quantitative Evidence from Top-5 and CSV Tables

5.1. Extremes: Top Profit Zone vs Top Cost Zone (2019–2024 Aggregate)

Table 1 isolates the most extreme partner–sector links. UK and Switzerland in Gold & Precious Stones have coverage ratios in the thousands, while the largest deficits are dominated by China in manufactured sectors (author calculations).

5.2. CSV Time-Series Excerpt: Donors vs DEFICIT Engines

Table 2 summarizes donor links (gold to UK/Switzerland) and the largest deficit engines (manufactured imports dominated by China).

Main takeaway. The surplus donor channel is strong but narrow: UK (+$2.92bn) and Switzerland (+$1.22bn) dominate positive net trade via gold (Table 2). The deficit engine is large and concentrated: China-driven manufactured imports dominate the negative balance with very low coverage (Table 1).

6. Limitations (Short)

Some flows appear under undetermined_country, and partner-attribution issues can shift measured donor shares. Mirror-statistics gaps and re-export dynamics are documented for the region, so results should be interpreted as concentration diagnostics rather than exact accounting identities [3,14].

7. Conclusion (Strong and Compact)

This paper shows that Kyrgyz goods trade in 2019–2024 is best described as a two-pole system.

1) Deficit engine (China-centered manufactures). The largest and most persistent deficits are concentrated in China-linked manufactured imports, especially Machinery & Electronics and Transport (Auto). These high-turnover links exhibit near-zero import coverage, meaning exports do not scale with the import requirement.

2) Surplus donors (UK and Switzerland via gold). The surplus side is disproportionately generated by Gold & Precious Stones exports, with the United Kingdom and Switzerland dominating positive net trade. Coverage ratios above 4,000–6,000× indicate that these partners act as surplus donors—they provide a concentrated foreign-exchange inflow channel that partially offsets the deficit engine.

3) Method value for monitoring. The main methodological contribution is a lightweight, reproducible pipeline (HS-to-sector mapping + partner-role labeling + coverage diagnostics) that can be rerun on new monthly/annual data. This enables routine monitoring of: (i) whether the donor channel becomes less concentrated (more partners/sectors generating surplus), and (ii) whether coverage improves in the largest deficit sectors.