Exploring Dogecoin: Investment Viability

This analysis delves into the intricate dynamics of Dogecoin as a cryptocurrency. By examining the impact of network externalities, investor sentiment, and social media influences, it evaluates Dogecoin's potential as a viable investment. The findings reveal significant correlations with investor attention, enhancing understanding of its market movements and explosive price behaviors, shaped by cultural roots and speculative interest.

DA: Blockchain and AI in Financial Data Science
IS DOGECOIN A VIABLE INVESTMENT?
INSIGHTS FROM NETWORK AND BUBBLE EFFECTS
Ruoxin Xiao
Xinyu Ying
Hengxu Li
Kexin Liu
(Equally contributed)
June 2022

1/16

Is Dogecoin a Viable Investment? Insights from Network and Bubble Effects
Ruoxin Xiao, Xinyu Ying, Hengxu Li and Kexin Liu
June 2022
ABSTRACT
We find that three factors—Dogecoin network externalities, momentum, and tweet sentiment—
capture the time-series expected Dogecoin returns. Dogecoin returns are exposed to Dogecoin
network factors. We construct the network factors to capture the user adoption of Dogecoin.
Moreover, there is a strong time-series momentum effect, and proxies for investor attention
strongly forecast future Dogecoin returns. Lastly, we examine potential underlying mechanisms
of the Dogecoin price bubble.

2/16

1 Introduction
Dogecoin is a peer-to-peer, open-source cryptocurrency created in 2013. Unlike other
cryptocurrencies, which have marketed themselves on differentiating technical features,
Dogecoin’s allure likely stems from its cultural roots as a meme (Albi Nani 2022). The community
of Dogecoin users have accomplished some rather odd feats, from sponsoring a National
Association for Stock Car Auto Racing (NASCAR) driver (Frum, 2014), to sending the Jamaican
bobsledding team to the 2014 Winter Olympics in Sochi, Russia (Abbruzzese, 2014) as an homage
to the 1993 movie Cool Runnings and launching initiatives called “SaveDogemas” after hacking
breach.
The cultural characteristic of Dogecoin is inseparable from its low price, unlimited supply and
rapid transactions. Dogecoin is issued with the price of 0.00863 USD and the capitalization of 600
million in 2014, while a member of cryptocurrency asset class with the price of 0.117 USD and
the capitalization of 90 billion in 2022. Although Dogecoin’s departure from a strictly financial
focus to one that was founded on meme culture contributed to its perception as a “joke” currency
in popular business media (Hackett, 2020; Jeong, 2017), it has achieved a price increase of more
than 50 times and an increase of 10,000 times the market volume in its evolution, among which
Elon Musk plays a very important role in the price fluctuations of Dogecoin. In January 2021,
Dogecoin went up over 800% in 24 hours as a result of attention from Reddit users, at least in part
encouraged by tech baron Elon Musk and the GameStop short squeeze (Usman W. Chohan 2021).
In May 2021, Dogecoin hit its all-time high price of 0.627 USD following a comic strip of Elon
Musk lifting Dogecoin to the sky posted by himself.
More notably, Dogecoin’s value seemed unaffected by other cryptocurrencies, rising even as
the rest of the crypto market collapsed (McGuire, 2013). However, the infrastructure and code for
Dogecoin was derived from Litecoin, which attempted to improve on Bitcoin’s transaction speed
and introduce a new hashing algorithm (Padamavathi and Suresh, 2019). Under the Merge Mining,
Litecoin miners would be allowed to mine Dogecoin (which is based on the same hashing
algorithm with Litecoin) with no need to put in additional computer power, which means they have
more chances to gain a reward for their work without much extra cost. As a result, Litecoin miners
would be incentivized to secure Dogecoin’s network by mining Litecoin and Dogecoin
simultaneously. Extraordinarily, the market volume of Dogecoin exceeded Litecoin in April, 2021,
which is hard to imagine especially when Dogecoin’s system is dependent on Litecoin.
Several recent papers document empirical facts related to cryptocurrency investments. Liu and
Tsyvinski (2021) are the first to comprehensively study valuations of the cryptocurrency market
in the aggregate time series. They show that cryptocurrency market returns have low exposures to
risk factors of the other markets. However, the aggregate market returns can be predicted by
cryptocurrency-specific factors such as time-series momentum and investor attention. This paper
studies the time-series of Dogecoin returns, considering Dogecoin network, market, and social
media messages. We show that Dogecoin returns are driven and can be predicted by these specific
factors.

3/16

The paper is organized as follows. Section 2 describes the data. Section 3 examines the
Dogecoin network and builds the time-series return predictors. Section 4 investigates potential
mechanisms behind the Dogecoin price bubble and social media effects. Finally, Section 5
concludes.
2 Data
We collect trading data on Dogecoin and Litecoin available from CoinGecko.com, a leading
source of cryptocurrency price and volume data suitable for scholarly use (Tomás, 2022). The
price of a cryptoasset is calculated based on the pairings available and collected by
CoinGecko.com from various exchanges. The price shown is calculated using a global volumeweighted average price formula.
We use daily close prices to construct weekly coin returns. Our sample includes the two coins
from the beginning of 2014 to March 2022. Trading volume data became available in the last week
of 2013, and thus our sample period starts from the beginning of 2014.
We use three primary measures to proxy the network effect of user adoption: the number of
active addresses, the number of transaction count, and the number of payment count. We obtain
data on them from Coinmetrics.io. Google search data series are downloaded from Google.
We now document the main statistical properties of the time series of Dogecoin and Litecoin
returns. Figures 1, and 2 show the distribution of market returns and market log returns for the two
coins at daily, weekly, and monthly frequencies. Figure 3 plots price movements of Dogecoin and
Litecoin having some comovements post 2021, but are not perfectly correlated. Because of a large
discrepancy in the magnitude of these prices, in Figure 3 we divide Litecoin price by 1000 to
improve exposition. Table 1 compares the properties of Dogecoin returns with those of Litecoin
returns.
For a particular coin, its excess market return is constructed as the difference between its
market return and the risk-free rate measured as the one-month Treasury bill rate from the Federal
Reserve Bank of St. Louis’ website. Summary statistics are presented in Panel A of Table 1. During
the sample period, the Dogecoin returns have a markedly higher mean (4.10% per week) and
standard deviation (0.282) compared to the Litecoin returns (2.00% per week with a SD of 0.148).
However, the Sharpe ratios of Dogecoin returns (0.23) are comparable to those of Litecoin (0.17)
at the monthly frequency. At the daily and weekly frequencies, the negative Sharpe ratios do not
convey any useful meaning.

4/16

Figure 1: Dogecoin Return Distributions
Figure 2: Litecoin Return Distributions
Figure 3: Price Movements

5/16

Both the Dogecoin and Litecoin returns are positively skewed at all frequencies and have high
kurtosis. For Litecoin returns, the skewness increases from 1.13 at the daily frequency to 2.35 at
the weekly frequency, and to 3.93 at the monthly frequency, while the skewness of Dogecoin
returns has shown the opposite, consistently decreasing from 16.28 at the daily frequency to 3.78
at the monthly frequency.
Both the Dogecoin and Litecoin returns have high probabilities of exceptional negative and
positive daily returns. The probability of a -5% Dogecoin daily return is as high as 48.03 percent,
quadruple of that of Litecoin.
Table 1: Dogecoin and Litecoin Returns Summary Statistics
3 Cryptocurrency Factors
The theoretical literature has proposed a number of cryptocurrency-specific factors as drivers
of cryptocurrency prices and predictors of cryptocurrency returns. In this section, we develop and
investigate the implications of cryptocurrency-specific factors. We first construct Dogecoin

6/16

network factors. We find that Dogecoin returns are strongly exposed to the Dogecoin network
factors. We then test whether Dogecoin returns are predictable by studying whether different
cryptocurrency factors predict future Dogecoin returns. We consider momentum and proxies for
investor attention. All of these variables are specific to each coin markets. We find that momentum
and proxies for investor attention can account for future Dogecoin returns.
3.1 Dogecoin Network
The theoretical literature on cryptocurrencies has emphasized the importance of network
factors in cryptocurrency valuation (e.g., Cong, Li, and Wang 2019; Sockin and Xiong 2019;
Pagnotta and Buraschi 2018; Biais et al. 2018). In particular, the network effect of user adoption
can potentially play a central role in the valuation of cryptocurrencies. Because user adoption of
cryptocurrencies generates positive network externality, cryptocurrency prices respond to user
adoptions. Hence, variations in user adoptions of cryptocurrency networks could contribute to
movements in cryptocurrency prices.
We construct network factors of Dogecoin and test whether these factors can account for
variations in Dogecoin prices. We use three measures to proxy for network effects: the number of
active addresses, the number of transaction count, and the number of payment count. Thus, we
measure Dogecoin network growth using the active address growth, transaction count growth, and
payment count growth. Panel A of Table 2 reports the correlation across the Dogecoin network
factors we consider. The three primary measures correlate with each other positively, with
correlations ranging from 0.39 to 0.69.
Table 2: Dogecoin and Litecoin Return Loadings to Network Factors
We regress the returns for each coin market on each of the three measures of changes in the
coin network. Panel B of Table 2 presents the results using the network factors. The Dogecoin
returns positively correlate with all three of the individual Dogecoin network factors. The
coefficient on the transaction count, payment count, and active address growth measure is

7/16

significant at the 0.1%, 1%, and 5% level respectively. Consistent with the theoretical models,
these results suggest that the Dogecoin network factors that measure the network effect of user
adoptions are important drivers of Dogecoin prices.
3.2 Are Dogecoin Returns Predictable?
In this section, we test whether the Dogecoin returns are predictable. Motivated by existing
theoretical developments and empirical findings in financial markets, we test whether Dogecoin
returns are predictable by momentum and investor attention.
3.2.1 Cryptocurrency Momentum. We start by establishing that there is strong evidence of timeseries momentum at various time horizons. Panel A in Table 3 documents the Dogecoin timeseries momentum results in the regression setting. Specifically, we regress cumulative future
Dogecoin returns on current Dogecoin returns from the one-week to eight-week horizons. The
current Dogecoin returns positively and statistically significantly predict cumulative future
Dogecoin returns at the 0.1% level for all eight horizons. For example, a one-standard-deviation
increase in the current Dogecoin return leads to increases in cumulative future Dogecoin returns
of 7.06%, 8.75%, 10.72%, and 11.29% increases at the one-week, two-week, three-week, and fourweek horizons, respectively. Specifically, the one-week-ahead weekly return is that of buying
Dogecoin at 11:59:59 UTD Sunday and selling Dogecoin at 11:59:59 UTD one week later.
Table 3: Time-series Momentum
Additionally, we test whether the time-series momentum effect is linked to network
externalities, as suggested in Cong, Li, and Wang (2019). In their dynamic cryptocurrency
valuation model, the momentum effect is generated by the positive externality of the network effect
that is not incorporated into cryptocurrency prices immediately. That is, their model implies that
controlling for cryptocurrency adoption growth would subsume the time-series momentum effect.
In Table 4, we show that there is evidence that Dogecoin adoption growth positively predicts future

8/16

Dogecoin returns. However, controlling for Dogecoin adoption growth does not subsume the timeseries momentum effect documented presented earlier.
Table 4: Momentum and Network Effect
3.2.2 Cryptocurrency Investor Attention. The theoretical literature of cryptocurrencies has also
suggested that investor attention could potentially be linked to future cryptocurrency returns (e.g.,
Sockin and Xiong 2019). In this section, we investigate the role of investor attention in predicting
Dogecoin returns. Specifically, we construct the deviation of Google searches for the word
“Dogecoin” in a given week compared with the average of those in the preceding four weeks. We
standardize the Google search measure to have a mean of zero and a standard deviation of one. In
Panel A of Table 5, we report the results of regressing cumulative future Dogecoin returns from
one-week to eight-week horizons on the Google search measure. The Google search measure
statistically significantly predicts the one-week to five-week ahead cumulative Dogecoin returns
at the 0.1% level. The coefficient estimates of the six-week and eight-week horizons are positive
but are less statistically significant. A one-standard-deviation increase in searches leads to
increases in weekly returns of about 2% for the one-week ahead cumulative Dogecoin returns and
about 2.5% for the two week-ahead cumulative Dogecoin returns.
In Panel B of Table 4, we also report results for Litecoin. Despite being by far the second
largest cryptocurrency in terms of capitalization, the current Litecoin returns positively but
insignificantly predict week-ahead returns.

9/16

Table 5: Google Searches
3.2.3 Interaction between Momentum and Attention. We have shown that there are strong
effects of time-series momentum and investor attention in the cryptocurrency market. The equity
market research (e.g., Hong, Lim, and Stein 2000; Hou, Xiong, and Peng 2009) shows that there
is a strong relationship between momentum and investor attention. It is possible that these two
results capture the same underlying phenomenon. For example, Sockin and Xiong (2019) propose
a potential channel to generate momentum. In their model, momentum arises because users have
incorrect expectations about future prices—a mechanism similar to De Long et al. (1990). Their
model suggests that cryptocurrency momentum and investor attention could potentially arise from
the same underlying mechanism. The cryptocurrency momentum and investor attention results
could also interact with each other. For example, the Dogecoin time-series momentum effect may
be weaker at times of high investor attention, because there is little information leakage at times
of high investor attention.
Table 6: Google Searches and Past Returns
First, we show that the current investor attention of Dogecoin is indeed associated with current
and past market performance. We regress the current deviation in the Google searches on the
contemporaneous and the Dogecoin returns of the previous four weeks. Table 6 documents the

10/16

results. We find that the deviations in Google searches are positively and significantly associated
with contemporaneous and the previous week’s Dogecoin returns. The Google search measures
do not significantly correlate with past Dogecoin returns beyond one week. Intuitively, these
results suggest that investor attention is elevated after superior Dogecoin market performance.
We further test the interaction between the time-series momentum and the investor attention
phenomena. The results are reported in Table 7. In Table 7, we regress cumulative future Dogecoin
returns on current Dogecoin returns and Google search measures. We find that the coefficients to
the current Dogecoin returns are significant for all the horizons, and the coefficients to the Google
search measures are significant at the one-week and three-week horizons. The magnitudes of the
coefficients to the current Dogecoin returns are similar to the standalone estimates, but the ones to
the Google search measures drop by one level. For example, the one-week-ahead coefficients
under the univariate regressions are 1.39 and 0.20 for the current Dogecoin returns and the Google
search measures, respectively, while they are 1.30 and 0.068 under the bivariate regressions. These
results show that the investor attention results actually subsume the time-series momentum.
Table 7: Interaction between Momentum and Attention
4 Speculative Interest and Sentiment
As argued by Pástor and Veronesi (2003), it takes time for investors to fully learn and
understand emerging technologies, which can lead to price bubbles. We have seven years of
Dogecoin market return data spanning from 2014 up to today. There is a great deal of uncertainty
and learning about Dogecoin as an asset class during the period. Moreover, the uncertainty has
increased since the beginning of 2021. The analysis on the volatility of the Dogecoin market
returns also supports this assumption.
In this section, we test whether speculation and investor sentiment may be important drivers
of Dogecoin prices. We construct three direct measures of Dogecoin investor sentiment, which are
the count of neutral, positive, and negative tweets of Dogecoin on Twitter. We test whether
explosive periods of Dogecoin price strongly respond to the contemporaneous and expectations of
future speculative price growth. We find sound evidence that the explosive periods positively load
on the contemporaneous speculative price growth with significant coefficient estimates in a
logistic regression setting.

11/16

$4.1 Bubble Detection The PSY procedure, known as a practical real-time bubble detection method, was proposed by Phillips, Shi and Yu (2015a, b; PSY hereafter) and was originally designed to identify and date stamp explosive periods in asset prices. Now it has been widely applied to financial, commodity, and real estate markets as an early warning alert system for exuberance. The method is based on an ADF model specification for the fitted regression equation but uses flexible window widths in its implementation to take time-varying dynamics and structural breaks into consideration. PSY (2015a, b) shows that this algorithm is superior to the forward expanding and rolling window algorithms in bubble identification, especially when the sample period contains multiple bubbles. The dynamics of asset prices during crisis periods may be modeled as a random drift martingale process (Phillips and Shi, 2018) log 𝑃𝑡 = −𝐿𝑡 + log 𝑃𝑡−1 + 𝑢𝑡 in which 𝑃𝑡is the price of cryptocurrency at time 𝑡, 𝐿𝑡is a random sequence independent of 𝑢𝑡. The sequence 𝐿𝑡 produces a random drift in the observed price process following an asymmetric scaled uniform distribution. 𝐿𝑡represent a positive scale quantity measuring shock intensity and hence the process exhibits an overall downward trend. Let 𝑇𝑐 = ⌊𝑇 ∙ 𝑟𝑐⌋ for sample fraction 𝑟𝑐 > 0. In the two-regime specification, the logarithmic asset price follows the data generating mechanism (DGP), 𝑓(𝑥) = { 𝑐𝑇 −𝜂 + log 𝑃𝑡−1 + 𝑢𝑡 , if 𝑡 ∈ 𝑁 ≡ [1, 𝑇𝑐] −𝐿𝑡 + log 𝑃𝑡−1 + 𝑢𝑡, if 𝑡 ∈ 𝐶 ≡ [𝑇𝑐, 𝑇] in which 𝑁 denotes the normal market and 𝐶 denotes collapse periods. We use backwards sup augmented Dicky Fuller (BSADF) statistics to date stamp the possibly multiple explosive and collapse episodes for Dogecoin by using the backwards sup augmented Dicky Fuller (BSADF) statistics. The BSADF test repeatedly estimates the supremum ADF statistic by considering the backwards expanding time windows ending in one day. Figure 4 provide the starting and ending dates of explosivity episodes and duration in days bases according to PSY analysis. As shown in Table 8, there were 6 explosive episodes containing 22 days in total in Dogecoin Price. The longest explosivity period was 6 days, whereas the shortest episode of explosivity lasted only 1 day. Following Etienne et al. (2014), who defined the minimum duration of a bubble episode to be three days, Dogecoin prices exhibited 3 bubble periods in April and March 2021. Notably, there is no explosivity periods observed before 2021, during which prices were relatively low.$

12/16

Figure 4: Dogecoin Price Bubble and Periods
4.2 Logistic Regression
After examining the explosive periods of Dogecoin prices from 2019 to 2021 through PSY
process, we apply logistic regression to see whether the sentiment analysis based on tweets is
correlated with the likelihood of explosivity in Dogecoin prices. Let 𝑌 be a dummy variable
representing the periods of price explosivity, and 𝑋𝑖,𝑡represents different kinds of tweets towards
Dogecoin, which are neutral, positive, and negative.
log (
𝑃(𝑌 = 1 | 𝑋)
1 − 𝑃(𝑌 = 1 | 𝑋)
) = 𝛽0 + 𝛽𝑖𝑋𝑖,𝑡 + 𝜀𝑡
where 𝑌 denotes the dummy variable that takes the value of 1 in the presence of explosive periods
of price (i.e., if explosivity happens) and 0 otherwise. 𝛽0 denotes the constant; 𝑋𝑖,𝑡is a set of three
dummy variables, i.e., one represents neutral tweets, one represents positive tweets and another
accounting for negative tweets (2019-2021); and 𝜀𝑡is the error term that follows a logistic
distribution. The correlation coefficient (i.e. 𝛽𝑖) between tweets and explosive periods of Dogecoin
price is as shown in Table 8.
Table 8: Correlation Coefficients for Logistic Regression
According to the correlation coefficient we obtain from the Logistic Regression Model, neural
tweets and positive tweets are positively associated to the period of explosivity while negative
tweets have a negative relation. Then we can arrive at a conclusion that Dogecoin-specific tweets

13/16

do impact the presence of explosivity in Dogecoin price. These results suggest that tweets
regarding cryptocurrency can drive explosivity in Dogecoin prices. According to Shiller (2015),
price explosivity occurs through psychological contagion where the news regarding the increase
in prices spreads contagiously, prompting investors’ enthusiasm and bringing a larger group of
investors by drawing in with excitement about the past price accelerations. Our results can validate
this idea to some extents. This view makes it possible to predict future price dynamics or
explosivity based on past information. Moreover, new traders tend to rely on the public than private
information like Dogecoin-specific tweets that leads to herding behavior (Banerjee, 1992). The
herding behavior can also be caused by some news or reports anticipating certain events, which
may be helpful in predicting the price explosivity in the cryptocurrency markets.
5 Conclusion
We find that Dogecoin returns strongly respond to Dogecoin network factors, as suggested by
the theoretical literature on cryptocurrencies. The returns of cryptocurrency can be predicted by
two factors specific to its markets: momentum and investor attention. However, our empirical
results do not support the notion that there is only limited interaction between the momentum result
and the investor attention result. Moreover, our findings cast doubt on popular explanations that
the behavior of Dogecoin returns is driven by its functions as a stake in the future of blockchain
technology similar to Litecoin, or as a unit of account similar to other cryptoassets. Finally, we
document that the sentiment embodied in social media has a potential to affect Dogecoin prices
and lead to a temporary bubble.

14/16

References
Aldo, and Jorellano. 2021. How Litecoin And Dogecoin Created One Of The Most Robust PoW
Network. Litecoin Foundation
Biais, B., C. Bisiere, M. Bouvard, C. Casamatta, A. J. Menkveld. 2018. Equilibrium bitcoin
pricing. Working Paper, Toulouse School of Economics
CEX.IO product team. 2014. Dogecoin and Litecoin as an Example of Merged Mining. News,
CEX.IO
Chohan, U. W. 2021. A History of Dogecoin. Discussion Series: Notes on the 21st Century.
Cong, L. W., Y. Li, and N. Wang. 2019. Tokenomics: Dynamic adoption and valuation. Working
Paper, Columbia University
De Long, J. B., A. Shleifer, L. H. Summers, and R. J. Waldmann. 1990. Noise trader risk in
financial markets. Journal of Political Economy 98:703–738
Hong, H., T. Lim, and J. C. Stein. 2000. Bad news travels slowly: Size, analyst coverage, and the
profitability of momentum strategies. Journal of Finance 55:265–295
Hou, K., W. Xiong, and L. Peng. 2009. A tale of two anomalies: The implications of investor
attention for price and earnings momentum. Working Paper, Princeton University
Kuhn, D., and Powers, B. 2020. ‘It’s part and parcel of crypto’: how memes drive narrative and
value. CoinDesk
Liu, Yukun, and Aleh Tsyvinski, 2021, Risks and returns of cryptocurrency, Review of Financial
Studies 34, 2689–2727
McGuire, P. 2013. Such weird: the founders of dogecoin see the meme currency’s tipping point.
Vice
Palmer, J. 2018. My Joke Cryptocurrency Hit $2 Billion and Something Is Very Wrong. Vice
Pagnotta, E., and A. Buraschi. 2018. An equilibrium valuation of bitcoin and decentralized
network assets. Working Paper, Imperial College Business School

15/16

Pástor, L., and P. Veronesi. 2003. Stock valuation and learning about profitability. Journal of
Finance 58:1749– 1789
Phillips, P., and Shi, S. 2018. Real time monitoring of asset markets: bubbles and crises. Cowles
Foundation Discussion Papers
Tomás, David Vidal. 2022. Which cryptocurrency data sources should scholars use?
International Review of Financial Analysis 81, 102061.
Shahzad, S., Anas, M., and Bouri, E. 2022. Price explosiveness in cryptocurrencies and Elon
Musk's tweets. Finance Research Letters 5, 102695.
Sockin, M., and W. Xiong. 2019. A model of cryptocurrencies. Working Paper, Princeton
University

16/16