Lessons from the World’s Largest Darknet Marketplace: Part 4

Death Star — November 22, 2023

We present an original translation of a foreign scientific study on the HYDRA darknet marketplace and its operations.

Authors: phØBoŠ & hŮmAnǿ1Đ

6. Conclusion

This work provides a unique quantitative overview of HYDRA, the illegal darknet marketplace that dominated Russia from 2015 to 2022. Using two new datasets, including drug sale listings and consumer reviews on HYDRA, we estimate online drug buyer behavior: the popularity of various substances, their prices, volumes, and geographic distribution both between Russian cities and within districts of the two largest cities—Moscow and St. Petersburg. We also present evidence that, prior to its shutdown in April 2022, HYDRA was the dominant method for acquiring drugs in Russia.

Perhaps the main feature that set HYDRA apart from darknet marketplaces in other countries and enabled its market dominance was its longevity. After RAMP closed in 2017, HYDRA was able to develop and strengthen its market position for years until its own closure in 2022. In contrast, of the 83 Western darknet marketplaces studied by Bhaskar et al. (2019), the original Silk Road lasted the longest—976 days. This stability seemingly allowed HYDRA to capture a market share far exceeding that of Western darknet marketplaces, where the vast majority of trade still occurs via street-level offline sales. By using cryptocurrency transactions, encrypted messaging, and eliminating physical contact between buyers, sellers, and couriers, HYDRA became the largest illegal online drug marketplace to date. It serves as an example of the potential evolution of the online drug market if a darknet marketplace is allowed to operate unchecked for several years.

This analysis has several limitations. First, we use listing data only from April 2020, which coincided with the peak of the COVID-19 pandemic and may have affected the composition and structure of Russia’s illegal drug market. Second, we operate with a very short time series, covering just one month. Future research could use a longer panel and analyze seasonal and other temporal effects in prices, supply, and other darknet marketplace indicators, which could lead to more nuanced, causally interpretable results.

Another limitation is the reliance on media materials and documents from international and government organizations for background information on HYDRA and the situation after its closure. Due to a lack of published peer-reviewed research on HYDRA and, more generally, on online drug markets in Russia, we had to use other sources such as journalistic investigations and government documents. However, we acknowledge that these may be biased and limited in their perspectives due to selective reporting, sensationalism, or specific political agendas.

We believe the HYDRA case illustrates the many trade-offs governments face when deciding whether to shut down darknet marketplaces. Quantitative assessment of these trade-offs is important for addressing the problem of drug abuse. We hope that future research on illegal drug markets will move in this direction.

Acknowledgments

All authors contributed equally to this work. Jonathan Caulkins and Aaron Chalfin provided valuable comments on an early version of this paper, for which we are grateful. We thank Alexey Lakhov, Andrey Kagansky, Maxim Gorbunov, journalists from Proekt Media, and anonymous interviewees for helpful discussions. All errors are our own. Alex Knorre acknowledges support from AFOSR MURI grant FA9550-22-1-0380.

Conflict of Interest

The authors confirm that they have no conflicts of interest to declare.

Notes

1. On April 5, 2022, German federal police announced that, in a joint operation with U.S. law enforcement, HYDRA’s servers in Germany were seized and the site was shut down. As a result, the U.S. Department of Justice charged Dmitry Olegovich Pavlov for his alleged role in managing the company that administered HYDRA’s servers. The indictment includes conspiracy to distribute drugs and conspiracy to launder money.
2. The marketplace discussed in this article should not be confused with another market of the same name that was closed in 2014.
3. While any sale to retailers for resale can be considered wholesale, for all quantitative analysis we follow Demant et al. (2018) and define it as any transaction over $1,000.
4. Pre-hidden drug stashes allowed transactions without any physical interaction between sellers and buyers.
5. The quality of these consultations, as well as the qualifications of those providing them, remains unclear.
6. To our knowledge, this is the first comprehensive, large-scale assessment of this market in the academic literature. Zvonka (2017) is an early analysis of data from HYDRA and RAMP, the marketplace HYDRA initially competed with until RAMP closed in 2017. Proekt (2019) is a journalistic investigation from 2019 and one of the first media discussions of HYDRA.
7. Some HYDRA vendors also used mail to sell drugs, though the most common delivery method was dead drops.
8. An “exit scam” refers to situations where darknet platform operators shut down the site and disappear with users’ funds held in escrow or account balances.
9. Section 4 discusses platforms that became popular after HYDRA’s closure. While all existed before 2022, none gained significant popularity until HYDRA was shut down (see darknet market share analysis in Chainalysis, 2021 and 2022).
10. HYDRA is currently offline. We have preserved HTML files for these pages.
11. All materials are available from the authors upon request.
12. The only information a potential user had to provide was a login, display nickname, and password.
13. HYDRA also had a “roulette” feature that allowed consumers to participate in a lottery. The lottery cost less than the product price, but the consumer would win the desired product with a probability less than 1.
14. All saved HTML files are available from the authors upon request.
15. The “snow burial” dead drop method appeared after our data was collected.
16. We found that the ratio of mail to instant positions is about 10% for LSD and about 5% for mushrooms, several times higher than for other drugs. Unlike instant positions, each mail position does not correspond to a specific district and may cover the entire country. Thus, this ratio likely underestimates the share of mail orders for these substances. This matches anecdotal data from our informants that mail was mainly used for delivering psychedelics on HYDRA.
17. During the first 36 hours, users could add additional comments to their reviews.
18. Here and below, we use the May 2020 exchange rate, approximately 74 rubles per U.S. dollar.
19. If we observe multiple reviews with identical text from the same user, we keep only one. Before removal, we had 414,700 reviews.
20. Similarly, we remove duplicates when merging listings by day.
21. Bhaskar et al. (2019) collected data over a year and noted fewer vendors on other major marketplaces: 902 on The Silk Road, 796 on Silk Road 2.0, 1,754 on Evolution, and 1,836 on Nucleus.
22. Some stores apparently use “holding” prices when they run out of stock. Instead of removing the listing, they set the price so high that no buyer will attempt to purchase it.
23. A large wholesale stash with ready-made dead drops is called a “master stash.”
24. We found that certain words were most often used to indicate that a dispute was resolved in favor of the store.
25. As mentioned in Section 2.2, selling goods outside the market was prohibited by HYDRA’s rules. Thus, before its closure, the largest HYDRA stores could not operate elsewhere under the same brand.
26. This analysis relies on known links between Bitcoin wallets and market participants, so there is some uncertainty in the exact number of transactions these darknet marketplaces receive. According to Chainalysis, which uses similar methods, total darknet revenue is still somewhat below pre-HYDRA levels (Chainalysis, 2023a). In any case, it’s clear that the drop in Russian darknet transactions was short-lived.
27. Darknet marketplace fixed costs are usually higher than for regular clearnet platforms. For example, software engineers must be paid more for illegal projects, and darknet hosting and overall site maintenance and protection are more expensive.

Appendix A: Intra-city Distribution in St. Petersburg

In Section 3.3.1, we discussed how the spatial distribution of dead drops within a city responds to changes in demand and thus reflects socioeconomic differences between city districts. Specifically, Figure 4 illustrated this for Moscow. Figure A.1 presents maps showing the spatial distribution of listings for St. Petersburg, HYDRA’s second-largest market. As in Moscow, cocaine is concentrated in the city center, while cheaper drugs like mephedrone and marijuana are found in peripheral districts.

Figure A1. Spatial distribution of dead drops in St. Petersburg

Note: Map of St. Petersburg showing the share of listings for each drug type as a percentage of all drug listings in each district. “Synthetics” includes methamphetamine, amphetamine, MDMA, alpha-PVP, MDPV, and mephedrone. Only districts with the highest number of dead drops are included: Admiralteysky, Vasileostrovsky, Vyborgsky, Kalininsky, Kirovsky, Krasnogvardeysky, Krasnoselsky, Moskovsky, Nevsky, Petrogradsky, Primorsky, Frunzensky, and Tsentralny.

Appendix B: Other Screenshots

Figure B1. Example job listing from a major vendor on KRAKEN

Table B1. Sample reviews on HYDRA.

Table B2. Sample reviews on HYDRA.

Table B3. Sample reviews on HYDRA.

Biographies

• Priyanka Gunetilleke is a PhD candidate in the Department of Economics at the University of Pennsylvania. His research focuses on law and economics, as well as the economics of crime, with an emphasis on racial inequality in the criminal justice system and the economics of illegal drug markets. He received a BA in Economics (First Class Honors) and a Bachelor of Laws (First Class Honors) from the University of Sydney.
• Alex Knorre is a postdoctoral research fellow in the Department of Computer Science at Boston College. His research interests include crime and victimization, illegal drugs, statistics, and causal inference.
• Artem Kuriksha is a PhD candidate in the Department of Economics at the University of Pennsylvania. His research interests are in the organization of production and empirical microeconomics. In particular, he is interested in the economics of illegal drug markets and the application of machine learning in economics. He received a master’s degree in economics from the New Economic School in Russia and a bachelor’s degree in mathematics from the Higher School of Economics.

Notes from Death Star

1. We were surprised not to read about HYDRA’s built-in mixer, called the Bitcoin Bank Mixer, which allowed users to withdraw clean crypto to exchanges for subsequent cash-out. This was one of the platform’s clear sources of income, as the service cost at least 5% of the withdrawal amount and collected dozens of reviews daily.
2. The same can be said about the roulette feature, which is only mentioned in the notes. While it’s nearly impossible to estimate the income from this activity, it would be naive to assume HYDRA ran it at a loss.
3. It’s also widely known that HYDRA had affiliated exchangers that offered clients BTC purchases with an average markup of about 18-20%.
4. It was odd not to read anything about their planned ICO called Eternos, which featured some truly revolutionary ideas not proposed by any other darknet marketplace or legitimate business.
5. The article mentions HYDRA’s protection against DDoS attacks but doesn’t mention that HYDRA itself regularly attacked other marketplaces, which helped it become a monopoly.

Otherwise, the material is good and interesting, with a warm and nostalgic feel. The data is fairly representative, though limited in time and lacking dynamics.

By phØBoŠ & hŮmAnǿ1Đ
Death Star