You are on page 1of 31

Data on Bitcoin-Exchange Closures Survival Analysis of Exchange Closure Regression Analysis of Exchange Breaches

Beware the Middleman: Empirical Analysis of Bitcoin-Exchange Risk


Tyler Moore1
1

Nicolas Christin2

Computer Science & Engineering, Southern Methodist University, USA, tylerm@smu.edu 2 INI & CyLab, Carnegie Mellon University, USA, nicolasc@cmu.edu

Financial Crypto 2013 Okinawa, Japan April 2, 2013

Tyler Moore & Nicolas Christin

Empirical Analysis of Bitcoin-Exchange Risk

Data on Bitcoin-Exchange Closures Survival Analysis of Exchange Closure Regression Analysis of Exchange Breaches

Motivation

Decentralization is a key feature of Bitcoins design Viewed as a security benet: protects against ination risk, sovereign risk, etc. Yet an extensive ecosystem of 3rd-party intermediaries now supports Bitcoin transactions: currency exchanges, escrow services, online wallets, mining pools, investment services, . . . Most risk Bitcoin holders face stems from interacting with these intermediaries, who act as de facto central authorities We focus on risk posed by failures of currency exchanges

Tyler Moore & Nicolas Christin

Empirical Analysis of Bitcoin-Exchange Risk

Data on Bitcoin-Exchange Closures Survival Analysis of Exchange Closure Regression Analysis of Exchange Breaches

Motivation

Decentralization is a key feature of Bitcoins design Viewed as a security benet: protects against ination risk, sovereign risk, etc. Yet an extensive ecosystem of 3rd-party intermediaries now supports Bitcoin transactions: currency exchanges, escrow services, online wallets, mining pools, investment services, . . . Most risk Bitcoin holders face stems from interacting with these intermediaries, who act as de facto central authorities We focus on risk posed by failures of currency exchanges

Tyler Moore & Nicolas Christin

Empirical Analysis of Bitcoin-Exchange Risk

Data on Bitcoin-Exchange Closures Survival Analysis of Exchange Closure Regression Analysis of Exchange Breaches

Motivation

Decentralization is a key feature of Bitcoins design Viewed as a security benet: protects against ination risk, sovereign risk, etc. Yet an extensive ecosystem of 3rd-party intermediaries now supports Bitcoin transactions: currency exchanges, escrow services, online wallets, mining pools, investment services, . . . Most risk Bitcoin holders face stems from interacting with these intermediaries, who act as de facto central authorities We focus on risk posed by failures of currency exchanges

Tyler Moore & Nicolas Christin

Empirical Analysis of Bitcoin-Exchange Risk

Data on Bitcoin-Exchange Closures Survival Analysis of Exchange Closure Regression Analysis of Exchange Breaches

Outline of todays talk


1

Data on Bitcoin-Exchange Closures Data Collection Methodology Summary Statistics Survival Analysis of Exchange Closure Statistical Model Results Risk Ratio for Bitcoin Exchanges Regression Analysis of Exchange Breaches Statistical Model Results

Tyler Moore & Nicolas Christin

Empirical Analysis of Bitcoin-Exchange Risk

Data on Bitcoin-Exchange Closures Survival Analysis of Exchange Closure Regression Analysis of Exchange Breaches

Data Collection Methodology Summary Statistics

Outline
1

Data on Bitcoin-Exchange Closures Data Collection Methodology Summary Statistics Survival Analysis of Exchange Closure Statistical Model Results Risk Ratio for Bitcoin Exchanges Regression Analysis of Exchange Breaches Statistical Model Results

Tyler Moore & Nicolas Christin

Empirical Analysis of Bitcoin-Exchange Risk

Data on Bitcoin-Exchange Closures Survival Analysis of Exchange Closure Regression Analysis of Exchange Breaches

Data Collection Methodology Summary Statistics

Data collection methodology


Data sources
1

Daily transaction volume data on 40 exchanges converting into 33 currencies from bitcoincharts.com Checked for closure, mention of security breaches and whether investors were repaid on Bitcoin Wiki and forums To assess impact of pressure from nancial regulators, we identied each exchanges country of incorporation and used a World Bank index on compliance with anti-money laundering regulations Time dierence between rst and last observed trade We deem an exchange closed if no transactions are observed at least 2 weeks before data collection nished

Key measure: exchange lifetime

Tyler Moore & Nicolas Christin

Empirical Analysis of Bitcoin-Exchange Risk

Data on Bitcoin-Exchange Closures Survival Analysis of Exchange Closure Regression Analysis of Exchange Breaches

Data Collection Methodology Summary Statistics

Some initial summary statistics

40 Bitcoin currency exchanges opened since 2010 18 have subsequently closed (45% failure rate)
Median lifetime is 381 days 45% of closed exchanges did not reimburse customers

9 exchanges were breached (5 closed)

Tyler Moore & Nicolas Christin

Empirical Analysis of Bitcoin-Exchange Risk

Data on Bitcoin-Exchange Closures Survival Analysis of Exchange Closure Regression Analysis of Exchange Breaches

Data Collection Methodology Summary Statistics

18 closed Bitcoin currency exchanges


Exchange BitcoinMarket Bitomat FreshBTC Bitcoin7 ExchangeBitCoins.com Bitchange.pl Brasil Bitcoin Market Aqoin Global Bitcoin Exchange Bitcoin2Cash TradeHill World Bitcoin Exchange Ruxum btctree btcex.com IMCEX.com Crypto X Change Bitmarket.eu Origin US PL PL US/BG US PL BR ES ? US US AU US US/CN RU SC AU PL Dates Active 4/10 6/11 4/11 8/11 8/11 9/11 6/11 10/11 6/11 10/11 8/11 10/11 9/11 11/11 9/11 11/11 9/11 1/12 4/11 - 1/12 6/11 - 2/12 8/11 2/12 6/11 4/12 5/12 7/12 9/10 7/12 7/11 10/12 11/11 11/12 4/11 12/12 Daily vol. 2454 758 3 528 551 380 0 11 14 18 5082 220 37 75 528 2 874 33 Closed? yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes Breached? yes yes no yes no no no no no no yes yes no no no no no no Repaid? yes no yes no yes yes no no AML 34.3 21.7 21.7 33.3 34.3 21.7 24.3 30.7 27.9 34.3 34.3 25.7 34.3 29.2 27.7 11.9 25.7 21.7

Tyler Moore & Nicolas Christin

Empirical Analysis of Bitcoin-Exchange Risk

Data on Bitcoin-Exchange Closures Survival Analysis of Exchange Closure Regression Analysis of Exchange Breaches

Data Collection Methodology Summary Statistics

22 open Bitcoin currency exchanges


Exchange bitNZ ICBIT Stock Exchange WeExchange Vircurex btc-e.com Mercado Bitcoin Canadian Virtual Exchange btcchina.com bitcoin-24.com VirWox Bitcoin.de Bitcoin Central Mt. Gox Bitcurex Kapiton bitstamp InterSango Bitoor Camp BX The Rock Trading Company bitme FYB-SG Origin NZ SE US/AU US? BG BR CA CN DE DE DE FR JP PL SE SL UK US US US US SG Dates Active 9/11 pres. 3/12 pres. 10/11 pres. 12/11 pres. 8/11 pres. 7/11 pres. 6/11 pres. 6/11 pres. 5/12 pres. 4/11 pres. 8/11 pres. 1/11 pres. 7/10 pres. 7/12 pres. 4/12 pres. 9/11 pres. 7/11 pres. 5/12 pres. 7/11 pres. 6/11 pres. 7/12 pres. 1/13 pres. Daily vol. 27 3 2 6 2604 67 832 473 924 1668 1204 118 43230 157 160 1274 2741 816 622 52 77 3 Closed? no no no no no no no no no no no no no no no no no no no no no no Breached? no no no yes yes no no no no no no no yes no no no no yes no no no no Repaid? yes yes no AML 21.3 27.0 30.0 27.9 32.3 24.3 25.0 24.0 26.0 26.0 26.0 31.7 22.7 21.7 27.0 35.3 35.3 34.3 34.3 34.3 34.3 33.7

Tyler Moore & Nicolas Christin

Empirical Analysis of Bitcoin-Exchange Risk

Data on Bitcoin-Exchange Closures Survival Analysis of Exchange Closure Regression Analysis of Exchange Breaches

Statistical Model Results Risk Ratio for Bitcoin Exchanges

Outline
1

Data on Bitcoin-Exchange Closures Data Collection Methodology Summary Statistics Survival Analysis of Exchange Closure Statistical Model Results Risk Ratio for Bitcoin Exchanges Regression Analysis of Exchange Breaches Statistical Model Results

Tyler Moore & Nicolas Christin

Empirical Analysis of Bitcoin-Exchange Risk

Data on Bitcoin-Exchange Closures Survival Analysis of Exchange Closure Regression Analysis of Exchange Breaches

Statistical Model Results Risk Ratio for Bitcoin Exchanges

What factors aect whether an exchange closes?

We hypothesize three variables aect survival time for a Bitcoin exchange


1 2 3

Average daily transaction volume (positive) Experiencing security breach (negative) AML/CFT compliance (negative)

Since lifetimes are censored, we construct a Cox proportional hazards model: hi (t) = h0 (t) exp(1 log(Daily vol.)i + 2 Breachedi + 3 AMLi ).

Tyler Moore & Nicolas Christin

Empirical Analysis of Bitcoin-Exchange Risk

Data on Bitcoin-Exchange Closures Survival Analysis of Exchange Closure Regression Analysis of Exchange Breaches

Statistical Model Results Risk Ratio for Bitcoin Exchanges

Cox proportional hazards model: results

log(Daily vol.)i Breachedi AMLi

1 2 3

coef. 0.173 0.857 0.004

exp(coef.) 0.840 2.36 1.004

Std. Err.) 0.072 0.572 0.042

Signicance p = 0.0156 p = 0.1338 p = 0.9221

log-rank test: Q=7.01 (p = 0.0715), R2 = 0.145

Higher daily transaction volumes associated with longer survival times (statistically signicant) Experiencing a breach associated with shorter survival times (not quite statistically signicant)
Tyler Moore & Nicolas Christin Empirical Analysis of Bitcoin-Exchange Risk

Data on Bitcoin-Exchange Closures Survival Analysis of Exchange Closure Regression Analysis of Exchange Breaches

Statistical Model Results Risk Ratio for Bitcoin Exchanges

Survival probability for Bitcoin exchanges


1.0
Average

Survival probability

0.0 0

0.2

0.4

0.6

0.8

200

400 Days

600

800

Tyler Moore & Nicolas Christin

Empirical Analysis of Bitcoin-Exchange Risk

Data on Bitcoin-Exchange Closures Survival Analysis of Exchange Closure Regression Analysis of Exchange Breaches

Statistical Model Results Risk Ratio for Bitcoin Exchanges

High-volume exchanges have better chance to survive


1.0
Mt. Gox Intersango Average

Survival probability

0.0 0

0.2

0.4

0.6

0.8

200

400 Days

600

800

Tyler Moore & Nicolas Christin

Empirical Analysis of Bitcoin-Exchange Risk

Data on Bitcoin-Exchange Closures Survival Analysis of Exchange Closure Regression Analysis of Exchange Breaches

Statistical Model Results Risk Ratio for Bitcoin Exchanges

Low-volume exchanges have worse chance to survive


1.0
Mt. Gox Intersango Bitcoin2Cash Average

Survival probability

0.0 0

0.2

0.4

0.6

0.8

200

400 Days

600

800

Tyler Moore & Nicolas Christin

Empirical Analysis of Bitcoin-Exchange Risk

Data on Bitcoin-Exchange Closures Survival Analysis of Exchange Closure Regression Analysis of Exchange Breaches

Statistical Model Results Risk Ratio for Bitcoin Exchanges

Yet some lower-risk exchanges collapse, high-risk survive


1.0
Mt. Gox Intersango

0.8

Bitcoin2Cash Vircurex Exchange BitCoins.com Average

Survival probability

0.0 0

0.2

0.4

0.6

200

400 Days

600

800

Tyler Moore & Nicolas Christin

Empirical Analysis of Bitcoin-Exchange Risk

Data on Bitcoin-Exchange Closures Survival Analysis of Exchange Closure Regression Analysis of Exchange Breaches

Statistical Model Results Risk Ratio for Bitcoin Exchanges

Risk ratio for all 40 exchanges


Survival Risk Ratio for Bitcoin Exchanges
Open (Breached)

G Open (Not Breached)

Closed (Breached)
G Closed (Not Breached)

risk ratio (1 = average)

G G G G G G G G GG G G GG G

G G G GG G G G G GG

0 1

10

100

1000

10000

average daily BTC transaction volume


Tyler Moore & Nicolas Christin Empirical Analysis of Bitcoin-Exchange Risk

Data on Bitcoin-Exchange Closures Survival Analysis of Exchange Closure Regression Analysis of Exchange Breaches

Statistical Model Results

Outline
1

Data on Bitcoin-Exchange Closures Data Collection Methodology Summary Statistics Survival Analysis of Exchange Closure Statistical Model Results Risk Ratio for Bitcoin Exchanges Regression Analysis of Exchange Breaches Statistical Model Results

Tyler Moore & Nicolas Christin

Empirical Analysis of Bitcoin-Exchange Risk

Data on Bitcoin-Exchange Closures Survival Analysis of Exchange Closure Regression Analysis of Exchange Breaches

Statistical Model Results

What factors aect whether an exchange is breached?

We hypothesize three variables aect whether a Bitcoin exchange loses money from a security breach
1 2

Average daily transaction volume (positive) Months operational (positive)

We use a logistic regression model with a dependent variable denoting whether or not an exchange experiences a breach: log (pb /(1 pb )) = c0 + c1 log(Daily vol.) + c2 months operational + .

Tyler Moore & Nicolas Christin

Empirical Analysis of Bitcoin-Exchange Risk

Data on Bitcoin-Exchange Closures Survival Analysis of Exchange Closure Regression Analysis of Exchange Breaches

Statistical Model Results

Logistic regression for exchange breaches

Intercept log(Daily vol.) Months operational

coef. 4.304 0.514 0.104

Odds-ratio 0.014 1.672 0.901

95% conf. int. (0.0002,0.211) (1.183,2.854) (0.771,1.025)

Signicance p = 0.0131 p = 0.0176 p = 0.1400

Model t: 2 = 10.3, p = 0.00579

Transaction volume is positively correlated with experiencing a breach (statistically signicant) Months operational is negatively correlated with being breached (not quite statistically signicant)
Tyler Moore & Nicolas Christin Empirical Analysis of Bitcoin-Exchange Risk

Data on Bitcoin-Exchange Closures Survival Analysis of Exchange Closure Regression Analysis of Exchange Breaches

Statistical Model Results

Breach probability as a function of daily transaction volume


Probability exchange has breach 1.0
Predicted probability 90% C.I.

0.0

0.2

0.4

0.6

0.8

50

500

5000

50000

Daily transaction volume at exchange


Tyler Moore & Nicolas Christin Empirical Analysis of Bitcoin-Exchange Risk

Data on Bitcoin-Exchange Closures Survival Analysis of Exchange Closure Regression Analysis of Exchange Breaches

Concluding remarks (1)


Currency exchanges pose substantial risk to Bitcoin holders: 45% of exchanges have closed, often leaving customers unable to withdraw stored funds Using survival analysis, we found that an exchanges average transaction volume is negatively correlated with the probability it will close prematurely Using regression, we found that transaction volume is positively correlated with experiencing a breach Hence, the continued operation of an exchange depends on running a high transaction volume, which makes the exchange a more valuable target to thieves

Tyler Moore & Nicolas Christin

Empirical Analysis of Bitcoin-Exchange Risk

Data on Bitcoin-Exchange Closures Survival Analysis of Exchange Closure Regression Analysis of Exchange Breaches

Concluding remarks (2)


Limitations to the statistical analysis
1

There is substantial randomness aecting when an exchange closes or is breached that is not captured by our mode Some of the explanatory variables did not achieve statistical signicance due to the datasets modest size

We focus on economic considerations, such as closure risk, that a rational actor should consider before transacting with an exchange But behavioral factors may explain participation better (e.g., Silk Road customers want exchanges that respect anonymity) Paper: http://lyle.smu.edu/~tylerm/fc13.pdf

Tyler Moore & Nicolas Christin

Empirical Analysis of Bitcoin-Exchange Risk