What Holds Its Value? Price Determinants in Indias Used Motorcycle Market and the Absence of the Yamaha RX100

Abstract

The Yamaha RX100 retains a strong reputation among Indian motorcycle enthusiasts, yet it does not appear once in 7,324 listings of used motorcycles offered for sale online in India. This paper starts from that absence and asks a broader question: what determines the asking price of a used two-wheeler in India? The data are a public, CC0-licensed file of listings scraped from the marketplace droom.in. The file holds 32,648 rows, but 25,324 of them are exact duplicates, leaving 7,324 distinct listings; the duplication is not disclosed by the publisher and is documented here. A hedonic regression explains 84.7 per cent of the variation in log asking price. Engine displacement, age and brand each carry a large partial effect, and asking price falls by about 7.1 per cent for each year of age. Ownership history matters much less than its place in every listing suggests: holding other attributes constant, a second-owner machine is listed 4.1 per cent below a comparable first-owner one (95 per cent confidence interval 1.4 to 6.7 per cent), and ownership has a partial ?² of 0.002. A deliberate check for Simpson’s paradox found none: the negative relationship between age and price holds within every brand and every displacement band examined. Depreciation slows with age, and the fitted curve turns upward at about 25 years (bootstrap 95 per cent interval 19.7 to 30.0 years), but that upturn rests on 17 to 20 listings, most of them Royal Enfields, and is reported as suggestive. Royal Enfield is listed 33.5 per cent above the reference brand, Bajaj, on otherwise comparable machines. A search of 173 further public datasets found only twelve RX100 sale listings among them. All results describe asking prices, not transaction prices.

Introduction

1. Introduction

The study examines the factors that influence the asking prices of used motorcycles in India. Two-wheelers account for 73.6% of vehicles registered in the Indian sample cited in the paper, but research on their resale prices is limited compared with used cars.

The research focuses on vehicle age, kilometres driven, engine displacement, ownership history, brand, and location. Its main purpose is to measure how strongly each factor is associated with asking price and to investigate whether ownership history has as much influence as commonly assumed.

The study initially aimed to investigate the resale value of the Yamaha RX100, but no RX100 listings were found in the main dataset. Therefore, the research was expanded to cover the broader used two-wheeler market.

2. Problem Statement and Research Gap

Used two-wheeler prices are difficult for buyers and sellers to evaluate because listings do not explain how much each vehicle characteristic contributes to the asking price.

Previous research mainly focuses on used-car price prediction using machine learning. Limited research has examined Indian used motorcycles using comparable effect sizes and statistical tests for aggregation bias.

The study addresses this gap by developing a hedonic price model and examining the relative contribution of vehicle characteristics.

3. Research Objectives

• Measure the relationship between asking price and age, mileage, engine displacement, and ownership history.

• Estimate brand-specific price differences after controlling for vehicle characteristics.

• Rank pricing factors according to their effect sizes.

• Test for Simpson’s paradox in age–price relationships across brands and engine categories.

• Investigate whether the oldest motorcycles show an increase in resale value.

• Document the availability of Yamaha RX100 listings in online resale data.

4. Research Methodology

The study uses a quantitative approach based entirely on secondary data collected from public sources.

Dataset and sample

The original dataset contained 77.6% duplicate rows, which were removed to prevent repeated listings from distorting the analysis.

The study uses Spearman correlation, Kruskal–Wallis and Mann–Whitney tests, and ordinary least squares hedonic regression. Effect sizes are reported alongside statistical significance.

5. Major Findings

The analysis identifies the following relationships:

The hedonic regression explains approximately 84.7% of the variation in log asking prices.

Each additional year of age is associated with a 7.1% lower asking price, while a 1% increase in kilometres covered is associated with approximately a 0.10% lower price, holding other variables constant.

An otherwise comparable 350cc motorcycle is estimated to have an asking price approximately 105.8% higher than a 100cc motorcycle.

6. Ownership History and Brand Effects

Ownership history has a smaller effect than the raw price differences suggest. First-owner motorcycles have a median asking price of ?55,000, compared with ?47,000 for later-owner motorcycles. However, later-owner vehicles are generally older and have higher mileage.

After controlling for age, mileage, displacement, and brand, second-owner motorcycles are listed approximately 4.1% below comparable first-owner motorcycles. The overall effect of ownership history is negligible in the model.

Brand differences also remain important after accounting for vehicle characteristics. Bajaj is used as the reference manufacturer for estimating brand effects.

7. Statistical Analysis and Challenges

The research examines Simpson’s paradox, where a relationship observed in combined data may differ within individual subgroups. It also emphasizes effect sizes rather than relying solely on p-values.

Important limitations include missing information about mechanical condition, service history, accident records, modifications, and actual transaction prices. The dataset represents online asking prices rather than completed sales.

Conclusion

This study set out to explain what determines the asking price of a used motorcycle in India, using 7,324 distinct listings recovered from a public file that was more than three-quarters duplicated. A hedonic model explains 84.7 per cent of the variation in log asking price. Engine displacement is the strongest factor, followed by age at about 7.1 per cent of asking price per year, then brand, with Royal Enfield listed 33.5 per cent above the reference manufacturer on comparable machines. Distance covered matters, but far less than age or displacement. Two of the results are useful precisely because they are small or negative. Ownership history is statistically significant but modest: once age, distance and brand are accounted for, a second-owner machine is listed about four per cent below a first-owner one, and ownership has a partial ?² of 0.002. The paper reports that plainly rather than letting a p-value of 0.000022 stand in for importance. A deliberate search for Simpson’s paradox found none: the negative age–price relationship holds within every brand and every displacement band, so the headline depreciation figure describes the market rather than an accident of aggregation. The evidence on whether old motorcycles recover value is mixed, and is reported that way. Depreciation clearly slows with age, and that result survives every check applied. The apparent upturn beyond twenty-five years is present in the fitted curve but rests on seventeen to twenty listings and is confounded with brand: the machines that remain listed at that age are mostly Royal Enfields. The study began with the Yamaha RX100 and ends with it, having found none. In 7,324 listings, among 651 Yamahas of 49 distinct models, in a dataset that contains a sixty-three-year-old Bullet and a forty-one-year-old Rajdoot, the RX100 does not appear once. A motorcycle that enthusiasts still remember is missing from a marketplace that lists almost everything else. Whatever value it holds is being exchanged somewhere the data cannot see. That is a finding of a kind, even if it is not one that can be given a p-value.

References

[1] H. G. Grubel, “International trade in used cars and problems of economic development,” World Development, vol. 8, no. 10, pp. 781?–?788, Oct. 1980. [2] Y. Nakamoto, “CO2 reduction potentials through the market expansion and lifetime extension of used cars,” Journal of Economic Structures, vol. 6, no. 1, Art. no. 17, 2017. [3] M. Koç and N. Kostak, “What is happening to used car prices in Turkey? Evidence from hedonic price model (HPM),” Selçuk Üniversitesi Sosyal Bilimler Enstitüsü Dergisi, no. 46, pp. 225?–?240, 2021. [4] M. Francke and A. Van de Minne, “Modeling unobserved heterogeneity in hedonic price models,” Real Estate Economics, vol. 49, no. 4, pp. 1315?–?1339, 2021. [5] E. Gegic, B. Isakovic, D. Keco, Z. Masetic, and J. Kevric, “Car price prediction using machine learning techniques,” TEM Journal, vol. 8, no. 1, pp. 113?–?118, Feb. 2019. [6] P. Bharambe, B. Bagul, S. Dandekar, and P. Ingle, “Used car price prediction using different machine learning algorithms,” International Journal for Research in Applied Science and Engineering Technology, vol. 10, no. 4, pp. 773?–?778, Apr. 2022. [7] G. Kalpana, A. Kanaka Durga, T. A. Reddy, and G. Karuna, “Predicting the price of pre-owned cars using machine learning and data science,” International Journal for Research in Applied Science and Engineering Technology, vol. 10, no. 7, pp. 1468?–?1476, Jul. 2022. [8] K. Ranjith, B. K. Rishi Sagar, and U. Sharma, “Prediction of resale value of pre-owned luxury cars in the Indian market employing machine learning techniques,” International Journal for Research in Applied Science and Engineering Technology, vol. 12, no. 7, pp. 1026?–?1037, Jul. 2024. [9] D. A. Budiono, K. S. Utomo, K. J. Wibowo, and M. J. Wiradinata, “Used car price prediction model: A machine learning approach,” International Journal of Computer and Information System, vol. 5, no. 1, pp. 59?–?66, Mar. 2024. [10] A. Barlybayev, A. Sankibayev, Y. Kadyr, N. Amangeldy, and T. Sabyrov, “Predicting used-vehicle resale value in developing markets: Application of machine learning models to the Kazakhstan car market,” Ingénierie des Systèmes d’Information, vol. 28, no. 5, pp. 1237?–?1246, Oct. 2023. [11] S. Das, R. Schmoyer, G. Harrison, and K. Hausker, “Prospects of inspection and maintenance of two-wheelers in India,” Journal of the Air & Waste Management Association, vol. 51, no. 10, pp. 1391?–?1400, Oct. 2001. [12] M. R. Shaharudin, S. Wan Mansor, A. Abu Hassan, M. Wan Omar, and E. H. Harun, “The relationship between product quality and purchase intention: The case of Malaysia’s national motorcycle/scooter manufacturer,” African Journal of Business Management, vol. 5, no. 20, pp. 8163?–?8176, Sep. 2011. [13] C. Balijepalli, S. Shepherd, R. Crastes dit Sourd, M. Farda, T. Praesha, and H. A.-R. Lubis, “Preferences for electric motorcycle adoption in Bandung, Indonesia,” Urban, Planning and Transport Research, vol. 11, no. 1, Art. no. 2238033, 2023. [14] S. Bakker, “Electric two-wheelers, sustainable mobility and the city,” in Sustainable Cities – Authenticity, Ambition and Dream. IntechOpen, 2019, ch. 6, doi: 10.5772/intechopen.81460. [15] L. R. Cirolia, R. Sitas, A. Pollio, A. G. Sebarenzi, and P. K. Guma, “Silicon Savannahs and motorcycle taxis: A Southern perspective on the frontiers of platform urbanism,” Environment and Planning A: Economy and Space, vol. 55, no. 8, pp. 1989?–?2008, 2023. [16] R. A. Kievit, W. E. Frankenhuis, L. J. Waldorp, and D. Borsboom, “Simpson’s paradox in psychological science: A practical guide,” Frontiers in Psychology, vol. 4, Art. no. 513, Aug. 2013. [17] Z. Armstrong and M. Wattenberg, “Visualizing statistical mix effects and Simpson’s paradox,” IEEE Transactions on Visualization and Computer Graphics, vol. 20, no. 12, pp. 2132?–?2141, Dec. 2014. [18] P. H. Walton and D. J. Walton, “Simpson’s paradox in the interpretation of ‘leaky pipeline’ data,” International Journal for Transformative Research, vol. 3, no. 2, pp. 1?–?7, 2016. [19] D. Lakens, “Calculating and reporting effect sizes to facilitate cumulative science: A practical primer for t-tests and ANOVAs,” Frontiers in Psychology, vol. 4, Art. no. 863, Nov. 2013. [20] P. H. P. Hanel and D. M. A. Mehler, “Beyond reporting statistical significance: Identifying informative effect sizes to improve scientific communication,” Public Understanding of Science, vol. 28, no. 4, pp. 468?–?485, 2019.

Copyright

Copyright © 2026 Perseus Bhavnagri. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.