Empirical associations between specific inventory practices and profitability have been documented mainly in large companies or urban settings; there is little evidence on micro and small enterprise (MSE) ecosystems in the Peruvian Amazon, where company size, informality, and logistical constraints may alter the magnitude of the effects (e.g., variable delivery times, higher risk of spoilage). In 2023, the Amazon region accounted for around 96% of micro-enterprises, with a predominance of commercial activities, conditions that amplify the operational impact of replenishment and record-keeping decisions ( Ministry of Production, 2024).

Why Bagua, Peru? Bagua is a representative province of the Amazon where the heterogeneity of the road surface and connectivity deficiencies complicate supplier delivery times and replenishment frequency, plausibly altering optimal targets (R, Q) and safety stock policies relative to coastal or metropolitan references. This context offers external validity for similar secondary cities, while providing policy-relevant insights ( Suárez et al., 2024).

The association between inventory control and measurement practices was quantified, emphasising accurate record keeping, safety stock policies and valuation methods consistent with IFRS/IAS 2 (i.e. FIFO and weighted average; LIFO is not permitted under IFRS) and profitability indicators (ROA, gross margin, ROE) in micro and small enterprises located in Bagua. By focusing on an underserved geographical area and aligning accounting treatment with IFRS, evidence was provided on how specific and applicable inventory control practices are linked to financial performance in constrained environments ( KPMG, 2021).

Inventory control refers to coordinated decisions about the timing and quantity of replenishment, commonly represented by continuous review policies (R, Q), which determine stock-out costs, carrying costs, and service levels; basic profitability channels operate through days of inventory outstanding (DIO), cash conversion cycle (CCC), and inventory profitability ratios such as GMROI; safety stock policies mitigate delivery time uncertainty but must balance capital tie-up and service performance ( Kotecha and del Rio, 2025).

Current studies have revealed a consistent positive relationship between CI practices and the profitability of micro and small enterprises, especially those where management skills and digital visibility remain scarce ( Li and Mizuno, 2022; Panigrahi A, et al., 2024). In terms of replenishment decisions, EOQ-type models and their extensions remain viable, low-cost solutions for companies with limited analytical capabilities, as they reduce ordering and storage costs ( Condeixa et al., 2020; Çalışkan, 2021). However, demand in small retailers is often intermittent or irregular, favouring Croston-type methods over naive smoothing to avoid systematic biases and service losses ( Syntetos et al., 2005; Erjiang et al., 2022).

Accurate record keeping and IFRS-aligned valuation methods (IAS 2), i.e. FIFO or weighted average, improve cost allocation, margin visibility and the quality of financial reporting ( Ala et al., 2023). Similarly, KPIs such as compliance rate, inventory turnover, and shrinkage rate have proven to be highly predictive of operating margins ( Sánchez-Córdova et al., 2024). Therefore, integrated inventory control systems, whether manual or digital, impact profitability not only through cost containment but also through strategic differentiation, especially in short-life-cycle categories susceptible to deterioration.

Despite these data, geographical coverage remains skewed towards the urban and well-digitised ecosystems of Asia and Europe, with limited evidence in secondary cities or Amazonian economies ( Granillo-Macías et al., 2024). The structural conditions of these environments—informality, long delivery times, concentration of suppliers—may moderate the link between inventory policies and profitability ( Singh et al., 2022). This contingency suggests that practices that are effective in metropolitan contexts may not produce similar effects in regions such as Bagua, Peru.

Previous studies corroborate that inventory capabilities create financial value, but few studies integrate RBV with dynamic capabilities to explain profitability outcomes in SMEs under the specific constraints of the Amazon; Consequently, this study examines how inventory control and measurement practices affect ROA, GM, and ROE in Bagua-based SMEs, addressing a geographical and theoretical gap with practical implications for similar contexts in developing economies.

A quantitative, descriptive and explanatory design was adopted, giving priority to inventory control ((R, Q), safety stock), valuation (FIFO or weighted average in line with IFRS/IAS 2), accuracy and timeliness in record keeping and inventory measurement (DIO, turnover, GMROI, compliance rate) as internal capabilities that link stock depletion, maintenance and replacement costs to profitability through the cash conversion cycle.

The target population consisted of micro and small enterprises registered and active in Bagua with ≥5 employees; 83 micro and small enterprises in the trade/services and construction sectors were contacted, inviting five informants per company (manager, warehouse manager, general accountant and two accounting assistants) to mitigate single source bias; inclusion required active tax registration, operational status, and willingness/commitment of time; the valid sample included 200 respondents from 40 micro-enterprises.

Non-probability convenience sampling was used due to access restrictions and management authorisation; this implies coverage/selection bias and limits statistical generalisation; future studies should apply probability-based stratified sampling (by sector/size), multi-wave panels, and triangulate survey responses with accounting/ERP records.

Inventory control comprised three dimensions: inventory valuation methods (P1-P4), inventory control records (P5-P8) and inventory measurement (P9-P12). Profitability included ROA (P13-P16), gross margin (P17-P20), and ROE (P21-P24). All items used 5-point Likert scales (Never = 1 to Always = 5). The complete correspondence is shown in Table X (instrument source: Questionnaires).

The items were adapted from operations/accounting sources and adjusted to the SMEs; content validity was based on expert judgement (CVI/CVR). A pilot test was conducted to assess clarity and timeliness. Reliability and dimensionality were assessed using Cronbach’s alpha and EFA (KMO, Bartlett); where possible, CFA was estimated and composite reliability (CR) and average variance extracted (AVE) were calculated (criteria CR ≥ 0.70; AVE ≥ 0.50).

Internal consistency was acceptable to excellent: α (DIM1) = 0.749; α (DIM2) = 0.858; α (DIM3) = 0.842; α (DIM4) = 0.888; α (DIM5) = 0.798; α (DIM6) = 0.829; sampling adequacy supported factor analysis (IV: KMO = 0.866; DV: KMO = 0.913; all items: KMO = 0.910); Bartlett’s tests rejected sphericity (IV: χ ² = 1111.59, df = 66; DV: χ ² = 1432.63, df = 66; all items: χ ² = 2818.27, df = 276).

Procedural corrective measures included anonymity, balanced wording, separate blocks, and a multiple informant approach; statistically, a Harman factor test was performed as exploratory screening, a marker variable adjustment, and, subject to the sample, MTMM/CFA.

Assumption checks (Shapiro-Wilk, Levene) preceded inference. Main models used OLS with industry/size/age controls; robustness checks included alternative specifications, exclusion of influential cases, and marker-adjusted correlations.

Descriptive statistics for the inventory management and profitability variables are presented in Table 1; overall performance was moderate across all dimensions, with inventory measurement receiving the highest mean score (M = 2.62), indicating an operational emphasis on tracking turnover and days of inventory. The average score of 3 across most variables reinforces a prevailing level of acceptable performance among the micro and small enterprises surveyed; a negative skew was observed across all dimensions, particularly in profitability metrics such as ROA (sk = -0.947) and GM (sk = -1.007), indicating a tendency for profitability to be lower than expected. A negative asymmetry was observed across all dimensions, particularly in profitability metrics such as ROA (sk = -0.947) and GM (sk = -1.007), suggesting that most firms are performing relatively well, while a smaller subset with poor performance is dragging down the mean. These results highlight that companies show greater maturity in metrics and monitoring than in planning and valuation, in line with studies in which MSMEs adopt visible control tools before strengthening their accounting bases ( Ahmad et al., 2022; Sultan, 2021).

The variables in Table 2 Inventory measurement was the strongest indicator and the only robust predictor of profitability (β = 0.383; p = 0.001), highlighting that companies capable of effectively tracking turnover, days of inventory, and obsolescence costs strengthen their margins and returns. This is consistent with Vahdani and Sazvar (2022), who argue that monitoring real-time inventory dynamics reduces capital immobilisation and improves ROA/ROE.

Inventory control also showed a statistically significant effect (β = 0.257; p = 0.013), suggesting that replenishment policies, safety stock, and order tracking help companies avoid stockouts and reduce emergency purchases, in line with Wilson (2023) and Panigrahi A, et al. (2024).

In contrast, inventory valuation methods and control records showed no direct statistical effects (p > 0.40), supporting contingency-based views: valuation decisions produce indirect effects mediated by the costs of goods sold (IAS 2; Chen and Kong, 2019); records may help internally, but they require integration with planning systems to be reflected in profitability ( Becerra et al., 2022). These findings reinforce the dynamic capabilities approach: capabilities must not only exist, but must be strategically deployed to convert operational data into economic benefits.

The model illustrates that inventory measurement (IM) has the strongest and only statistically significant relationship with profitability (β = 0.383, p < 0.001); MSEs that continuously evaluate turnover, ageing, stockouts, and replenishment rates demonstrate greater operational responsiveness, leading to better cost control and higher financial returns.

In contrast, inventory valuation methods (IVM) (β = 0.053, p > 0.05) and inventory control records (ICR) (β = 0.037, p > 0.05) show no significant direct effect on profitability; these areas, while relevant to accounting accuracy and compliance, appear insufficient on their own to drive financial results without being connected to tactical decision-making.

These data reinforce the idea that measurement capability is a dynamic capability that strengthens strategic decision-making in purchasing and replenishment; in line with the resource-based view (RBV), inventory management becomes a competitive resource when it generates financial benefits and cannot be easily imitated by competitors.

The results are also consistent with contingency theory, which indicates that operational practices must adapt to the environmental uncertainty typical of micro and small enterprises in Bagua, where high demand variability and limited administrative experience increase vulnerability to financial losses.

From a management perspective, the model suggests that micro-enterprises should prioritise inventory analysis dashboards over purely documentary control. In addition, they should strengthen forecasting, reorder points and safety stock policies related to measured KPIs. Similarly, they should integrate IVM and ICR into pricing, purchasing, and logistics workflows to unlock their indirect effect on returns.

This research was approved by the Ethics Committee of César Vallejo University (Resolution No. 447-2023-VI-UCV of the Vice-Rectorate for Research), in accordance with the principles of the Declaration of Helsinki. All procedures complied with national and international ethical standards applicable to research involving human participants.

Written informed consent was obtained from all participants before administering the questionnaire. All respondents were over 18 years of age and were fully informed about the objectives of the study, the voluntary nature of their participation, the confidentiality of their data, and the exclusive use of the information for academic and scientific purposes.