Collectable Cars as an Alternative Investment: Examination of car traits for investment selection & comparison of investment performance against traditional investments
Translated title
Collectable Cars as an Alternative Investment
Author
Møller Svarrer, Christopher
Term
4. semester
Publication year
2026
Submitted on
2026-06-01
Pages
605
Abstract
Interest in alternative assets like collectibles has grown. This thesis examines whether collectible cars are attractive investments. Using data from 300 auction sales, we apply a hedonic regression—a statistical method that estimates how specific features influence prices—and select the final model with the Bayesian Information Criterion (BIC), which balances fit and simplicity. We find that higher mileage, poorer condition, and large production volumes reduce value, while manual transmission, convertible body style, racing pedigree, higher power, and greater age increase value. We then build a portfolio that favors value-adding traits and avoids value-detracting ones, and track it from 2010 to 2024. To mimic real ownership, ongoing costs are included, providing a more realistic view of returns. Because data are limited, the analysis relies on in-sample estimates and interpolation, which can introduce bias and smooth volatility. Compared with financial benchmarks, the portfolio underperforms most, especially after accounting for costs. Over the period, only the conservative US aggregate bond ETF (AGG) performs worse. Stocks perform much better, helped by positive yields, whereas cars have a negative cost of carry (they cost money to hold). Risk-adjusted metrics tell the same story: the portfolio’s Sharpe ratio is -0.16 versus 0.97 for the Dow Jones Industrial Average and 0.86 for the S&P 500; smoothing likely means true risk-adjusted performance is slightly worse. Even so, collectible cars show weak negative correlation with traditional assets and negative betas to the S&P 500, suggesting diversification benefits. Ownership costs significantly reduce financial returns, but owners may see non-financial benefits (enjoyment, networking). Overall, the constructed portfolio underperforms comparable benchmarks; annual price volatility appears lower, and cars less than 25 years old are not optimal for superior risk-adjusted returns. Results are explanatory and may change with different assumptions or car selections.
Interessen for alternative aktiver som samlerobjekter er steget. Dette speciale undersøger, om samlerbiler er en attraktiv investering. Med auktionsdata fra 300 salg anvendes en hedonisk regressionsmodel—en statistisk metode, der estimerer, hvordan konkrete egenskaber påvirker priser—og den endelige model vælges med det bayesianske informationskriterium (BIC), som afvejer tilpasning mod enkelhed. Vi finder, at højere kilometertal, ringere stand og store produktionsvolumener trækker værdien ned, mens manuel gearkasse, cabriolet, racerløbshistorik, højere effekt og højere alder trækker op. Herefter konstrueres en portefølje, der maksimerer værdiskabende træk og minimerer værdiforringende kendetegn, og den følges fra 2010 til 2024. For at efterligne virkeligheden indregnes løbende ejerskabsomkostninger, hvilket giver mere realistiske afkast. Begrænset datatilgængelighed betyder, at analysen bygger på in-sample-estimater og interpolation, som kan give bias og udglatte udsving. Sammenlignet med finansielle benchmarks underpræsterer porteføljen de fleste, især når omkostninger medregnes. Over perioden klarer kun det konservative amerikanske obligations-ETF (AGG) sig dårligere. Aktier klarer sig langt bedre, hjulpet af positive udbytter, mens biler har en negativ “carry”-omkostning (det koster at eje dem). Risikojusterede mål peger samme vej: porteføljens Sharpe-ratio er -0,16 mod 0,97 for Dow Jones Industrial Average og 0,86 for S&P 500; udglatning indebærer, at den reelle risikojusterede performance sandsynligvis er lidt svagere. Alligevel viser samlerbiler svag negativ korrelation med traditionelle aktiver og negative betaer mod S&P 500, hvilket kan give diversifikationsfordele. Ejerskabsomkostninger påvirker afkastet markant, men der kan være indirekte gevinster (glæde, netværk). Samlet set underpræsterer den konstruerede portefølje sammenlignelige benchmarks; den årlige prisvolatilitet synes lavere, og biler under 25 år er ikke optimale til at opnå overlegne risikojusterede afkast. Resultaterne er forklarende og kan variere med andre antagelser eller bilvalg.
[This apstract has been rewritten with the help of AI based on the project's original abstract]
Keywords